diff options
Diffstat (limited to 'arch')
393 files changed, 28813 insertions, 3151 deletions
diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index f7c96635d3b..c5739d6309d 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -509,7 +509,7 @@ config NR_CPUS depends on SMP default "64" -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool "Discontiguous Memory Support (EXPERIMENTAL)" depends on EXPERIMENTAL help @@ -518,6 +518,8 @@ config DISCONTIGMEM or have huge holes in the physical address space for other reasons. See <file:Documentation/vm/numa> for more. +source "mm/Kconfig" + config NUMA bool "NUMA Support (EXPERIMENTAL)" depends on DISCONTIGMEM diff --git a/arch/alpha/defconfig b/arch/alpha/defconfig index 5e39b7a7c8f..6da9c3dbde4 100644 --- a/arch/alpha/defconfig +++ b/arch/alpha/defconfig @@ -96,7 +96,7 @@ CONFIG_ALPHA_CORE_AGP=y CONFIG_ALPHA_BROKEN_IRQ_MASK=y CONFIG_EISA=y # CONFIG_SMP is not set -# CONFIG_DISCONTIGMEM is not set +# CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set CONFIG_VERBOSE_MCHECK=y CONFIG_VERBOSE_MCHECK_ON=1 CONFIG_PCI_LEGACY_PROC=y diff --git a/arch/alpha/mm/numa.c b/arch/alpha/mm/numa.c index ba81c4422aa..c7481d59b6d 100644 --- a/arch/alpha/mm/numa.c +++ b/arch/alpha/mm/numa.c @@ -327,8 +327,6 @@ void __init mem_init(void) extern char _text, _etext, _data, _edata; extern char __init_begin, __init_end; unsigned long nid, i; - struct page * lmem_map; - high_memory = (void *) __va(max_low_pfn << PAGE_SHIFT); reservedpages = 0; @@ -338,10 +336,10 @@ void __init mem_init(void) */ totalram_pages += free_all_bootmem_node(NODE_DATA(nid)); - lmem_map = node_mem_map(nid); pfn = NODE_DATA(nid)->node_start_pfn; for (i = 0; i < node_spanned_pages(nid); i++, pfn++) - if (page_is_ram(pfn) && PageReserved(lmem_map+i)) + if (page_is_ram(pfn) && + PageReserved(nid_page_nr(nid, i))) reservedpages++; } @@ -373,18 +371,18 @@ show_mem(void) show_free_areas(); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_online_node(nid) { - struct page * lmem_map = node_mem_map(nid); i = node_spanned_pages(nid); while (i-- > 0) { + struct page *page = nid_page_nr(nid, i); total++; - if (PageReserved(lmem_map+i)) + if (PageReserved(page)) reserved++; - else if (PageSwapCache(lmem_map+i)) + else if (PageSwapCache(page)) cached++; - else if (!page_count(lmem_map+i)) + else if (!page_count(page)) free++; else - shared += page_count(lmem_map + i) - 1; + shared += page_count(page) - 1; } } printk("%ld pages of RAM\n",total); diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index ee8a9ad7bbd..c8d94dcd8ef 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -157,7 +157,7 @@ config ARCH_RPC config ARCH_SA1100 bool "SA1100-based" select ISA - select DISCONTIGMEM + select ARCH_DISCONTIGMEM_ENABLE config ARCH_S3C2410 bool "Samsung S3C2410" @@ -346,7 +346,22 @@ config PREEMPT Say Y here if you are building a kernel for a desktop, embedded or real-time system. Say N if you are unsure. -config DISCONTIGMEM +config NO_IDLE_HZ + bool "Dynamic tick timer" + help + Select this option if you want to disable continuous timer ticks + and have them programmed to occur as required. This option saves + power as the system can remain in idle state for longer. + + By default dynamic tick is disabled during the boot, and can be + manually enabled with: + + echo 1 > /sys/devices/system/timer/timer0/dyn_tick + + Alternatively, if you want dynamic tick automatically enabled + during boot, pass "dyntick=enable" via the kernel command string. + +config ARCH_DISCONTIGMEM_ENABLE bool default (ARCH_LH7A40X && !LH7A40X_CONTIGMEM) help @@ -355,6 +370,8 @@ config DISCONTIGMEM or have huge holes in the physical address space for other reasons. See <file:Documentation/vm/numa> for more. +source "mm/Kconfig" + config LEDS bool "Timer and CPU usage LEDs" depends on ARCH_CDB89712 || ARCH_CO285 || ARCH_EBSA110 || \ diff --git a/arch/arm/boot/install.sh b/arch/arm/boot/install.sh index 935bb27369e..9f9bed20734 100644 --- a/arch/arm/boot/install.sh +++ b/arch/arm/boot/install.sh @@ -21,8 +21,8 @@ # # User may have a custom install script -if [ -x ~/bin/installkernel ]; then exec ~/bin/installkernel "$@"; fi -if [ -x /sbin/installkernel ]; then exec /sbin/installkernel "$@"; fi +if [ -x ~/bin/${CROSS_COMPILE}installkernel ]; then exec ~/bin/${CROSS_COMPILE}installkernel "$@"; fi +if [ -x /sbin/${CROSS_COMPILE}installkernel ]; then exec /sbin/${CROSS_COMPILE}installkernel "$@"; fi if [ "$(basename $2)" = "zImage" ]; then # Compressed install diff --git a/arch/arm/configs/enp2611_defconfig b/arch/arm/configs/enp2611_defconfig index 06fae4b6277..b8c51ee7f1b 100644 --- a/arch/arm/configs/enp2611_defconfig +++ b/arch/arm/configs/enp2611_defconfig @@ -1,14 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk2 -# Sun Mar 27 22:08:24 2005 +# Linux kernel version: 2.6.12-git6 +# Sat Jun 25 00:57:29 2005 # CONFIG_ARM=y CONFIG_MMU=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y CONFIG_GENERIC_CALIBRATE_DELAY=y -CONFIG_GENERIC_IOMAP=y # # Code maturity level options @@ -16,6 +15,7 @@ CONFIG_GENERIC_IOMAP=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -35,6 +35,8 @@ CONFIG_EMBEDDED=y CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_PRINTK=y +CONFIG_BUG=y CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y @@ -82,6 +84,7 @@ CONFIG_ARCH_IXP2000=y # CONFIG_ARCH_VERSATILE is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set +# CONFIG_ARCH_AAEC2000 is not set CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y # @@ -96,6 +99,7 @@ CONFIG_ARCH_ENP2611=y # CONFIG_ARCH_IXDP2800 is not set # CONFIG_ARCH_IXDP2401 is not set # CONFIG_ARCH_IXDP2801 is not set +# CONFIG_IXP2000_SUPPORT_BROKEN_PCI_IO is not set # # Processor Type @@ -106,7 +110,6 @@ CONFIG_CPU_32v5=y CONFIG_CPU_ABRT_EV5T=y CONFIG_CPU_CACHE_VIVT=y CONFIG_CPU_TLB_V4WBI=y -CONFIG_CPU_MINICACHE=y # # Processor Features @@ -118,9 +121,11 @@ CONFIG_XSCALE_PMU=y # # Bus support # +CONFIG_ISA_DMA_API=y CONFIG_PCI=y CONFIG_PCI_LEGACY_PROC=y CONFIG_PCI_NAMES=y +# CONFIG_PCI_DEBUG is not set # # PCCARD (PCMCIA/CardBus) support @@ -130,7 +135,15 @@ CONFIG_PCI_NAMES=y # # Kernel Features # +# CONFIG_SMP is not set # CONFIG_PREEMPT is not set +# CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set +CONFIG_SELECT_MEMORY_MODEL=y +CONFIG_FLATMEM_MANUAL=y +# CONFIG_DISCONTIGMEM_MANUAL is not set +# CONFIG_SPARSEMEM_MANUAL is not set +CONFIG_FLATMEM=y +CONFIG_FLAT_NODE_MEM_MAP=y CONFIG_ALIGNMENT_TRAP=y # @@ -269,7 +282,6 @@ CONFIG_MTD_IXP2000=y # # Block devices # -# CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_CPQ_CISS_DA is not set # CONFIG_BLK_DEV_DAC960 is not set @@ -308,6 +320,7 @@ CONFIG_IOSCHED_CFQ=y # # Fusion MPT device support # +# CONFIG_FUSION is not set # # IEEE 1394 (FireWire) support @@ -329,10 +342,11 @@ CONFIG_NET=y # CONFIG_PACKET=y CONFIG_PACKET_MMAP=y -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y +CONFIG_IP_FIB_HASH=y +# CONFIG_IP_FIB_TRIE is not set # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set CONFIG_IP_PNP=y @@ -349,6 +363,17 @@ CONFIG_SYN_COOKIES=y # CONFIG_INET_TUNNEL is not set # CONFIG_IP_TCPDIAG is not set # CONFIG_IP_TCPDIAG_IPV6 is not set + +# +# TCP congestion control +# +CONFIG_TCP_CONG_BIC=y +CONFIG_TCP_CONG_WESTWOOD=m +CONFIG_TCP_CONG_HTCP=m +# CONFIG_TCP_CONG_HSTCP is not set +# CONFIG_TCP_CONG_HYBLA is not set +# CONFIG_TCP_CONG_VEGAS is not set +# CONFIG_TCP_CONG_SCALABLE is not set # CONFIG_IPV6 is not set # CONFIG_NETFILTER is not set @@ -404,6 +429,7 @@ CONFIG_MII=y # CONFIG_SUNGEM is not set # CONFIG_NET_VENDOR_3COM is not set # CONFIG_SMC91X is not set +# CONFIG_DM9000 is not set # # Tulip family network device support @@ -440,9 +466,11 @@ CONFIG_EEPRO100=y # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set # CONFIG_R8169 is not set +# CONFIG_SKGE is not set # CONFIG_SK98LIN is not set # CONFIG_VIA_VELOCITY is not set # CONFIG_TIGON3 is not set +# CONFIG_BNX2 is not set # # Ethernet (10000 Mbit) @@ -464,6 +492,7 @@ CONFIG_EEPRO100=y # Wan interfaces # CONFIG_WAN=y +# CONFIG_DSCC4 is not set # CONFIG_LANMEDIA is not set # CONFIG_SYNCLINK_SYNCPPP is not set CONFIG_HDLC=y @@ -526,7 +555,6 @@ CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768 # # CONFIG_SERIO is not set # CONFIG_GAMEPORT is not set -CONFIG_SOUND_GAMEPORT=y # # Character devices @@ -547,6 +575,7 @@ CONFIG_SERIAL_8250_NR_UARTS=2 # CONFIG_SERIAL_CORE=y CONFIG_SERIAL_CORE_CONSOLE=y +# CONFIG_SERIAL_JSM is not set CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -613,17 +642,18 @@ CONFIG_I2C_ALGOBIT=y # CONFIG_I2C_AMD8111 is not set # CONFIG_I2C_I801 is not set # CONFIG_I2C_I810 is not set +# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_ISA is not set # CONFIG_I2C_IXP2000 is not set # CONFIG_I2C_NFORCE2 is not set # CONFIG_I2C_PARPORT_LIGHT is not set -# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_PROSAVAGE is not set # CONFIG_I2C_SAVAGE4 is not set # CONFIG_SCx200_ACB is not set # CONFIG_I2C_SIS5595 is not set # CONFIG_I2C_SIS630 is not set # CONFIG_I2C_SIS96X is not set +# CONFIG_I2C_STUB is not set # CONFIG_I2C_VIA is not set # CONFIG_I2C_VIAPRO is not set # CONFIG_I2C_VOODOO3 is not set @@ -637,7 +667,9 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_ADM1025 is not set # CONFIG_SENSORS_ADM1026 is not set # CONFIG_SENSORS_ADM1031 is not set +# CONFIG_SENSORS_ADM9240 is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set @@ -653,6 +685,7 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_LM85 is not set # CONFIG_SENSORS_LM87 is not set # CONFIG_SENSORS_LM90 is not set +# CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_SMSC47B397 is not set @@ -662,14 +695,19 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_W83781D is not set # CONFIG_SENSORS_W83L785TS is not set # CONFIG_SENSORS_W83627HF is not set +# CONFIG_SENSORS_W83627EHF is not set # # Other I2C Chip support # +# CONFIG_SENSORS_DS1337 is not set +# CONFIG_SENSORS_DS1374 is not set CONFIG_SENSORS_EEPROM=y # CONFIG_SENSORS_PCF8574 is not set +# CONFIG_SENSORS_PCA9539 is not set # CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_RTC8564 is not set +# CONFIG_SENSORS_MAX6875 is not set # CONFIG_I2C_DEBUG_CORE is not set # CONFIG_I2C_DEBUG_ALGO is not set # CONFIG_I2C_DEBUG_BUS is not set @@ -723,6 +761,7 @@ CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_POSIX_ACL=y # CONFIG_EXT2_FS_SECURITY is not set +# CONFIG_EXT2_FS_XIP is not set CONFIG_EXT3_FS=y CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y @@ -763,7 +802,6 @@ CONFIG_DNOTIFY=y # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVFS_FS is not set # CONFIG_DEVPTS_FS_XATTR is not set CONFIG_TMPFS=y # CONFIG_TMPFS_XATTR is not set @@ -801,12 +839,14 @@ CONFIG_JFFS2_RTIME=y # CONFIG_NFS_FS=y CONFIG_NFS_V3=y +# CONFIG_NFS_V3_ACL is not set # CONFIG_NFS_V4 is not set # CONFIG_NFS_DIRECTIO is not set # CONFIG_NFSD is not set CONFIG_ROOT_NFS=y CONFIG_LOCKD=y CONFIG_LOCKD_V4=y +CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y # CONFIG_RPCSEC_GSS_KRB5 is not set # CONFIG_RPCSEC_GSS_SPKM3 is not set @@ -891,3 +931,4 @@ CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y CONFIG_ZLIB_DEFLATE=y +# CONFIG_TEXTSEARCH is not set diff --git a/arch/arm/configs/ixdp2400_defconfig b/arch/arm/configs/ixdp2400_defconfig index 810a450a55d..3cfbe2ec29c 100644 --- a/arch/arm/configs/ixdp2400_defconfig +++ b/arch/arm/configs/ixdp2400_defconfig @@ -1,14 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk2 -# Sun Mar 27 21:13:38 2005 +# Linux kernel version: 2.6.12-git6 +# Sat Jun 25 00:58:38 2005 # CONFIG_ARM=y CONFIG_MMU=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y CONFIG_GENERIC_CALIBRATE_DELAY=y -CONFIG_GENERIC_IOMAP=y # # Code maturity level options @@ -16,6 +15,7 @@ CONFIG_GENERIC_IOMAP=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -35,6 +35,8 @@ CONFIG_EMBEDDED=y CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_PRINTK=y +CONFIG_BUG=y CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y @@ -82,6 +84,7 @@ CONFIG_ARCH_IXP2000=y # CONFIG_ARCH_VERSATILE is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set +# CONFIG_ARCH_AAEC2000 is not set CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y # @@ -97,6 +100,7 @@ CONFIG_ARCH_IXDP2400=y CONFIG_ARCH_IXDP2X00=y # CONFIG_ARCH_IXDP2401 is not set # CONFIG_ARCH_IXDP2801 is not set +# CONFIG_IXP2000_SUPPORT_BROKEN_PCI_IO is not set # # Processor Type @@ -107,7 +111,6 @@ CONFIG_CPU_32v5=y CONFIG_CPU_ABRT_EV5T=y CONFIG_CPU_CACHE_VIVT=y CONFIG_CPU_TLB_V4WBI=y -CONFIG_CPU_MINICACHE=y # # Processor Features @@ -119,9 +122,11 @@ CONFIG_XSCALE_PMU=y # # Bus support # +CONFIG_ISA_DMA_API=y CONFIG_PCI=y CONFIG_PCI_LEGACY_PROC=y CONFIG_PCI_NAMES=y +# CONFIG_PCI_DEBUG is not set # # PCCARD (PCMCIA/CardBus) support @@ -131,7 +136,15 @@ CONFIG_PCI_NAMES=y # # Kernel Features # +# CONFIG_SMP is not set # CONFIG_PREEMPT is not set +# CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set +CONFIG_SELECT_MEMORY_MODEL=y +CONFIG_FLATMEM_MANUAL=y +# CONFIG_DISCONTIGMEM_MANUAL is not set +# CONFIG_SPARSEMEM_MANUAL is not set +CONFIG_FLATMEM=y +CONFIG_FLAT_NODE_MEM_MAP=y CONFIG_ALIGNMENT_TRAP=y # @@ -270,7 +283,6 @@ CONFIG_MTD_IXP2000=y # # Block devices # -# CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_CPQ_CISS_DA is not set # CONFIG_BLK_DEV_DAC960 is not set @@ -309,6 +321,7 @@ CONFIG_IOSCHED_CFQ=y # # Fusion MPT device support # +# CONFIG_FUSION is not set # # IEEE 1394 (FireWire) support @@ -330,10 +343,11 @@ CONFIG_NET=y # CONFIG_PACKET=y CONFIG_PACKET_MMAP=y -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y +CONFIG_IP_FIB_HASH=y +# CONFIG_IP_FIB_TRIE is not set # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set CONFIG_IP_PNP=y @@ -350,6 +364,17 @@ CONFIG_SYN_COOKIES=y # CONFIG_INET_TUNNEL is not set # CONFIG_IP_TCPDIAG is not set # CONFIG_IP_TCPDIAG_IPV6 is not set + +# +# TCP congestion control +# +CONFIG_TCP_CONG_BIC=y +CONFIG_TCP_CONG_WESTWOOD=m +CONFIG_TCP_CONG_HTCP=m +# CONFIG_TCP_CONG_HSTCP is not set +# CONFIG_TCP_CONG_HYBLA is not set +# CONFIG_TCP_CONG_VEGAS is not set +# CONFIG_TCP_CONG_SCALABLE is not set # CONFIG_IPV6 is not set # CONFIG_NETFILTER is not set @@ -405,6 +430,7 @@ CONFIG_MII=y # CONFIG_SUNGEM is not set # CONFIG_NET_VENDOR_3COM is not set # CONFIG_SMC91X is not set +# CONFIG_DM9000 is not set # # Tulip family network device support @@ -441,9 +467,11 @@ CONFIG_EEPRO100=y # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set # CONFIG_R8169 is not set +# CONFIG_SKGE is not set # CONFIG_SK98LIN is not set # CONFIG_VIA_VELOCITY is not set # CONFIG_TIGON3 is not set +# CONFIG_BNX2 is not set # # Ethernet (10000 Mbit) @@ -465,6 +493,7 @@ CONFIG_EEPRO100=y # Wan interfaces # CONFIG_WAN=y +# CONFIG_DSCC4 is not set # CONFIG_LANMEDIA is not set # CONFIG_SYNCLINK_SYNCPPP is not set CONFIG_HDLC=y @@ -527,7 +556,6 @@ CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768 # # CONFIG_SERIO is not set # CONFIG_GAMEPORT is not set -CONFIG_SOUND_GAMEPORT=y # # Character devices @@ -548,6 +576,7 @@ CONFIG_SERIAL_8250_NR_UARTS=2 # CONFIG_SERIAL_CORE=y CONFIG_SERIAL_CORE_CONSOLE=y +# CONFIG_SERIAL_JSM is not set CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -614,17 +643,18 @@ CONFIG_I2C_ALGOBIT=y # CONFIG_I2C_AMD8111 is not set # CONFIG_I2C_I801 is not set # CONFIG_I2C_I810 is not set +# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_ISA is not set # CONFIG_I2C_IXP2000 is not set # CONFIG_I2C_NFORCE2 is not set # CONFIG_I2C_PARPORT_LIGHT is not set -# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_PROSAVAGE is not set # CONFIG_I2C_SAVAGE4 is not set # CONFIG_SCx200_ACB is not set # CONFIG_I2C_SIS5595 is not set # CONFIG_I2C_SIS630 is not set # CONFIG_I2C_SIS96X is not set +# CONFIG_I2C_STUB is not set # CONFIG_I2C_VIA is not set # CONFIG_I2C_VIAPRO is not set # CONFIG_I2C_VOODOO3 is not set @@ -638,7 +668,9 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_ADM1025 is not set # CONFIG_SENSORS_ADM1026 is not set # CONFIG_SENSORS_ADM1031 is not set +# CONFIG_SENSORS_ADM9240 is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set @@ -654,6 +686,7 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_LM85 is not set # CONFIG_SENSORS_LM87 is not set # CONFIG_SENSORS_LM90 is not set +# CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_SMSC47B397 is not set @@ -663,14 +696,19 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_W83781D is not set # CONFIG_SENSORS_W83L785TS is not set # CONFIG_SENSORS_W83627HF is not set +# CONFIG_SENSORS_W83627EHF is not set # # Other I2C Chip support # +# CONFIG_SENSORS_DS1337 is not set +# CONFIG_SENSORS_DS1374 is not set CONFIG_SENSORS_EEPROM=y # CONFIG_SENSORS_PCF8574 is not set +# CONFIG_SENSORS_PCA9539 is not set # CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_RTC8564 is not set +# CONFIG_SENSORS_MAX6875 is not set # CONFIG_I2C_DEBUG_CORE is not set # CONFIG_I2C_DEBUG_ALGO is not set # CONFIG_I2C_DEBUG_BUS is not set @@ -724,6 +762,7 @@ CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_POSIX_ACL=y # CONFIG_EXT2_FS_SECURITY is not set +# CONFIG_EXT2_FS_XIP is not set CONFIG_EXT3_FS=y CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y @@ -764,7 +803,6 @@ CONFIG_DNOTIFY=y # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVFS_FS is not set # CONFIG_DEVPTS_FS_XATTR is not set CONFIG_TMPFS=y # CONFIG_TMPFS_XATTR is not set @@ -802,12 +840,14 @@ CONFIG_JFFS2_RTIME=y # CONFIG_NFS_FS=y CONFIG_NFS_V3=y +# CONFIG_NFS_V3_ACL is not set # CONFIG_NFS_V4 is not set # CONFIG_NFS_DIRECTIO is not set # CONFIG_NFSD is not set CONFIG_ROOT_NFS=y CONFIG_LOCKD=y CONFIG_LOCKD_V4=y +CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y # CONFIG_RPCSEC_GSS_KRB5 is not set # CONFIG_RPCSEC_GSS_SPKM3 is not set @@ -892,3 +932,4 @@ CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y CONFIG_ZLIB_DEFLATE=y +# CONFIG_TEXTSEARCH is not set diff --git a/arch/arm/configs/ixdp2401_defconfig b/arch/arm/configs/ixdp2401_defconfig index 72e1b940e97..5c87e8e6969 100644 --- a/arch/arm/configs/ixdp2401_defconfig +++ b/arch/arm/configs/ixdp2401_defconfig @@ -1,14 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk2 -# Sun Mar 27 21:53:55 2005 +# Linux kernel version: 2.6.12-git6 +# Sat Jun 25 00:59:35 2005 # CONFIG_ARM=y CONFIG_MMU=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y CONFIG_GENERIC_CALIBRATE_DELAY=y -CONFIG_GENERIC_IOMAP=y # # Code maturity level options @@ -16,6 +15,7 @@ CONFIG_GENERIC_IOMAP=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -35,6 +35,8 @@ CONFIG_EMBEDDED=y CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_PRINTK=y +CONFIG_BUG=y CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y @@ -82,6 +84,7 @@ CONFIG_ARCH_IXP2000=y # CONFIG_ARCH_VERSATILE is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set +# CONFIG_ARCH_AAEC2000 is not set CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y # @@ -97,6 +100,7 @@ CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y CONFIG_ARCH_IXDP2401=y # CONFIG_ARCH_IXDP2801 is not set CONFIG_ARCH_IXDP2X01=y +# CONFIG_IXP2000_SUPPORT_BROKEN_PCI_IO is not set # # Processor Type @@ -107,7 +111,6 @@ CONFIG_CPU_32v5=y CONFIG_CPU_ABRT_EV5T=y CONFIG_CPU_CACHE_VIVT=y CONFIG_CPU_TLB_V4WBI=y -CONFIG_CPU_MINICACHE=y # # Processor Features @@ -119,9 +122,11 @@ CONFIG_XSCALE_PMU=y # # Bus support # +CONFIG_ISA_DMA_API=y CONFIG_PCI=y CONFIG_PCI_LEGACY_PROC=y CONFIG_PCI_NAMES=y +# CONFIG_PCI_DEBUG is not set # # PCCARD (PCMCIA/CardBus) support @@ -131,7 +136,15 @@ CONFIG_PCI_NAMES=y # # Kernel Features # +# CONFIG_SMP is not set # CONFIG_PREEMPT is not set +# CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set +CONFIG_SELECT_MEMORY_MODEL=y +CONFIG_FLATMEM_MANUAL=y +# CONFIG_DISCONTIGMEM_MANUAL is not set +# CONFIG_SPARSEMEM_MANUAL is not set +CONFIG_FLATMEM=y +CONFIG_FLAT_NODE_MEM_MAP=y CONFIG_ALIGNMENT_TRAP=y # @@ -270,7 +283,6 @@ CONFIG_MTD_IXP2000=y # # Block devices # -# CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_CPQ_CISS_DA is not set # CONFIG_BLK_DEV_DAC960 is not set @@ -309,6 +321,7 @@ CONFIG_IOSCHED_CFQ=y # # Fusion MPT device support # +# CONFIG_FUSION is not set # # IEEE 1394 (FireWire) support @@ -330,10 +343,11 @@ CONFIG_NET=y # CONFIG_PACKET=y CONFIG_PACKET_MMAP=y -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y +CONFIG_IP_FIB_HASH=y +# CONFIG_IP_FIB_TRIE is not set # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set CONFIG_IP_PNP=y @@ -350,6 +364,17 @@ CONFIG_SYN_COOKIES=y # CONFIG_INET_TUNNEL is not set CONFIG_IP_TCPDIAG=y # CONFIG_IP_TCPDIAG_IPV6 is not set + +# +# TCP congestion control +# +CONFIG_TCP_CONG_BIC=y +CONFIG_TCP_CONG_WESTWOOD=m +CONFIG_TCP_CONG_HTCP=m +# CONFIG_TCP_CONG_HSTCP is not set +# CONFIG_TCP_CONG_HYBLA is not set +# CONFIG_TCP_CONG_VEGAS is not set +# CONFIG_TCP_CONG_SCALABLE is not set # CONFIG_IPV6 is not set # CONFIG_NETFILTER is not set @@ -405,6 +430,7 @@ CONFIG_MII=y # CONFIG_SUNGEM is not set # CONFIG_NET_VENDOR_3COM is not set # CONFIG_SMC91X is not set +# CONFIG_DM9000 is not set # # Tulip family network device support @@ -442,9 +468,11 @@ CONFIG_EEPRO100=y # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set # CONFIG_R8169 is not set +# CONFIG_SKGE is not set # CONFIG_SK98LIN is not set # CONFIG_VIA_VELOCITY is not set # CONFIG_TIGON3 is not set +# CONFIG_BNX2 is not set # # Ethernet (10000 Mbit) @@ -466,6 +494,7 @@ CONFIG_EEPRO100=y # Wan interfaces # CONFIG_WAN=y +# CONFIG_DSCC4 is not set # CONFIG_LANMEDIA is not set # CONFIG_SYNCLINK_SYNCPPP is not set CONFIG_HDLC=y @@ -528,7 +557,6 @@ CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768 # # CONFIG_SERIO is not set # CONFIG_GAMEPORT is not set -CONFIG_SOUND_GAMEPORT=y # # Character devices @@ -549,6 +577,7 @@ CONFIG_SERIAL_8250_NR_UARTS=2 # CONFIG_SERIAL_CORE=y CONFIG_SERIAL_CORE_CONSOLE=y +# CONFIG_SERIAL_JSM is not set CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -615,17 +644,18 @@ CONFIG_I2C_ALGOBIT=y # CONFIG_I2C_AMD8111 is not set # CONFIG_I2C_I801 is not set # CONFIG_I2C_I810 is not set +# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_ISA is not set # CONFIG_I2C_IXP2000 is not set # CONFIG_I2C_NFORCE2 is not set # CONFIG_I2C_PARPORT_LIGHT is not set -# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_PROSAVAGE is not set # CONFIG_I2C_SAVAGE4 is not set # CONFIG_SCx200_ACB is not set # CONFIG_I2C_SIS5595 is not set # CONFIG_I2C_SIS630 is not set # CONFIG_I2C_SIS96X is not set +# CONFIG_I2C_STUB is not set # CONFIG_I2C_VIA is not set # CONFIG_I2C_VIAPRO is not set # CONFIG_I2C_VOODOO3 is not set @@ -639,7 +669,9 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_ADM1025 is not set # CONFIG_SENSORS_ADM1026 is not set # CONFIG_SENSORS_ADM1031 is not set +# CONFIG_SENSORS_ADM9240 is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set @@ -655,6 +687,7 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_LM85 is not set # CONFIG_SENSORS_LM87 is not set # CONFIG_SENSORS_LM90 is not set +# CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_SMSC47B397 is not set @@ -664,14 +697,19 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_W83781D is not set # CONFIG_SENSORS_W83L785TS is not set # CONFIG_SENSORS_W83627HF is not set +# CONFIG_SENSORS_W83627EHF is not set # # Other I2C Chip support # +# CONFIG_SENSORS_DS1337 is not set +# CONFIG_SENSORS_DS1374 is not set CONFIG_SENSORS_EEPROM=y # CONFIG_SENSORS_PCF8574 is not set +# CONFIG_SENSORS_PCA9539 is not set # CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_RTC8564 is not set +# CONFIG_SENSORS_MAX6875 is not set # CONFIG_I2C_DEBUG_CORE is not set # CONFIG_I2C_DEBUG_ALGO is not set # CONFIG_I2C_DEBUG_BUS is not set @@ -725,6 +763,7 @@ CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_POSIX_ACL=y # CONFIG_EXT2_FS_SECURITY is not set +# CONFIG_EXT2_FS_XIP is not set CONFIG_EXT3_FS=y CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y @@ -765,7 +804,6 @@ CONFIG_DNOTIFY=y # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVFS_FS is not set # CONFIG_DEVPTS_FS_XATTR is not set CONFIG_TMPFS=y # CONFIG_TMPFS_XATTR is not set @@ -803,12 +841,14 @@ CONFIG_JFFS2_RTIME=y # CONFIG_NFS_FS=y CONFIG_NFS_V3=y +# CONFIG_NFS_V3_ACL is not set # CONFIG_NFS_V4 is not set # CONFIG_NFS_DIRECTIO is not set # CONFIG_NFSD is not set CONFIG_ROOT_NFS=y CONFIG_LOCKD=y CONFIG_LOCKD_V4=y +CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y # CONFIG_RPCSEC_GSS_KRB5 is not set # CONFIG_RPCSEC_GSS_SPKM3 is not set @@ -893,3 +933,4 @@ CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y CONFIG_ZLIB_DEFLATE=y +# CONFIG_TEXTSEARCH is not set diff --git a/arch/arm/configs/ixdp2800_defconfig b/arch/arm/configs/ixdp2800_defconfig index 1592e45f027..3cb561a551c 100644 --- a/arch/arm/configs/ixdp2800_defconfig +++ b/arch/arm/configs/ixdp2800_defconfig @@ -1,14 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk2 -# Sun Mar 27 22:15:23 2005 +# Linux kernel version: 2.6.12-git6 +# Sat Jun 25 01:00:27 2005 # CONFIG_ARM=y CONFIG_MMU=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y CONFIG_GENERIC_CALIBRATE_DELAY=y -CONFIG_GENERIC_IOMAP=y # # Code maturity level options @@ -16,6 +15,7 @@ CONFIG_GENERIC_IOMAP=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -35,6 +35,8 @@ CONFIG_EMBEDDED=y CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_PRINTK=y +CONFIG_BUG=y CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y @@ -82,6 +84,7 @@ CONFIG_ARCH_IXP2000=y # CONFIG_ARCH_VERSATILE is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set +# CONFIG_ARCH_AAEC2000 is not set CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y # @@ -97,6 +100,7 @@ CONFIG_ARCH_IXDP2800=y CONFIG_ARCH_IXDP2X00=y # CONFIG_ARCH_IXDP2401 is not set # CONFIG_ARCH_IXDP2801 is not set +# CONFIG_IXP2000_SUPPORT_BROKEN_PCI_IO is not set # # Processor Type @@ -107,7 +111,6 @@ CONFIG_CPU_32v5=y CONFIG_CPU_ABRT_EV5T=y CONFIG_CPU_CACHE_VIVT=y CONFIG_CPU_TLB_V4WBI=y -CONFIG_CPU_MINICACHE=y # # Processor Features @@ -119,9 +122,11 @@ CONFIG_XSCALE_PMU=y # # Bus support # +CONFIG_ISA_DMA_API=y CONFIG_PCI=y CONFIG_PCI_LEGACY_PROC=y CONFIG_PCI_NAMES=y +# CONFIG_PCI_DEBUG is not set # # PCCARD (PCMCIA/CardBus) support @@ -131,7 +136,15 @@ CONFIG_PCI_NAMES=y # # Kernel Features # +# CONFIG_SMP is not set # CONFIG_PREEMPT is not set +# CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set +CONFIG_SELECT_MEMORY_MODEL=y +CONFIG_FLATMEM_MANUAL=y +# CONFIG_DISCONTIGMEM_MANUAL is not set +# CONFIG_SPARSEMEM_MANUAL is not set +CONFIG_FLATMEM=y +CONFIG_FLAT_NODE_MEM_MAP=y CONFIG_ALIGNMENT_TRAP=y # @@ -270,7 +283,6 @@ CONFIG_MTD_IXP2000=y # # Block devices # -# CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_CPQ_CISS_DA is not set # CONFIG_BLK_DEV_DAC960 is not set @@ -309,6 +321,7 @@ CONFIG_IOSCHED_CFQ=y # # Fusion MPT device support # +# CONFIG_FUSION is not set # # IEEE 1394 (FireWire) support @@ -330,10 +343,11 @@ CONFIG_NET=y # CONFIG_PACKET=y CONFIG_PACKET_MMAP=y -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y +CONFIG_IP_FIB_HASH=y +# CONFIG_IP_FIB_TRIE is not set # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set CONFIG_IP_PNP=y @@ -350,6 +364,17 @@ CONFIG_SYN_COOKIES=y # CONFIG_INET_TUNNEL is not set # CONFIG_IP_TCPDIAG is not set # CONFIG_IP_TCPDIAG_IPV6 is not set + +# +# TCP congestion control +# +CONFIG_TCP_CONG_BIC=y +CONFIG_TCP_CONG_WESTWOOD=m +CONFIG_TCP_CONG_HTCP=m +# CONFIG_TCP_CONG_HSTCP is not set +# CONFIG_TCP_CONG_HYBLA is not set +# CONFIG_TCP_CONG_VEGAS is not set +# CONFIG_TCP_CONG_SCALABLE is not set # CONFIG_IPV6 is not set # CONFIG_NETFILTER is not set @@ -405,6 +430,7 @@ CONFIG_MII=y # CONFIG_SUNGEM is not set # CONFIG_NET_VENDOR_3COM is not set # CONFIG_SMC91X is not set +# CONFIG_DM9000 is not set # # Tulip family network device support @@ -441,9 +467,11 @@ CONFIG_EEPRO100=y # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set # CONFIG_R8169 is not set +# CONFIG_SKGE is not set # CONFIG_SK98LIN is not set # CONFIG_VIA_VELOCITY is not set # CONFIG_TIGON3 is not set +# CONFIG_BNX2 is not set # # Ethernet (10000 Mbit) @@ -465,6 +493,7 @@ CONFIG_EEPRO100=y # Wan interfaces # CONFIG_WAN=y +# CONFIG_DSCC4 is not set # CONFIG_LANMEDIA is not set # CONFIG_SYNCLINK_SYNCPPP is not set CONFIG_HDLC=y @@ -527,7 +556,6 @@ CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768 # # CONFIG_SERIO is not set # CONFIG_GAMEPORT is not set -CONFIG_SOUND_GAMEPORT=y # # Character devices @@ -548,6 +576,7 @@ CONFIG_SERIAL_8250_NR_UARTS=2 # CONFIG_SERIAL_CORE=y CONFIG_SERIAL_CORE_CONSOLE=y +# CONFIG_SERIAL_JSM is not set CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -614,17 +643,18 @@ CONFIG_I2C_ALGOBIT=y # CONFIG_I2C_AMD8111 is not set # CONFIG_I2C_I801 is not set # CONFIG_I2C_I810 is not set +# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_ISA is not set # CONFIG_I2C_IXP2000 is not set # CONFIG_I2C_NFORCE2 is not set # CONFIG_I2C_PARPORT_LIGHT is not set -# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_PROSAVAGE is not set # CONFIG_I2C_SAVAGE4 is not set # CONFIG_SCx200_ACB is not set # CONFIG_I2C_SIS5595 is not set # CONFIG_I2C_SIS630 is not set # CONFIG_I2C_SIS96X is not set +# CONFIG_I2C_STUB is not set # CONFIG_I2C_VIA is not set # CONFIG_I2C_VIAPRO is not set # CONFIG_I2C_VOODOO3 is not set @@ -638,7 +668,9 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_ADM1025 is not set # CONFIG_SENSORS_ADM1026 is not set # CONFIG_SENSORS_ADM1031 is not set +# CONFIG_SENSORS_ADM9240 is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set @@ -654,6 +686,7 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_LM85 is not set # CONFIG_SENSORS_LM87 is not set # CONFIG_SENSORS_LM90 is not set +# CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_SMSC47B397 is not set @@ -663,14 +696,19 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_W83781D is not set # CONFIG_SENSORS_W83L785TS is not set # CONFIG_SENSORS_W83627HF is not set +# CONFIG_SENSORS_W83627EHF is not set # # Other I2C Chip support # +# CONFIG_SENSORS_DS1337 is not set +# CONFIG_SENSORS_DS1374 is not set CONFIG_SENSORS_EEPROM=y # CONFIG_SENSORS_PCF8574 is not set +# CONFIG_SENSORS_PCA9539 is not set # CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_RTC8564 is not set +# CONFIG_SENSORS_MAX6875 is not set # CONFIG_I2C_DEBUG_CORE is not set # CONFIG_I2C_DEBUG_ALGO is not set # CONFIG_I2C_DEBUG_BUS is not set @@ -724,6 +762,7 @@ CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_POSIX_ACL=y # CONFIG_EXT2_FS_SECURITY is not set +# CONFIG_EXT2_FS_XIP is not set CONFIG_EXT3_FS=y CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y @@ -764,7 +803,6 @@ CONFIG_DNOTIFY=y # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVFS_FS is not set # CONFIG_DEVPTS_FS_XATTR is not set CONFIG_TMPFS=y # CONFIG_TMPFS_XATTR is not set @@ -802,12 +840,14 @@ CONFIG_JFFS2_RTIME=y # CONFIG_NFS_FS=y CONFIG_NFS_V3=y +# CONFIG_NFS_V3_ACL is not set # CONFIG_NFS_V4 is not set # CONFIG_NFS_DIRECTIO is not set # CONFIG_NFSD is not set CONFIG_ROOT_NFS=y CONFIG_LOCKD=y CONFIG_LOCKD_V4=y +CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y # CONFIG_RPCSEC_GSS_KRB5 is not set # CONFIG_RPCSEC_GSS_SPKM3 is not set @@ -892,3 +932,4 @@ CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y CONFIG_ZLIB_DEFLATE=y +# CONFIG_TEXTSEARCH is not set diff --git a/arch/arm/configs/ixdp2801_defconfig b/arch/arm/configs/ixdp2801_defconfig index f1afe3d09ec..b1e162f29cb 100644 --- a/arch/arm/configs/ixdp2801_defconfig +++ b/arch/arm/configs/ixdp2801_defconfig @@ -1,14 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk2 -# Sun Mar 27 22:39:19 2005 +# Linux kernel version: 2.6.12-git6 +# Sat Jun 25 01:01:18 2005 # CONFIG_ARM=y CONFIG_MMU=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y CONFIG_GENERIC_CALIBRATE_DELAY=y -CONFIG_GENERIC_IOMAP=y # # Code maturity level options @@ -16,6 +15,7 @@ CONFIG_GENERIC_IOMAP=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -35,6 +35,8 @@ CONFIG_EMBEDDED=y CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_PRINTK=y +CONFIG_BUG=y CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y @@ -82,6 +84,7 @@ CONFIG_ARCH_IXP2000=y # CONFIG_ARCH_VERSATILE is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set +# CONFIG_ARCH_AAEC2000 is not set CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y # @@ -97,6 +100,7 @@ CONFIG_ARCH_SUPPORTS_BIG_ENDIAN=y # CONFIG_ARCH_IXDP2401 is not set CONFIG_ARCH_IXDP2801=y CONFIG_ARCH_IXDP2X01=y +# CONFIG_IXP2000_SUPPORT_BROKEN_PCI_IO is not set # # Processor Type @@ -107,7 +111,6 @@ CONFIG_CPU_32v5=y CONFIG_CPU_ABRT_EV5T=y CONFIG_CPU_CACHE_VIVT=y CONFIG_CPU_TLB_V4WBI=y -CONFIG_CPU_MINICACHE=y # # Processor Features @@ -119,9 +122,11 @@ CONFIG_XSCALE_PMU=y # # Bus support # +CONFIG_ISA_DMA_API=y CONFIG_PCI=y CONFIG_PCI_LEGACY_PROC=y CONFIG_PCI_NAMES=y +# CONFIG_PCI_DEBUG is not set # # PCCARD (PCMCIA/CardBus) support @@ -131,7 +136,15 @@ CONFIG_PCI_NAMES=y # # Kernel Features # +# CONFIG_SMP is not set # CONFIG_PREEMPT is not set +# CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set +CONFIG_SELECT_MEMORY_MODEL=y +CONFIG_FLATMEM_MANUAL=y +# CONFIG_DISCONTIGMEM_MANUAL is not set +# CONFIG_SPARSEMEM_MANUAL is not set +CONFIG_FLATMEM=y +CONFIG_FLAT_NODE_MEM_MAP=y CONFIG_ALIGNMENT_TRAP=y # @@ -270,7 +283,6 @@ CONFIG_MTD_IXP2000=y # # Block devices # -# CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_CPQ_CISS_DA is not set # CONFIG_BLK_DEV_DAC960 is not set @@ -309,6 +321,7 @@ CONFIG_IOSCHED_CFQ=y # # Fusion MPT device support # +# CONFIG_FUSION is not set # # IEEE 1394 (FireWire) support @@ -330,10 +343,11 @@ CONFIG_NET=y # CONFIG_PACKET=y CONFIG_PACKET_MMAP=y -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y +CONFIG_IP_FIB_HASH=y +# CONFIG_IP_FIB_TRIE is not set # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set CONFIG_IP_PNP=y @@ -350,6 +364,17 @@ CONFIG_SYN_COOKIES=y # CONFIG_INET_TUNNEL is not set # CONFIG_IP_TCPDIAG is not set # CONFIG_IP_TCPDIAG_IPV6 is not set + +# +# TCP congestion control +# +CONFIG_TCP_CONG_BIC=y +CONFIG_TCP_CONG_WESTWOOD=m +CONFIG_TCP_CONG_HTCP=m +# CONFIG_TCP_CONG_HSTCP is not set +# CONFIG_TCP_CONG_HYBLA is not set +# CONFIG_TCP_CONG_VEGAS is not set +# CONFIG_TCP_CONG_SCALABLE is not set # CONFIG_IPV6 is not set # CONFIG_NETFILTER is not set @@ -405,6 +430,7 @@ CONFIG_MII=y # CONFIG_SUNGEM is not set # CONFIG_NET_VENDOR_3COM is not set # CONFIG_SMC91X is not set +# CONFIG_DM9000 is not set # # Tulip family network device support @@ -442,9 +468,11 @@ CONFIG_EEPRO100=y # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set # CONFIG_R8169 is not set +# CONFIG_SKGE is not set # CONFIG_SK98LIN is not set # CONFIG_VIA_VELOCITY is not set # CONFIG_TIGON3 is not set +# CONFIG_BNX2 is not set # # Ethernet (10000 Mbit) @@ -466,6 +494,7 @@ CONFIG_EEPRO100=y # Wan interfaces # CONFIG_WAN=y +# CONFIG_DSCC4 is not set # CONFIG_LANMEDIA is not set # CONFIG_SYNCLINK_SYNCPPP is not set CONFIG_HDLC=y @@ -528,7 +557,6 @@ CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768 # # CONFIG_SERIO is not set # CONFIG_GAMEPORT is not set -CONFIG_SOUND_GAMEPORT=y # # Character devices @@ -549,6 +577,7 @@ CONFIG_SERIAL_8250_NR_UARTS=2 # CONFIG_SERIAL_CORE=y CONFIG_SERIAL_CORE_CONSOLE=y +# CONFIG_SERIAL_JSM is not set CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -615,17 +644,18 @@ CONFIG_I2C_ALGOBIT=y # CONFIG_I2C_AMD8111 is not set # CONFIG_I2C_I801 is not set # CONFIG_I2C_I810 is not set +# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_ISA is not set # CONFIG_I2C_IXP2000 is not set # CONFIG_I2C_NFORCE2 is not set # CONFIG_I2C_PARPORT_LIGHT is not set -# CONFIG_I2C_PIIX4 is not set # CONFIG_I2C_PROSAVAGE is not set # CONFIG_I2C_SAVAGE4 is not set # CONFIG_SCx200_ACB is not set # CONFIG_I2C_SIS5595 is not set # CONFIG_I2C_SIS630 is not set # CONFIG_I2C_SIS96X is not set +# CONFIG_I2C_STUB is not set # CONFIG_I2C_VIA is not set # CONFIG_I2C_VIAPRO is not set # CONFIG_I2C_VOODOO3 is not set @@ -639,7 +669,9 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_ADM1025 is not set # CONFIG_SENSORS_ADM1026 is not set # CONFIG_SENSORS_ADM1031 is not set +# CONFIG_SENSORS_ADM9240 is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set @@ -655,6 +687,7 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_LM85 is not set # CONFIG_SENSORS_LM87 is not set # CONFIG_SENSORS_LM90 is not set +# CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_SMSC47B397 is not set @@ -664,14 +697,19 @@ CONFIG_I2C_SENSOR=y # CONFIG_SENSORS_W83781D is not set # CONFIG_SENSORS_W83L785TS is not set # CONFIG_SENSORS_W83627HF is not set +# CONFIG_SENSORS_W83627EHF is not set # # Other I2C Chip support # +# CONFIG_SENSORS_DS1337 is not set +# CONFIG_SENSORS_DS1374 is not set CONFIG_SENSORS_EEPROM=y # CONFIG_SENSORS_PCF8574 is not set +# CONFIG_SENSORS_PCA9539 is not set # CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_RTC8564 is not set +# CONFIG_SENSORS_MAX6875 is not set # CONFIG_I2C_DEBUG_CORE is not set # CONFIG_I2C_DEBUG_ALGO is not set # CONFIG_I2C_DEBUG_BUS is not set @@ -725,6 +763,7 @@ CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_POSIX_ACL=y # CONFIG_EXT2_FS_SECURITY is not set +# CONFIG_EXT2_FS_XIP is not set CONFIG_EXT3_FS=y CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y @@ -765,7 +804,6 @@ CONFIG_DNOTIFY=y # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVFS_FS is not set # CONFIG_DEVPTS_FS_XATTR is not set CONFIG_TMPFS=y # CONFIG_TMPFS_XATTR is not set @@ -803,12 +841,14 @@ CONFIG_JFFS2_RTIME=y # CONFIG_NFS_FS=y CONFIG_NFS_V3=y +# CONFIG_NFS_V3_ACL is not set # CONFIG_NFS_V4 is not set # CONFIG_NFS_DIRECTIO is not set # CONFIG_NFSD is not set CONFIG_ROOT_NFS=y CONFIG_LOCKD=y CONFIG_LOCKD_V4=y +CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y # CONFIG_RPCSEC_GSS_KRB5 is not set # CONFIG_RPCSEC_GSS_SPKM3 is not set @@ -893,3 +933,4 @@ CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y CONFIG_ZLIB_DEFLATE=y +# CONFIG_TEXTSEARCH is not set diff --git a/arch/arm/configs/s3c2410_defconfig b/arch/arm/configs/s3c2410_defconfig index 2a63fb27719..98b72ff3883 100644 --- a/arch/arm/configs/s3c2410_defconfig +++ b/arch/arm/configs/s3c2410_defconfig @@ -1,14 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk2 -# Sun Mar 27 17:47:45 2005 +# Linux kernel version: 2.6.12-git4 +# Wed Jun 22 15:56:42 2005 # CONFIG_ARM=y CONFIG_MMU=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y CONFIG_GENERIC_CALIBRATE_DELAY=y -CONFIG_GENERIC_IOMAP=y # # Code maturity level options @@ -17,6 +16,7 @@ CONFIG_EXPERIMENTAL=y # CONFIG_CLEAN_COMPILE is not set CONFIG_BROKEN=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -35,6 +35,8 @@ CONFIG_KOBJECT_UEVENT=y CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_PRINTK=y +CONFIG_BUG=y CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y @@ -81,6 +83,7 @@ CONFIG_ARCH_S3C2410=y # CONFIG_ARCH_VERSATILE is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set +# CONFIG_ARCH_AAEC2000 is not set # # S3C24XX Implementations @@ -134,6 +137,7 @@ CONFIG_CPU_TLB_V4WBI=y # # Bus support # +CONFIG_ISA_DMA_API=y # # PCCARD (PCMCIA/CardBus) support @@ -143,7 +147,9 @@ CONFIG_CPU_TLB_V4WBI=y # # Kernel Features # +# CONFIG_SMP is not set # CONFIG_PREEMPT is not set +# CONFIG_DISCONTIGMEM is not set CONFIG_ALIGNMENT_TRAP=y # @@ -297,7 +303,6 @@ CONFIG_PARPORT_1284=y # # Block devices # -# CONFIG_BLK_DEV_FD is not set # CONFIG_PARIDE is not set # CONFIG_BLK_DEV_COW_COMMON is not set CONFIG_BLK_DEV_LOOP=y @@ -359,6 +364,7 @@ CONFIG_BLK_DEV_IDE_BAST=y # # Fusion MPT device support # +# CONFIG_FUSION is not set # # IEEE 1394 (FireWire) support @@ -378,10 +384,11 @@ CONFIG_NET=y # Networking options # # CONFIG_PACKET is not set -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y +CONFIG_IP_FIB_HASH=y +# CONFIG_IP_FIB_TRIE is not set # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set CONFIG_IP_PNP=y @@ -443,8 +450,9 @@ CONFIG_NETDEVICES=y # Ethernet (10 or 100Mbit) # CONFIG_NET_ETHERNET=y -# CONFIG_MII is not set +CONFIG_MII=m # CONFIG_SMC91X is not set +CONFIG_DM9000=m # # Ethernet (1000 Mbit) @@ -521,7 +529,6 @@ CONFIG_SERIO_SERPORT=y CONFIG_SERIO_LIBPS2=y # CONFIG_SERIO_RAW is not set # CONFIG_GAMEPORT is not set -CONFIG_SOUND_GAMEPORT=y # # Character devices @@ -605,7 +612,6 @@ CONFIG_S3C2410_RTC=y # # TPM devices # -# CONFIG_TCG_TPM is not set # # I2C support @@ -654,6 +660,7 @@ CONFIG_SENSORS_LM78=m CONFIG_SENSORS_LM85=m # CONFIG_SENSORS_LM87 is not set # CONFIG_SENSORS_LM90 is not set +# CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_SMSC47B397 is not set @@ -665,6 +672,7 @@ CONFIG_SENSORS_LM85=m # # Other I2C Chip support # +# CONFIG_SENSORS_DS1337 is not set CONFIG_SENSORS_EEPROM=m # CONFIG_SENSORS_PCF8574 is not set # CONFIG_SENSORS_PCF8591 is not set @@ -696,8 +704,10 @@ CONFIG_FB=y # CONFIG_FB_CFB_COPYAREA is not set # CONFIG_FB_CFB_IMAGEBLIT is not set # CONFIG_FB_SOFT_CURSOR is not set +# CONFIG_FB_MACMODES is not set CONFIG_FB_MODE_HELPERS=y # CONFIG_FB_TILEBLITTING is not set +# CONFIG_FB_S1D13XXX is not set # CONFIG_FB_VIRTUAL is not set # @@ -782,7 +792,6 @@ CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1" # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVFS_FS is not set # CONFIG_DEVPTS_FS_XATTR is not set # CONFIG_TMPFS is not set # CONFIG_HUGETLBFS is not set diff --git a/arch/arm/kernel/irq.c b/arch/arm/kernel/irq.c index ff187f4308f..395137a8fad 100644 --- a/arch/arm/kernel/irq.c +++ b/arch/arm/kernel/irq.c @@ -4,6 +4,10 @@ * Copyright (C) 1992 Linus Torvalds * Modifications for ARM processor Copyright (C) 1995-2000 Russell King. * + * Support for Dynamic Tick Timer Copyright (C) 2004-2005 Nokia Corporation. + * Dynamic Tick Timer written by Tony Lindgren <tony@atomide.com> and + * Tuukka Tikkanen <tuukka.tikkanen@elektrobit.com>. + * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. @@ -37,6 +41,7 @@ #include <asm/irq.h> #include <asm/system.h> #include <asm/mach/irq.h> +#include <asm/mach/time.h> /* * Maximum IRQ count. Currently, this is arbitary. However, it should @@ -329,6 +334,15 @@ __do_irq(unsigned int irq, struct irqaction *action, struct pt_regs *regs) spin_unlock(&irq_controller_lock); +#ifdef CONFIG_NO_IDLE_HZ + if (!(action->flags & SA_TIMER) && system_timer->dyn_tick != NULL) { + write_seqlock(&xtime_lock); + if (system_timer->dyn_tick->state & DYN_TICK_ENABLED) + system_timer->dyn_tick->handler(irq, 0, regs); + write_sequnlock(&xtime_lock); + } +#endif + if (!(action->flags & SA_INTERRUPT)) local_irq_enable(); diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c index 07ddeed6176..5e435e42dac 100644 --- a/arch/arm/kernel/signal.c +++ b/arch/arm/kernel/signal.c @@ -697,7 +697,7 @@ static int do_signal(sigset_t *oldset, struct pt_regs *regs, int syscall) if (!user_mode(regs)) return 0; - if (try_to_freeze(0)) + if (try_to_freeze()) goto no_signal; if (current->ptrace & PT_SINGLESTEP) diff --git a/arch/arm/kernel/time.c b/arch/arm/kernel/time.c index c232f24f4a6..06054c9ba07 100644 --- a/arch/arm/kernel/time.c +++ b/arch/arm/kernel/time.c @@ -381,6 +381,95 @@ static struct sysdev_class timer_sysclass = { .resume = timer_resume, }; +#ifdef CONFIG_NO_IDLE_HZ +static int timer_dyn_tick_enable(void) +{ + struct dyn_tick_timer *dyn_tick = system_timer->dyn_tick; + unsigned long flags; + int ret = -ENODEV; + + if (dyn_tick) { + write_seqlock_irqsave(&xtime_lock, flags); + ret = 0; + if (!(dyn_tick->state & DYN_TICK_ENABLED)) { + ret = dyn_tick->enable(); + + if (ret == 0) + dyn_tick->state |= DYN_TICK_ENABLED; + } + write_sequnlock_irqrestore(&xtime_lock, flags); + } + + return ret; +} + +static int timer_dyn_tick_disable(void) +{ + struct dyn_tick_timer *dyn_tick = system_timer->dyn_tick; + unsigned long flags; + int ret = -ENODEV; + + if (dyn_tick) { + write_seqlock_irqsave(&xtime_lock, flags); + ret = 0; + if (dyn_tick->state & DYN_TICK_ENABLED) { + ret = dyn_tick->disable(); + + if (ret == 0) + dyn_tick->state &= ~DYN_TICK_ENABLED; + } + write_sequnlock_irqrestore(&xtime_lock, flags); + } + + return ret; +} + +void timer_dyn_reprogram(void) +{ + struct dyn_tick_timer *dyn_tick = system_timer->dyn_tick; + unsigned long flags; + + write_seqlock_irqsave(&xtime_lock, flags); + if (dyn_tick->state & DYN_TICK_ENABLED) + dyn_tick->reprogram(next_timer_interrupt() - jiffies); + write_sequnlock_irqrestore(&xtime_lock, flags); +} + +static ssize_t timer_show_dyn_tick(struct sys_device *dev, char *buf) +{ + return sprintf(buf, "%i\n", + (system_timer->dyn_tick->state & DYN_TICK_ENABLED) >> 1); +} + +static ssize_t timer_set_dyn_tick(struct sys_device *dev, const char *buf, + size_t count) +{ + unsigned int enable = simple_strtoul(buf, NULL, 2); + + if (enable) + timer_dyn_tick_enable(); + else + timer_dyn_tick_disable(); + + return count; +} +static SYSDEV_ATTR(dyn_tick, 0644, timer_show_dyn_tick, timer_set_dyn_tick); + +/* + * dyntick=enable|disable + */ +static char dyntick_str[4] __initdata = ""; + +static int __init dyntick_setup(char *str) +{ + if (str) + strlcpy(dyntick_str, str, sizeof(dyntick_str)); + return 1; +} + +__setup("dyntick=", dyntick_setup); +#endif + static int __init timer_init_sysfs(void) { int ret = sysdev_class_register(&timer_sysclass); @@ -388,6 +477,20 @@ static int __init timer_init_sysfs(void) system_timer->dev.cls = &timer_sysclass; ret = sysdev_register(&system_timer->dev); } + +#ifdef CONFIG_NO_IDLE_HZ + if (ret == 0 && system_timer->dyn_tick) { + ret = sysdev_create_file(&system_timer->dev, &attr_dyn_tick); + + /* + * Turn on dynamic tick after calibrate delay + * for correct bogomips + */ + if (ret == 0 && dyntick_str[0] == 'e') + ret = timer_dyn_tick_enable(); + } +#endif + return ret; } diff --git a/arch/arm/mach-clps711x/Kconfig b/arch/arm/mach-clps711x/Kconfig index 45c930ccd06..0793dcf54f2 100644 --- a/arch/arm/mach-clps711x/Kconfig +++ b/arch/arm/mach-clps711x/Kconfig @@ -28,7 +28,7 @@ config ARCH_CLEP7312 config ARCH_EDB7211 bool "EDB7211" select ISA - select DISCONTIGMEM + select ARCH_DISCONTIGMEM_ENABLE help Say Y here if you intend to run this kernel on a Cirrus Logic EDB-7211 evaluation board. diff --git a/arch/arm/mach-ixp2000/Kconfig b/arch/arm/mach-ixp2000/Kconfig index 9361e05f6fa..ecb58d83478 100644 --- a/arch/arm/mach-ixp2000/Kconfig +++ b/arch/arm/mach-ixp2000/Kconfig @@ -54,6 +54,14 @@ config ARCH_IXDP2X01 depends on ARCH_IXDP2401 || ARCH_IXDP2801 default y +config IXP2000_SUPPORT_BROKEN_PCI_IO + bool "Support broken PCI I/O on older IXP2000s" + default y + help + Say 'N' here if you only intend to run your kernel on an + IXP2000 B0 or later model and do not need the PCI I/O + byteswap workaround. Say 'Y' otherwise. + endmenu endif diff --git a/arch/arm/mach-ixp2000/core.c b/arch/arm/mach-ixp2000/core.c index fc0555596d6..0ee34acb8d7 100644 --- a/arch/arm/mach-ixp2000/core.c +++ b/arch/arm/mach-ixp2000/core.c @@ -40,6 +40,8 @@ #include <asm/mach/time.h> #include <asm/mach/irq.h> +#include <asm/arch/gpio.h> + static DEFINE_SPINLOCK(ixp2000_slowport_lock); static unsigned long ixp2000_slowport_irq_flags; @@ -179,7 +181,7 @@ static int ixp2000_timer_interrupt(int irq, void *dev_id, struct pt_regs *regs) /* clear timer 1 */ ixp2000_reg_write(IXP2000_T1_CLR, 1); - + while ((next_jiffy_time - *missing_jiffy_timer_csr) > ticks_per_jiffy) { timer_tick(regs); next_jiffy_time -= ticks_per_jiffy; @@ -238,35 +240,40 @@ void __init ixp2000_init_time(unsigned long tick_rate) /************************************************************************* * GPIO helpers *************************************************************************/ -static unsigned long GPIO_IRQ_rising_edge; static unsigned long GPIO_IRQ_falling_edge; +static unsigned long GPIO_IRQ_rising_edge; static unsigned long GPIO_IRQ_level_low; static unsigned long GPIO_IRQ_level_high; -void gpio_line_config(int line, int style) +static void update_gpio_int_csrs(void) +{ + ixp2000_reg_write(IXP2000_GPIO_FEDR, GPIO_IRQ_falling_edge); + ixp2000_reg_write(IXP2000_GPIO_REDR, GPIO_IRQ_rising_edge); + ixp2000_reg_write(IXP2000_GPIO_LSLR, GPIO_IRQ_level_low); + ixp2000_reg_write(IXP2000_GPIO_LSHR, GPIO_IRQ_level_high); +} + +void gpio_line_config(int line, int direction) { unsigned long flags; local_irq_save(flags); + if (direction == GPIO_OUT) { + irq_desc[line + IRQ_IXP2000_GPIO0].valid = 0; - if(style == GPIO_OUT) { /* if it's an output, it ain't an interrupt anymore */ - ixp2000_reg_write(IXP2000_GPIO_PDSR, (1 << line)); GPIO_IRQ_falling_edge &= ~(1 << line); GPIO_IRQ_rising_edge &= ~(1 << line); GPIO_IRQ_level_low &= ~(1 << line); GPIO_IRQ_level_high &= ~(1 << line); - ixp2000_reg_write(IXP2000_GPIO_FEDR, GPIO_IRQ_falling_edge); - ixp2000_reg_write(IXP2000_GPIO_REDR, GPIO_IRQ_rising_edge); - ixp2000_reg_write(IXP2000_GPIO_LSHR, GPIO_IRQ_level_high); - ixp2000_reg_write(IXP2000_GPIO_LSLR, GPIO_IRQ_level_low); - irq_desc[line+IRQ_IXP2000_GPIO0].valid = 0; - } else if(style == GPIO_IN) { - ixp2000_reg_write(IXP2000_GPIO_PDCR, (1 << line)); + update_gpio_int_csrs(); + + ixp2000_reg_write(IXP2000_GPIO_PDSR, 1 << line); + } else if (direction == GPIO_IN) { + ixp2000_reg_write(IXP2000_GPIO_PDCR, 1 << line); } - local_irq_restore(flags); -} +} /************************************************************************* @@ -285,9 +292,50 @@ static void ixp2000_GPIO_irq_handler(unsigned int irq, struct irqdesc *desc, str } } +static int ixp2000_GPIO_irq_type(unsigned int irq, unsigned int type) +{ + int line = irq - IRQ_IXP2000_GPIO0; + + /* + * First, configure this GPIO line as an input. + */ + ixp2000_reg_write(IXP2000_GPIO_PDCR, 1 << line); + + /* + * Then, set the proper trigger type. + */ + if (type & IRQT_FALLING) + GPIO_IRQ_falling_edge |= 1 << line; + else + GPIO_IRQ_falling_edge &= ~(1 << line); + if (type & IRQT_RISING) + GPIO_IRQ_rising_edge |= 1 << line; + else + GPIO_IRQ_rising_edge &= ~(1 << line); + if (type & IRQT_LOW) + GPIO_IRQ_level_low |= 1 << line; + else + GPIO_IRQ_level_low &= ~(1 << line); + if (type & IRQT_HIGH) + GPIO_IRQ_level_high |= 1 << line; + else + GPIO_IRQ_level_high &= ~(1 << line); + update_gpio_int_csrs(); + + /* + * Finally, mark the corresponding IRQ as valid. + */ + irq_desc[irq].valid = 1; + + return 0; +} + static void ixp2000_GPIO_irq_mask_ack(unsigned int irq) { ixp2000_reg_write(IXP2000_GPIO_INCR, (1 << (irq - IRQ_IXP2000_GPIO0))); + + ixp2000_reg_write(IXP2000_GPIO_EDSR, (1 << (irq - IRQ_IXP2000_GPIO0))); + ixp2000_reg_write(IXP2000_GPIO_LDSR, (1 << (irq - IRQ_IXP2000_GPIO0))); ixp2000_reg_write(IXP2000_GPIO_INST, (1 << (irq - IRQ_IXP2000_GPIO0))); } @@ -302,6 +350,7 @@ static void ixp2000_GPIO_irq_unmask(unsigned int irq) } static struct irqchip ixp2000_GPIO_irq_chip = { + .type = ixp2000_GPIO_irq_type, .ack = ixp2000_GPIO_irq_mask_ack, .mask = ixp2000_GPIO_irq_mask, .unmask = ixp2000_GPIO_irq_unmask @@ -338,7 +387,7 @@ static void ixp2000_irq_mask(unsigned int irq) static void ixp2000_irq_unmask(unsigned int irq) { - ixp2000_reg_write(IXP2000_IRQ_ENABLE_SET, (1 << irq)); + ixp2000_reg_write(IXP2000_IRQ_ENABLE_SET, (1 << irq)); } static struct irqchip ixp2000_irq_chip = { @@ -375,16 +424,16 @@ void __init ixp2000_init_irq(void) * our mask/unmask code much simpler. */ for (irq = IRQ_IXP2000_SOFT_INT; irq <= IRQ_IXP2000_THDB3; irq++) { - if((1 << irq) & IXP2000_VALID_IRQ_MASK) { + if ((1 << irq) & IXP2000_VALID_IRQ_MASK) { set_irq_chip(irq, &ixp2000_irq_chip); set_irq_handler(irq, do_level_IRQ); set_irq_flags(irq, IRQF_VALID); } else set_irq_flags(irq, 0); } - + /* * GPIO IRQs are invalid until someone sets the interrupt mode - * by calling gpio_line_set(); + * by calling set_irq_type(). */ for (irq = IRQ_IXP2000_GPIO0; irq <= IRQ_IXP2000_GPIO7; irq++) { set_irq_chip(irq, &ixp2000_GPIO_irq_chip); diff --git a/arch/arm/mach-ixp2000/enp2611.c b/arch/arm/mach-ixp2000/enp2611.c index 04b38bcf9aa..f3a291b6a9f 100644 --- a/arch/arm/mach-ixp2000/enp2611.c +++ b/arch/arm/mach-ixp2000/enp2611.c @@ -197,8 +197,23 @@ static struct platform_device enp2611_flash = { .resource = &enp2611_flash_resource, }; +static struct ixp2000_i2c_pins enp2611_i2c_gpio_pins = { + .sda_pin = ENP2611_GPIO_SDA, + .scl_pin = ENP2611_GPIO_SCL, +}; + +static struct platform_device enp2611_i2c_controller = { + .name = "IXP2000-I2C", + .id = 0, + .dev = { + .platform_data = &enp2611_i2c_gpio_pins + }, + .num_resources = 0 +}; + static struct platform_device *enp2611_devices[] __initdata = { - &enp2611_flash + &enp2611_flash, + &enp2611_i2c_controller }; static void __init enp2611_init_machine(void) diff --git a/arch/arm/mach-ixp2000/ixdp2x00.c b/arch/arm/mach-ixp2000/ixdp2x00.c index 21c41fe15b9..5e4380747b5 100644 --- a/arch/arm/mach-ixp2000/ixdp2x00.c +++ b/arch/arm/mach-ixp2000/ixdp2x00.c @@ -42,6 +42,9 @@ #include <asm/mach/flash.h> #include <asm/mach/arch.h> +#include <asm/arch/gpio.h> + + /************************************************************************* * IXDP2x00 IRQ Initialization *************************************************************************/ diff --git a/arch/arm/mach-ixp2000/pci.c b/arch/arm/mach-ixp2000/pci.c index 5ff2f2718c5..0788fb2b5c1 100644 --- a/arch/arm/mach-ixp2000/pci.c +++ b/arch/arm/mach-ixp2000/pci.c @@ -198,6 +198,19 @@ clear_master_aborts(void) void __init ixp2000_pci_preinit(void) { +#ifndef CONFIG_IXP2000_SUPPORT_BROKEN_PCI_IO + /* + * Configure the PCI unit to properly byteswap I/O transactions, + * and verify that it worked. + */ + ixp2000_reg_write(IXP2000_PCI_CONTROL, + (*IXP2000_PCI_CONTROL | PCI_CONTROL_IEE)); + + if ((*IXP2000_PCI_CONTROL & PCI_CONTROL_IEE) == 0) + panic("IXP2000: PCI I/O is broken on this ixp model, and " + "the needed workaround has not been configured in"); +#endif + hook_fault_code(16+6, ixp2000_pci_abort_handler, SIGBUS, "PCI config cycle to non-existent device"); } diff --git a/arch/arm/mach-ixp4xx/common.c b/arch/arm/mach-ixp4xx/common.c index 267ba02d77d..f39e8408488 100644 --- a/arch/arm/mach-ixp4xx/common.c +++ b/arch/arm/mach-ixp4xx/common.c @@ -141,7 +141,15 @@ static struct map_desc ixp4xx_io_desc[] __initdata = { .physical = IXP4XX_PCI_CFG_BASE_PHYS, .length = IXP4XX_PCI_CFG_REGION_SIZE, .type = MT_DEVICE + }, +#ifdef CONFIG_DEBUG_LL + { /* Debug UART mapping */ + .virtual = IXP4XX_DEBUG_UART_BASE_VIRT, + .physical = IXP4XX_DEBUG_UART_BASE_PHYS, + .length = IXP4XX_DEBUG_UART_REGION_SIZE, + .type = MT_DEVICE } +#endif }; void __init ixp4xx_map_io(void) diff --git a/arch/arm/mach-s3c2410/mach-bast.c b/arch/arm/mach-s3c2410/mach-bast.c index 3bb97eb6e69..f3e970039b6 100644 --- a/arch/arm/mach-s3c2410/mach-bast.c +++ b/arch/arm/mach-s3c2410/mach-bast.c @@ -26,6 +26,7 @@ * 03-Mar-2005 BJD Ensured that bast-cpld.h is included * 10-Mar-2005 LCVR Changed S3C2410_VA to S3C24XX_VA * 14-Mar-2006 BJD Updated for __iomem changes + * 22-Jun-2006 BJD Added DM9000 platform information */ #include <linux/kernel.h> @@ -35,6 +36,7 @@ #include <linux/timer.h> #include <linux/init.h> #include <linux/device.h> +#include <linux/dm9000.h> #include <asm/mach/arch.h> #include <asm/mach/map.h> @@ -53,6 +55,7 @@ #include <asm/arch/regs-serial.h> #include <asm/arch/regs-gpio.h> #include <asm/arch/regs-mem.h> +#include <asm/arch/regs-lcd.h> #include <asm/arch/nand.h> #include <linux/mtd/mtd.h> @@ -112,7 +115,6 @@ static struct map_desc bast_iodesc[] __initdata = { { VA_C2(BAST_VA_ISAMEM), PA_CS2(BAST_PA_ISAMEM), SZ_16M, MT_DEVICE }, { VA_C2(BAST_VA_ASIXNET), PA_CS3(BAST_PA_ASIXNET), SZ_1M, MT_DEVICE }, { VA_C2(BAST_VA_SUPERIO), PA_CS2(BAST_PA_SUPERIO), SZ_1M, MT_DEVICE }, - { VA_C2(BAST_VA_DM9000), PA_CS2(BAST_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C2(BAST_VA_IDEPRI), PA_CS3(BAST_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C2(BAST_VA_IDESEC), PA_CS3(BAST_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C2(BAST_VA_IDEPRIAUX), PA_CS3(BAST_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, @@ -123,7 +125,6 @@ static struct map_desc bast_iodesc[] __initdata = { { VA_C3(BAST_VA_ISAMEM), PA_CS3(BAST_PA_ISAMEM), SZ_16M, MT_DEVICE }, { VA_C3(BAST_VA_ASIXNET), PA_CS3(BAST_PA_ASIXNET), SZ_1M, MT_DEVICE }, { VA_C3(BAST_VA_SUPERIO), PA_CS3(BAST_PA_SUPERIO), SZ_1M, MT_DEVICE }, - { VA_C3(BAST_VA_DM9000), PA_CS3(BAST_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C3(BAST_VA_IDEPRI), PA_CS3(BAST_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C3(BAST_VA_IDESEC), PA_CS3(BAST_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C3(BAST_VA_IDEPRIAUX), PA_CS3(BAST_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, @@ -134,7 +135,6 @@ static struct map_desc bast_iodesc[] __initdata = { { VA_C4(BAST_VA_ISAMEM), PA_CS4(BAST_PA_ISAMEM), SZ_16M, MT_DEVICE }, { VA_C4(BAST_VA_ASIXNET), PA_CS5(BAST_PA_ASIXNET), SZ_1M, MT_DEVICE }, { VA_C4(BAST_VA_SUPERIO), PA_CS4(BAST_PA_SUPERIO), SZ_1M, MT_DEVICE }, - { VA_C4(BAST_VA_DM9000), PA_CS4(BAST_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C4(BAST_VA_IDEPRI), PA_CS5(BAST_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C4(BAST_VA_IDESEC), PA_CS5(BAST_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C4(BAST_VA_IDEPRIAUX), PA_CS5(BAST_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, @@ -145,7 +145,6 @@ static struct map_desc bast_iodesc[] __initdata = { { VA_C5(BAST_VA_ISAMEM), PA_CS5(BAST_PA_ISAMEM), SZ_16M, MT_DEVICE }, { VA_C5(BAST_VA_ASIXNET), PA_CS5(BAST_PA_ASIXNET), SZ_1M, MT_DEVICE }, { VA_C5(BAST_VA_SUPERIO), PA_CS5(BAST_PA_SUPERIO), SZ_1M, MT_DEVICE }, - { VA_C5(BAST_VA_DM9000), PA_CS5(BAST_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C5(BAST_VA_IDEPRI), PA_CS5(BAST_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C5(BAST_VA_IDESEC), PA_CS5(BAST_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C5(BAST_VA_IDEPRIAUX), PA_CS5(BAST_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, @@ -313,6 +312,45 @@ static struct s3c2410_platform_nand bast_nand_info = { .select_chip = bast_nand_select, }; +/* DM9000 */ + +static struct resource bast_dm9k_resource[] = { + [0] = { + .start = S3C2410_CS5 + BAST_PA_DM9000, + .end = S3C2410_CS5 + BAST_PA_DM9000 + 3, + .flags = IORESOURCE_MEM + }, + [1] = { + .start = S3C2410_CS5 + BAST_PA_DM9000 + 0x40, + .end = S3C2410_CS5 + BAST_PA_DM9000 + 0x40 + 0x3f, + .flags = IORESOURCE_MEM + }, + [2] = { + .start = IRQ_DM9000, + .end = IRQ_DM9000, + .flags = IORESOURCE_IRQ + } + +}; + +/* for the moment we limit ourselves to 16bit IO until some + * better IO routines can be written and tested +*/ + +struct dm9000_plat_data bast_dm9k_platdata = { + .flags = DM9000_PLATF_16BITONLY +}; + +static struct platform_device bast_device_dm9k = { + .name = "dm9000", + .id = 0, + .num_resources = ARRAY_SIZE(bast_dm9k_resource), + .resource = bast_dm9k_resource, + .dev = { + .platform_data = &bast_dm9k_platdata, + } +}; + /* Standard BAST devices */ @@ -324,7 +362,8 @@ static struct platform_device *bast_devices[] __initdata = { &s3c_device_iis, &s3c_device_rtc, &s3c_device_nand, - &bast_device_nor + &bast_device_nor, + &bast_device_dm9k, }; static struct clk *bast_clocks[] = { diff --git a/arch/arm/mach-s3c2410/mach-vr1000.c b/arch/arm/mach-s3c2410/mach-vr1000.c index 5512146b1ce..76be074944a 100644 --- a/arch/arm/mach-s3c2410/mach-vr1000.c +++ b/arch/arm/mach-s3c2410/mach-vr1000.c @@ -27,6 +27,7 @@ * 10-Feb-2005 BJD Added power-off capability * 10-Mar-2005 LCVR Changed S3C2410_VA to S3C24XX_VA * 14-Mar-2006 BJD void __iomem fixes + * 22-Jun-2006 BJD Added DM9000 platform information */ #include <linux/kernel.h> @@ -35,6 +36,7 @@ #include <linux/list.h> #include <linux/timer.h> #include <linux/init.h> +#include <linux/dm9000.h> #include <linux/serial.h> #include <linux/tty.h> @@ -98,28 +100,24 @@ static struct map_desc vr1000_iodesc[] __initdata = { * are only 8bit */ /* slow, byte */ - { VA_C2(VR1000_VA_DM9000), PA_CS2(VR1000_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C2(VR1000_VA_IDEPRI), PA_CS3(VR1000_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C2(VR1000_VA_IDESEC), PA_CS3(VR1000_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C2(VR1000_VA_IDEPRIAUX), PA_CS3(VR1000_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, { VA_C2(VR1000_VA_IDESECAUX), PA_CS3(VR1000_PA_IDESECAUX), SZ_1M, MT_DEVICE }, /* slow, word */ - { VA_C3(VR1000_VA_DM9000), PA_CS3(VR1000_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C3(VR1000_VA_IDEPRI), PA_CS3(VR1000_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C3(VR1000_VA_IDESEC), PA_CS3(VR1000_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C3(VR1000_VA_IDEPRIAUX), PA_CS3(VR1000_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, { VA_C3(VR1000_VA_IDESECAUX), PA_CS3(VR1000_PA_IDESECAUX), SZ_1M, MT_DEVICE }, /* fast, byte */ - { VA_C4(VR1000_VA_DM9000), PA_CS4(VR1000_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C4(VR1000_VA_IDEPRI), PA_CS5(VR1000_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C4(VR1000_VA_IDESEC), PA_CS5(VR1000_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C4(VR1000_VA_IDEPRIAUX), PA_CS5(VR1000_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, { VA_C4(VR1000_VA_IDESECAUX), PA_CS5(VR1000_PA_IDESECAUX), SZ_1M, MT_DEVICE }, /* fast, word */ - { VA_C5(VR1000_VA_DM9000), PA_CS5(VR1000_PA_DM9000), SZ_1M, MT_DEVICE }, { VA_C5(VR1000_VA_IDEPRI), PA_CS5(VR1000_PA_IDEPRI), SZ_1M, MT_DEVICE }, { VA_C5(VR1000_VA_IDESEC), PA_CS5(VR1000_PA_IDESEC), SZ_1M, MT_DEVICE }, { VA_C5(VR1000_VA_IDEPRIAUX), PA_CS5(VR1000_PA_IDEPRIAUX), SZ_1M, MT_DEVICE }, @@ -246,6 +244,74 @@ static struct platform_device vr1000_nor = { .resource = vr1000_nor_resource, }; +/* DM9000 ethernet devices */ + +static struct resource vr1000_dm9k0_resource[] = { + [0] = { + .start = S3C2410_CS5 + VR1000_PA_DM9000, + .end = S3C2410_CS5 + VR1000_PA_DM9000 + 3, + .flags = IORESOURCE_MEM + }, + [1] = { + .start = S3C2410_CS5 + VR1000_PA_DM9000 + 0x40, + .end = S3C2410_CS5 + VR1000_PA_DM9000 + 0x7f, + .flags = IORESOURCE_MEM + }, + [2] = { + .start = IRQ_VR1000_DM9000A, + .end = IRQ_VR1000_DM9000A, + .flags = IORESOURCE_IRQ + } + +}; + +static struct resource vr1000_dm9k1_resource[] = { + [0] = { + .start = S3C2410_CS5 + VR1000_PA_DM9000 + 0x80, + .end = S3C2410_CS5 + VR1000_PA_DM9000 + 0x83, + .flags = IORESOURCE_MEM + }, + [1] = { + .start = S3C2410_CS5 + VR1000_PA_DM9000 + 0xC0, + .end = S3C2410_CS5 + VR1000_PA_DM9000 + 0xFF, + .flags = IORESOURCE_MEM + }, + [2] = { + .start = IRQ_VR1000_DM9000N, + .end = IRQ_VR1000_DM9000N, + .flags = IORESOURCE_IRQ + } +}; + +/* for the moment we limit ourselves to 16bit IO until some + * better IO routines can be written and tested +*/ + +struct dm9000_plat_data vr1000_dm9k_platdata = { + .flags = DM9000_PLATF_16BITONLY, +}; + +static struct platform_device vr1000_dm9k0 = { + .name = "dm9000", + .id = 0, + .num_resources = ARRAY_SIZE(vr1000_dm9k0_resource), + .resource = vr1000_dm9k0_resource, + .dev = { + .platform_data = &vr1000_dm9k_platdata, + } +}; + +static struct platform_device vr1000_dm9k1 = { + .name = "dm9000", + .id = 1, + .num_resources = ARRAY_SIZE(vr1000_dm9k1_resource), + .resource = vr1000_dm9k1_resource, + .dev = { + .platform_data = &vr1000_dm9k_platdata, + } +}; + +/* devices for this board */ static struct platform_device *vr1000_devices[] __initdata = { &s3c_device_usb, @@ -253,8 +319,11 @@ static struct platform_device *vr1000_devices[] __initdata = { &s3c_device_wdt, &s3c_device_i2c, &s3c_device_iis, + &s3c_device_adc, &serial_device, &vr1000_nor, + &vr1000_dm9k0, + &vr1000_dm9k1 }; static struct clk *vr1000_clocks[] = { diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S index 0aa73d41478..e3d8510f434 100644 --- a/arch/arm/mm/proc-v6.S +++ b/arch/arm/mm/proc-v6.S @@ -132,8 +132,8 @@ ENTRY(cpu_v6_switch_mm) * 100x 1 0 1 r/o no acc * 10x0 1 0 1 r/o no acc * 1011 0 0 1 r/w no acc - * 110x 1 1 0 r/o r/o - * 11x0 1 1 0 r/o r/o + * 110x 0 1 0 r/w r/o + * 11x0 0 1 0 r/w r/o * 1111 0 1 1 r/w r/w */ ENTRY(cpu_v6_set_pte) @@ -150,7 +150,7 @@ ENTRY(cpu_v6_set_pte) tst r1, #L_PTE_USER orrne r2, r2, #AP1 | nG tstne r2, #APX - eorne r2, r2, #AP0 + bicne r2, r2, #APX | AP0 tst r1, #L_PTE_YOUNG biceq r2, r2, #APX | AP1 | AP0 diff --git a/arch/arm/nwfpe/softfloat-macros b/arch/arm/nwfpe/softfloat-macros index 5469989f2c5..5a060f95a58 100644 --- a/arch/arm/nwfpe/softfloat-macros +++ b/arch/arm/nwfpe/softfloat-macros @@ -563,8 +563,14 @@ static bits64 estimateDiv128To64( bits64 a0, bits64 a1, bits64 b ) bits64 rem0, rem1, term0, term1; bits64 z; if ( b <= a0 ) return LIT64( 0xFFFFFFFFFFFFFFFF ); - b0 = b>>32; - z = ( b0<<32 <= a0 ) ? LIT64( 0xFFFFFFFF00000000 ) : ( a0 / b0 )<<32; + b0 = b>>32; /* hence b0 is 32 bits wide now */ + if ( b0<<32 <= a0 ) { + z = LIT64( 0xFFFFFFFF00000000 ); + } else { + z = a0; + do_div( z, b0 ); + z <<= 32; + } mul64To128( b, z, &term0, &term1 ); sub128( a0, a1, term0, term1, &rem0, &rem1 ); while ( ( (sbits64) rem0 ) < 0 ) { @@ -573,7 +579,12 @@ static bits64 estimateDiv128To64( bits64 a0, bits64 a1, bits64 b ) add128( rem0, rem1, b0, b1, &rem0, &rem1 ); } rem0 = ( rem0<<32 ) | ( rem1>>32 ); - z |= ( b0<<32 <= rem0 ) ? 0xFFFFFFFF : rem0 / b0; + if ( b0<<32 <= rem0 ) { + z |= 0xFFFFFFFF; + } else { + do_div( rem0, b0 ); + z |= rem0; + } return z; } @@ -601,6 +612,7 @@ static bits32 estimateSqrt32( int16 aExp, bits32 a ) }; int8 index; bits32 z; + bits64 A; index = ( a>>27 ) & 15; if ( aExp & 1 ) { @@ -614,7 +626,9 @@ static bits32 estimateSqrt32( int16 aExp, bits32 a ) z = ( 0x20000 <= z ) ? 0xFFFF8000 : ( z<<15 ); if ( z <= a ) return (bits32) ( ( (sbits32) a )>>1 ); } - return ( (bits32) ( ( ( (bits64) a )<<31 ) / z ) ) + ( z>>1 ); + A = ( (bits64) a )<<31; + do_div( A, z ); + return ( (bits32) A ) + ( z>>1 ); } diff --git a/arch/arm/nwfpe/softfloat.c b/arch/arm/nwfpe/softfloat.c index 9d743ae2906..e038dd3be9b 100644 --- a/arch/arm/nwfpe/softfloat.c +++ b/arch/arm/nwfpe/softfloat.c @@ -28,6 +28,8 @@ this code that are retained. =============================================================================== */ +#include <asm/div64.h> + #include "fpa11.h" //#include "milieu.h" //#include "softfloat.h" @@ -1331,7 +1333,11 @@ float32 float32_div( float32 a, float32 b ) aSig >>= 1; ++zExp; } - zSig = ( ( (bits64) aSig )<<32 ) / bSig; + { + bits64 tmp = ( (bits64) aSig )<<32; + do_div( tmp, bSig ); + zSig = tmp; + } if ( ( zSig & 0x3F ) == 0 ) { zSig |= ( ( (bits64) bSig ) * zSig != ( (bits64) aSig )<<32 ); } @@ -1397,7 +1403,9 @@ float32 float32_rem( float32 a, float32 b ) q = ( bSig <= aSig ); if ( q ) aSig -= bSig; if ( 0 < expDiff ) { - q = ( ( (bits64) aSig )<<32 ) / bSig; + bits64 tmp = ( (bits64) aSig )<<32; + do_div( tmp, bSig ); + q = tmp; q >>= 32 - expDiff; bSig >>= 2; aSig = ( ( aSig>>1 )<<( expDiff - 1 ) ) - bSig * q; diff --git a/arch/arm26/Kconfig b/arch/arm26/Kconfig index 6caed90661f..dc0c1936969 100644 --- a/arch/arm26/Kconfig +++ b/arch/arm26/Kconfig @@ -179,6 +179,8 @@ config CMDLINE time by entering them here. As a minimum, you should specify the memory size and the root device (e.g., mem=64M root=/dev/nfs). +source "mm/Kconfig" + endmenu source "drivers/base/Kconfig" diff --git a/arch/arm26/boot/install.sh b/arch/arm26/boot/install.sh index c628328dd9e..8a8399b26cf 100644 --- a/arch/arm26/boot/install.sh +++ b/arch/arm26/boot/install.sh @@ -23,8 +23,8 @@ # User may have a custom install script -if [ -x /sbin/installkernel ]; then - exec /sbin/installkernel "$@" +if [ -x /sbin/${CROSS_COMPILE}installkernel ]; then + exec /sbin/${CROSS_COMPILE}installkernel "$@" fi if [ "$2" = "zImage" ]; then diff --git a/arch/cris/Kconfig b/arch/cris/Kconfig index 4332ca348d5..f848e376149 100644 --- a/arch/cris/Kconfig +++ b/arch/cris/Kconfig @@ -74,6 +74,8 @@ config PREEMPT Say Y here if you are building a kernel for a desktop, embedded or real-time system. Say N if you are unsure. +source mm/Kconfig + endmenu menu "Hardware setup" diff --git a/arch/frv/Kconfig b/arch/frv/Kconfig index 2b19372767e..c93f95146cc 100644 --- a/arch/frv/Kconfig +++ b/arch/frv/Kconfig @@ -74,6 +74,8 @@ config HIGHPTE with a lot of RAM, this can be wasteful of precious low memory. Setting this option will put user-space page tables in high memory. +source "mm/Kconfig" + choice prompt "uClinux kernel load address" depends on !MMU diff --git a/arch/frv/kernel/setup.c b/arch/frv/kernel/setup.c index ef6865f0b97..767ebb55bd8 100644 --- a/arch/frv/kernel/setup.c +++ b/arch/frv/kernel/setup.c @@ -790,12 +790,10 @@ void __init setup_arch(char **cmdline_p) #ifndef CONFIG_GDBSTUB_UART0 __reg(UART0_BASE + UART_IER * 8) = 0; early_serial_setup(&__frv_uart0); -// register_serial(&__frv_uart0); #endif #ifndef CONFIG_GDBSTUB_UART1 __reg(UART1_BASE + UART_IER * 8) = 0; early_serial_setup(&__frv_uart1); -// register_serial(&__frv_uart1); #endif #if defined(CONFIG_CHR_DEV_FLASH) || defined(CONFIG_BLK_DEV_FLASH) diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c index d8d8f3d4304..36a2dffc8eb 100644 --- a/arch/frv/kernel/signal.c +++ b/arch/frv/kernel/signal.c @@ -536,10 +536,8 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset) if (!user_mode(regs)) return 1; - if (current->flags & PF_FREEZE) { - refrigerator(0); + if (try_to_freeze()) goto no_signal; - } if (!oldset) oldset = ¤t->blocked; diff --git a/arch/h8300/Kconfig.cpu b/arch/h8300/Kconfig.cpu index d9dd62a565a..a380167a13c 100644 --- a/arch/h8300/Kconfig.cpu +++ b/arch/h8300/Kconfig.cpu @@ -180,4 +180,7 @@ config CPU_H8S config PREEMPT bool "Preemptible Kernel" default n + +source "mm/Kconfig" + endmenu diff --git a/arch/h8300/kernel/signal.c b/arch/h8300/kernel/signal.c index a4799d633ef..5aab87eae1f 100644 --- a/arch/h8300/kernel/signal.c +++ b/arch/h8300/kernel/signal.c @@ -517,10 +517,8 @@ asmlinkage int do_signal(struct pt_regs *regs, sigset_t *oldset) if ((regs->ccr & 0x10)) return 1; - if (current->flags & PF_FREEZE) { - refrigerator(0); + if (try_to_freeze()) goto no_signal; - } current->thread.esp0 = (unsigned long) regs; diff --git a/arch/h8300/platform/h8300h/ptrace_h8300h.c b/arch/h8300/platform/h8300h/ptrace_h8300h.c index 18e51a7167d..6ac93c05a1a 100644 --- a/arch/h8300/platform/h8300h/ptrace_h8300h.c +++ b/arch/h8300/platform/h8300h/ptrace_h8300h.c @@ -245,12 +245,12 @@ static unsigned short *getnextpc(struct task_struct *child, unsigned short *pc) addr = h8300_get_reg(child, regno-1+PT_ER1); return (unsigned short *)addr; case relb: - if ((inst = 0x55) || isbranch(child,inst & 0x0f)) + if (inst == 0x55 || isbranch(child,inst & 0x0f)) pc = (unsigned short *)((unsigned long)pc + ((signed char)(*fetch_p))); return pc+1; /* skip myself */ case relw: - if ((inst = 0x5c) || isbranch(child,(*fetch_p & 0xf0) >> 4)) + if (inst == 0x5c || isbranch(child,(*fetch_p & 0xf0) >> 4)) pc = (unsigned short *)((unsigned long)pc + ((signed short)(*(pc+1)))); return pc+2; /* skip myself */ diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig index dfd904f6883..6c02336fe2e 100644 --- a/arch/i386/Kconfig +++ b/arch/i386/Kconfig @@ -68,7 +68,6 @@ config X86_VOYAGER config X86_NUMAQ bool "NUMAQ (IBM/Sequent)" - select DISCONTIGMEM select NUMA help This option is used for getting Linux to run on a (IBM/Sequent) NUMA @@ -511,28 +510,7 @@ config SCHED_SMT cost of slightly increased overhead in some places. If unsure say N here. -config PREEMPT - bool "Preemptible Kernel" - help - This option reduces the latency of the kernel when reacting to - real-time or interactive events by allowing a low priority process to - be preempted even if it is in kernel mode executing a system call. - This allows applications to run more reliably even when the system is - under load. - - Say Y here if you are building a kernel for a desktop, embedded - or real-time system. Say N if you are unsure. - -config PREEMPT_BKL - bool "Preempt The Big Kernel Lock" - depends on PREEMPT - default y - help - This option reduces the latency of the kernel by making the - big kernel lock preemptible. - - Say Y here if you are building a kernel for a desktop system. - Say N if you are unsure. +source "kernel/Kconfig.preempt" config X86_UP_APIC bool "Local APIC support on uniprocessors" @@ -783,25 +761,48 @@ comment "NUMA (NUMA-Q) requires SMP, 64GB highmem support" comment "NUMA (Summit) requires SMP, 64GB highmem support, ACPI" depends on X86_SUMMIT && (!HIGHMEM64G || !ACPI) -config DISCONTIGMEM - bool - depends on NUMA - default y - config HAVE_ARCH_BOOTMEM_NODE bool depends on NUMA default y -config HAVE_MEMORY_PRESENT +config ARCH_HAVE_MEMORY_PRESENT bool depends on DISCONTIGMEM default y config NEED_NODE_MEMMAP_SIZE bool - depends on DISCONTIGMEM + depends on DISCONTIGMEM || SPARSEMEM + default y + +config HAVE_ARCH_ALLOC_REMAP + bool + depends on NUMA + default y + +config ARCH_DISCONTIGMEM_ENABLE + def_bool y + depends on NUMA + +config ARCH_DISCONTIGMEM_DEFAULT + def_bool y + depends on NUMA + +config ARCH_SPARSEMEM_ENABLE + def_bool y + depends on NUMA + +config ARCH_SELECT_MEMORY_MODEL + def_bool y + depends on ARCH_SPARSEMEM_ENABLE + +source "mm/Kconfig" + +config HAVE_ARCH_EARLY_PFN_TO_NID + bool default y + depends on NUMA config HIGHPTE bool "Allocate 3rd-level pagetables from highmem" @@ -939,6 +940,43 @@ config SECCOMP If unsure, say Y. Only embedded should say N here. +source kernel/Kconfig.hz + +config PHYSICAL_START + hex "Physical address where the kernel is loaded" if EMBEDDED + default "0x100000" + help + This gives the physical address where the kernel is loaded. + Primarily used in the case of kexec on panic where the + fail safe kernel needs to run at a different address than + the panic-ed kernel. + + Don't change this unless you know what you are doing. + +config KEXEC + bool "kexec system call (EXPERIMENTAL)" + depends on EXPERIMENTAL + help + kexec is a system call that implements the ability to shutdown your + current kernel, and to start another kernel. It is like a reboot + but it is indepedent of the system firmware. And like a reboot + you can start any kernel with it, not just Linux. + + The name comes from the similiarity to the exec system call. + + It is an ongoing process to be certain the hardware in a machine + is properly shutdown, so do not be surprised if this code does not + initially work for you. It may help to enable device hotplugging + support. As of this writing the exact hardware interface is + strongly in flux, so no good recommendation can be made. + +config CRASH_DUMP + bool "kernel crash dumps (EXPERIMENTAL)" + depends on EMBEDDED + depends on EXPERIMENTAL + depends on HIGHMEM + help + Generate crash dump after being started by kexec. endmenu @@ -1226,6 +1264,15 @@ config SCx200 This support is also available as a module. If compiled as a module, it will be called scx200. +config HOTPLUG_CPU + bool "Support for hot-pluggable CPUs (EXPERIMENTAL)" + depends on SMP && HOTPLUG && EXPERIMENTAL + ---help--- + Say Y here to experiment with turning CPUs off and on. CPUs + can be controlled through /sys/devices/system/cpu. + + Say N. + source "drivers/pcmcia/Kconfig" source "drivers/pci/hotplug/Kconfig" diff --git a/arch/i386/Makefile b/arch/i386/Makefile index 1c36ca332a9..bf7c9ba709f 100644 --- a/arch/i386/Makefile +++ b/arch/i386/Makefile @@ -17,6 +17,13 @@ # 20050320 Kianusch Sayah Karadji <kianusch@sk-tech.net> # Added support for GEODE CPU +HAS_BIARCH := $(call cc-option-yn, -m32) +ifeq ($(HAS_BIARCH),y) +AS := $(AS) --32 +LD := $(LD) -m elf_i386 +CC := $(CC) -m32 +endif + LDFLAGS := -m elf_i386 OBJCOPYFLAGS := -O binary -R .note -R .comment -S LDFLAGS_vmlinux := diff --git a/arch/i386/boot/Makefile b/arch/i386/boot/Makefile index 43cd6220ee4..1e71382d413 100644 --- a/arch/i386/boot/Makefile +++ b/arch/i386/boot/Makefile @@ -25,8 +25,8 @@ SVGA_MODE := -DSVGA_MODE=NORMAL_VGA #RAMDISK := -DRAMDISK=512 -targets := vmlinux.bin bootsect bootsect.o setup setup.o \ - zImage bzImage +targets := vmlinux.bin bootsect bootsect.o \ + setup setup.o zImage bzImage subdir- := compressed hostprogs-y := tools/build diff --git a/arch/i386/boot/compressed/head.S b/arch/i386/boot/compressed/head.S index c5e80b69e7d..b5893e4ecd3 100644 --- a/arch/i386/boot/compressed/head.S +++ b/arch/i386/boot/compressed/head.S @@ -25,6 +25,7 @@ #include <linux/linkage.h> #include <asm/segment.h> +#include <asm/page.h> .globl startup_32 @@ -74,7 +75,7 @@ startup_32: popl %esi # discard address popl %esi # real mode pointer xorl %ebx,%ebx - ljmp $(__BOOT_CS), $0x100000 + ljmp $(__BOOT_CS), $__PHYSICAL_START /* * We come here, if we were loaded high. @@ -99,7 +100,7 @@ startup_32: popl %ecx # lcount popl %edx # high_buffer_start popl %eax # hcount - movl $0x100000,%edi + movl $__PHYSICAL_START,%edi cli # make sure we don't get interrupted ljmp $(__BOOT_CS), $0x1000 # and jump to the move routine @@ -124,5 +125,5 @@ move_routine_start: movsl movl %ebx,%esi # Restore setup pointer xorl %ebx,%ebx - ljmp $(__BOOT_CS), $0x100000 + ljmp $(__BOOT_CS), $__PHYSICAL_START move_routine_end: diff --git a/arch/i386/boot/compressed/misc.c b/arch/i386/boot/compressed/misc.c index cedc55cc47d..82a807f9f5e 100644 --- a/arch/i386/boot/compressed/misc.c +++ b/arch/i386/boot/compressed/misc.c @@ -13,6 +13,7 @@ #include <linux/vmalloc.h> #include <linux/tty.h> #include <asm/io.h> +#include <asm/page.h> /* * gzip declarations @@ -308,7 +309,7 @@ static void setup_normal_output_buffer(void) #else if ((RM_ALT_MEM_K > RM_EXT_MEM_K ? RM_ALT_MEM_K : RM_EXT_MEM_K) < 1024) error("Less than 2MB of memory"); #endif - output_data = (char *)0x100000; /* Points to 1M */ + output_data = (char *)__PHYSICAL_START; /* Normally Points to 1M */ free_mem_end_ptr = (long)real_mode; } @@ -333,8 +334,8 @@ static void setup_output_buffer_if_we_run_high(struct moveparams *mv) low_buffer_size = low_buffer_end - LOW_BUFFER_START; high_loaded = 1; free_mem_end_ptr = (long)high_buffer_start; - if ( (0x100000 + low_buffer_size) > ((ulg)high_buffer_start)) { - high_buffer_start = (uch *)(0x100000 + low_buffer_size); + if ( (__PHYSICAL_START + low_buffer_size) > ((ulg)high_buffer_start)) { + high_buffer_start = (uch *)(__PHYSICAL_START + low_buffer_size); mv->hcount = 0; /* say: we need not to move high_buffer */ } else mv->hcount = -1; @@ -353,7 +354,6 @@ static void close_output_buffer_if_we_run_high(struct moveparams *mv) } } - asmlinkage int decompress_kernel(struct moveparams *mv, void *rmode) { real_mode = rmode; diff --git a/arch/i386/boot/edd.S b/arch/i386/boot/edd.S index 027d6b354ff..d8d69f2b911 100644 --- a/arch/i386/boot/edd.S +++ b/arch/i386/boot/edd.S @@ -6,7 +6,7 @@ * projects 1572D, 1484D, 1386D, 1226DT * disk signature read by Matt Domsch <Matt_Domsch@dell.com> * and Andrew Wilks <Andrew_Wilks@dell.com> September 2003, June 2004 - * legacy CHS retreival by Patrick J. LoPresti <patl@users.sourceforge.net> + * legacy CHS retrieval by Patrick J. LoPresti <patl@users.sourceforge.net> * March 2004 * Command line option parsing, Matt Domsch, November 2004 */ diff --git a/arch/i386/boot/install.sh b/arch/i386/boot/install.sh index 90f2452b3b9..f17b40dfc0f 100644 --- a/arch/i386/boot/install.sh +++ b/arch/i386/boot/install.sh @@ -21,8 +21,8 @@ # User may have a custom install script -if [ -x ~/bin/installkernel ]; then exec ~/bin/installkernel "$@"; fi -if [ -x /sbin/installkernel ]; then exec /sbin/installkernel "$@"; fi +if [ -x ~/bin/${CROSS_COMPILE}installkernel ]; then exec ~/bin/${CROSS_COMPILE}installkernel "$@"; fi +if [ -x /sbin/${CROSS_COMPILE}installkernel ]; then exec /sbin/${CROSS_COMPILE}installkernel "$@"; fi # Default install - same as make zlilo diff --git a/arch/i386/boot/setup.S b/arch/i386/boot/setup.S index caa1fde6904..8cb420f40c5 100644 --- a/arch/i386/boot/setup.S +++ b/arch/i386/boot/setup.S @@ -33,7 +33,7 @@ * Transcribed from Intel (as86) -> AT&T (gas) by Chris Noe, May 1999. * <stiker@northlink.com> * - * Fix to work around buggy BIOSes which dont use carry bit correctly + * Fix to work around buggy BIOSes which don't use carry bit correctly * and/or report extended memory in CX/DX for e801h memory size detection * call. As a result the kernel got wrong figures. The int15/e801h docs * from Ralf Brown interrupt list seem to indicate AX/BX should be used @@ -357,7 +357,7 @@ bail820: meme801: stc # fix to work around buggy - xorw %cx,%cx # BIOSes which dont clear/set + xorw %cx,%cx # BIOSes which don't clear/set xorw %dx,%dx # carry on pass/error of # e801h memory size call # or merely pass cx,dx though @@ -847,7 +847,7 @@ flush_instr: # # but we yet haven't reloaded the CS register, so the default size # of the target offset still is 16 bit. -# However, using an operand prefix (0x66), the CPU will properly +# However, using an operand prefix (0x66), the CPU will properly # take our 48 bit far pointer. (INTeL 80386 Programmer's Reference # Manual, Mixing 16-bit and 32-bit code, page 16-6) diff --git a/arch/i386/boot/tools/build.c b/arch/i386/boot/tools/build.c index 26509b826ae..4a17956512e 100644 --- a/arch/i386/boot/tools/build.c +++ b/arch/i386/boot/tools/build.c @@ -1,6 +1,4 @@ /* - * $Id: build.c,v 1.5 1997/05/19 12:29:58 mj Exp $ - * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright (C) 1997 Martin Mares */ @@ -8,7 +6,8 @@ /* * This file builds a disk-image from three different files: * - * - bootsect: exactly 512 bytes of 8086 machine code, loads the rest + * - bootsect: compatibility mbr which prints an error message if + * someone tries to boot the kernel directly. * - setup: 8086 machine code, sets up system parm * - system: 80386 code for actual system * diff --git a/arch/i386/crypto/aes.c b/arch/i386/crypto/aes.c index 1019430fc1f..88ee85c3b43 100644 --- a/arch/i386/crypto/aes.c +++ b/arch/i386/crypto/aes.c @@ -59,7 +59,7 @@ struct aes_ctx { }; #define WPOLY 0x011b -#define u32_in(x) le32_to_cpu(*(const u32 *)(x)) +#define u32_in(x) le32_to_cpup((const __le32 *)(x)) #define bytes2word(b0, b1, b2, b3) \ (((u32)(b3) << 24) | ((u32)(b2) << 16) | ((u32)(b1) << 8) | (b0)) diff --git a/arch/i386/defconfig b/arch/i386/defconfig index 28e62038379..ca07b95c06b 100644 --- a/arch/i386/defconfig +++ b/arch/i386/defconfig @@ -126,7 +126,6 @@ CONFIG_HAVE_DEC_LOCK=y # CONFIG_PM=y CONFIG_SOFTWARE_SUSPEND=y -# CONFIG_PM_DISK is not set # # ACPI (Advanced Configuration and Power Interface) Support diff --git a/arch/i386/kernel/Makefile b/arch/i386/kernel/Makefile index 51ecd512603..4cc83b322b3 100644 --- a/arch/i386/kernel/Makefile +++ b/arch/i386/kernel/Makefile @@ -24,6 +24,7 @@ obj-$(CONFIG_X86_MPPARSE) += mpparse.o obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o obj-$(CONFIG_X86_IO_APIC) += io_apic.o obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups.o +obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o obj-$(CONFIG_X86_NUMAQ) += numaq.o obj-$(CONFIG_X86_SUMMIT_NUMA) += summit.o obj-$(CONFIG_KPROBES) += kprobes.o diff --git a/arch/i386/kernel/acpi/boot.c b/arch/i386/kernel/acpi/boot.c index 848bb97af7c..9f63ae0f404 100644 --- a/arch/i386/kernel/acpi/boot.c +++ b/arch/i386/kernel/acpi/boot.c @@ -29,6 +29,7 @@ #include <linux/efi.h> #include <linux/irq.h> #include <linux/module.h> +#include <linux/dmi.h> #include <asm/pgtable.h> #include <asm/io_apic.h> @@ -815,6 +816,219 @@ acpi_process_madt(void) return; } +extern int acpi_force; + +#ifdef __i386__ + +#ifdef CONFIG_ACPI_PCI +static int __init disable_acpi_irq(struct dmi_system_id *d) +{ + if (!acpi_force) { + printk(KERN_NOTICE "%s detected: force use of acpi=noirq\n", + d->ident); + acpi_noirq_set(); + } + return 0; +} + +static int __init disable_acpi_pci(struct dmi_system_id *d) +{ + if (!acpi_force) { + printk(KERN_NOTICE "%s detected: force use of pci=noacpi\n", + d->ident); + acpi_disable_pci(); + } + return 0; +} +#endif + +static int __init dmi_disable_acpi(struct dmi_system_id *d) +{ + if (!acpi_force) { + printk(KERN_NOTICE "%s detected: acpi off\n",d->ident); + disable_acpi(); + } else { + printk(KERN_NOTICE + "Warning: DMI blacklist says broken, but acpi forced\n"); + } + return 0; +} + +/* + * Limit ACPI to CPU enumeration for HT + */ +static int __init force_acpi_ht(struct dmi_system_id *d) +{ + if (!acpi_force) { + printk(KERN_NOTICE "%s detected: force use of acpi=ht\n", d->ident); + disable_acpi(); + acpi_ht = 1; + } else { + printk(KERN_NOTICE + "Warning: acpi=force overrules DMI blacklist: acpi=ht\n"); + } + return 0; +} + +/* + * If your system is blacklisted here, but you find that acpi=force + * works for you, please contact acpi-devel@sourceforge.net + */ +static struct dmi_system_id __initdata acpi_dmi_table[] = { + /* + * Boxes that need ACPI disabled + */ + { + .callback = dmi_disable_acpi, + .ident = "IBM Thinkpad", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_BOARD_NAME, "2629H1G"), + }, + }, + + /* + * Boxes that need acpi=ht + */ + { + .callback = force_acpi_ht, + .ident = "FSC Primergy T850", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "FUJITSU SIEMENS"), + DMI_MATCH(DMI_PRODUCT_NAME, "PRIMERGY T850"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "DELL GX240", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "Dell Computer Corporation"), + DMI_MATCH(DMI_BOARD_NAME, "OptiPlex GX240"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "HP VISUALIZE NT Workstation", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "Hewlett-Packard"), + DMI_MATCH(DMI_PRODUCT_NAME, "HP VISUALIZE NT Workstation"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "Compaq Workstation W8000", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Compaq"), + DMI_MATCH(DMI_PRODUCT_NAME, "Workstation W8000"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "ASUS P4B266", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), + DMI_MATCH(DMI_BOARD_NAME, "P4B266"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "ASUS P2B-DS", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), + DMI_MATCH(DMI_BOARD_NAME, "P2B-DS"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "ASUS CUR-DLS", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), + DMI_MATCH(DMI_BOARD_NAME, "CUR-DLS"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "ABIT i440BX-W83977", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ABIT <http://www.abit.com>"), + DMI_MATCH(DMI_BOARD_NAME, "i440BX-W83977 (BP6)"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "IBM Bladecenter", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_BOARD_NAME, "IBM eServer BladeCenter HS20"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "IBM eServer xSeries 360", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_BOARD_NAME, "eServer xSeries 360"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "IBM eserver xSeries 330", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_BOARD_NAME, "eserver xSeries 330"), + }, + }, + { + .callback = force_acpi_ht, + .ident = "IBM eserver xSeries 440", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_PRODUCT_NAME, "eserver xSeries 440"), + }, + }, + +#ifdef CONFIG_ACPI_PCI + /* + * Boxes that need ACPI PCI IRQ routing disabled + */ + { + .callback = disable_acpi_irq, + .ident = "ASUS A7V", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC"), + DMI_MATCH(DMI_BOARD_NAME, "<A7V>"), + /* newer BIOS, Revision 1011, does work */ + DMI_MATCH(DMI_BIOS_VERSION, "ASUS A7V ACPI BIOS Revision 1007"), + }, + }, + + /* + * Boxes that need ACPI PCI IRQ routing and PCI scan disabled + */ + { /* _BBN 0 bug */ + .callback = disable_acpi_pci, + .ident = "ASUS PR-DLS", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), + DMI_MATCH(DMI_BOARD_NAME, "PR-DLS"), + DMI_MATCH(DMI_BIOS_VERSION, "ASUS PR-DLS ACPI BIOS Revision 1010"), + DMI_MATCH(DMI_BIOS_DATE, "03/21/2003") + }, + }, + { + .callback = disable_acpi_pci, + .ident = "Acer TravelMate 36x Laptop", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Acer"), + DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 360"), + }, + }, +#endif + { } +}; + +#endif /* __i386__ */ + /* * acpi_boot_table_init() and acpi_boot_init() * called from setup_arch(), always. @@ -843,6 +1057,10 @@ acpi_boot_table_init(void) { int error; +#ifdef __i386__ + dmi_check_system(acpi_dmi_table); +#endif + /* * If acpi_disabled, bail out * One exception: acpi=ht continues far enough to enumerate LAPICs @@ -870,8 +1088,6 @@ acpi_boot_table_init(void) */ error = acpi_blacklisted(); if (error) { - extern int acpi_force; - if (acpi_force) { printk(KERN_WARNING PREFIX "acpi=force override\n"); } else { diff --git a/arch/i386/kernel/acpi/sleep.c b/arch/i386/kernel/acpi/sleep.c index 28bb0514bb6..c1af93032ff 100644 --- a/arch/i386/kernel/acpi/sleep.c +++ b/arch/i386/kernel/acpi/sleep.c @@ -7,6 +7,7 @@ #include <linux/acpi.h> #include <linux/bootmem.h> +#include <linux/dmi.h> #include <asm/smp.h> #include <asm/tlbflush.h> @@ -91,3 +92,29 @@ static int __init acpi_sleep_setup(char *str) __setup("acpi_sleep=", acpi_sleep_setup); + + +static __init int reset_videomode_after_s3(struct dmi_system_id *d) +{ + acpi_video_flags |= 2; + return 0; +} + +static __initdata struct dmi_system_id acpisleep_dmi_table[] = { + { /* Reset video mode after returning from ACPI S3 sleep */ + .callback = reset_videomode_after_s3, + .ident = "Toshiba Satellite 4030cdt", + .matches = { + DMI_MATCH(DMI_PRODUCT_NAME, "S4030CDT/4.3"), + }, + }, + { } +}; + +static int __init acpisleep_dmi_init(void) +{ + dmi_check_system(acpisleep_dmi_table); + return 0; +} + +core_initcall(acpisleep_dmi_init); diff --git a/arch/i386/kernel/apic.c b/arch/i386/kernel/apic.c index d509836b70c..93df90bbb87 100644 --- a/arch/i386/kernel/apic.c +++ b/arch/i386/kernel/apic.c @@ -26,6 +26,7 @@ #include <linux/mc146818rtc.h> #include <linux/kernel_stat.h> #include <linux/sysdev.h> +#include <linux/cpu.h> #include <asm/atomic.h> #include <asm/smp.h> @@ -40,6 +41,11 @@ #include "io_ports.h" /* + * Knob to control our willingness to enable the local APIC. + */ +int enable_local_apic __initdata = 0; /* -1=force-disable, +1=force-enable */ + +/* * Debug level */ int apic_verbosity; @@ -205,7 +211,7 @@ void __init connect_bsp_APIC(void) enable_apic_mode(); } -void disconnect_bsp_APIC(void) +void disconnect_bsp_APIC(int virt_wire_setup) { if (pic_mode) { /* @@ -219,6 +225,42 @@ void disconnect_bsp_APIC(void) outb(0x70, 0x22); outb(0x00, 0x23); } + else { + /* Go back to Virtual Wire compatibility mode */ + unsigned long value; + + /* For the spurious interrupt use vector F, and enable it */ + value = apic_read(APIC_SPIV); + value &= ~APIC_VECTOR_MASK; + value |= APIC_SPIV_APIC_ENABLED; + value |= 0xf; + apic_write_around(APIC_SPIV, value); + + if (!virt_wire_setup) { + /* For LVT0 make it edge triggered, active high, external and enabled */ + value = apic_read(APIC_LVT0); + value &= ~(APIC_MODE_MASK | APIC_SEND_PENDING | + APIC_INPUT_POLARITY | APIC_LVT_REMOTE_IRR | + APIC_LVT_LEVEL_TRIGGER | APIC_LVT_MASKED ); + value |= APIC_LVT_REMOTE_IRR | APIC_SEND_PENDING; + value = SET_APIC_DELIVERY_MODE(value, APIC_MODE_EXTINT); + apic_write_around(APIC_LVT0, value); + } + else { + /* Disable LVT0 */ + apic_write_around(APIC_LVT0, APIC_LVT_MASKED); + } + + /* For LVT1 make it edge triggered, active high, nmi and enabled */ + value = apic_read(APIC_LVT1); + value &= ~( + APIC_MODE_MASK | APIC_SEND_PENDING | + APIC_INPUT_POLARITY | APIC_LVT_REMOTE_IRR | + APIC_LVT_LEVEL_TRIGGER | APIC_LVT_MASKED); + value |= APIC_LVT_REMOTE_IRR | APIC_SEND_PENDING; + value = SET_APIC_DELIVERY_MODE(value, APIC_MODE_NMI); + apic_write_around(APIC_LVT1, value); + } } void disable_local_APIC(void) @@ -363,7 +405,7 @@ void __init init_bsp_APIC(void) apic_write_around(APIC_LVT1, value); } -void __init setup_local_APIC (void) +void __devinit setup_local_APIC(void) { unsigned long oldvalue, value, ver, maxlvt; @@ -634,7 +676,7 @@ static struct sys_device device_lapic = { .cls = &lapic_sysclass, }; -static void __init apic_pm_activate(void) +static void __devinit apic_pm_activate(void) { apic_pm_state.active = 1; } @@ -665,26 +707,6 @@ static void apic_pm_activate(void) { } * Original code written by Keir Fraser. */ -/* - * Knob to control our willingness to enable the local APIC. - */ -int enable_local_apic __initdata = 0; /* -1=force-disable, +1=force-enable */ - -static int __init lapic_disable(char *str) -{ - enable_local_apic = -1; - clear_bit(X86_FEATURE_APIC, boot_cpu_data.x86_capability); - return 0; -} -__setup("nolapic", lapic_disable); - -static int __init lapic_enable(char *str) -{ - enable_local_apic = 1; - return 0; -} -__setup("lapic", lapic_enable); - static int __init apic_set_verbosity(char *str) { if (strcmp("debug", str) == 0) @@ -855,7 +877,7 @@ fake_ioapic_page: * but we do not accept timer interrupts yet. We only allow the BP * to calibrate. */ -static unsigned int __init get_8254_timer_count(void) +static unsigned int __devinit get_8254_timer_count(void) { extern spinlock_t i8253_lock; unsigned long flags; @@ -874,7 +896,7 @@ static unsigned int __init get_8254_timer_count(void) } /* next tick in 8254 can be caught by catching timer wraparound */ -static void __init wait_8254_wraparound(void) +static void __devinit wait_8254_wraparound(void) { unsigned int curr_count, prev_count; @@ -894,7 +916,7 @@ static void __init wait_8254_wraparound(void) * Default initialization for 8254 timers. If we use other timers like HPET, * we override this later */ -void (*wait_timer_tick)(void) __initdata = wait_8254_wraparound; +void (*wait_timer_tick)(void) __devinitdata = wait_8254_wraparound; /* * This function sets up the local APIC timer, with a timeout of @@ -930,7 +952,7 @@ static void __setup_APIC_LVTT(unsigned int clocks) apic_write_around(APIC_TMICT, clocks/APIC_DIVISOR); } -static void __init setup_APIC_timer(unsigned int clocks) +static void __devinit setup_APIC_timer(unsigned int clocks) { unsigned long flags; @@ -1043,12 +1065,12 @@ void __init setup_boot_APIC_clock(void) local_irq_enable(); } -void __init setup_secondary_APIC_clock(void) +void __devinit setup_secondary_APIC_clock(void) { setup_APIC_timer(calibration_result); } -void __init disable_APIC_timer(void) +void __devinit disable_APIC_timer(void) { if (using_apic_timer) { unsigned long v; @@ -1133,7 +1155,7 @@ inline void smp_local_timer_interrupt(struct pt_regs * regs) } #ifdef CONFIG_SMP - update_process_times(user_mode(regs)); + update_process_times(user_mode_vm(regs)); #endif } diff --git a/arch/i386/kernel/apm.c b/arch/i386/kernel/apm.c index 0ff65abcd56..d48ce929096 100644 --- a/arch/i386/kernel/apm.c +++ b/arch/i386/kernel/apm.c @@ -346,10 +346,10 @@ extern int (*console_blank_hook)(int); struct apm_user { int magic; struct apm_user * next; - int suser: 1; - int writer: 1; - int reader: 1; - int suspend_wait: 1; + unsigned int suser: 1; + unsigned int writer: 1; + unsigned int reader: 1; + unsigned int suspend_wait: 1; int suspend_result; int suspends_pending; int standbys_pending; diff --git a/arch/i386/kernel/cpu/common.c b/arch/i386/kernel/cpu/common.c index d199e525680..2203a9d2021 100644 --- a/arch/i386/kernel/cpu/common.c +++ b/arch/i386/kernel/cpu/common.c @@ -24,9 +24,9 @@ EXPORT_PER_CPU_SYMBOL(cpu_gdt_table); DEFINE_PER_CPU(unsigned char, cpu_16bit_stack[CPU_16BIT_STACK_SIZE]); EXPORT_PER_CPU_SYMBOL(cpu_16bit_stack); -static int cachesize_override __initdata = -1; -static int disable_x86_fxsr __initdata = 0; -static int disable_x86_serial_nr __initdata = 1; +static int cachesize_override __devinitdata = -1; +static int disable_x86_fxsr __devinitdata = 0; +static int disable_x86_serial_nr __devinitdata = 1; struct cpu_dev * cpu_devs[X86_VENDOR_NUM] = {}; @@ -59,7 +59,7 @@ static int __init cachesize_setup(char *str) } __setup("cachesize=", cachesize_setup); -int __init get_model_name(struct cpuinfo_x86 *c) +int __devinit get_model_name(struct cpuinfo_x86 *c) { unsigned int *v; char *p, *q; @@ -89,7 +89,7 @@ int __init get_model_name(struct cpuinfo_x86 *c) } -void __init display_cacheinfo(struct cpuinfo_x86 *c) +void __devinit display_cacheinfo(struct cpuinfo_x86 *c) { unsigned int n, dummy, ecx, edx, l2size; @@ -130,7 +130,7 @@ void __init display_cacheinfo(struct cpuinfo_x86 *c) /* in particular, if CPUID levels 0x80000002..4 are supported, this isn't used */ /* Look up CPU names by table lookup. */ -static char __init *table_lookup_model(struct cpuinfo_x86 *c) +static char __devinit *table_lookup_model(struct cpuinfo_x86 *c) { struct cpu_model_info *info; @@ -151,7 +151,7 @@ static char __init *table_lookup_model(struct cpuinfo_x86 *c) } -void __init get_cpu_vendor(struct cpuinfo_x86 *c, int early) +void __devinit get_cpu_vendor(struct cpuinfo_x86 *c, int early) { char *v = c->x86_vendor_id; int i; @@ -202,7 +202,7 @@ static inline int flag_is_changeable_p(u32 flag) /* Probe for the CPUID instruction */ -static int __init have_cpuid_p(void) +static int __devinit have_cpuid_p(void) { return flag_is_changeable_p(X86_EFLAGS_ID); } @@ -249,7 +249,7 @@ static void __init early_cpu_detect(void) #endif } -void __init generic_identify(struct cpuinfo_x86 * c) +void __devinit generic_identify(struct cpuinfo_x86 * c) { u32 tfms, xlvl; int junk; @@ -296,7 +296,7 @@ void __init generic_identify(struct cpuinfo_x86 * c) } } -static void __init squash_the_stupid_serial_number(struct cpuinfo_x86 *c) +static void __devinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c) { if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr ) { /* Disable processor serial number */ @@ -324,7 +324,7 @@ __setup("serialnumber", x86_serial_nr_setup); /* * This does the hard work of actually picking apart the CPU stuff... */ -void __init identify_cpu(struct cpuinfo_x86 *c) +void __devinit identify_cpu(struct cpuinfo_x86 *c) { int i; @@ -432,10 +432,13 @@ void __init identify_cpu(struct cpuinfo_x86 *c) #ifdef CONFIG_X86_MCE mcheck_init(c); #endif + if (c == &boot_cpu_data) + sysenter_setup(); + enable_sep_cpu(); } #ifdef CONFIG_X86_HT -void __init detect_ht(struct cpuinfo_x86 *c) +void __devinit detect_ht(struct cpuinfo_x86 *c) { u32 eax, ebx, ecx, edx; int index_msb, tmp; @@ -490,7 +493,7 @@ void __init detect_ht(struct cpuinfo_x86 *c) } #endif -void __init print_cpu_info(struct cpuinfo_x86 *c) +void __devinit print_cpu_info(struct cpuinfo_x86 *c) { char *vendor = NULL; @@ -513,7 +516,7 @@ void __init print_cpu_info(struct cpuinfo_x86 *c) printk("\n"); } -cpumask_t cpu_initialized __initdata = CPU_MASK_NONE; +cpumask_t cpu_initialized __devinitdata = CPU_MASK_NONE; /* This is hacky. :) * We're emulating future behavior. @@ -560,7 +563,7 @@ void __init early_cpu_init(void) * and IDT. We reload them nevertheless, this function acts as a * 'CPU state barrier', nothing should get across. */ -void __init cpu_init (void) +void __devinit cpu_init(void) { int cpu = smp_processor_id(); struct tss_struct * t = &per_cpu(init_tss, cpu); @@ -635,7 +638,7 @@ void __init cpu_init (void) /* Clear all 6 debug registers: */ -#define CD(register) __asm__("movl %0,%%db" #register ::"r"(0) ); +#define CD(register) set_debugreg(0, register) CD(0); CD(1); CD(2); CD(3); /* no db4 and db5 */; CD(6); CD(7); @@ -648,3 +651,15 @@ void __init cpu_init (void) clear_used_math(); mxcsr_feature_mask_init(); } + +#ifdef CONFIG_HOTPLUG_CPU +void __devinit cpu_uninit(void) +{ + int cpu = raw_smp_processor_id(); + cpu_clear(cpu, cpu_initialized); + + /* lazy TLB state */ + per_cpu(cpu_tlbstate, cpu).state = 0; + per_cpu(cpu_tlbstate, cpu).active_mm = &init_mm; +} +#endif diff --git a/arch/i386/kernel/cpu/cpufreq/powernow-k7.c b/arch/i386/kernel/cpu/cpufreq/powernow-k7.c index 5c530064eb7..73a5dc5b26b 100644 --- a/arch/i386/kernel/cpu/cpufreq/powernow-k7.c +++ b/arch/i386/kernel/cpu/cpufreq/powernow-k7.c @@ -648,9 +648,7 @@ static int powernow_cpu_exit (struct cpufreq_policy *policy) { } #endif - if (powernow_table) - kfree(powernow_table); - + kfree(powernow_table); return 0; } diff --git a/arch/i386/kernel/cpu/intel.c b/arch/i386/kernel/cpu/intel.c index 121aa2176e6..96a75d04583 100644 --- a/arch/i386/kernel/cpu/intel.c +++ b/arch/i386/kernel/cpu/intel.c @@ -28,7 +28,7 @@ extern int trap_init_f00f_bug(void); struct movsl_mask movsl_mask; #endif -void __init early_intel_workaround(struct cpuinfo_x86 *c) +void __devinit early_intel_workaround(struct cpuinfo_x86 *c) { if (c->x86_vendor != X86_VENDOR_INTEL) return; @@ -43,7 +43,7 @@ void __init early_intel_workaround(struct cpuinfo_x86 *c) * This is called before we do cpu ident work */ -int __init ppro_with_ram_bug(void) +int __devinit ppro_with_ram_bug(void) { /* Uses data from early_cpu_detect now */ if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && @@ -61,7 +61,7 @@ int __init ppro_with_ram_bug(void) * P4 Xeon errata 037 workaround. * Hardware prefetcher may cause stale data to be loaded into the cache. */ -static void __init Intel_errata_workarounds(struct cpuinfo_x86 *c) +static void __devinit Intel_errata_workarounds(struct cpuinfo_x86 *c) { unsigned long lo, hi; @@ -80,7 +80,7 @@ static void __init Intel_errata_workarounds(struct cpuinfo_x86 *c) /* * find out the number of processor cores on the die */ -static int __init num_cpu_cores(struct cpuinfo_x86 *c) +static int __devinit num_cpu_cores(struct cpuinfo_x86 *c) { unsigned int eax; @@ -98,7 +98,7 @@ static int __init num_cpu_cores(struct cpuinfo_x86 *c) return 1; } -static void __init init_intel(struct cpuinfo_x86 *c) +static void __devinit init_intel(struct cpuinfo_x86 *c) { unsigned int l2 = 0; char *p = NULL; @@ -204,7 +204,7 @@ static unsigned int intel_size_cache(struct cpuinfo_x86 * c, unsigned int size) return size; } -static struct cpu_dev intel_cpu_dev __initdata = { +static struct cpu_dev intel_cpu_dev __devinitdata = { .c_vendor = "Intel", .c_ident = { "GenuineIntel" }, .c_models = { diff --git a/arch/i386/kernel/cpu/intel_cacheinfo.c b/arch/i386/kernel/cpu/intel_cacheinfo.c index a710dc4eb20..1d768b26326 100644 --- a/arch/i386/kernel/cpu/intel_cacheinfo.c +++ b/arch/i386/kernel/cpu/intel_cacheinfo.c @@ -28,7 +28,7 @@ struct _cache_table }; /* all the cache descriptor types we care about (no TLB or trace cache entries) */ -static struct _cache_table cache_table[] __initdata = +static struct _cache_table cache_table[] __devinitdata = { { 0x06, LVL_1_INST, 8 }, /* 4-way set assoc, 32 byte line size */ { 0x08, LVL_1_INST, 16 }, /* 4-way set assoc, 32 byte line size */ @@ -160,7 +160,7 @@ static int __init find_num_cache_leaves(void) return retval; } -unsigned int __init init_intel_cacheinfo(struct cpuinfo_x86 *c) +unsigned int __devinit init_intel_cacheinfo(struct cpuinfo_x86 *c) { unsigned int trace = 0, l1i = 0, l1d = 0, l2 = 0, l3 = 0; /* Cache sizes */ unsigned int new_l1d = 0, new_l1i = 0; /* Cache sizes from cpuid(4) */ diff --git a/arch/i386/kernel/cpu/mcheck/k7.c b/arch/i386/kernel/cpu/mcheck/k7.c index 8df52e86c4d..c4abe765739 100644 --- a/arch/i386/kernel/cpu/mcheck/k7.c +++ b/arch/i386/kernel/cpu/mcheck/k7.c @@ -69,7 +69,7 @@ static fastcall void k7_machine_check(struct pt_regs * regs, long error_code) /* AMD K7 machine check is Intel like */ -void __init amd_mcheck_init(struct cpuinfo_x86 *c) +void __devinit amd_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; int i; diff --git a/arch/i386/kernel/cpu/mcheck/mce.c b/arch/i386/kernel/cpu/mcheck/mce.c index bf6d1aefafc..2cf25d2ba0f 100644 --- a/arch/i386/kernel/cpu/mcheck/mce.c +++ b/arch/i386/kernel/cpu/mcheck/mce.c @@ -16,7 +16,7 @@ #include "mce.h" -int mce_disabled __initdata = 0; +int mce_disabled __devinitdata = 0; int nr_mce_banks; EXPORT_SYMBOL_GPL(nr_mce_banks); /* non-fatal.o */ @@ -31,7 +31,7 @@ static fastcall void unexpected_machine_check(struct pt_regs * regs, long error_ void fastcall (*machine_check_vector)(struct pt_regs *, long error_code) = unexpected_machine_check; /* This has to be run for each processor */ -void __init mcheck_init(struct cpuinfo_x86 *c) +void __devinit mcheck_init(struct cpuinfo_x86 *c) { if (mce_disabled==1) return; diff --git a/arch/i386/kernel/cpu/mcheck/p4.c b/arch/i386/kernel/cpu/mcheck/p4.c index 8b16ceb929b..0abccb6fdf9 100644 --- a/arch/i386/kernel/cpu/mcheck/p4.c +++ b/arch/i386/kernel/cpu/mcheck/p4.c @@ -78,7 +78,7 @@ fastcall void smp_thermal_interrupt(struct pt_regs *regs) } /* P4/Xeon Thermal regulation detect and init */ -static void __init intel_init_thermal(struct cpuinfo_x86 *c) +static void __devinit intel_init_thermal(struct cpuinfo_x86 *c) { u32 l, h; unsigned int cpu = smp_processor_id(); @@ -232,7 +232,7 @@ static fastcall void intel_machine_check(struct pt_regs * regs, long error_code) } -void __init intel_p4_mcheck_init(struct cpuinfo_x86 *c) +void __devinit intel_p4_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; int i; diff --git a/arch/i386/kernel/cpu/mcheck/p5.c b/arch/i386/kernel/cpu/mcheck/p5.c index c45a1b485c8..ec0614cd292 100644 --- a/arch/i386/kernel/cpu/mcheck/p5.c +++ b/arch/i386/kernel/cpu/mcheck/p5.c @@ -29,7 +29,7 @@ static fastcall void pentium_machine_check(struct pt_regs * regs, long error_cod } /* Set up machine check reporting for processors with Intel style MCE */ -void __init intel_p5_mcheck_init(struct cpuinfo_x86 *c) +void __devinit intel_p5_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; diff --git a/arch/i386/kernel/cpu/mcheck/p6.c b/arch/i386/kernel/cpu/mcheck/p6.c index 46640f8c249..f01b73f947e 100644 --- a/arch/i386/kernel/cpu/mcheck/p6.c +++ b/arch/i386/kernel/cpu/mcheck/p6.c @@ -80,7 +80,7 @@ static fastcall void intel_machine_check(struct pt_regs * regs, long error_code) } /* Set up machine check reporting for processors with Intel style MCE */ -void __init intel_p6_mcheck_init(struct cpuinfo_x86 *c) +void __devinit intel_p6_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; int i; diff --git a/arch/i386/kernel/cpu/mcheck/winchip.c b/arch/i386/kernel/cpu/mcheck/winchip.c index 753fa7acb98..7bae68fa168 100644 --- a/arch/i386/kernel/cpu/mcheck/winchip.c +++ b/arch/i386/kernel/cpu/mcheck/winchip.c @@ -23,7 +23,7 @@ static fastcall void winchip_machine_check(struct pt_regs * regs, long error_cod } /* Set up machine check reporting on the Winchip C6 series */ -void __init winchip_mcheck_init(struct cpuinfo_x86 *c) +void __devinit winchip_mcheck_init(struct cpuinfo_x86 *c) { u32 lo, hi; machine_check_vector = winchip_machine_check; diff --git a/arch/i386/kernel/cpu/mtrr/generic.c b/arch/i386/kernel/cpu/mtrr/generic.c index f468a979e9a..64d91f73a0a 100644 --- a/arch/i386/kernel/cpu/mtrr/generic.c +++ b/arch/i386/kernel/cpu/mtrr/generic.c @@ -70,8 +70,7 @@ void __init get_mtrr_state(void) /* Free resources associated with a struct mtrr_state */ void __init finalize_mtrr_state(void) { - if (mtrr_state.var_ranges) - kfree(mtrr_state.var_ranges); + kfree(mtrr_state.var_ranges); mtrr_state.var_ranges = NULL; } diff --git a/arch/i386/kernel/cpu/mtrr/main.c b/arch/i386/kernel/cpu/mtrr/main.c index e1c2042b9b7..d66b09e0c82 100644 --- a/arch/i386/kernel/cpu/mtrr/main.c +++ b/arch/i386/kernel/cpu/mtrr/main.c @@ -375,6 +375,19 @@ int mtrr_add_page(unsigned long base, unsigned long size, return error; } +static int mtrr_check(unsigned long base, unsigned long size) +{ + if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) { + printk(KERN_WARNING + "mtrr: size and base must be multiples of 4 kiB\n"); + printk(KERN_DEBUG + "mtrr: size: 0x%lx base: 0x%lx\n", size, base); + dump_stack(); + return -1; + } + return 0; +} + /** * mtrr_add - Add a memory type region * @base: Physical base address of region @@ -415,11 +428,8 @@ int mtrr_add(unsigned long base, unsigned long size, unsigned int type, char increment) { - if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) { - printk(KERN_WARNING "mtrr: size and base must be multiples of 4 kiB\n"); - printk(KERN_DEBUG "mtrr: size: 0x%lx base: 0x%lx\n", size, base); + if (mtrr_check(base, size)) return -EINVAL; - } return mtrr_add_page(base >> PAGE_SHIFT, size >> PAGE_SHIFT, type, increment); } @@ -511,11 +521,8 @@ int mtrr_del_page(int reg, unsigned long base, unsigned long size) int mtrr_del(int reg, unsigned long base, unsigned long size) { - if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) { - printk(KERN_INFO "mtrr: size and base must be multiples of 4 kiB\n"); - printk(KERN_DEBUG "mtrr: size: 0x%lx base: 0x%lx\n", size, base); + if (mtrr_check(base, size)) return -EINVAL; - } return mtrr_del_page(reg, base >> PAGE_SHIFT, size >> PAGE_SHIFT); } diff --git a/arch/i386/kernel/cpu/proc.c b/arch/i386/kernel/cpu/proc.c index 7323c19f354..8bd77d948a8 100644 --- a/arch/i386/kernel/cpu/proc.c +++ b/arch/i386/kernel/cpu/proc.c @@ -86,7 +86,7 @@ static int show_cpuinfo(struct seq_file *m, void *v) seq_printf(m, "stepping\t: unknown\n"); if ( cpu_has(c, X86_FEATURE_TSC) ) { - seq_printf(m, "cpu MHz\t\t: %lu.%03lu\n", + seq_printf(m, "cpu MHz\t\t: %u.%03u\n", cpu_khz / 1000, (cpu_khz % 1000)); } diff --git a/arch/i386/kernel/crash.c b/arch/i386/kernel/crash.c new file mode 100644 index 00000000000..e5fab12f792 --- /dev/null +++ b/arch/i386/kernel/crash.c @@ -0,0 +1,223 @@ +/* + * Architecture specific (i386) functions for kexec based crash dumps. + * + * Created by: Hariprasad Nellitheertha (hari@in.ibm.com) + * + * Copyright (C) IBM Corporation, 2004. All rights reserved. + * + */ + +#include <linux/init.h> +#include <linux/types.h> +#include <linux/kernel.h> +#include <linux/smp.h> +#include <linux/irq.h> +#include <linux/reboot.h> +#include <linux/kexec.h> +#include <linux/irq.h> +#include <linux/delay.h> +#include <linux/elf.h> +#include <linux/elfcore.h> + +#include <asm/processor.h> +#include <asm/hardirq.h> +#include <asm/nmi.h> +#include <asm/hw_irq.h> +#include <asm/apic.h> +#include <mach_ipi.h> + + +note_buf_t crash_notes[NR_CPUS]; +/* This keeps a track of which one is crashing cpu. */ +static int crashing_cpu; + +static u32 *append_elf_note(u32 *buf, char *name, unsigned type, void *data, + size_t data_len) +{ + struct elf_note note; + + note.n_namesz = strlen(name) + 1; + note.n_descsz = data_len; + note.n_type = type; + memcpy(buf, ¬e, sizeof(note)); + buf += (sizeof(note) +3)/4; + memcpy(buf, name, note.n_namesz); + buf += (note.n_namesz + 3)/4; + memcpy(buf, data, note.n_descsz); + buf += (note.n_descsz + 3)/4; + + return buf; +} + +static void final_note(u32 *buf) +{ + struct elf_note note; + + note.n_namesz = 0; + note.n_descsz = 0; + note.n_type = 0; + memcpy(buf, ¬e, sizeof(note)); +} + +static void crash_save_this_cpu(struct pt_regs *regs, int cpu) +{ + struct elf_prstatus prstatus; + u32 *buf; + + if ((cpu < 0) || (cpu >= NR_CPUS)) + return; + + /* Using ELF notes here is opportunistic. + * I need a well defined structure format + * for the data I pass, and I need tags + * on the data to indicate what information I have + * squirrelled away. ELF notes happen to provide + * all of that that no need to invent something new. + */ + buf = &crash_notes[cpu][0]; + memset(&prstatus, 0, sizeof(prstatus)); + prstatus.pr_pid = current->pid; + elf_core_copy_regs(&prstatus.pr_reg, regs); + buf = append_elf_note(buf, "CORE", NT_PRSTATUS, &prstatus, + sizeof(prstatus)); + final_note(buf); +} + +static void crash_get_current_regs(struct pt_regs *regs) +{ + __asm__ __volatile__("movl %%ebx,%0" : "=m"(regs->ebx)); + __asm__ __volatile__("movl %%ecx,%0" : "=m"(regs->ecx)); + __asm__ __volatile__("movl %%edx,%0" : "=m"(regs->edx)); + __asm__ __volatile__("movl %%esi,%0" : "=m"(regs->esi)); + __asm__ __volatile__("movl %%edi,%0" : "=m"(regs->edi)); + __asm__ __volatile__("movl %%ebp,%0" : "=m"(regs->ebp)); + __asm__ __volatile__("movl %%eax,%0" : "=m"(regs->eax)); + __asm__ __volatile__("movl %%esp,%0" : "=m"(regs->esp)); + __asm__ __volatile__("movw %%ss, %%ax;" :"=a"(regs->xss)); + __asm__ __volatile__("movw %%cs, %%ax;" :"=a"(regs->xcs)); + __asm__ __volatile__("movw %%ds, %%ax;" :"=a"(regs->xds)); + __asm__ __volatile__("movw %%es, %%ax;" :"=a"(regs->xes)); + __asm__ __volatile__("pushfl; popl %0" :"=m"(regs->eflags)); + + regs->eip = (unsigned long)current_text_addr(); +} + +/* CPU does not save ss and esp on stack if execution is already + * running in kernel mode at the time of NMI occurrence. This code + * fixes it. + */ +static void crash_setup_regs(struct pt_regs *newregs, struct pt_regs *oldregs) +{ + memcpy(newregs, oldregs, sizeof(*newregs)); + newregs->esp = (unsigned long)&(oldregs->esp); + __asm__ __volatile__("xorl %eax, %eax;"); + __asm__ __volatile__ ("movw %%ss, %%ax;" :"=a"(newregs->xss)); +} + +/* We may have saved_regs from where the error came from + * or it is NULL if via a direct panic(). + */ +static void crash_save_self(struct pt_regs *saved_regs) +{ + struct pt_regs regs; + int cpu; + + cpu = smp_processor_id(); + if (saved_regs) + crash_setup_regs(®s, saved_regs); + else + crash_get_current_regs(®s); + crash_save_this_cpu(®s, cpu); +} + +#ifdef CONFIG_SMP +static atomic_t waiting_for_crash_ipi; + +static int crash_nmi_callback(struct pt_regs *regs, int cpu) +{ + struct pt_regs fixed_regs; + + /* Don't do anything if this handler is invoked on crashing cpu. + * Otherwise, system will completely hang. Crashing cpu can get + * an NMI if system was initially booted with nmi_watchdog parameter. + */ + if (cpu == crashing_cpu) + return 1; + local_irq_disable(); + + if (!user_mode(regs)) { + crash_setup_regs(&fixed_regs, regs); + regs = &fixed_regs; + } + crash_save_this_cpu(regs, cpu); + disable_local_APIC(); + atomic_dec(&waiting_for_crash_ipi); + /* Assume hlt works */ + __asm__("hlt"); + for(;;); + + return 1; +} + +/* + * By using the NMI code instead of a vector we just sneak thru the + * word generator coming out with just what we want. AND it does + * not matter if clustered_apic_mode is set or not. + */ +static void smp_send_nmi_allbutself(void) +{ + send_IPI_allbutself(APIC_DM_NMI); +} + +static void nmi_shootdown_cpus(void) +{ + unsigned long msecs; + + atomic_set(&waiting_for_crash_ipi, num_online_cpus() - 1); + /* Would it be better to replace the trap vector here? */ + set_nmi_callback(crash_nmi_callback); + /* Ensure the new callback function is set before sending + * out the NMI + */ + wmb(); + + smp_send_nmi_allbutself(); + + msecs = 1000; /* Wait at most a second for the other cpus to stop */ + while ((atomic_read(&waiting_for_crash_ipi) > 0) && msecs) { + mdelay(1); + msecs--; + } + + /* Leave the nmi callback set */ + disable_local_APIC(); +} +#else +static void nmi_shootdown_cpus(void) +{ + /* There are no cpus to shootdown */ +} +#endif + +void machine_crash_shutdown(struct pt_regs *regs) +{ + /* This function is only called after the system + * has paniced or is otherwise in a critical state. + * The minimum amount of code to allow a kexec'd kernel + * to run successfully needs to happen here. + * + * In practice this means shooting down the other cpus in + * an SMP system. + */ + /* The kernel is broken so disable interrupts */ + local_irq_disable(); + + /* Make a note of crashing cpu. Will be used in NMI callback.*/ + crashing_cpu = smp_processor_id(); + nmi_shootdown_cpus(); + lapic_shutdown(); +#if defined(CONFIG_X86_IO_APIC) + disable_IO_APIC(); +#endif + crash_save_self(regs); +} diff --git a/arch/i386/kernel/dmi_scan.c b/arch/i386/kernel/dmi_scan.c index 6ed7e28f306..a3cdf894302 100644 --- a/arch/i386/kernel/dmi_scan.c +++ b/arch/i386/kernel/dmi_scan.c @@ -1,22 +1,15 @@ #include <linux/types.h> -#include <linux/kernel.h> #include <linux/string.h> #include <linux/init.h> #include <linux/module.h> -#include <linux/slab.h> -#include <linux/acpi.h> -#include <asm/io.h> -#include <linux/pm.h> -#include <asm/system.h> #include <linux/dmi.h> #include <linux/bootmem.h> -struct dmi_header -{ - u8 type; - u8 length; - u16 handle; +struct dmi_header { + u8 type; + u8 length; + u16 handle; }; #undef DMI_DEBUG @@ -29,15 +22,13 @@ struct dmi_header static char * __init dmi_string(struct dmi_header *dm, u8 s) { - u8 *bp=(u8 *)dm; - bp+=dm->length; - if(!s) + u8 *bp = ((u8 *) dm) + dm->length; + + if (!s) return ""; s--; - while(s>0 && *bp) - { - bp+=strlen(bp); - bp++; + while (s > 0 && *bp) { + bp += strlen(bp) + 1; s--; } return bp; @@ -47,16 +38,14 @@ static char * __init dmi_string(struct dmi_header *dm, u8 s) * We have to be cautious here. We have seen BIOSes with DMI pointers * pointing to completely the wrong place for example */ - -static int __init dmi_table(u32 base, int len, int num, void (*decode)(struct dmi_header *)) +static int __init dmi_table(u32 base, int len, int num, + void (*decode)(struct dmi_header *)) { - u8 *buf; - struct dmi_header *dm; - u8 *data; - int i=0; + u8 *buf, *data; + int i = 0; buf = bt_ioremap(base, len); - if(buf==NULL) + if (buf == NULL) return -1; data = buf; @@ -65,36 +54,34 @@ static int __init dmi_table(u32 base, int len, int num, void (*decode)(struct dm * Stop when we see all the items the table claimed to have * OR we run off the end of the table (also happens) */ - - while(i<num && data-buf+sizeof(struct dmi_header)<=len) - { - dm=(struct dmi_header *)data; + while ((i < num) && (data - buf + sizeof(struct dmi_header)) <= len) { + struct dmi_header *dm = (struct dmi_header *)data; /* * We want to know the total length (formated area and strings) * before decoding to make sure we won't run off the table in * dmi_decode or dmi_string */ - data+=dm->length; - while(data-buf<len-1 && (data[0] || data[1])) + data += dm->length; + while ((data - buf < len - 1) && (data[0] || data[1])) data++; - if(data-buf<len-1) + if (data - buf < len - 1) decode(dm); - data+=2; + data += 2; i++; } bt_iounmap(buf, len); return 0; } - -inline static int __init dmi_checksum(u8 *buf) +static int __init dmi_checksum(u8 *buf) { - u8 sum=0; + u8 sum = 0; int a; - for(a=0; a<15; a++) - sum+=buf[a]; - return (sum==0); + for (a = 0; a < 15; a++) + sum += buf[a]; + + return sum == 0; } static int __init dmi_iterate(void (*decode)(struct dmi_header *)) @@ -110,28 +97,30 @@ static int __init dmi_iterate(void (*decode)(struct dmi_header *)) p = ioremap(0xF0000, 0x10000); if (p == NULL) return -1; + for (q = p; q < p + 0x10000; q += 16) { memcpy_fromio(buf, q, 15); - if(memcmp(buf, "_DMI_", 5)==0 && dmi_checksum(buf)) - { - u16 num=buf[13]<<8|buf[12]; - u16 len=buf[7]<<8|buf[6]; - u32 base=buf[11]<<24|buf[10]<<16|buf[9]<<8|buf[8]; + if ((memcmp(buf, "_DMI_", 5) == 0) && dmi_checksum(buf)) { + u16 num = (buf[13] << 8) | buf[12]; + u16 len = (buf[7] << 8) | buf[6]; + u32 base = (buf[11] << 24) | (buf[10] << 16) | + (buf[9] << 8) | buf[8]; /* * DMI version 0.0 means that the real version is taken from * the SMBIOS version, which we don't know at this point. */ - if(buf[14]!=0) + if (buf[14] != 0) printk(KERN_INFO "DMI %d.%d present.\n", - buf[14]>>4, buf[14]&0x0F); + buf[14] >> 4, buf[14] & 0xF); else printk(KERN_INFO "DMI present.\n"); + dmi_printk((KERN_INFO "%d structures occupying %d bytes.\n", num, len)); - dmi_printk((KERN_INFO "DMI table at 0x%08X.\n", - base)); - if(dmi_table(base,len, num, decode)==0) + dmi_printk((KERN_INFO "DMI table at 0x%08X.\n", base)); + + if (dmi_table(base,len, num, decode) == 0) return 0; } } @@ -143,16 +132,17 @@ static char *dmi_ident[DMI_STRING_MAX]; /* * Save a DMI string */ - static void __init dmi_save_ident(struct dmi_header *dm, int slot, int string) { char *d = (char*)dm; char *p = dmi_string(dm, d[string]); - if(p==NULL || *p == 0) + + if (p == NULL || *p == 0) return; if (dmi_ident[slot]) return; - dmi_ident[slot] = alloc_bootmem(strlen(p)+1); + + dmi_ident[slot] = alloc_bootmem(strlen(p) + 1); if(dmi_ident[slot]) strcpy(dmi_ident[slot], p); else @@ -160,281 +150,47 @@ static void __init dmi_save_ident(struct dmi_header *dm, int slot, int string) } /* - * Ugly compatibility crap. - */ -#define dmi_blacklist dmi_system_id -#define NO_MATCH { DMI_NONE, NULL} -#define MATCH DMI_MATCH - -/* - * Toshiba keyboard likes to repeat keys when they are not repeated. - */ - -static __init int broken_toshiba_keyboard(struct dmi_blacklist *d) -{ - printk(KERN_WARNING "Toshiba with broken keyboard detected. If your keyboard sometimes generates 3 keypresses instead of one, see http://davyd.ucc.asn.au/projects/toshiba/README\n"); - return 0; -} - - -#ifdef CONFIG_ACPI_SLEEP -static __init int reset_videomode_after_s3(struct dmi_blacklist *d) -{ - /* See acpi_wakeup.S */ - extern long acpi_video_flags; - acpi_video_flags |= 2; - return 0; -} -#endif - - -#ifdef CONFIG_ACPI_BOOT -extern int acpi_force; - -static __init __attribute__((unused)) int dmi_disable_acpi(struct dmi_blacklist *d) -{ - if (!acpi_force) { - printk(KERN_NOTICE "%s detected: acpi off\n",d->ident); - disable_acpi(); - } else { - printk(KERN_NOTICE - "Warning: DMI blacklist says broken, but acpi forced\n"); - } - return 0; -} - -/* - * Limit ACPI to CPU enumeration for HT - */ -static __init __attribute__((unused)) int force_acpi_ht(struct dmi_blacklist *d) -{ - if (!acpi_force) { - printk(KERN_NOTICE "%s detected: force use of acpi=ht\n", d->ident); - disable_acpi(); - acpi_ht = 1; - } else { - printk(KERN_NOTICE - "Warning: acpi=force overrules DMI blacklist: acpi=ht\n"); - } - return 0; -} -#endif - -#ifdef CONFIG_ACPI_PCI -static __init int disable_acpi_irq(struct dmi_blacklist *d) -{ - if (!acpi_force) { - printk(KERN_NOTICE "%s detected: force use of acpi=noirq\n", - d->ident); - acpi_noirq_set(); - } - return 0; -} -static __init int disable_acpi_pci(struct dmi_blacklist *d) -{ - if (!acpi_force) { - printk(KERN_NOTICE "%s detected: force use of pci=noacpi\n", - d->ident); - acpi_disable_pci(); - } - return 0; -} -#endif - -/* - * Process the DMI blacklists - */ - - -/* - * This will be expanded over time to force things like the APM - * interrupt mask settings according to the laptop - */ - -static __initdata struct dmi_blacklist dmi_blacklist[]={ - - { broken_toshiba_keyboard, "Toshiba Satellite 4030cdt", { /* Keyboard generates spurious repeats */ - MATCH(DMI_PRODUCT_NAME, "S4030CDT/4.3"), - NO_MATCH, NO_MATCH, NO_MATCH - } }, -#ifdef CONFIG_ACPI_SLEEP - { reset_videomode_after_s3, "Toshiba Satellite 4030cdt", { /* Reset video mode after returning from ACPI S3 sleep */ - MATCH(DMI_PRODUCT_NAME, "S4030CDT/4.3"), - NO_MATCH, NO_MATCH, NO_MATCH - } }, -#endif - -#ifdef CONFIG_ACPI_BOOT - /* - * If your system is blacklisted here, but you find that acpi=force - * works for you, please contact acpi-devel@sourceforge.net - */ - - /* - * Boxes that need ACPI disabled - */ - - { dmi_disable_acpi, "IBM Thinkpad", { - MATCH(DMI_BOARD_VENDOR, "IBM"), - MATCH(DMI_BOARD_NAME, "2629H1G"), - NO_MATCH, NO_MATCH }}, - - /* - * Boxes that need acpi=ht - */ - - { force_acpi_ht, "FSC Primergy T850", { - MATCH(DMI_SYS_VENDOR, "FUJITSU SIEMENS"), - MATCH(DMI_PRODUCT_NAME, "PRIMERGY T850"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "DELL GX240", { - MATCH(DMI_BOARD_VENDOR, "Dell Computer Corporation"), - MATCH(DMI_BOARD_NAME, "OptiPlex GX240"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "HP VISUALIZE NT Workstation", { - MATCH(DMI_BOARD_VENDOR, "Hewlett-Packard"), - MATCH(DMI_PRODUCT_NAME, "HP VISUALIZE NT Workstation"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "Compaq Workstation W8000", { - MATCH(DMI_SYS_VENDOR, "Compaq"), - MATCH(DMI_PRODUCT_NAME, "Workstation W8000"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "ASUS P4B266", { - MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), - MATCH(DMI_BOARD_NAME, "P4B266"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "ASUS P2B-DS", { - MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), - MATCH(DMI_BOARD_NAME, "P2B-DS"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "ASUS CUR-DLS", { - MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), - MATCH(DMI_BOARD_NAME, "CUR-DLS"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "ABIT i440BX-W83977", { - MATCH(DMI_BOARD_VENDOR, "ABIT <http://www.abit.com>"), - MATCH(DMI_BOARD_NAME, "i440BX-W83977 (BP6)"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "IBM Bladecenter", { - MATCH(DMI_BOARD_VENDOR, "IBM"), - MATCH(DMI_BOARD_NAME, "IBM eServer BladeCenter HS20"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "IBM eServer xSeries 360", { - MATCH(DMI_BOARD_VENDOR, "IBM"), - MATCH(DMI_BOARD_NAME, "eServer xSeries 360"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "IBM eserver xSeries 330", { - MATCH(DMI_BOARD_VENDOR, "IBM"), - MATCH(DMI_BOARD_NAME, "eserver xSeries 330"), - NO_MATCH, NO_MATCH }}, - - { force_acpi_ht, "IBM eserver xSeries 440", { - MATCH(DMI_BOARD_VENDOR, "IBM"), - MATCH(DMI_PRODUCT_NAME, "eserver xSeries 440"), - NO_MATCH, NO_MATCH }}, - -#endif // CONFIG_ACPI_BOOT - -#ifdef CONFIG_ACPI_PCI - /* - * Boxes that need ACPI PCI IRQ routing disabled - */ - - { disable_acpi_irq, "ASUS A7V", { - MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC"), - MATCH(DMI_BOARD_NAME, "<A7V>"), - /* newer BIOS, Revision 1011, does work */ - MATCH(DMI_BIOS_VERSION, "ASUS A7V ACPI BIOS Revision 1007"), - NO_MATCH }}, - - /* - * Boxes that need ACPI PCI IRQ routing and PCI scan disabled - */ - { disable_acpi_pci, "ASUS PR-DLS", { /* _BBN 0 bug */ - MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), - MATCH(DMI_BOARD_NAME, "PR-DLS"), - MATCH(DMI_BIOS_VERSION, "ASUS PR-DLS ACPI BIOS Revision 1010"), - MATCH(DMI_BIOS_DATE, "03/21/2003") }}, - - { disable_acpi_pci, "Acer TravelMate 36x Laptop", { - MATCH(DMI_SYS_VENDOR, "Acer"), - MATCH(DMI_PRODUCT_NAME, "TravelMate 360"), - NO_MATCH, NO_MATCH - } }, - -#endif - - { NULL, } -}; - -/* * Process a DMI table entry. Right now all we care about are the BIOS * and machine entries. For 2.5 we should pull the smbus controller info * out of here. */ - static void __init dmi_decode(struct dmi_header *dm) { -#ifdef DMI_DEBUG - u8 *data = (u8 *)dm; -#endif + u8 *data __attribute__((__unused__)) = (u8 *)dm; - switch(dm->type) - { - case 0: - dmi_printk(("BIOS Vendor: %s\n", - dmi_string(dm, data[4]))); - dmi_save_ident(dm, DMI_BIOS_VENDOR, 4); - dmi_printk(("BIOS Version: %s\n", - dmi_string(dm, data[5]))); - dmi_save_ident(dm, DMI_BIOS_VERSION, 5); - dmi_printk(("BIOS Release: %s\n", - dmi_string(dm, data[8]))); - dmi_save_ident(dm, DMI_BIOS_DATE, 8); - break; - case 1: - dmi_printk(("System Vendor: %s\n", - dmi_string(dm, data[4]))); - dmi_save_ident(dm, DMI_SYS_VENDOR, 4); - dmi_printk(("Product Name: %s\n", - dmi_string(dm, data[5]))); - dmi_save_ident(dm, DMI_PRODUCT_NAME, 5); - dmi_printk(("Version: %s\n", - dmi_string(dm, data[6]))); - dmi_save_ident(dm, DMI_PRODUCT_VERSION, 6); - dmi_printk(("Serial Number: %s\n", - dmi_string(dm, data[7]))); - break; - case 2: - dmi_printk(("Board Vendor: %s\n", - dmi_string(dm, data[4]))); - dmi_save_ident(dm, DMI_BOARD_VENDOR, 4); - dmi_printk(("Board Name: %s\n", - dmi_string(dm, data[5]))); - dmi_save_ident(dm, DMI_BOARD_NAME, 5); - dmi_printk(("Board Version: %s\n", - dmi_string(dm, data[6]))); - dmi_save_ident(dm, DMI_BOARD_VERSION, 6); - break; + switch(dm->type) { + case 0: + dmi_printk(("BIOS Vendor: %s\n", dmi_string(dm, data[4]))); + dmi_save_ident(dm, DMI_BIOS_VENDOR, 4); + dmi_printk(("BIOS Version: %s\n", dmi_string(dm, data[5]))); + dmi_save_ident(dm, DMI_BIOS_VERSION, 5); + dmi_printk(("BIOS Release: %s\n", dmi_string(dm, data[8]))); + dmi_save_ident(dm, DMI_BIOS_DATE, 8); + break; + case 1: + dmi_printk(("System Vendor: %s\n", dmi_string(dm, data[4]))); + dmi_save_ident(dm, DMI_SYS_VENDOR, 4); + dmi_printk(("Product Name: %s\n", dmi_string(dm, data[5]))); + dmi_save_ident(dm, DMI_PRODUCT_NAME, 5); + dmi_printk(("Version: %s\n", dmi_string(dm, data[6]))); + dmi_save_ident(dm, DMI_PRODUCT_VERSION, 6); + dmi_printk(("Serial Number: %s\n", dmi_string(dm, data[7]))); + dmi_save_ident(dm, DMI_PRODUCT_SERIAL, 7); + break; + case 2: + dmi_printk(("Board Vendor: %s\n", dmi_string(dm, data[4]))); + dmi_save_ident(dm, DMI_BOARD_VENDOR, 4); + dmi_printk(("Board Name: %s\n", dmi_string(dm, data[5]))); + dmi_save_ident(dm, DMI_BOARD_NAME, 5); + dmi_printk(("Board Version: %s\n", dmi_string(dm, data[6]))); + dmi_save_ident(dm, DMI_BOARD_VERSION, 6); + break; } } void __init dmi_scan_machine(void) { - int err = dmi_iterate(dmi_decode); - if(err == 0) - dmi_check_system(dmi_blacklist); - else + if (dmi_iterate(dmi_decode)) printk(KERN_INFO "DMI not present.\n"); } @@ -470,7 +226,6 @@ fail: d++; return count; } - EXPORT_SYMBOL(dmi_check_system); /** @@ -480,8 +235,8 @@ EXPORT_SYMBOL(dmi_check_system); * Returns one DMI data value, can be used to perform * complex DMI data checks. */ -char * dmi_get_system_info(int field) +char *dmi_get_system_info(int field) { return dmi_ident[field]; } - +EXPORT_SYMBOL(dmi_get_system_info); diff --git a/arch/i386/kernel/efi.c b/arch/i386/kernel/efi.c index f732f427b41..385883ea8c1 100644 --- a/arch/i386/kernel/efi.c +++ b/arch/i386/kernel/efi.c @@ -30,6 +30,7 @@ #include <linux/ioport.h> #include <linux/module.h> #include <linux/efi.h> +#include <linux/kexec.h> #include <asm/setup.h> #include <asm/io.h> @@ -598,6 +599,9 @@ efi_initialize_iomem_resources(struct resource *code_resource, if (md->type == EFI_CONVENTIONAL_MEMORY) { request_resource(res, code_resource); request_resource(res, data_resource); +#ifdef CONFIG_KEXEC + request_resource(res, &crashk_res); +#endif } } } diff --git a/arch/i386/kernel/head.S b/arch/i386/kernel/head.S index e966fc8c44c..4477bb10709 100644 --- a/arch/i386/kernel/head.S +++ b/arch/i386/kernel/head.S @@ -299,7 +299,6 @@ is386: movl $2,%ecx # set MP movl %eax,%cr0 call check_x87 - incb ready lgdt cpu_gdt_descr lidt idt_descr ljmp $(__KERNEL_CS),$1f @@ -316,8 +315,9 @@ is386: movl $2,%ecx # set MP lldt %ax cld # gcc2 wants the direction flag cleared at all times #ifdef CONFIG_SMP - movb ready, %cl - cmpb $1,%cl + movb ready, %cl + movb $1, ready + cmpb $0,%cl je 1f # the first CPU calls start_kernel # all other CPUs call initialize_secondary call initialize_secondary diff --git a/arch/i386/kernel/i386_ksyms.c b/arch/i386/kernel/i386_ksyms.c index 903190a4b3f..180f070d03c 100644 --- a/arch/i386/kernel/i386_ksyms.c +++ b/arch/i386/kernel/i386_ksyms.c @@ -1,97 +1,17 @@ #include <linux/config.h> #include <linux/module.h> -#include <linux/smp.h> -#include <linux/user.h> -#include <linux/elfcore.h> -#include <linux/mca.h> -#include <linux/sched.h> -#include <linux/in6.h> -#include <linux/interrupt.h> -#include <linux/smp_lock.h> -#include <linux/pm.h> -#include <linux/pci.h> -#include <linux/apm_bios.h> -#include <linux/kernel.h> -#include <linux/string.h> -#include <linux/tty.h> -#include <linux/highmem.h> -#include <linux/time.h> - -#include <asm/semaphore.h> -#include <asm/processor.h> -#include <asm/i387.h> -#include <asm/uaccess.h> #include <asm/checksum.h> -#include <asm/io.h> -#include <asm/delay.h> -#include <asm/irq.h> -#include <asm/mmx.h> #include <asm/desc.h> -#include <asm/pgtable.h> -#include <asm/tlbflush.h> -#include <asm/nmi.h> -#include <asm/ist.h> -#include <asm/kdebug.h> - -extern void dump_thread(struct pt_regs *, struct user *); -extern spinlock_t rtc_lock; /* This is definitely a GPL-only symbol */ EXPORT_SYMBOL_GPL(cpu_gdt_table); -#if defined(CONFIG_APM_MODULE) -extern void machine_real_restart(unsigned char *, int); -EXPORT_SYMBOL(machine_real_restart); -extern void default_idle(void); -EXPORT_SYMBOL(default_idle); -#endif - -#ifdef CONFIG_SMP -extern void FASTCALL( __write_lock_failed(rwlock_t *rw)); -extern void FASTCALL( __read_lock_failed(rwlock_t *rw)); -#endif - -#if defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_HD) || defined(CONFIG_BLK_DEV_IDE_MODULE) || defined(CONFIG_BLK_DEV_HD_MODULE) -extern struct drive_info_struct drive_info; -EXPORT_SYMBOL(drive_info); -#endif - -extern unsigned long cpu_khz; -extern unsigned long get_cmos_time(void); - -/* platform dependent support */ -EXPORT_SYMBOL(boot_cpu_data); -#ifdef CONFIG_DISCONTIGMEM -EXPORT_SYMBOL(node_data); -EXPORT_SYMBOL(physnode_map); -#endif -#ifdef CONFIG_X86_NUMAQ -EXPORT_SYMBOL(xquad_portio); -#endif -EXPORT_SYMBOL(dump_thread); -EXPORT_SYMBOL(dump_fpu); -EXPORT_SYMBOL_GPL(kernel_fpu_begin); -EXPORT_SYMBOL(__ioremap); -EXPORT_SYMBOL(ioremap_nocache); -EXPORT_SYMBOL(iounmap); -EXPORT_SYMBOL(kernel_thread); -EXPORT_SYMBOL(pm_idle); -EXPORT_SYMBOL(pm_power_off); -EXPORT_SYMBOL(get_cmos_time); -EXPORT_SYMBOL(cpu_khz); -EXPORT_SYMBOL(apm_info); - EXPORT_SYMBOL(__down_failed); EXPORT_SYMBOL(__down_failed_interruptible); EXPORT_SYMBOL(__down_failed_trylock); EXPORT_SYMBOL(__up_wakeup); /* Networking helper routines. */ EXPORT_SYMBOL(csum_partial_copy_generic); -/* Delay loops */ -EXPORT_SYMBOL(__ndelay); -EXPORT_SYMBOL(__udelay); -EXPORT_SYMBOL(__delay); -EXPORT_SYMBOL(__const_udelay); EXPORT_SYMBOL(__get_user_1); EXPORT_SYMBOL(__get_user_2); @@ -105,87 +25,11 @@ EXPORT_SYMBOL(__put_user_8); EXPORT_SYMBOL(strpbrk); EXPORT_SYMBOL(strstr); -EXPORT_SYMBOL(strncpy_from_user); -EXPORT_SYMBOL(__strncpy_from_user); -EXPORT_SYMBOL(clear_user); -EXPORT_SYMBOL(__clear_user); -EXPORT_SYMBOL(__copy_from_user_ll); -EXPORT_SYMBOL(__copy_to_user_ll); -EXPORT_SYMBOL(strnlen_user); - -EXPORT_SYMBOL(dma_alloc_coherent); -EXPORT_SYMBOL(dma_free_coherent); - -#ifdef CONFIG_PCI -EXPORT_SYMBOL(pci_mem_start); -#endif - -#ifdef CONFIG_PCI_BIOS -EXPORT_SYMBOL(pcibios_set_irq_routing); -EXPORT_SYMBOL(pcibios_get_irq_routing_table); -#endif - -#ifdef CONFIG_X86_USE_3DNOW -EXPORT_SYMBOL(_mmx_memcpy); -EXPORT_SYMBOL(mmx_clear_page); -EXPORT_SYMBOL(mmx_copy_page); -#endif - -#ifdef CONFIG_X86_HT -EXPORT_SYMBOL(smp_num_siblings); -EXPORT_SYMBOL(cpu_sibling_map); -#endif - #ifdef CONFIG_SMP -EXPORT_SYMBOL(cpu_data); -EXPORT_SYMBOL(cpu_online_map); -EXPORT_SYMBOL(cpu_callout_map); +extern void FASTCALL( __write_lock_failed(rwlock_t *rw)); +extern void FASTCALL( __read_lock_failed(rwlock_t *rw)); EXPORT_SYMBOL(__write_lock_failed); EXPORT_SYMBOL(__read_lock_failed); - -/* Global SMP stuff */ -EXPORT_SYMBOL(smp_call_function); - -/* TLB flushing */ -EXPORT_SYMBOL(flush_tlb_page); -#endif - -#ifdef CONFIG_X86_IO_APIC -EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); -#endif - -#ifdef CONFIG_MCA -EXPORT_SYMBOL(machine_id); -#endif - -#ifdef CONFIG_VT -EXPORT_SYMBOL(screen_info); -#endif - -EXPORT_SYMBOL(get_wchan); - -EXPORT_SYMBOL(rtc_lock); - -EXPORT_SYMBOL_GPL(set_nmi_callback); -EXPORT_SYMBOL_GPL(unset_nmi_callback); - -EXPORT_SYMBOL(register_die_notifier); -#ifdef CONFIG_HAVE_DEC_LOCK -EXPORT_SYMBOL(_atomic_dec_and_lock); -#endif - -EXPORT_SYMBOL(__PAGE_KERNEL); - -#ifdef CONFIG_HIGHMEM -EXPORT_SYMBOL(kmap); -EXPORT_SYMBOL(kunmap); -EXPORT_SYMBOL(kmap_atomic); -EXPORT_SYMBOL(kunmap_atomic); -EXPORT_SYMBOL(kmap_atomic_to_page); -#endif - -#if defined(CONFIG_X86_SPEEDSTEP_SMI) || defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE) -EXPORT_SYMBOL(ist_info); #endif EXPORT_SYMBOL(csum_partial); diff --git a/arch/i386/kernel/i387.c b/arch/i386/kernel/i387.c index c55e037f08f..b817168d9c6 100644 --- a/arch/i386/kernel/i387.c +++ b/arch/i386/kernel/i387.c @@ -10,6 +10,7 @@ #include <linux/config.h> #include <linux/sched.h> +#include <linux/module.h> #include <asm/processor.h> #include <asm/i387.h> #include <asm/math_emu.h> @@ -79,6 +80,7 @@ void kernel_fpu_begin(void) } clts(); } +EXPORT_SYMBOL_GPL(kernel_fpu_begin); void restore_fpu( struct task_struct *tsk ) { @@ -526,6 +528,7 @@ int dump_fpu( struct pt_regs *regs, struct user_i387_struct *fpu ) return fpvalid; } +EXPORT_SYMBOL(dump_fpu); int dump_task_fpu(struct task_struct *tsk, struct user_i387_struct *fpu) { diff --git a/arch/i386/kernel/i8259.c b/arch/i386/kernel/i8259.c index 2c4813b47e5..178f4e9bac9 100644 --- a/arch/i386/kernel/i8259.c +++ b/arch/i386/kernel/i8259.c @@ -268,10 +268,22 @@ static int i8259A_suspend(struct sys_device *dev, pm_message_t state) return 0; } +static int i8259A_shutdown(struct sys_device *dev) +{ + /* Put the i8259A into a quiescent state that + * the kernel initialization code can get it + * out of. + */ + outb(0xff, 0x21); /* mask all of 8259A-1 */ + outb(0xff, 0xA1); /* mask all of 8259A-1 */ + return 0; +} + static struct sysdev_class i8259_sysdev_class = { set_kset_name("i8259"), .suspend = i8259A_suspend, .resume = i8259A_resume, + .shutdown = i8259A_shutdown, }; static struct sys_device device_i8259A = { diff --git a/arch/i386/kernel/io_apic.c b/arch/i386/kernel/io_apic.c index 7a324e8b86f..35eb8e29c48 100644 --- a/arch/i386/kernel/io_apic.c +++ b/arch/i386/kernel/io_apic.c @@ -31,7 +31,7 @@ #include <linux/mc146818rtc.h> #include <linux/compiler.h> #include <linux/acpi.h> - +#include <linux/module.h> #include <linux/sysdev.h> #include <asm/io.h> #include <asm/smp.h> @@ -573,12 +573,14 @@ static int balanced_irq(void *unused) for ( ; ; ) { set_current_state(TASK_INTERRUPTIBLE); time_remaining = schedule_timeout(time_remaining); - try_to_freeze(PF_FREEZE); + try_to_freeze(); if (time_after(jiffies, prev_balance_time+balanced_irq_interval)) { + preempt_disable(); do_irq_balance(); prev_balance_time = jiffies; time_remaining = balanced_irq_interval; + preempt_enable(); } } return 0; @@ -630,10 +632,8 @@ static int __init balanced_irq_init(void) printk(KERN_ERR "balanced_irq_init: failed to spawn balanced_irq"); failed: for (i = 0; i < NR_CPUS; i++) { - if(irq_cpu_data[i].irq_delta) - kfree(irq_cpu_data[i].irq_delta); - if(irq_cpu_data[i].last_irq) - kfree(irq_cpu_data[i].last_irq); + kfree(irq_cpu_data[i].irq_delta); + kfree(irq_cpu_data[i].last_irq); } return 0; } @@ -812,6 +812,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) } return best_guess; } +EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); /* * This function currently is only a helper for the i386 smp boot process where @@ -1633,12 +1634,43 @@ static void __init enable_IO_APIC(void) */ void disable_IO_APIC(void) { + int pin; /* * Clear the IO-APIC before rebooting: */ clear_IO_APIC(); - disconnect_bsp_APIC(); + /* + * If the i82559 is routed through an IOAPIC + * Put that IOAPIC in virtual wire mode + * so legacy interrups can be delivered. + */ + pin = find_isa_irq_pin(0, mp_ExtINT); + if (pin != -1) { + struct IO_APIC_route_entry entry; + unsigned long flags; + + memset(&entry, 0, sizeof(entry)); + entry.mask = 0; /* Enabled */ + entry.trigger = 0; /* Edge */ + entry.irr = 0; + entry.polarity = 0; /* High */ + entry.delivery_status = 0; + entry.dest_mode = 0; /* Physical */ + entry.delivery_mode = 7; /* ExtInt */ + entry.vector = 0; + entry.dest.physical.physical_dest = 0; + + + /* + * Add it to the IO-APIC irq-routing table: + */ + spin_lock_irqsave(&ioapic_lock, flags); + io_apic_write(0, 0x11+2*pin, *(((int *)&entry)+1)); + io_apic_write(0, 0x10+2*pin, *(((int *)&entry)+0)); + spin_unlock_irqrestore(&ioapic_lock, flags); + } + disconnect_bsp_APIC(pin != -1); } /* @@ -1659,6 +1691,12 @@ static void __init setup_ioapic_ids_from_mpc(void) unsigned long flags; /* + * Don't check I/O APIC IDs for xAPIC systems. They have + * no meaning without the serial APIC bus. + */ + if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && boot_cpu_data.x86 < 15)) + return; + /* * This is broken; anything with a real cpu count has to * circumvent this idiocy regardless. */ @@ -1684,10 +1722,6 @@ static void __init setup_ioapic_ids_from_mpc(void) mp_ioapics[apic].mpc_apicid = reg_00.bits.ID; } - /* Don't check I/O APIC IDs for some xAPIC systems. They have - * no meaning without the serial APIC bus. */ - if (NO_IOAPIC_CHECK) - continue; /* * Sanity check, is the ID really free? Every APIC in a * system must have a unique ID or we get lots of nice diff --git a/arch/i386/kernel/irq.c b/arch/i386/kernel/irq.c index 73945a3c53c..ce66dcc26d9 100644 --- a/arch/i386/kernel/irq.c +++ b/arch/i386/kernel/irq.c @@ -15,6 +15,9 @@ #include <linux/seq_file.h> #include <linux/interrupt.h> #include <linux/kernel_stat.h> +#include <linux/notifier.h> +#include <linux/cpu.h> +#include <linux/delay.h> DEFINE_PER_CPU(irq_cpustat_t, irq_stat) ____cacheline_maxaligned_in_smp; EXPORT_PER_CPU_SYMBOL(irq_stat); @@ -153,6 +156,11 @@ void irq_ctx_init(int cpu) cpu,hardirq_ctx[cpu],softirq_ctx[cpu]); } +void irq_ctx_exit(int cpu) +{ + hardirq_ctx[cpu] = NULL; +} + extern asmlinkage void __do_softirq(void); asmlinkage void do_softirq(void) @@ -210,9 +218,8 @@ int show_interrupts(struct seq_file *p, void *v) if (i == 0) { seq_printf(p, " "); - for (j=0; j<NR_CPUS; j++) - if (cpu_online(j)) - seq_printf(p, "CPU%d ",j); + for_each_cpu(j) + seq_printf(p, "CPU%d ",j); seq_putc(p, '\n'); } @@ -225,9 +232,8 @@ int show_interrupts(struct seq_file *p, void *v) #ifndef CONFIG_SMP seq_printf(p, "%10u ", kstat_irqs(i)); #else - for (j = 0; j < NR_CPUS; j++) - if (cpu_online(j)) - seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]); + for_each_cpu(j) + seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]); #endif seq_printf(p, " %14s", irq_desc[i].handler->typename); seq_printf(p, " %s", action->name); @@ -240,16 +246,14 @@ skip: spin_unlock_irqrestore(&irq_desc[i].lock, flags); } else if (i == NR_IRQS) { seq_printf(p, "NMI: "); - for (j = 0; j < NR_CPUS; j++) - if (cpu_online(j)) - seq_printf(p, "%10u ", nmi_count(j)); + for_each_cpu(j) + seq_printf(p, "%10u ", nmi_count(j)); seq_putc(p, '\n'); #ifdef CONFIG_X86_LOCAL_APIC seq_printf(p, "LOC: "); - for (j = 0; j < NR_CPUS; j++) - if (cpu_online(j)) - seq_printf(p, "%10u ", - per_cpu(irq_stat,j).apic_timer_irqs); + for_each_cpu(j) + seq_printf(p, "%10u ", + per_cpu(irq_stat,j).apic_timer_irqs); seq_putc(p, '\n'); #endif seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count)); @@ -259,3 +263,45 @@ skip: } return 0; } + +#ifdef CONFIG_HOTPLUG_CPU +#include <mach_apic.h> + +void fixup_irqs(cpumask_t map) +{ + unsigned int irq; + static int warned; + + for (irq = 0; irq < NR_IRQS; irq++) { + cpumask_t mask; + if (irq == 2) + continue; + + cpus_and(mask, irq_affinity[irq], map); + if (any_online_cpu(mask) == NR_CPUS) { + printk("Breaking affinity for irq %i\n", irq); + mask = map; + } + if (irq_desc[irq].handler->set_affinity) + irq_desc[irq].handler->set_affinity(irq, mask); + else if (irq_desc[irq].action && !(warned++)) + printk("Cannot set affinity for irq %i\n", irq); + } + +#if 0 + barrier(); + /* Ingo Molnar says: "after the IO-APIC masks have been redirected + [note the nop - the interrupt-enable boundary on x86 is two + instructions from sti] - to flush out pending hardirqs and + IPIs. After this point nothing is supposed to reach this CPU." */ + __asm__ __volatile__("sti; nop; cli"); + barrier(); +#else + /* That doesn't seem sufficient. Give it 1ms. */ + local_irq_enable(); + mdelay(1); + local_irq_disable(); +#endif +} +#endif + diff --git a/arch/i386/kernel/kprobes.c b/arch/i386/kernel/kprobes.c index 59ff9b45506..3762f6b35ab 100644 --- a/arch/i386/kernel/kprobes.c +++ b/arch/i386/kernel/kprobes.c @@ -23,6 +23,9 @@ * Rusty Russell). * 2004-July Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes * interface to access function arguments. + * 2005-May Hien Nguyen <hien@us.ibm.com>, Jim Keniston + * <jkenisto@us.ibm.com> and Prasanna S Panchamukhi + * <prasanna@in.ibm.com> added function-return probes. */ #include <linux/config.h> @@ -30,15 +33,14 @@ #include <linux/ptrace.h> #include <linux/spinlock.h> #include <linux/preempt.h> +#include <asm/cacheflush.h> #include <asm/kdebug.h> #include <asm/desc.h> -/* kprobe_status settings */ -#define KPROBE_HIT_ACTIVE 0x00000001 -#define KPROBE_HIT_SS 0x00000002 - static struct kprobe *current_kprobe; static unsigned long kprobe_status, kprobe_old_eflags, kprobe_saved_eflags; +static struct kprobe *kprobe_prev; +static unsigned long kprobe_status_prev, kprobe_old_eflags_prev, kprobe_saved_eflags_prev; static struct pt_regs jprobe_saved_regs; static long *jprobe_saved_esp; /* copy of the kernel stack at the probe fire time */ @@ -68,16 +70,50 @@ int arch_prepare_kprobe(struct kprobe *p) void arch_copy_kprobe(struct kprobe *p) { memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + p->opcode = *p->addr; } -void arch_remove_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { + *p->addr = BREAKPOINT_INSTRUCTION; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) +void arch_disarm_kprobe(struct kprobe *p) { *p->addr = p->opcode; - regs->eip = (unsigned long)p->addr; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); +} + +void arch_remove_kprobe(struct kprobe *p) +{ +} + +static inline void save_previous_kprobe(void) +{ + kprobe_prev = current_kprobe; + kprobe_status_prev = kprobe_status; + kprobe_old_eflags_prev = kprobe_old_eflags; + kprobe_saved_eflags_prev = kprobe_saved_eflags; +} + +static inline void restore_previous_kprobe(void) +{ + current_kprobe = kprobe_prev; + kprobe_status = kprobe_status_prev; + kprobe_old_eflags = kprobe_old_eflags_prev; + kprobe_saved_eflags = kprobe_saved_eflags_prev; +} + +static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs) +{ + current_kprobe = p; + kprobe_saved_eflags = kprobe_old_eflags + = (regs->eflags & (TF_MASK | IF_MASK)); + if (is_IF_modifier(p->opcode)) + kprobe_saved_eflags &= ~IF_MASK; } static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -91,6 +127,50 @@ static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) regs->eip = (unsigned long)&p->ainsn.insn; } +struct task_struct *arch_get_kprobe_task(void *ptr) +{ + return ((struct thread_info *) (((unsigned long) ptr) & + (~(THREAD_SIZE -1))))->task; +} + +void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs) +{ + unsigned long *sara = (unsigned long *)®s->esp; + struct kretprobe_instance *ri; + static void *orig_ret_addr; + + /* + * Save the return address when the return probe hits + * the first time, and use it to populate the (krprobe + * instance)->ret_addr for subsequent return probes at + * the same addrress since stack address would have + * the kretprobe_trampoline by then. + */ + if (((void*) *sara) != kretprobe_trampoline) + orig_ret_addr = (void*) *sara; + + if ((ri = get_free_rp_inst(rp)) != NULL) { + ri->rp = rp; + ri->stack_addr = sara; + ri->ret_addr = orig_ret_addr; + add_rp_inst(ri); + /* Replace the return addr with trampoline addr */ + *sara = (unsigned long) &kretprobe_trampoline; + } else { + rp->nmissed++; + } +} + +void arch_kprobe_flush_task(struct task_struct *tk) +{ + struct kretprobe_instance *ri; + while ((ri = get_rp_inst_tsk(tk)) != NULL) { + *((unsigned long *)(ri->stack_addr)) = + (unsigned long) ri->ret_addr; + recycle_rp_inst(ri); + } +} + /* * Interrupts are disabled on entry as trap3 is an interrupt gate and they * remain disabled thorough out this function. @@ -127,8 +207,18 @@ static int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); - ret = 1; + /* We have reentered the kprobe_handler(), since + * another probe was hit while within the handler. + * We here save the original kprobes variables and + * just single step on the instruction of the new probe + * without calling any user handlers. + */ + save_previous_kprobe(); + set_current_kprobe(p, regs); + p->nmissed++; + prepare_singlestep(p, regs); + kprobe_status = KPROBE_REENTER; + return 1; } else { p = current_kprobe; if (p->break_handler && p->break_handler(p, regs)) { @@ -163,11 +253,7 @@ static int kprobe_handler(struct pt_regs *regs) } kprobe_status = KPROBE_HIT_ACTIVE; - current_kprobe = p; - kprobe_saved_eflags = kprobe_old_eflags - = (regs->eflags & (TF_MASK | IF_MASK)); - if (is_IF_modifier(p->opcode)) - kprobe_saved_eflags &= ~IF_MASK; + set_current_kprobe(p, regs); if (p->pre_handler && p->pre_handler(p, regs)) /* handler has already set things up, so skip ss setup */ @@ -184,6 +270,55 @@ no_kprobe: } /* + * For function-return probes, init_kprobes() establishes a probepoint + * here. When a retprobed function returns, this probe is hit and + * trampoline_probe_handler() runs, calling the kretprobe's handler. + */ + void kretprobe_trampoline_holder(void) + { + asm volatile ( ".global kretprobe_trampoline\n" + "kretprobe_trampoline: \n" + "nop\n"); + } + +/* + * Called when we hit the probe point at kretprobe_trampoline + */ +int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct task_struct *tsk; + struct kretprobe_instance *ri; + struct hlist_head *head; + struct hlist_node *node; + unsigned long *sara = ((unsigned long *) ®s->esp) - 1; + + tsk = arch_get_kprobe_task(sara); + head = kretprobe_inst_table_head(tsk); + + hlist_for_each_entry(ri, node, head, hlist) { + if (ri->stack_addr == sara && ri->rp) { + if (ri->rp->handler) + ri->rp->handler(ri, regs); + } + } + return 0; +} + +void trampoline_post_handler(struct kprobe *p, struct pt_regs *regs, + unsigned long flags) +{ + struct kretprobe_instance *ri; + /* RA already popped */ + unsigned long *sara = ((unsigned long *)®s->esp) - 1; + + while ((ri = get_rp_inst(sara))) { + regs->eip = (unsigned long)ri->ret_addr; + recycle_rp_inst(ri); + } + regs->eflags &= ~TF_MASK; +} + +/* * Called after single-stepping. p->addr is the address of the * instruction whose first byte has been replaced by the "int 3" * instruction. To avoid the SMP problems that can occur when we @@ -263,13 +398,22 @@ static inline int post_kprobe_handler(struct pt_regs *regs) if (!kprobe_running()) return 0; - if (current_kprobe->post_handler) + if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) { + kprobe_status = KPROBE_HIT_SSDONE; current_kprobe->post_handler(current_kprobe, regs, 0); + } - resume_execution(current_kprobe, regs); + if (current_kprobe->post_handler != trampoline_post_handler) + resume_execution(current_kprobe, regs); regs->eflags |= kprobe_saved_eflags; + /*Restore back the original saved kprobes variables and continue. */ + if (kprobe_status == KPROBE_REENTER) { + restore_previous_kprobe(); + goto out; + } unlock_kprobes(); +out: preempt_enable_no_resched(); /* diff --git a/arch/i386/kernel/machine_kexec.c b/arch/i386/kernel/machine_kexec.c new file mode 100644 index 00000000000..52ed18d8b51 --- /dev/null +++ b/arch/i386/kernel/machine_kexec.c @@ -0,0 +1,226 @@ +/* + * machine_kexec.c - handle transition of Linux booting another kernel + * Copyright (C) 2002-2005 Eric Biederman <ebiederm@xmission.com> + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include <linux/mm.h> +#include <linux/kexec.h> +#include <linux/delay.h> +#include <asm/pgtable.h> +#include <asm/pgalloc.h> +#include <asm/tlbflush.h> +#include <asm/mmu_context.h> +#include <asm/io.h> +#include <asm/apic.h> +#include <asm/cpufeature.h> + +static inline unsigned long read_cr3(void) +{ + unsigned long cr3; + asm volatile("movl %%cr3,%0": "=r"(cr3)); + return cr3; +} + +#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE))) + +#define L0_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define L1_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define L2_ATTR (_PAGE_PRESENT) + +#define LEVEL0_SIZE (1UL << 12UL) + +#ifndef CONFIG_X86_PAE +#define LEVEL1_SIZE (1UL << 22UL) +static u32 pgtable_level1[1024] PAGE_ALIGNED; + +static void identity_map_page(unsigned long address) +{ + unsigned long level1_index, level2_index; + u32 *pgtable_level2; + + /* Find the current page table */ + pgtable_level2 = __va(read_cr3()); + + /* Find the indexes of the physical address to identity map */ + level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE; + level2_index = address / LEVEL1_SIZE; + + /* Identity map the page table entry */ + pgtable_level1[level1_index] = address | L0_ATTR; + pgtable_level2[level2_index] = __pa(pgtable_level1) | L1_ATTR; + + /* Flush the tlb so the new mapping takes effect. + * Global tlb entries are not flushed but that is not an issue. + */ + load_cr3(pgtable_level2); +} + +#else +#define LEVEL1_SIZE (1UL << 21UL) +#define LEVEL2_SIZE (1UL << 30UL) +static u64 pgtable_level1[512] PAGE_ALIGNED; +static u64 pgtable_level2[512] PAGE_ALIGNED; + +static void identity_map_page(unsigned long address) +{ + unsigned long level1_index, level2_index, level3_index; + u64 *pgtable_level3; + + /* Find the current page table */ + pgtable_level3 = __va(read_cr3()); + + /* Find the indexes of the physical address to identity map */ + level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE; + level2_index = (address % LEVEL2_SIZE)/LEVEL1_SIZE; + level3_index = address / LEVEL2_SIZE; + + /* Identity map the page table entry */ + pgtable_level1[level1_index] = address | L0_ATTR; + pgtable_level2[level2_index] = __pa(pgtable_level1) | L1_ATTR; + set_64bit(&pgtable_level3[level3_index], + __pa(pgtable_level2) | L2_ATTR); + + /* Flush the tlb so the new mapping takes effect. + * Global tlb entries are not flushed but that is not an issue. + */ + load_cr3(pgtable_level3); +} +#endif + + +static void set_idt(void *newidt, __u16 limit) +{ + unsigned char curidt[6]; + + /* ia32 supports unaliged loads & stores */ + (*(__u16 *)(curidt)) = limit; + (*(__u32 *)(curidt +2)) = (unsigned long)(newidt); + + __asm__ __volatile__ ( + "lidt %0\n" + : "=m" (curidt) + ); +}; + + +static void set_gdt(void *newgdt, __u16 limit) +{ + unsigned char curgdt[6]; + + /* ia32 supports unaligned loads & stores */ + (*(__u16 *)(curgdt)) = limit; + (*(__u32 *)(curgdt +2)) = (unsigned long)(newgdt); + + __asm__ __volatile__ ( + "lgdt %0\n" + : "=m" (curgdt) + ); +}; + +static void load_segments(void) +{ +#define __STR(X) #X +#define STR(X) __STR(X) + + __asm__ __volatile__ ( + "\tljmp $"STR(__KERNEL_CS)",$1f\n" + "\t1:\n" + "\tmovl $"STR(__KERNEL_DS)",%eax\n" + "\tmovl %eax,%ds\n" + "\tmovl %eax,%es\n" + "\tmovl %eax,%fs\n" + "\tmovl %eax,%gs\n" + "\tmovl %eax,%ss\n" + ); +#undef STR +#undef __STR +} + +typedef asmlinkage NORET_TYPE void (*relocate_new_kernel_t)( + unsigned long indirection_page, + unsigned long reboot_code_buffer, + unsigned long start_address, + unsigned int has_pae) ATTRIB_NORET; + +const extern unsigned char relocate_new_kernel[]; +extern void relocate_new_kernel_end(void); +const extern unsigned int relocate_new_kernel_size; + +/* + * A architecture hook called to validate the + * proposed image and prepare the control pages + * as needed. The pages for KEXEC_CONTROL_CODE_SIZE + * have been allocated, but the segments have yet + * been copied into the kernel. + * + * Do what every setup is needed on image and the + * reboot code buffer to allow us to avoid allocations + * later. + * + * Currently nothing. + */ +int machine_kexec_prepare(struct kimage *image) +{ + return 0; +} + +/* + * Undo anything leftover by machine_kexec_prepare + * when an image is freed. + */ +void machine_kexec_cleanup(struct kimage *image) +{ +} + +/* + * Do not allocate memory (or fail in any way) in machine_kexec(). + * We are past the point of no return, committed to rebooting now. + */ +NORET_TYPE void machine_kexec(struct kimage *image) +{ + unsigned long page_list; + unsigned long reboot_code_buffer; + + relocate_new_kernel_t rnk; + + /* Interrupts aren't acceptable while we reboot */ + local_irq_disable(); + + /* Compute some offsets */ + reboot_code_buffer = page_to_pfn(image->control_code_page) + << PAGE_SHIFT; + page_list = image->head; + + /* Set up an identity mapping for the reboot_code_buffer */ + identity_map_page(reboot_code_buffer); + + /* copy it out */ + memcpy((void *)reboot_code_buffer, relocate_new_kernel, + relocate_new_kernel_size); + + /* The segment registers are funny things, they are + * automatically loaded from a table, in memory wherever you + * set them to a specific selector, but this table is never + * accessed again you set the segment to a different selector. + * + * The more common model is are caches where the behide + * the scenes work is done, but is also dropped at arbitrary + * times. + * + * I take advantage of this here by force loading the + * segments, before I zap the gdt with an invalid value. + */ + load_segments(); + /* The gdt & idt are now invalid. + * If you want to load them you must set up your own idt & gdt. + */ + set_gdt(phys_to_virt(0),0); + set_idt(phys_to_virt(0),0); + + /* now call it */ + rnk = (relocate_new_kernel_t) reboot_code_buffer; + (*rnk)(page_list, reboot_code_buffer, image->start, cpu_has_pae); +} diff --git a/arch/i386/kernel/mpparse.c b/arch/i386/kernel/mpparse.c index 1347ab4939e..af917f609c7 100644 --- a/arch/i386/kernel/mpparse.c +++ b/arch/i386/kernel/mpparse.c @@ -67,7 +67,6 @@ unsigned long mp_lapic_addr; /* Processor that is doing the boot up */ unsigned int boot_cpu_physical_apicid = -1U; -unsigned int boot_cpu_logical_apicid = -1U; /* Internal processor count */ static unsigned int __initdata num_processors; @@ -180,7 +179,6 @@ static void __init MP_processor_info (struct mpc_config_processor *m) if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) { Dprintk(" Bootup CPU\n"); boot_cpu_physical_apicid = m->mpc_apicid; - boot_cpu_logical_apicid = apicid; } if (num_processors >= NR_CPUS) { @@ -914,7 +912,10 @@ void __init mp_register_ioapic ( mp_ioapics[idx].mpc_apicaddr = address; set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); - mp_ioapics[idx].mpc_apicid = io_apic_get_unique_id(idx, id); + if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && (boot_cpu_data.x86 < 15)) + mp_ioapics[idx].mpc_apicid = io_apic_get_unique_id(idx, id); + else + mp_ioapics[idx].mpc_apicid = id; mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx); /* @@ -1055,11 +1056,20 @@ void __init mp_config_acpi_legacy_irqs (void) } } +#define MAX_GSI_NUM 4096 + int mp_register_gsi (u32 gsi, int edge_level, int active_high_low) { int ioapic = -1; int ioapic_pin = 0; int idx, bit = 0; + static int pci_irq = 16; + /* + * Mapping between Global System Interrups, which + * represent all possible interrupts, and IRQs + * assigned to actual devices. + */ + static int gsi_to_irq[MAX_GSI_NUM]; #ifdef CONFIG_ACPI_BUS /* Don't set up the ACPI SCI because it's already set up */ @@ -1094,11 +1104,26 @@ int mp_register_gsi (u32 gsi, int edge_level, int active_high_low) if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) { Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", mp_ioapic_routing[ioapic].apic_id, ioapic_pin); - return gsi; + return gsi_to_irq[gsi]; } mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit); + if (edge_level) { + /* + * For PCI devices assign IRQs in order, avoiding gaps + * due to unused I/O APIC pins. + */ + int irq = gsi; + if (gsi < MAX_GSI_NUM) { + gsi = pci_irq++; + gsi_to_irq[irq] = gsi; + } else { + printk(KERN_ERR "GSI %u is too high\n", gsi); + return gsi; + } + } + io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, edge_level == ACPI_EDGE_SENSITIVE ? 0 : 1, active_high_low == ACPI_ACTIVE_HIGH ? 0 : 1); diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c index 2c0ee9c2d02..da6c46d667c 100644 --- a/arch/i386/kernel/nmi.c +++ b/arch/i386/kernel/nmi.c @@ -28,8 +28,7 @@ #include <linux/sysctl.h> #include <asm/smp.h> -#include <asm/mtrr.h> -#include <asm/mpspec.h> +#include <asm/div64.h> #include <asm/nmi.h> #include "mach_traps.h" @@ -324,6 +323,16 @@ static void clear_msr_range(unsigned int base, unsigned int n) wrmsr(base+i, 0, 0); } +static inline void write_watchdog_counter(const char *descr) +{ + u64 count = (u64)cpu_khz * 1000; + + do_div(count, nmi_hz); + if(descr) + Dprintk("setting %s to -0x%08Lx\n", descr, count); + wrmsrl(nmi_perfctr_msr, 0 - count); +} + static void setup_k7_watchdog(void) { unsigned int evntsel; @@ -339,8 +348,7 @@ static void setup_k7_watchdog(void) | K7_NMI_EVENT; wrmsr(MSR_K7_EVNTSEL0, evntsel, 0); - Dprintk("setting K7_PERFCTR0 to %08lx\n", -(cpu_khz/nmi_hz*1000)); - wrmsr(MSR_K7_PERFCTR0, -(cpu_khz/nmi_hz*1000), -1); + write_watchdog_counter("K7_PERFCTR0"); apic_write(APIC_LVTPC, APIC_DM_NMI); evntsel |= K7_EVNTSEL_ENABLE; wrmsr(MSR_K7_EVNTSEL0, evntsel, 0); @@ -361,8 +369,7 @@ static void setup_p6_watchdog(void) | P6_NMI_EVENT; wrmsr(MSR_P6_EVNTSEL0, evntsel, 0); - Dprintk("setting P6_PERFCTR0 to %08lx\n", -(cpu_khz/nmi_hz*1000)); - wrmsr(MSR_P6_PERFCTR0, -(cpu_khz/nmi_hz*1000), 0); + write_watchdog_counter("P6_PERFCTR0"); apic_write(APIC_LVTPC, APIC_DM_NMI); evntsel |= P6_EVNTSEL0_ENABLE; wrmsr(MSR_P6_EVNTSEL0, evntsel, 0); @@ -402,8 +409,7 @@ static int setup_p4_watchdog(void) wrmsr(MSR_P4_CRU_ESCR0, P4_NMI_CRU_ESCR0, 0); wrmsr(MSR_P4_IQ_CCCR0, P4_NMI_IQ_CCCR0 & ~P4_CCCR_ENABLE, 0); - Dprintk("setting P4_IQ_COUNTER0 to 0x%08lx\n", -(cpu_khz/nmi_hz*1000)); - wrmsr(MSR_P4_IQ_COUNTER0, -(cpu_khz/nmi_hz*1000), -1); + write_watchdog_counter("P4_IQ_COUNTER0"); apic_write(APIC_LVTPC, APIC_DM_NMI); wrmsr(MSR_P4_IQ_CCCR0, nmi_p4_cccr_val, 0); return 1; @@ -518,7 +524,7 @@ void nmi_watchdog_tick (struct pt_regs * regs) * other P6 variant */ apic_write(APIC_LVTPC, APIC_DM_NMI); } - wrmsr(nmi_perfctr_msr, -(cpu_khz/nmi_hz*1000), -1); + write_watchdog_counter(NULL); } } diff --git a/arch/i386/kernel/pci-dma.c b/arch/i386/kernel/pci-dma.c index 4de2e03c7b4..1e51427cc9e 100644 --- a/arch/i386/kernel/pci-dma.c +++ b/arch/i386/kernel/pci-dma.c @@ -11,6 +11,7 @@ #include <linux/mm.h> #include <linux/string.h> #include <linux/pci.h> +#include <linux/module.h> #include <asm/io.h> struct dma_coherent_mem { @@ -54,6 +55,7 @@ void *dma_alloc_coherent(struct device *dev, size_t size, } return ret; } +EXPORT_SYMBOL(dma_alloc_coherent); void dma_free_coherent(struct device *dev, size_t size, void *vaddr, dma_addr_t dma_handle) @@ -68,6 +70,7 @@ void dma_free_coherent(struct device *dev, size_t size, } else free_pages((unsigned long)vaddr, order); } +EXPORT_SYMBOL(dma_free_coherent); int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr, dma_addr_t device_addr, size_t size, int flags) diff --git a/arch/i386/kernel/process.c b/arch/i386/kernel/process.c index 96e3ea6b17c..5f8cfa6b794 100644 --- a/arch/i386/kernel/process.c +++ b/arch/i386/kernel/process.c @@ -13,6 +13,7 @@ #include <stdarg.h> +#include <linux/cpu.h> #include <linux/errno.h> #include <linux/sched.h> #include <linux/fs.h> @@ -37,6 +38,7 @@ #include <linux/kallsyms.h> #include <linux/ptrace.h> #include <linux/random.h> +#include <linux/kprobes.h> #include <asm/uaccess.h> #include <asm/pgtable.h> @@ -54,6 +56,9 @@ #include <linux/irq.h> #include <linux/err.h> +#include <asm/tlbflush.h> +#include <asm/cpu.h> + asmlinkage void ret_from_fork(void) __asm__("ret_from_fork"); static int hlt_counter; @@ -73,6 +78,7 @@ unsigned long thread_saved_pc(struct task_struct *tsk) * Powermanagement idle function, if any.. */ void (*pm_idle)(void); +EXPORT_SYMBOL(pm_idle); static DEFINE_PER_CPU(unsigned int, cpu_idle_state); void disable_hlt(void) @@ -105,6 +111,9 @@ void default_idle(void) cpu_relax(); } } +#ifdef CONFIG_APM_MODULE +EXPORT_SYMBOL(default_idle); +#endif /* * On SMP it's slightly faster (but much more power-consuming!) @@ -138,14 +147,42 @@ static void poll_idle (void) } } +#ifdef CONFIG_HOTPLUG_CPU +#include <asm/nmi.h> +/* We don't actually take CPU down, just spin without interrupts. */ +static inline void play_dead(void) +{ + /* This must be done before dead CPU ack */ + cpu_exit_clear(); + wbinvd(); + mb(); + /* Ack it */ + __get_cpu_var(cpu_state) = CPU_DEAD; + + /* + * With physical CPU hotplug, we should halt the cpu + */ + local_irq_disable(); + while (1) + __asm__ __volatile__("hlt":::"memory"); +} +#else +static inline void play_dead(void) +{ + BUG(); +} +#endif /* CONFIG_HOTPLUG_CPU */ + /* * The idle thread. There's no useful work to be * done, so just try to conserve power and have a * low exit latency (ie sit in a loop waiting for * somebody to say that they'd like to reschedule) */ -void cpu_idle (void) +void cpu_idle(void) { + int cpu = raw_smp_processor_id(); + /* endless idle loop with no priority at all */ while (1) { while (!need_resched()) { @@ -160,6 +197,9 @@ void cpu_idle (void) if (!idle) idle = default_idle; + if (cpu_is_offline(cpu)) + play_dead(); + __get_cpu_var(irq_stat).idle_timestamp = jiffies; idle(); } @@ -218,7 +258,7 @@ static void mwait_idle(void) } } -void __init select_idle_routine(const struct cpuinfo_x86 *c) +void __devinit select_idle_routine(const struct cpuinfo_x86 *c) { if (cpu_has(c, X86_FEATURE_MWAIT)) { printk("monitor/mwait feature present.\n"); @@ -262,7 +302,7 @@ void show_regs(struct pt_regs * regs) printk("EIP: %04x:[<%08lx>] CPU: %d\n",0xffff & regs->xcs,regs->eip, smp_processor_id()); print_symbol("EIP is at %s\n", regs->eip); - if (regs->xcs & 3) + if (user_mode(regs)) printk(" ESP: %04x:%08lx",0xffff & regs->xss,regs->esp); printk(" EFLAGS: %08lx %s (%s)\n", regs->eflags, print_tainted(), system_utsname.release); @@ -325,6 +365,7 @@ int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags) /* Ok, create the new process.. */ return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, ®s, 0, NULL, NULL); } +EXPORT_SYMBOL(kernel_thread); /* * Free current thread data structures etc.. @@ -334,6 +375,13 @@ void exit_thread(void) struct task_struct *tsk = current; struct thread_struct *t = &tsk->thread; + /* + * Remove function-return probe instances associated with this task + * and put them back on the free list. Do not insert an exit probe for + * this function, it will be disabled by kprobe_flush_task if you do. + */ + kprobe_flush_task(tsk); + /* The process may have allocated an io port bitmap... nuke it. */ if (unlikely(NULL != t->io_bitmap_ptr)) { int cpu = get_cpu(); @@ -357,6 +405,13 @@ void flush_thread(void) { struct task_struct *tsk = current; + /* + * Remove function-return probe instances associated with this task + * and put them back on the free list. Do not insert an exit probe for + * this function, it will be disabled by kprobe_flush_task if you do. + */ + kprobe_flush_task(tsk); + memset(tsk->thread.debugreg, 0, sizeof(unsigned long)*8); memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array)); /* @@ -508,6 +563,7 @@ void dump_thread(struct pt_regs * regs, struct user * dump) dump->u_fpvalid = dump_fpu (regs, &dump->i387); } +EXPORT_SYMBOL(dump_thread); /* * Capture the user space registers if the task is not running (in user space) @@ -627,13 +683,13 @@ struct task_struct fastcall * __switch_to(struct task_struct *prev_p, struct tas * Now maybe reload the debug registers */ if (unlikely(next->debugreg[7])) { - loaddebug(next, 0); - loaddebug(next, 1); - loaddebug(next, 2); - loaddebug(next, 3); + set_debugreg(current->thread.debugreg[0], 0); + set_debugreg(current->thread.debugreg[1], 1); + set_debugreg(current->thread.debugreg[2], 2); + set_debugreg(current->thread.debugreg[3], 3); /* no 4 and 5 */ - loaddebug(next, 6); - loaddebug(next, 7); + set_debugreg(current->thread.debugreg[6], 6); + set_debugreg(current->thread.debugreg[7], 7); } if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) @@ -731,6 +787,7 @@ unsigned long get_wchan(struct task_struct *p) } while (count++ < 16); return 0; } +EXPORT_SYMBOL(get_wchan); /* * sys_alloc_thread_area: get a yet unused TLS descriptor index. diff --git a/arch/i386/kernel/ptrace.c b/arch/i386/kernel/ptrace.c index e34f651fa13..0da59b42843 100644 --- a/arch/i386/kernel/ptrace.c +++ b/arch/i386/kernel/ptrace.c @@ -668,7 +668,7 @@ void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code) info.si_code = TRAP_BRKPT; /* User-mode eip? */ - info.si_addr = user_mode(regs) ? (void __user *) regs->eip : NULL; + info.si_addr = user_mode_vm(regs) ? (void __user *) regs->eip : NULL; /* Send us the fakey SIGTRAP */ force_sig_info(SIGTRAP, &info, tsk); diff --git a/arch/i386/kernel/reboot.c b/arch/i386/kernel/reboot.c index 6dc27eb70ee..b3e58484996 100644 --- a/arch/i386/kernel/reboot.c +++ b/arch/i386/kernel/reboot.c @@ -2,6 +2,7 @@ * linux/arch/i386/kernel/reboot.c */ +#include <linux/config.h> #include <linux/mm.h> #include <linux/module.h> #include <linux/delay.h> @@ -19,12 +20,12 @@ * Power off function, if any */ void (*pm_power_off)(void); +EXPORT_SYMBOL(pm_power_off); static int reboot_mode; static int reboot_thru_bios; #ifdef CONFIG_SMP -int reboot_smp = 0; static int reboot_cpu = -1; /* shamelessly grabbed from lib/vsprintf.c for readability */ #define is_digit(c) ((c) >= '0' && (c) <= '9') @@ -47,7 +48,6 @@ static int __init reboot_setup(char *str) break; #ifdef CONFIG_SMP case 's': /* "smp" reboot by executing reset on BSP or other CPU*/ - reboot_smp = 1; if (is_digit(*(str+1))) { reboot_cpu = (int) (*(str+1) - '0'); if (is_digit(*(str+2))) @@ -86,33 +86,9 @@ static int __init set_bios_reboot(struct dmi_system_id *d) return 0; } -/* - * Some machines require the "reboot=s" commandline option, this quirk makes that automatic. - */ -static int __init set_smp_reboot(struct dmi_system_id *d) -{ -#ifdef CONFIG_SMP - if (!reboot_smp) { - reboot_smp = 1; - printk(KERN_INFO "%s series board detected. Selecting SMP-method for reboots.\n", d->ident); - } -#endif - return 0; -} - -/* - * Some machines require the "reboot=b,s" commandline option, this quirk makes that automatic. - */ -static int __init set_smp_bios_reboot(struct dmi_system_id *d) -{ - set_smp_reboot(d); - set_bios_reboot(d); - return 0; -} - static struct dmi_system_id __initdata reboot_dmi_table[] = { { /* Handle problems with rebooting on Dell 1300's */ - .callback = set_smp_bios_reboot, + .callback = set_bios_reboot, .ident = "Dell PowerEdge 1300", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), @@ -295,42 +271,36 @@ void machine_real_restart(unsigned char *code, int length) : : "i" ((void *) (0x1000 - sizeof (real_mode_switch) - 100))); } +#ifdef CONFIG_APM_MODULE +EXPORT_SYMBOL(machine_real_restart); +#endif -void machine_restart(char * __unused) +void machine_shutdown(void) { #ifdef CONFIG_SMP - int cpuid; - - cpuid = GET_APIC_ID(apic_read(APIC_ID)); - - if (reboot_smp) { - - /* check to see if reboot_cpu is valid - if its not, default to the BSP */ - if ((reboot_cpu == -1) || - (reboot_cpu > (NR_CPUS -1)) || - !physid_isset(cpuid, phys_cpu_present_map)) - reboot_cpu = boot_cpu_physical_apicid; - - reboot_smp = 0; /* use this as a flag to only go through this once*/ - /* re-run this function on the other CPUs - it will fall though this section since we have - cleared reboot_smp, and do the reboot if it is the - correct CPU, otherwise it halts. */ - if (reboot_cpu != cpuid) - smp_call_function((void *)machine_restart , NULL, 1, 0); + int reboot_cpu_id; + + /* The boot cpu is always logical cpu 0 */ + reboot_cpu_id = 0; + + /* See if there has been given a command line override */ + if ((reboot_cpu_id != -1) && (reboot_cpu < NR_CPUS) && + cpu_isset(reboot_cpu, cpu_online_map)) { + reboot_cpu_id = reboot_cpu; } - /* if reboot_cpu is still -1, then we want a tradional reboot, - and if we are not running on the reboot_cpu,, halt */ - if ((reboot_cpu != -1) && (cpuid != reboot_cpu)) { - for (;;) - __asm__ __volatile__ ("hlt"); + /* Make certain the cpu I'm rebooting on is online */ + if (!cpu_isset(reboot_cpu_id, cpu_online_map)) { + reboot_cpu_id = smp_processor_id(); } - /* - * Stop all CPUs and turn off local APICs and the IO-APIC, so - * other OSs see a clean IRQ state. + + /* Make certain I only run on the appropriate processor */ + set_cpus_allowed(current, cpumask_of_cpu(reboot_cpu_id)); + + /* O.K. Now that I'm on the appropriate processor, stop + * all of the others, and disable their local APICs. */ + smp_send_stop(); #endif /* CONFIG_SMP */ @@ -339,6 +309,11 @@ void machine_restart(char * __unused) #ifdef CONFIG_X86_IO_APIC disable_IO_APIC(); #endif +} + +void machine_restart(char * __unused) +{ + machine_shutdown(); if (!reboot_thru_bios) { if (efi_enabled) { diff --git a/arch/i386/kernel/relocate_kernel.S b/arch/i386/kernel/relocate_kernel.S new file mode 100644 index 00000000000..d312616effa --- /dev/null +++ b/arch/i386/kernel/relocate_kernel.S @@ -0,0 +1,120 @@ +/* + * relocate_kernel.S - put the kernel image in place to boot + * Copyright (C) 2002-2004 Eric Biederman <ebiederm@xmission.com> + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include <linux/linkage.h> + + /* + * Must be relocatable PIC code callable as a C function, that once + * it starts can not use the previous processes stack. + */ + .globl relocate_new_kernel +relocate_new_kernel: + /* read the arguments and say goodbye to the stack */ + movl 4(%esp), %ebx /* page_list */ + movl 8(%esp), %ebp /* reboot_code_buffer */ + movl 12(%esp), %edx /* start address */ + movl 16(%esp), %ecx /* cpu_has_pae */ + + /* zero out flags, and disable interrupts */ + pushl $0 + popfl + + /* set a new stack at the bottom of our page... */ + lea 4096(%ebp), %esp + + /* store the parameters back on the stack */ + pushl %edx /* store the start address */ + + /* Set cr0 to a known state: + * 31 0 == Paging disabled + * 18 0 == Alignment check disabled + * 16 0 == Write protect disabled + * 3 0 == No task switch + * 2 0 == Don't do FP software emulation. + * 0 1 == Proctected mode enabled + */ + movl %cr0, %eax + andl $~((1<<31)|(1<<18)|(1<<16)|(1<<3)|(1<<2)), %eax + orl $(1<<0), %eax + movl %eax, %cr0 + + /* clear cr4 if applicable */ + testl %ecx, %ecx + jz 1f + /* Set cr4 to a known state: + * Setting everything to zero seems safe. + */ + movl %cr4, %eax + andl $0, %eax + movl %eax, %cr4 + + jmp 1f +1: + + /* Flush the TLB (needed?) */ + xorl %eax, %eax + movl %eax, %cr3 + + /* Do the copies */ + movl %ebx, %ecx + jmp 1f + +0: /* top, read another word from the indirection page */ + movl (%ebx), %ecx + addl $4, %ebx +1: + testl $0x1, %ecx /* is it a destination page */ + jz 2f + movl %ecx, %edi + andl $0xfffff000, %edi + jmp 0b +2: + testl $0x2, %ecx /* is it an indirection page */ + jz 2f + movl %ecx, %ebx + andl $0xfffff000, %ebx + jmp 0b +2: + testl $0x4, %ecx /* is it the done indicator */ + jz 2f + jmp 3f +2: + testl $0x8, %ecx /* is it the source indicator */ + jz 0b /* Ignore it otherwise */ + movl %ecx, %esi /* For every source page do a copy */ + andl $0xfffff000, %esi + + movl $1024, %ecx + rep ; movsl + jmp 0b + +3: + + /* To be certain of avoiding problems with self-modifying code + * I need to execute a serializing instruction here. + * So I flush the TLB, it's handy, and not processor dependent. + */ + xorl %eax, %eax + movl %eax, %cr3 + + /* set all of the registers to known values */ + /* leave %esp alone */ + + xorl %eax, %eax + xorl %ebx, %ebx + xorl %ecx, %ecx + xorl %edx, %edx + xorl %esi, %esi + xorl %edi, %edi + xorl %ebp, %ebp + ret +relocate_new_kernel_end: + + .globl relocate_new_kernel_size +relocate_new_kernel_size: + .long relocate_new_kernel_end - relocate_new_kernel diff --git a/arch/i386/kernel/setup.c b/arch/i386/kernel/setup.c index 2bfbddebdbf..7306353c520 100644 --- a/arch/i386/kernel/setup.c +++ b/arch/i386/kernel/setup.c @@ -23,8 +23,10 @@ * This file handles the architecture-dependent parts of initialization */ +#include <linux/config.h> #include <linux/sched.h> #include <linux/mm.h> +#include <linux/mmzone.h> #include <linux/tty.h> #include <linux/ioport.h> #include <linux/acpi.h> @@ -41,7 +43,12 @@ #include <linux/init.h> #include <linux/edd.h> #include <linux/nodemask.h> +#include <linux/kexec.h> +#include <linux/crash_dump.h> + #include <video/edid.h> + +#include <asm/apic.h> #include <asm/e820.h> #include <asm/mpspec.h> #include <asm/setup.h> @@ -53,12 +60,15 @@ #include "setup_arch_pre.h" #include <bios_ebda.h> +/* Forward Declaration. */ +void __init find_max_pfn(void); + /* This value is set up by the early boot code to point to the value immediately after the boot time page tables. It contains a *physical* address, and must not be in the .bss segment! */ unsigned long init_pg_tables_end __initdata = ~0UL; -int disable_pse __initdata = 0; +int disable_pse __devinitdata = 0; /* * Machine setup.. @@ -73,6 +83,7 @@ EXPORT_SYMBOL(efi_enabled); struct cpuinfo_x86 new_cpu_data __initdata = { 0, 0, 0, 0, -1, 1, 0, 0, -1 }; /* common cpu data for all cpus */ struct cpuinfo_x86 boot_cpu_data = { 0, 0, 0, 0, -1, 1, 0, 0, -1 }; +EXPORT_SYMBOL(boot_cpu_data); unsigned long mmu_cr4_features; @@ -90,12 +101,18 @@ extern acpi_interrupt_flags acpi_sci_flags; /* for MCA, but anyone else can use it if they want */ unsigned int machine_id; +#ifdef CONFIG_MCA +EXPORT_SYMBOL(machine_id); +#endif unsigned int machine_submodel_id; unsigned int BIOS_revision; unsigned int mca_pentium_flag; /* For PCI or other memory-mapped resources */ unsigned long pci_mem_start = 0x10000000; +#ifdef CONFIG_PCI +EXPORT_SYMBOL(pci_mem_start); +#endif /* Boot loader ID as an integer, for the benefit of proc_dointvec */ int bootloader_type; @@ -107,14 +124,26 @@ static unsigned int highmem_pages = -1; * Setup options */ struct drive_info_struct { char dummy[32]; } drive_info; +#if defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_HD) || \ + defined(CONFIG_BLK_DEV_IDE_MODULE) || defined(CONFIG_BLK_DEV_HD_MODULE) +EXPORT_SYMBOL(drive_info); +#endif struct screen_info screen_info; +#ifdef CONFIG_VT +EXPORT_SYMBOL(screen_info); +#endif struct apm_info apm_info; +EXPORT_SYMBOL(apm_info); struct sys_desc_table_struct { unsigned short length; unsigned char table[0]; }; struct edid_info edid_info; struct ist_info ist_info; +#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \ + defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE) +EXPORT_SYMBOL(ist_info); +#endif struct e820map e820; extern void early_cpu_init(void); @@ -711,6 +740,15 @@ static void __init parse_cmdline_early (char ** cmdline_p) if (to != command_line) to--; if (!memcmp(from+7, "exactmap", 8)) { +#ifdef CONFIG_CRASH_DUMP + /* If we are doing a crash dump, we + * still need to know the real mem + * size before original memory map is + * reset. + */ + find_max_pfn(); + saved_max_pfn = max_pfn; +#endif from += 8+7; e820.nr_map = 0; userdef = 1; @@ -814,6 +852,44 @@ static void __init parse_cmdline_early (char ** cmdline_p) #endif /* CONFIG_X86_LOCAL_APIC */ #endif /* CONFIG_ACPI_BOOT */ +#ifdef CONFIG_X86_LOCAL_APIC + /* enable local APIC */ + else if (!memcmp(from, "lapic", 5)) + lapic_enable(); + + /* disable local APIC */ + else if (!memcmp(from, "nolapic", 6)) + lapic_disable(); +#endif /* CONFIG_X86_LOCAL_APIC */ + +#ifdef CONFIG_KEXEC + /* crashkernel=size@addr specifies the location to reserve for + * a crash kernel. By reserving this memory we guarantee + * that linux never set's it up as a DMA target. + * Useful for holding code to do something appropriate + * after a kernel panic. + */ + else if (!memcmp(from, "crashkernel=", 12)) { + unsigned long size, base; + size = memparse(from+12, &from); + if (*from == '@') { + base = memparse(from+1, &from); + /* FIXME: Do I want a sanity check + * to validate the memory range? + */ + crashk_res.start = base; + crashk_res.end = base + size - 1; + } + } +#endif +#ifdef CONFIG_CRASH_DUMP + /* elfcorehdr= specifies the location of elf core header + * stored by the crashed kernel. + */ + else if (!memcmp(from, "elfcorehdr=", 11)) + elfcorehdr_addr = memparse(from+11, &from); +#endif + /* * highmem=size forces highmem to be exactly 'size' bytes. * This works even on boxes that have no highmem otherwise. @@ -1022,7 +1098,7 @@ static void __init reserve_ebda_region(void) reserve_bootmem(addr, PAGE_SIZE); } -#ifndef CONFIG_DISCONTIGMEM +#ifndef CONFIG_NEED_MULTIPLE_NODES void __init setup_bootmem_allocator(void); static unsigned long __init setup_memory(void) { @@ -1072,9 +1148,9 @@ void __init zone_sizes_init(void) free_area_init(zones_size); } #else -extern unsigned long setup_memory(void); +extern unsigned long __init setup_memory(void); extern void zone_sizes_init(void); -#endif /* !CONFIG_DISCONTIGMEM */ +#endif /* !CONFIG_NEED_MULTIPLE_NODES */ void __init setup_bootmem_allocator(void) { @@ -1092,8 +1168,8 @@ void __init setup_bootmem_allocator(void) * the (very unlikely) case of us accidentally initializing the * bootmem allocator with an invalid RAM area. */ - reserve_bootmem(HIGH_MEMORY, (PFN_PHYS(min_low_pfn) + - bootmap_size + PAGE_SIZE-1) - (HIGH_MEMORY)); + reserve_bootmem(__PHYSICAL_START, (PFN_PHYS(min_low_pfn) + + bootmap_size + PAGE_SIZE-1) - (__PHYSICAL_START)); /* * reserve physical page 0 - it's a special BIOS page on many boxes, @@ -1149,6 +1225,11 @@ void __init setup_bootmem_allocator(void) } } #endif +#ifdef CONFIG_KEXEC + if (crashk_res.start != crashk_res.end) + reserve_bootmem(crashk_res.start, + crashk_res.end - crashk_res.start + 1); +#endif } /* @@ -1202,6 +1283,9 @@ legacy_init_iomem_resources(struct resource *code_resource, struct resource *dat */ request_resource(res, code_resource); request_resource(res, data_resource); +#ifdef CONFIG_KEXEC + request_resource(res, &crashk_res); +#endif } } } @@ -1475,6 +1559,7 @@ void __init setup_arch(char **cmdline_p) #endif paging_init(); remapped_pgdat_init(); + sparse_init(); zone_sizes_init(); /* diff --git a/arch/i386/kernel/signal.c b/arch/i386/kernel/signal.c index ea46d028af0..89ef7adc63a 100644 --- a/arch/i386/kernel/signal.c +++ b/arch/i386/kernel/signal.c @@ -346,8 +346,8 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) extern void __user __kernel_sigreturn; extern void __user __kernel_rt_sigreturn; -static void setup_frame(int sig, struct k_sigaction *ka, - sigset_t *set, struct pt_regs * regs) +static int setup_frame(int sig, struct k_sigaction *ka, + sigset_t *set, struct pt_regs * regs) { void __user *restorer; struct sigframe __user *frame; @@ -429,13 +429,14 @@ static void setup_frame(int sig, struct k_sigaction *ka, current->comm, current->pid, frame, regs->eip, frame->pretcode); #endif - return; + return 1; give_sigsegv: force_sigsegv(sig, current); + return 0; } -static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, +static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set, struct pt_regs * regs) { void __user *restorer; @@ -522,20 +523,23 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, current->comm, current->pid, frame, regs->eip, frame->pretcode); #endif - return; + return 1; give_sigsegv: force_sigsegv(sig, current); + return 0; } /* * OK, we're invoking a handler */ -static void +static int handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, sigset_t *oldset, struct pt_regs * regs) { + int ret; + /* Are we from a system call? */ if (regs->orig_eax >= 0) { /* If so, check system call restarting.. */ @@ -569,17 +573,19 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, /* Set up the stack frame */ if (ka->sa.sa_flags & SA_SIGINFO) - setup_rt_frame(sig, ka, info, oldset, regs); + ret = setup_rt_frame(sig, ka, info, oldset, regs); else - setup_frame(sig, ka, oldset, regs); + ret = setup_frame(sig, ka, oldset, regs); - if (!(ka->sa.sa_flags & SA_NODEFER)) { + if (ret && !(ka->sa.sa_flags & SA_NODEFER)) { spin_lock_irq(¤t->sighand->siglock); sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask); sigaddset(¤t->blocked,sig); recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); } + + return ret; } /* @@ -599,13 +605,11 @@ int fastcall do_signal(struct pt_regs *regs, sigset_t *oldset) * kernel mode. Just return without doing anything * if so. */ - if ((regs->xcs & 3) != 3) + if (!user_mode(regs)) return 1; - if (current->flags & PF_FREEZE) { - refrigerator(0); + if (try_to_freeze()) goto no_signal; - } if (!oldset) oldset = ¤t->blocked; @@ -618,12 +622,11 @@ int fastcall do_signal(struct pt_regs *regs, sigset_t *oldset) * inside the kernel. */ if (unlikely(current->thread.debugreg[7])) { - loaddebug(¤t->thread, 7); + set_debugreg(current->thread.debugreg[7], 7); } /* Whee! Actually deliver the signal. */ - handle_signal(signr, &info, &ka, oldset, regs); - return 1; + return handle_signal(signr, &info, &ka, oldset, regs); } no_signal: diff --git a/arch/i386/kernel/smp.c b/arch/i386/kernel/smp.c index 6223c33ac91..cec4bde6716 100644 --- a/arch/i386/kernel/smp.c +++ b/arch/i386/kernel/smp.c @@ -19,6 +19,8 @@ #include <linux/mc146818rtc.h> #include <linux/cache.h> #include <linux/interrupt.h> +#include <linux/cpu.h> +#include <linux/module.h> #include <asm/mtrr.h> #include <asm/tlbflush.h> @@ -163,7 +165,7 @@ void send_IPI_mask_bitmask(cpumask_t cpumask, int vector) unsigned long flags; local_irq_save(flags); - + WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]); /* * Wait for idle. */ @@ -345,21 +347,21 @@ out: static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm, unsigned long va) { - cpumask_t tmp; /* * A couple of (to be removed) sanity checks: * - * - we do not send IPIs to not-yet booted CPUs. * - current CPU must not be in mask * - mask must exist :) */ BUG_ON(cpus_empty(cpumask)); - - cpus_and(tmp, cpumask, cpu_online_map); - BUG_ON(!cpus_equal(cpumask, tmp)); BUG_ON(cpu_isset(smp_processor_id(), cpumask)); BUG_ON(!mm); + /* If a CPU which we ran on has gone down, OK. */ + cpus_and(cpumask, cpumask, cpu_online_map); + if (cpus_empty(cpumask)) + return; + /* * i'm not happy about this global shared spinlock in the * MM hot path, but we'll see how contended it is. @@ -452,6 +454,7 @@ void flush_tlb_page(struct vm_area_struct * vma, unsigned long va) preempt_enable(); } +EXPORT_SYMBOL(flush_tlb_page); static void do_flush_tlb_all(void* info) { @@ -474,6 +477,7 @@ void flush_tlb_all(void) */ void smp_send_reschedule(int cpu) { + WARN_ON(cpu_is_offline(cpu)); send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR); } @@ -491,6 +495,16 @@ struct call_data_struct { int wait; }; +void lock_ipi_call_lock(void) +{ + spin_lock_irq(&call_lock); +} + +void unlock_ipi_call_lock(void) +{ + spin_unlock_irq(&call_lock); +} + static struct call_data_struct * call_data; /* @@ -514,10 +528,15 @@ int smp_call_function (void (*func) (void *info), void *info, int nonatomic, */ { struct call_data_struct data; - int cpus = num_online_cpus()-1; + int cpus; - if (!cpus) + /* Holding any lock stops cpus from going down. */ + spin_lock(&call_lock); + cpus = num_online_cpus() - 1; + if (!cpus) { + spin_unlock(&call_lock); return 0; + } /* Can deadlock when called with interrupts disabled */ WARN_ON(irqs_disabled()); @@ -529,7 +548,6 @@ int smp_call_function (void (*func) (void *info), void *info, int nonatomic, if (wait) atomic_set(&data.finished, 0); - spin_lock(&call_lock); call_data = &data; mb(); @@ -547,6 +565,7 @@ int smp_call_function (void (*func) (void *info), void *info, int nonatomic, return 0; } +EXPORT_SYMBOL(smp_call_function); static void stop_this_cpu (void * dummy) { diff --git a/arch/i386/kernel/smpboot.c b/arch/i386/kernel/smpboot.c index bc1bb6919e6..d66bf489a2e 100644 --- a/arch/i386/kernel/smpboot.c +++ b/arch/i386/kernel/smpboot.c @@ -44,6 +44,9 @@ #include <linux/smp_lock.h> #include <linux/irq.h> #include <linux/bootmem.h> +#include <linux/notifier.h> +#include <linux/cpu.h> +#include <linux/percpu.h> #include <linux/delay.h> #include <linux/mc146818rtc.h> @@ -56,24 +59,46 @@ #include <smpboot_hooks.h> /* Set if we find a B stepping CPU */ -static int __initdata smp_b_stepping; +static int __devinitdata smp_b_stepping; /* Number of siblings per CPU package */ int smp_num_siblings = 1; -int phys_proc_id[NR_CPUS]; /* Package ID of each logical CPU */ +#ifdef CONFIG_X86_HT +EXPORT_SYMBOL(smp_num_siblings); +#endif + +/* Package ID of each logical CPU */ +int phys_proc_id[NR_CPUS] = {[0 ... NR_CPUS-1] = BAD_APICID}; EXPORT_SYMBOL(phys_proc_id); -int cpu_core_id[NR_CPUS]; /* Core ID of each logical CPU */ + +/* Core ID of each logical CPU */ +int cpu_core_id[NR_CPUS] = {[0 ... NR_CPUS-1] = BAD_APICID}; EXPORT_SYMBOL(cpu_core_id); +cpumask_t cpu_sibling_map[NR_CPUS]; +EXPORT_SYMBOL(cpu_sibling_map); + +cpumask_t cpu_core_map[NR_CPUS]; +EXPORT_SYMBOL(cpu_core_map); + /* bitmap of online cpus */ cpumask_t cpu_online_map; +EXPORT_SYMBOL(cpu_online_map); cpumask_t cpu_callin_map; cpumask_t cpu_callout_map; +EXPORT_SYMBOL(cpu_callout_map); static cpumask_t smp_commenced_mask; +/* TSC's upper 32 bits can't be written in eariler CPU (before prescott), there + * is no way to resync one AP against BP. TBD: for prescott and above, we + * should use IA64's algorithm + */ +static int __devinitdata tsc_sync_disabled; + /* Per CPU bogomips and other parameters */ struct cpuinfo_x86 cpu_data[NR_CPUS] __cacheline_aligned; +EXPORT_SYMBOL(cpu_data); u8 x86_cpu_to_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = 0xff }; @@ -90,13 +115,16 @@ static int trampoline_exec; static void map_cpu_to_logical_apicid(void); +/* State of each CPU. */ +DEFINE_PER_CPU(int, cpu_state) = { 0 }; + /* * Currently trivial. Write the real->protected mode * bootstrap into the page concerned. The caller * has made sure it's suitably aligned. */ -static unsigned long __init setup_trampoline(void) +static unsigned long __devinit setup_trampoline(void) { memcpy(trampoline_base, trampoline_data, trampoline_end - trampoline_data); return virt_to_phys(trampoline_base); @@ -126,7 +154,7 @@ void __init smp_alloc_memory(void) * a given CPU */ -static void __init smp_store_cpu_info(int id) +static void __devinit smp_store_cpu_info(int id) { struct cpuinfo_x86 *c = cpu_data + id; @@ -199,7 +227,7 @@ static void __init synchronize_tsc_bp (void) unsigned long long t0; unsigned long long sum, avg; long long delta; - unsigned long one_usec; + unsigned int one_usec; int buggy = 0; printk(KERN_INFO "checking TSC synchronization across %u CPUs: ", num_booting_cpus()); @@ -320,7 +348,7 @@ extern void calibrate_delay(void); static atomic_t init_deasserted; -static void __init smp_callin(void) +static void __devinit smp_callin(void) { int cpuid, phys_id; unsigned long timeout; @@ -405,16 +433,48 @@ static void __init smp_callin(void) /* * Synchronize the TSC with the BP */ - if (cpu_has_tsc && cpu_khz) + if (cpu_has_tsc && cpu_khz && !tsc_sync_disabled) synchronize_tsc_ap(); } static int cpucount; +static inline void +set_cpu_sibling_map(int cpu) +{ + int i; + + if (smp_num_siblings > 1) { + for (i = 0; i < NR_CPUS; i++) { + if (!cpu_isset(i, cpu_callout_map)) + continue; + if (cpu_core_id[cpu] == cpu_core_id[i]) { + cpu_set(i, cpu_sibling_map[cpu]); + cpu_set(cpu, cpu_sibling_map[i]); + } + } + } else { + cpu_set(cpu, cpu_sibling_map[cpu]); + } + + if (current_cpu_data.x86_num_cores > 1) { + for (i = 0; i < NR_CPUS; i++) { + if (!cpu_isset(i, cpu_callout_map)) + continue; + if (phys_proc_id[cpu] == phys_proc_id[i]) { + cpu_set(i, cpu_core_map[cpu]); + cpu_set(cpu, cpu_core_map[i]); + } + } + } else { + cpu_core_map[cpu] = cpu_sibling_map[cpu]; + } +} + /* * Activate a secondary processor. */ -static void __init start_secondary(void *unused) +static void __devinit start_secondary(void *unused) { /* * Dont put anything before smp_callin(), SMP @@ -437,7 +497,23 @@ static void __init start_secondary(void *unused) * the local TLBs too. */ local_flush_tlb(); + + /* This must be done before setting cpu_online_map */ + set_cpu_sibling_map(raw_smp_processor_id()); + wmb(); + + /* + * We need to hold call_lock, so there is no inconsistency + * between the time smp_call_function() determines number of + * IPI receipients, and the time when the determination is made + * for which cpus receive the IPI. Holding this + * lock helps us to not include this cpu in a currently in progress + * smp_call_function(). + */ + lock_ipi_call_lock(); cpu_set(smp_processor_id(), cpu_online_map); + unlock_ipi_call_lock(); + per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; /* We can take interrupts now: we're officially "up". */ local_irq_enable(); @@ -452,7 +528,7 @@ static void __init start_secondary(void *unused) * from the task structure * This function must not return. */ -void __init initialize_secondary(void) +void __devinit initialize_secondary(void) { /* * We don't actually need to load the full TSS, @@ -566,7 +642,7 @@ static inline void __inquire_remote_apic(int apicid) * INIT, INIT, STARTUP sequence will reset the chip hard for us, and this * won't ... remember to clear down the APIC, etc later. */ -static int __init +static int __devinit wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip) { unsigned long send_status = 0, accept_status = 0; @@ -612,7 +688,7 @@ wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip) #endif /* WAKE_SECONDARY_VIA_NMI */ #ifdef WAKE_SECONDARY_VIA_INIT -static int __init +static int __devinit wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) { unsigned long send_status = 0, accept_status = 0; @@ -747,8 +823,43 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) #endif /* WAKE_SECONDARY_VIA_INIT */ extern cpumask_t cpu_initialized; +static inline int alloc_cpu_id(void) +{ + cpumask_t tmp_map; + int cpu; + cpus_complement(tmp_map, cpu_present_map); + cpu = first_cpu(tmp_map); + if (cpu >= NR_CPUS) + return -ENODEV; + return cpu; +} + +#ifdef CONFIG_HOTPLUG_CPU +static struct task_struct * __devinitdata cpu_idle_tasks[NR_CPUS]; +static inline struct task_struct * alloc_idle_task(int cpu) +{ + struct task_struct *idle; + + if ((idle = cpu_idle_tasks[cpu]) != NULL) { + /* initialize thread_struct. we really want to avoid destroy + * idle tread + */ + idle->thread.esp = (unsigned long)(((struct pt_regs *) + (THREAD_SIZE + (unsigned long) idle->thread_info)) - 1); + init_idle(idle, cpu); + return idle; + } + idle = fork_idle(cpu); -static int __init do_boot_cpu(int apicid) + if (!IS_ERR(idle)) + cpu_idle_tasks[cpu] = idle; + return idle; +} +#else +#define alloc_idle_task(cpu) fork_idle(cpu) +#endif + +static int __devinit do_boot_cpu(int apicid, int cpu) /* * NOTE - on most systems this is a PHYSICAL apic ID, but on multiquad * (ie clustered apic addressing mode), this is a LOGICAL apic ID. @@ -757,16 +868,17 @@ static int __init do_boot_cpu(int apicid) { struct task_struct *idle; unsigned long boot_error; - int timeout, cpu; + int timeout; unsigned long start_eip; unsigned short nmi_high = 0, nmi_low = 0; - cpu = ++cpucount; + ++cpucount; + /* * We can't use kernel_thread since we must avoid to * reschedule the child. */ - idle = fork_idle(cpu); + idle = alloc_idle_task(cpu); if (IS_ERR(idle)) panic("failed fork for CPU %d", cpu); idle->thread.eip = (unsigned long) start_secondary; @@ -833,13 +945,16 @@ static int __init do_boot_cpu(int apicid) inquire_remote_apic(apicid); } } - x86_cpu_to_apicid[cpu] = apicid; + if (boot_error) { /* Try to put things back the way they were before ... */ unmap_cpu_to_logical_apicid(cpu); cpu_clear(cpu, cpu_callout_map); /* was set here (do_boot_cpu()) */ cpu_clear(cpu, cpu_initialized); /* was set by cpu_init() */ cpucount--; + } else { + x86_cpu_to_apicid[cpu] = apicid; + cpu_set(cpu, cpu_present_map); } /* mark "stuck" area as not stuck */ @@ -848,6 +963,75 @@ static int __init do_boot_cpu(int apicid) return boot_error; } +#ifdef CONFIG_HOTPLUG_CPU +void cpu_exit_clear(void) +{ + int cpu = raw_smp_processor_id(); + + idle_task_exit(); + + cpucount --; + cpu_uninit(); + irq_ctx_exit(cpu); + + cpu_clear(cpu, cpu_callout_map); + cpu_clear(cpu, cpu_callin_map); + cpu_clear(cpu, cpu_present_map); + + cpu_clear(cpu, smp_commenced_mask); + unmap_cpu_to_logical_apicid(cpu); +} + +struct warm_boot_cpu_info { + struct completion *complete; + int apicid; + int cpu; +}; + +static void __devinit do_warm_boot_cpu(void *p) +{ + struct warm_boot_cpu_info *info = p; + do_boot_cpu(info->apicid, info->cpu); + complete(info->complete); +} + +int __devinit smp_prepare_cpu(int cpu) +{ + DECLARE_COMPLETION(done); + struct warm_boot_cpu_info info; + struct work_struct task; + int apicid, ret; + + lock_cpu_hotplug(); + apicid = x86_cpu_to_apicid[cpu]; + if (apicid == BAD_APICID) { + ret = -ENODEV; + goto exit; + } + + info.complete = &done; + info.apicid = apicid; + info.cpu = cpu; + INIT_WORK(&task, do_warm_boot_cpu, &info); + + tsc_sync_disabled = 1; + + /* init low mem mapping */ + memcpy(swapper_pg_dir, swapper_pg_dir + USER_PGD_PTRS, + sizeof(swapper_pg_dir[0]) * KERNEL_PGD_PTRS); + flush_tlb_all(); + schedule_work(&task); + wait_for_completion(&done); + + tsc_sync_disabled = 0; + zap_low_mappings(); + ret = 0; +exit: + unlock_cpu_hotplug(); + return ret; +} +#endif + static void smp_tune_scheduling (void) { unsigned long cachesize; /* kB */ @@ -885,10 +1069,9 @@ static void smp_tune_scheduling (void) static int boot_cpu_logical_apicid; /* Where the IO area was mapped on multiquad, always 0 otherwise */ void *xquad_portio; - -cpumask_t cpu_sibling_map[NR_CPUS] __cacheline_aligned; -cpumask_t cpu_core_map[NR_CPUS] __cacheline_aligned; -EXPORT_SYMBOL(cpu_core_map); +#ifdef CONFIG_X86_NUMAQ +EXPORT_SYMBOL(xquad_portio); +#endif static void __init smp_boot_cpus(unsigned int max_cpus) { @@ -1001,7 +1184,7 @@ static void __init smp_boot_cpus(unsigned int max_cpus) if (max_cpus <= cpucount+1) continue; - if (do_boot_cpu(apicid)) + if (((cpu = alloc_cpu_id()) <= 0) || do_boot_cpu(apicid, cpu)) printk("CPU #%d not responding - cannot use it.\n", apicid); else @@ -1053,44 +1236,8 @@ static void __init smp_boot_cpus(unsigned int max_cpus) cpus_clear(cpu_core_map[cpu]); } - for (cpu = 0; cpu < NR_CPUS; cpu++) { - struct cpuinfo_x86 *c = cpu_data + cpu; - int siblings = 0; - int i; - if (!cpu_isset(cpu, cpu_callout_map)) - continue; - - if (smp_num_siblings > 1) { - for (i = 0; i < NR_CPUS; i++) { - if (!cpu_isset(i, cpu_callout_map)) - continue; - if (cpu_core_id[cpu] == cpu_core_id[i]) { - siblings++; - cpu_set(i, cpu_sibling_map[cpu]); - } - } - } else { - siblings++; - cpu_set(cpu, cpu_sibling_map[cpu]); - } - - if (siblings != smp_num_siblings) { - printk(KERN_WARNING "WARNING: %d siblings found for CPU%d, should be %d\n", siblings, cpu, smp_num_siblings); - smp_num_siblings = siblings; - } - - if (c->x86_num_cores > 1) { - for (i = 0; i < NR_CPUS; i++) { - if (!cpu_isset(i, cpu_callout_map)) - continue; - if (phys_proc_id[cpu] == phys_proc_id[i]) { - cpu_set(i, cpu_core_map[cpu]); - } - } - } else { - cpu_core_map[cpu] = cpu_sibling_map[cpu]; - } - } + cpu_set(0, cpu_sibling_map[0]); + cpu_set(0, cpu_core_map[0]); smpboot_setup_io_apic(); @@ -1107,6 +1254,9 @@ static void __init smp_boot_cpus(unsigned int max_cpus) who understands all this stuff should rewrite it properly. --RR 15/Jul/02 */ void __init smp_prepare_cpus(unsigned int max_cpus) { + smp_commenced_mask = cpumask_of_cpu(0); + cpu_callin_map = cpumask_of_cpu(0); + mb(); smp_boot_cpus(max_cpus); } @@ -1114,23 +1264,98 @@ void __devinit smp_prepare_boot_cpu(void) { cpu_set(smp_processor_id(), cpu_online_map); cpu_set(smp_processor_id(), cpu_callout_map); + cpu_set(smp_processor_id(), cpu_present_map); + per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; } -int __devinit __cpu_up(unsigned int cpu) +#ifdef CONFIG_HOTPLUG_CPU +static void +remove_siblinginfo(int cpu) { - /* This only works at boot for x86. See "rewrite" above. */ - if (cpu_isset(cpu, smp_commenced_mask)) { - local_irq_enable(); - return -ENOSYS; + int sibling; + + for_each_cpu_mask(sibling, cpu_sibling_map[cpu]) + cpu_clear(cpu, cpu_sibling_map[sibling]); + for_each_cpu_mask(sibling, cpu_core_map[cpu]) + cpu_clear(cpu, cpu_core_map[sibling]); + cpus_clear(cpu_sibling_map[cpu]); + cpus_clear(cpu_core_map[cpu]); + phys_proc_id[cpu] = BAD_APICID; + cpu_core_id[cpu] = BAD_APICID; +} + +int __cpu_disable(void) +{ + cpumask_t map = cpu_online_map; + int cpu = smp_processor_id(); + + /* + * Perhaps use cpufreq to drop frequency, but that could go + * into generic code. + * + * We won't take down the boot processor on i386 due to some + * interrupts only being able to be serviced by the BSP. + * Especially so if we're not using an IOAPIC -zwane + */ + if (cpu == 0) + return -EBUSY; + + /* We enable the timer again on the exit path of the death loop */ + disable_APIC_timer(); + /* Allow any queued timer interrupts to get serviced */ + local_irq_enable(); + mdelay(1); + local_irq_disable(); + + remove_siblinginfo(cpu); + + cpu_clear(cpu, map); + fixup_irqs(map); + /* It's now safe to remove this processor from the online map */ + cpu_clear(cpu, cpu_online_map); + return 0; +} + +void __cpu_die(unsigned int cpu) +{ + /* We don't do anything here: idle task is faking death itself. */ + unsigned int i; + + for (i = 0; i < 10; i++) { + /* They ack this in play_dead by setting CPU_DEAD */ + if (per_cpu(cpu_state, cpu) == CPU_DEAD) { + printk ("CPU %d is now offline\n", cpu); + return; + } + current->state = TASK_UNINTERRUPTIBLE; + schedule_timeout(HZ/10); } + printk(KERN_ERR "CPU %u didn't die...\n", cpu); +} +#else /* ... !CONFIG_HOTPLUG_CPU */ +int __cpu_disable(void) +{ + return -ENOSYS; +} +void __cpu_die(unsigned int cpu) +{ + /* We said "no" in __cpu_disable */ + BUG(); +} +#endif /* CONFIG_HOTPLUG_CPU */ + +int __devinit __cpu_up(unsigned int cpu) +{ /* In case one didn't come up */ if (!cpu_isset(cpu, cpu_callin_map)) { + printk(KERN_DEBUG "skipping cpu%d, didn't come online\n", cpu); local_irq_enable(); return -EIO; } local_irq_enable(); + per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; /* Unleash the CPU! */ cpu_set(cpu, smp_commenced_mask); while (!cpu_isset(cpu, cpu_online_map)) @@ -1144,10 +1369,12 @@ void __init smp_cpus_done(unsigned int max_cpus) setup_ioapic_dest(); #endif zap_low_mappings(); +#ifndef CONFIG_HOTPLUG_CPU /* * Disable executability of the SMP trampoline: */ set_kernel_exec((unsigned long)trampoline_base, trampoline_exec); +#endif } void __init smp_intr_init(void) diff --git a/arch/i386/kernel/syscall_table.S b/arch/i386/kernel/syscall_table.S index d408afaf649..442a6e937b1 100644 --- a/arch/i386/kernel/syscall_table.S +++ b/arch/i386/kernel/syscall_table.S @@ -283,7 +283,7 @@ ENTRY(sys_call_table) .long sys_mq_timedreceive /* 280 */ .long sys_mq_notify .long sys_mq_getsetattr - .long sys_ni_syscall /* reserved for kexec */ + .long sys_kexec_load .long sys_waitid .long sys_ni_syscall /* 285 */ /* available */ .long sys_add_key diff --git a/arch/i386/kernel/sysenter.c b/arch/i386/kernel/sysenter.c index 960d8bd137d..0bada1870bd 100644 --- a/arch/i386/kernel/sysenter.c +++ b/arch/i386/kernel/sysenter.c @@ -21,11 +21,16 @@ extern asmlinkage void sysenter_entry(void); -void enable_sep_cpu(void *info) +void enable_sep_cpu(void) { int cpu = get_cpu(); struct tss_struct *tss = &per_cpu(init_tss, cpu); + if (!boot_cpu_has(X86_FEATURE_SEP)) { + put_cpu(); + return; + } + tss->ss1 = __KERNEL_CS; tss->esp1 = sizeof(struct tss_struct) + (unsigned long) tss; wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0); @@ -41,7 +46,7 @@ void enable_sep_cpu(void *info) extern const char vsyscall_int80_start, vsyscall_int80_end; extern const char vsyscall_sysenter_start, vsyscall_sysenter_end; -static int __init sysenter_setup(void) +int __init sysenter_setup(void) { void *page = (void *)get_zeroed_page(GFP_ATOMIC); @@ -58,8 +63,5 @@ static int __init sysenter_setup(void) &vsyscall_sysenter_start, &vsyscall_sysenter_end - &vsyscall_sysenter_start); - on_each_cpu(enable_sep_cpu, NULL, 1, 1); return 0; } - -__initcall(sysenter_setup); diff --git a/arch/i386/kernel/time.c b/arch/i386/kernel/time.c index a0dcb7c87c3..e68d9fdb075 100644 --- a/arch/i386/kernel/time.c +++ b/arch/i386/kernel/time.c @@ -77,11 +77,13 @@ u64 jiffies_64 = INITIAL_JIFFIES; EXPORT_SYMBOL(jiffies_64); -unsigned long cpu_khz; /* Detected as we calibrate the TSC */ +unsigned int cpu_khz; /* Detected as we calibrate the TSC */ +EXPORT_SYMBOL(cpu_khz); extern unsigned long wall_jiffies; DEFINE_SPINLOCK(rtc_lock); +EXPORT_SYMBOL(rtc_lock); DEFINE_SPINLOCK(i8253_lock); EXPORT_SYMBOL(i8253_lock); @@ -324,6 +326,8 @@ unsigned long get_cmos_time(void) return retval; } +EXPORT_SYMBOL(get_cmos_time); + static void sync_cmos_clock(unsigned long dummy); static struct timer_list sync_cmos_timer = diff --git a/arch/i386/kernel/time_hpet.c b/arch/i386/kernel/time_hpet.c index 10a0cbb88e7..658c0629ba6 100644 --- a/arch/i386/kernel/time_hpet.c +++ b/arch/i386/kernel/time_hpet.c @@ -50,7 +50,7 @@ static void hpet_writel(unsigned long d, unsigned long a) * comparator value and continue. Next tick can be caught by checking * for a change in the comparator value. Used in apic.c. */ -static void __init wait_hpet_tick(void) +static void __devinit wait_hpet_tick(void) { unsigned int start_cmp_val, end_cmp_val; diff --git a/arch/i386/kernel/timers/common.c b/arch/i386/kernel/timers/common.c index 8e201219f52..8163fe0cf1f 100644 --- a/arch/i386/kernel/timers/common.c +++ b/arch/i386/kernel/timers/common.c @@ -86,7 +86,7 @@ bad_ctc: #define CALIBRATE_CNT_HPET (5 * hpet_tick) #define CALIBRATE_TIME_HPET (5 * KERNEL_TICK_USEC) -unsigned long __init calibrate_tsc_hpet(unsigned long *tsc_hpet_quotient_ptr) +unsigned long __devinit calibrate_tsc_hpet(unsigned long *tsc_hpet_quotient_ptr) { unsigned long tsc_startlow, tsc_starthigh; unsigned long tsc_endlow, tsc_endhigh; @@ -139,6 +139,15 @@ bad_calibration: } #endif + +unsigned long read_timer_tsc(void) +{ + unsigned long retval; + rdtscl(retval); + return retval; +} + + /* calculate cpu_khz */ void init_cpu_khz(void) { @@ -154,7 +163,8 @@ void init_cpu_khz(void) :"=a" (cpu_khz), "=d" (edx) :"r" (tsc_quotient), "0" (eax), "1" (edx)); - printk("Detected %lu.%03lu MHz processor.\n", cpu_khz / 1000, cpu_khz % 1000); + printk("Detected %u.%03u MHz processor.\n", + cpu_khz / 1000, cpu_khz % 1000); } } } diff --git a/arch/i386/kernel/timers/timer.c b/arch/i386/kernel/timers/timer.c index a3d6a288088..7e39ed8e33f 100644 --- a/arch/i386/kernel/timers/timer.c +++ b/arch/i386/kernel/timers/timer.c @@ -64,3 +64,12 @@ struct timer_opts* __init select_timer(void) panic("select_timer: Cannot find a suitable timer\n"); return NULL; } + +int read_current_timer(unsigned long *timer_val) +{ + if (cur_timer->read_timer) { + *timer_val = cur_timer->read_timer(); + return 0; + } + return -1; +} diff --git a/arch/i386/kernel/timers/timer_hpet.c b/arch/i386/kernel/timers/timer_hpet.c index f778f471a09..d766e0963ac 100644 --- a/arch/i386/kernel/timers/timer_hpet.c +++ b/arch/i386/kernel/timers/timer_hpet.c @@ -158,7 +158,7 @@ static int __init init_hpet(char* override) { unsigned long eax=0, edx=1000; ASM_DIV64_REG(cpu_khz, edx, tsc_quotient, eax, edx); - printk("Detected %lu.%03lu MHz processor.\n", + printk("Detected %u.%03u MHz processor.\n", cpu_khz / 1000, cpu_khz % 1000); } set_cyc2ns_scale(cpu_khz/1000); @@ -186,6 +186,7 @@ static struct timer_opts timer_hpet = { .get_offset = get_offset_hpet, .monotonic_clock = monotonic_clock_hpet, .delay = delay_hpet, + .read_timer = read_timer_tsc, }; struct init_timer_opts __initdata timer_hpet_init = { diff --git a/arch/i386/kernel/timers/timer_pm.c b/arch/i386/kernel/timers/timer_pm.c index d77f22030fe..4ef20e66349 100644 --- a/arch/i386/kernel/timers/timer_pm.c +++ b/arch/i386/kernel/timers/timer_pm.c @@ -246,6 +246,7 @@ static struct timer_opts timer_pmtmr = { .get_offset = get_offset_pmtmr, .monotonic_clock = monotonic_clock_pmtmr, .delay = delay_pmtmr, + .read_timer = read_timer_tsc, }; struct init_timer_opts __initdata timer_pmtmr_init = { diff --git a/arch/i386/kernel/timers/timer_tsc.c b/arch/i386/kernel/timers/timer_tsc.c index 180444d8782..f46e625bab6 100644 --- a/arch/i386/kernel/timers/timer_tsc.c +++ b/arch/i386/kernel/timers/timer_tsc.c @@ -33,7 +33,7 @@ static struct timer_opts timer_tsc; static inline void cpufreq_delayed_get(void); -int tsc_disable __initdata = 0; +int tsc_disable __devinitdata = 0; extern spinlock_t i8253_lock; @@ -256,7 +256,7 @@ static unsigned long loops_per_jiffy_ref = 0; #ifndef CONFIG_SMP static unsigned long fast_gettimeoffset_ref = 0; -static unsigned long cpu_khz_ref = 0; +static unsigned int cpu_khz_ref = 0; #endif static int @@ -323,7 +323,7 @@ static inline void cpufreq_delayed_get(void) { return; } int recalibrate_cpu_khz(void) { #ifndef CONFIG_SMP - unsigned long cpu_khz_old = cpu_khz; + unsigned int cpu_khz_old = cpu_khz; if (cpu_has_tsc) { init_cpu_khz(); @@ -534,7 +534,8 @@ static int __init init_tsc(char* override) :"=a" (cpu_khz), "=d" (edx) :"r" (tsc_quotient), "0" (eax), "1" (edx)); - printk("Detected %lu.%03lu MHz processor.\n", cpu_khz / 1000, cpu_khz % 1000); + printk("Detected %u.%03u MHz processor.\n", + cpu_khz / 1000, cpu_khz % 1000); } set_cyc2ns_scale(cpu_khz/1000); return 0; @@ -572,6 +573,7 @@ static struct timer_opts timer_tsc = { .get_offset = get_offset_tsc, .monotonic_clock = monotonic_clock_tsc, .delay = delay_tsc, + .read_timer = read_timer_tsc, }; struct init_timer_opts __initdata timer_tsc_init = { diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 83c579e82a8..a61f33d06ea 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -27,6 +27,7 @@ #include <linux/ptrace.h> #include <linux/utsname.h> #include <linux/kprobes.h> +#include <linux/kexec.h> #ifdef CONFIG_EISA #include <linux/ioport.h> @@ -104,6 +105,7 @@ int register_die_notifier(struct notifier_block *nb) spin_unlock_irqrestore(&die_notifier_lock, flags); return err; } +EXPORT_SYMBOL(register_die_notifier); static inline int valid_stack_ptr(struct thread_info *tinfo, void *p) { @@ -209,7 +211,7 @@ void show_registers(struct pt_regs *regs) esp = (unsigned long) (®s->esp); ss = __KERNEL_DS; - if (regs->xcs & 3) { + if (user_mode(regs)) { in_kernel = 0; esp = regs->esp; ss = regs->xss & 0xffff; @@ -233,22 +235,22 @@ void show_registers(struct pt_regs *regs) * time of the fault.. */ if (in_kernel) { - u8 *eip; + u8 __user *eip; printk("\nStack: "); show_stack(NULL, (unsigned long*)esp); printk("Code: "); - eip = (u8 *)regs->eip - 43; + eip = (u8 __user *)regs->eip - 43; for (i = 0; i < 64; i++, eip++) { unsigned char c; - if (eip < (u8 *)PAGE_OFFSET || __get_user(c, eip)) { + if (eip < (u8 __user *)PAGE_OFFSET || __get_user(c, eip)) { printk(" Bad EIP value."); break; } - if (eip == (u8 *)regs->eip) + if (eip == (u8 __user *)regs->eip) printk("<%02x> ", c); else printk("%02x ", c); @@ -265,20 +267,20 @@ static void handle_BUG(struct pt_regs *regs) char c; unsigned long eip; - if (regs->xcs & 3) + if (user_mode(regs)) goto no_bug; /* Not in kernel */ eip = regs->eip; if (eip < PAGE_OFFSET) goto no_bug; - if (__get_user(ud2, (unsigned short *)eip)) + if (__get_user(ud2, (unsigned short __user *)eip)) goto no_bug; if (ud2 != 0x0b0f) goto no_bug; - if (__get_user(line, (unsigned short *)(eip + 2))) + if (__get_user(line, (unsigned short __user *)(eip + 2))) goto bug; - if (__get_user(file, (char **)(eip + 4)) || + if (__get_user(file, (char * __user *)(eip + 4)) || (unsigned long)file < PAGE_OFFSET || __get_user(c, file)) file = "<bad filename>"; @@ -293,6 +295,9 @@ bug: printk("Kernel BUG\n"); } +/* This is gone through when something in the kernel + * has done something bad and is about to be terminated. +*/ void die(const char * str, struct pt_regs * regs, long err) { static struct { @@ -340,6 +345,10 @@ void die(const char * str, struct pt_regs * regs, long err) bust_spinlocks(0); die.lock_owner = -1; spin_unlock_irq(&die.lock); + + if (kexec_should_crash(current)) + crash_kexec(regs); + if (in_interrupt()) panic("Fatal exception in interrupt"); @@ -353,26 +362,27 @@ void die(const char * str, struct pt_regs * regs, long err) static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err) { - if (!(regs->eflags & VM_MASK) && !(3 & regs->xcs)) + if (!user_mode_vm(regs)) die(str, regs, err); } static void do_trap(int trapnr, int signr, char *str, int vm86, struct pt_regs * regs, long error_code, siginfo_t *info) { + struct task_struct *tsk = current; + tsk->thread.error_code = error_code; + tsk->thread.trap_no = trapnr; + if (regs->eflags & VM_MASK) { if (vm86) goto vm86_trap; goto trap_signal; } - if (!(regs->xcs & 3)) + if (!user_mode(regs)) goto kernel_trap; trap_signal: { - struct task_struct *tsk = current; - tsk->thread.error_code = error_code; - tsk->thread.trap_no = trapnr; if (info) force_sig_info(signr, info, tsk); else @@ -485,10 +495,13 @@ fastcall void do_general_protection(struct pt_regs * regs, long error_code) } put_cpu(); + current->thread.error_code = error_code; + current->thread.trap_no = 13; + if (regs->eflags & VM_MASK) goto gp_in_vm86; - if (!(regs->xcs & 3)) + if (!user_mode(regs)) goto gp_in_kernel; current->thread.error_code = error_code; @@ -569,6 +582,15 @@ void die_nmi (struct pt_regs *regs, const char *msg) console_silent(); spin_unlock(&nmi_print_lock); bust_spinlocks(0); + + /* If we are in kernel we are probably nested up pretty bad + * and might aswell get out now while we still can. + */ + if (!user_mode(regs)) { + current->thread.trap_no = 2; + crash_kexec(regs); + } + do_exit(SIGSEGV); } @@ -624,6 +646,14 @@ fastcall void do_nmi(struct pt_regs * regs, long error_code) nmi_enter(); cpu = smp_processor_id(); + +#ifdef CONFIG_HOTPLUG_CPU + if (!cpu_online(cpu)) { + nmi_exit(); + return; + } +#endif + ++nmi_count(cpu); if (!nmi_callback(regs, cpu)) @@ -636,11 +666,13 @@ void set_nmi_callback(nmi_callback_t callback) { nmi_callback = callback; } +EXPORT_SYMBOL_GPL(set_nmi_callback); void unset_nmi_callback(void) { nmi_callback = dummy_nmi_callback; } +EXPORT_SYMBOL_GPL(unset_nmi_callback); #ifdef CONFIG_KPROBES fastcall void do_int3(struct pt_regs *regs, long error_code) @@ -682,7 +714,7 @@ fastcall void do_debug(struct pt_regs * regs, long error_code) unsigned int condition; struct task_struct *tsk = current; - __asm__ __volatile__("movl %%db6,%0" : "=r" (condition)); + get_debugreg(condition, 6); if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code, SIGTRAP) == NOTIFY_STOP) @@ -713,7 +745,7 @@ fastcall void do_debug(struct pt_regs * regs, long error_code) * check for kernel mode by just checking the CPL * of CS. */ - if ((regs->xcs & 3) == 0) + if (!user_mode(regs)) goto clear_TF_reenable; } @@ -724,9 +756,7 @@ fastcall void do_debug(struct pt_regs * regs, long error_code) * the signal is delivered. */ clear_dr7: - __asm__("movl %0,%%db7" - : /* no output */ - : "r" (0)); + set_debugreg(0, 7); return; debug_vm86: @@ -871,9 +901,9 @@ fastcall void do_simd_coprocessor_error(struct pt_regs * regs, error_code); return; } - die_if_kernel("cache flush denied", regs, error_code); current->thread.trap_no = 19; current->thread.error_code = error_code; + die_if_kernel("cache flush denied", regs, error_code); force_sig(SIGSEGV, current); } } diff --git a/arch/i386/kernel/vmlinux.lds.S b/arch/i386/kernel/vmlinux.lds.S index e0512cc8bea..7e01a528a83 100644 --- a/arch/i386/kernel/vmlinux.lds.S +++ b/arch/i386/kernel/vmlinux.lds.S @@ -2,20 +2,23 @@ * Written by Martin Mares <mj@atrey.karlin.mff.cuni.cz>; */ +#define LOAD_OFFSET __PAGE_OFFSET + #include <asm-generic/vmlinux.lds.h> #include <asm/thread_info.h> #include <asm/page.h> OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") OUTPUT_ARCH(i386) -ENTRY(startup_32) +ENTRY(phys_startup_32) jiffies = jiffies_64; SECTIONS { - . = __PAGE_OFFSET + 0x100000; + . = __KERNEL_START; + phys_startup_32 = startup_32 - LOAD_OFFSET; /* read-only */ _text = .; /* Text and read-only data */ - .text : { + .text : AT(ADDR(.text) - LOAD_OFFSET) { *(.text) SCHED_TEXT LOCK_TEXT @@ -27,49 +30,55 @@ SECTIONS . = ALIGN(16); /* Exception table */ __start___ex_table = .; - __ex_table : { *(__ex_table) } + __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { *(__ex_table) } __stop___ex_table = .; RODATA /* writeable */ - .data : { /* Data */ + .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */ *(.data) CONSTRUCTORS } . = ALIGN(4096); __nosave_begin = .; - .data_nosave : { *(.data.nosave) } + .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { *(.data.nosave) } . = ALIGN(4096); __nosave_end = .; . = ALIGN(4096); - .data.page_aligned : { *(.data.idt) } + .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { + *(.data.idt) + } . = ALIGN(32); - .data.cacheline_aligned : { *(.data.cacheline_aligned) } + .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { + *(.data.cacheline_aligned) + } _edata = .; /* End of data section */ . = ALIGN(THREAD_SIZE); /* init_task */ - .data.init_task : { *(.data.init_task) } + .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { + *(.data.init_task) + } /* will be freed after init */ . = ALIGN(4096); /* Init code and data */ __init_begin = .; - .init.text : { + .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { _sinittext = .; *(.init.text) _einittext = .; } - .init.data : { *(.init.data) } + .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { *(.init.data) } . = ALIGN(16); __setup_start = .; - .init.setup : { *(.init.setup) } + .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { *(.init.setup) } __setup_end = .; __initcall_start = .; - .initcall.init : { + .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) { *(.initcall1.init) *(.initcall2.init) *(.initcall3.init) @@ -80,33 +89,41 @@ SECTIONS } __initcall_end = .; __con_initcall_start = .; - .con_initcall.init : { *(.con_initcall.init) } + .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { + *(.con_initcall.init) + } __con_initcall_end = .; SECURITY_INIT . = ALIGN(4); __alt_instructions = .; - .altinstructions : { *(.altinstructions) } + .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { + *(.altinstructions) + } __alt_instructions_end = .; - .altinstr_replacement : { *(.altinstr_replacement) } + .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) { + *(.altinstr_replacement) + } /* .exit.text is discard at runtime, not link time, to deal with references from .altinstructions and .eh_frame */ - .exit.text : { *(.exit.text) } - .exit.data : { *(.exit.data) } + .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { *(.exit.text) } + .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { *(.exit.data) } . = ALIGN(4096); __initramfs_start = .; - .init.ramfs : { *(.init.ramfs) } + .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { *(.init.ramfs) } __initramfs_end = .; . = ALIGN(32); __per_cpu_start = .; - .data.percpu : { *(.data.percpu) } + .data.percpu : AT(ADDR(.data.percpu) - LOAD_OFFSET) { *(.data.percpu) } __per_cpu_end = .; . = ALIGN(4096); __init_end = .; /* freed after init ends here */ __bss_start = .; /* BSS */ - .bss : { + .bss.page_aligned : AT(ADDR(.bss.page_aligned) - LOAD_OFFSET) { *(.bss.page_aligned) + } + .bss : AT(ADDR(.bss) - LOAD_OFFSET) { *(.bss) } . = ALIGN(4); diff --git a/arch/i386/lib/dec_and_lock.c b/arch/i386/lib/dec_and_lock.c index ab43394dc77..8b81b2524fa 100644 --- a/arch/i386/lib/dec_and_lock.c +++ b/arch/i386/lib/dec_and_lock.c @@ -8,6 +8,7 @@ */ #include <linux/spinlock.h> +#include <linux/module.h> #include <asm/atomic.h> int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock) @@ -38,3 +39,4 @@ slow_path: spin_unlock(lock); return 0; } +EXPORT_SYMBOL(_atomic_dec_and_lock); diff --git a/arch/i386/lib/delay.c b/arch/i386/lib/delay.c index eb0cdfe9280..c49a6acbee5 100644 --- a/arch/i386/lib/delay.c +++ b/arch/i386/lib/delay.c @@ -13,6 +13,7 @@ #include <linux/config.h> #include <linux/sched.h> #include <linux/delay.h> +#include <linux/module.h> #include <asm/processor.h> #include <asm/delay.h> #include <asm/timer.h> @@ -47,3 +48,8 @@ void __ndelay(unsigned long nsecs) { __const_udelay(nsecs * 0x00005); /* 2**32 / 1000000000 (rounded up) */ } + +EXPORT_SYMBOL(__delay); +EXPORT_SYMBOL(__const_udelay); +EXPORT_SYMBOL(__udelay); +EXPORT_SYMBOL(__ndelay); diff --git a/arch/i386/lib/mmx.c b/arch/i386/lib/mmx.c index 01f8b1a2cc8..2afda94dffd 100644 --- a/arch/i386/lib/mmx.c +++ b/arch/i386/lib/mmx.c @@ -3,6 +3,7 @@ #include <linux/string.h> #include <linux/sched.h> #include <linux/hardirq.h> +#include <linux/module.h> #include <asm/i387.h> @@ -397,3 +398,7 @@ void mmx_copy_page(void *to, void *from) else fast_copy_page(to, from); } + +EXPORT_SYMBOL(_mmx_memcpy); +EXPORT_SYMBOL(mmx_clear_page); +EXPORT_SYMBOL(mmx_copy_page); diff --git a/arch/i386/lib/usercopy.c b/arch/i386/lib/usercopy.c index 51aa2bbb026..4cf981d70f4 100644 --- a/arch/i386/lib/usercopy.c +++ b/arch/i386/lib/usercopy.c @@ -84,6 +84,7 @@ __strncpy_from_user(char *dst, const char __user *src, long count) __do_strncpy_from_user(dst, src, count, res); return res; } +EXPORT_SYMBOL(__strncpy_from_user); /** * strncpy_from_user: - Copy a NUL terminated string from userspace. @@ -111,7 +112,7 @@ strncpy_from_user(char *dst, const char __user *src, long count) __do_strncpy_from_user(dst, src, count, res); return res; } - +EXPORT_SYMBOL(strncpy_from_user); /* * Zero Userspace @@ -157,6 +158,7 @@ clear_user(void __user *to, unsigned long n) __do_clear_user(to, n); return n; } +EXPORT_SYMBOL(clear_user); /** * __clear_user: - Zero a block of memory in user space, with less checking. @@ -175,6 +177,7 @@ __clear_user(void __user *to, unsigned long n) __do_clear_user(to, n); return n; } +EXPORT_SYMBOL(__clear_user); /** * strlen_user: - Get the size of a string in user space. @@ -218,6 +221,7 @@ long strnlen_user(const char __user *s, long n) :"cc"); return res & mask; } +EXPORT_SYMBOL(strnlen_user); #ifdef CONFIG_X86_INTEL_USERCOPY static unsigned long @@ -570,6 +574,7 @@ survive: n = __copy_user_intel(to, from, n); return n; } +EXPORT_SYMBOL(__copy_to_user_ll); unsigned long __copy_from_user_ll(void *to, const void __user *from, unsigned long n) @@ -581,6 +586,7 @@ __copy_from_user_ll(void *to, const void __user *from, unsigned long n) n = __copy_user_zeroing_intel(to, from, n); return n; } +EXPORT_SYMBOL(__copy_from_user_ll); /** * copy_to_user: - Copy a block of data into user space. diff --git a/arch/i386/mach-default/setup.c b/arch/i386/mach-default/setup.c index 0aa08eaa893..e5a1a83d09e 100644 --- a/arch/i386/mach-default/setup.c +++ b/arch/i386/mach-default/setup.c @@ -10,6 +10,14 @@ #include <asm/acpi.h> #include <asm/arch_hooks.h> +#ifdef CONFIG_HOTPLUG_CPU +#define DEFAULT_SEND_IPI (1) +#else +#define DEFAULT_SEND_IPI (0) +#endif + +int no_broadcast=DEFAULT_SEND_IPI; + /** * pre_intr_init_hook - initialisation prior to setting up interrupt vectors * @@ -104,3 +112,22 @@ void __init mca_nmi_hook(void) printk("NMI generated from unknown source!\n"); } #endif + +static __init int no_ipi_broadcast(char *str) +{ + get_option(&str, &no_broadcast); + printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" : + "IPI Broadcast"); + return 1; +} + +__setup("no_ipi_broadcast", no_ipi_broadcast); + +static int __init print_ipi_mode(void) +{ + printk ("Using IPI %s mode\n", no_broadcast ? "No-Shortcut" : + "Shortcut"); + return 0; +} + +late_initcall(print_ipi_mode); diff --git a/arch/i386/mach-default/topology.c b/arch/i386/mach-default/topology.c index 5b3e8817dae..23395fff35d 100644 --- a/arch/i386/mach-default/topology.c +++ b/arch/i386/mach-default/topology.c @@ -73,12 +73,11 @@ static int __init topology_init(void) { int i; - for (i = 0; i < MAX_NUMNODES; i++) { - if (node_online(i)) - arch_register_node(i); - } - for (i = 0; i < NR_CPUS; i++) - if (cpu_possible(i)) arch_register_cpu(i); + for_each_online_node(i) + arch_register_node(i); + + for_each_cpu(i) + arch_register_cpu(i); return 0; } @@ -88,8 +87,8 @@ static int __init topology_init(void) { int i; - for (i = 0; i < NR_CPUS; i++) - if (cpu_possible(i)) arch_register_cpu(i); + for_each_cpu(i) + arch_register_cpu(i); return 0; } diff --git a/arch/i386/mach-visws/mpparse.c b/arch/i386/mach-visws/mpparse.c index 5a22082147f..5f3d7e6de37 100644 --- a/arch/i386/mach-visws/mpparse.c +++ b/arch/i386/mach-visws/mpparse.c @@ -23,7 +23,6 @@ unsigned long mp_lapic_addr; /* Processor that is doing the boot up */ unsigned int boot_cpu_physical_apicid = -1U; -unsigned int boot_cpu_logical_apicid = -1U; /* Bitmask of physically existing CPUs */ physid_mask_t phys_cpu_present_map; @@ -52,10 +51,8 @@ static void __init MP_processor_info (struct mpc_config_processor *m) (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, m->mpc_apicver); - if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) { + if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) boot_cpu_physical_apicid = m->mpc_apicid; - boot_cpu_logical_apicid = logical_apicid; - } ver = m->mpc_apicver; if ((ver >= 0x14 && m->mpc_apicid >= 0xff) || m->mpc_apicid >= 0xf) { diff --git a/arch/i386/mach-voyager/voyager_smp.c b/arch/i386/mach-voyager/voyager_smp.c index a6e0ddd65bd..8c8527593da 100644 --- a/arch/i386/mach-voyager/voyager_smp.c +++ b/arch/i386/mach-voyager/voyager_smp.c @@ -1288,7 +1288,7 @@ smp_local_timer_interrupt(struct pt_regs * regs) per_cpu(prof_counter, cpu); } - update_process_times(user_mode(regs)); + update_process_times(user_mode_vm(regs)); } if( ((1<<cpu) & voyager_extended_vic_processors) == 0) diff --git a/arch/i386/mm/Makefile b/arch/i386/mm/Makefile index fc327250684..80908b5aa60 100644 --- a/arch/i386/mm/Makefile +++ b/arch/i386/mm/Makefile @@ -4,7 +4,7 @@ obj-y := init.o pgtable.o fault.o ioremap.o extable.o pageattr.o mmap.o -obj-$(CONFIG_DISCONTIGMEM) += discontig.o +obj-$(CONFIG_NUMA) += discontig.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o obj-$(CONFIG_HIGHMEM) += highmem.o obj-$(CONFIG_BOOT_IOREMAP) += boot_ioremap.o diff --git a/arch/i386/mm/discontig.c b/arch/i386/mm/discontig.c index 1726b4096b1..b358f0702a4 100644 --- a/arch/i386/mm/discontig.c +++ b/arch/i386/mm/discontig.c @@ -29,12 +29,16 @@ #include <linux/highmem.h> #include <linux/initrd.h> #include <linux/nodemask.h> +#include <linux/module.h> +#include <linux/kexec.h> + #include <asm/e820.h> #include <asm/setup.h> #include <asm/mmzone.h> #include <bios_ebda.h> struct pglist_data *node_data[MAX_NUMNODES]; +EXPORT_SYMBOL(node_data); bootmem_data_t node0_bdata; /* @@ -42,12 +46,16 @@ bootmem_data_t node0_bdata; * populated the following initialisation. * * 1) node_online_map - the map of all nodes configured (online) in the system - * 2) physnode_map - the mapping between a pfn and owning node - * 3) node_start_pfn - the starting page frame number for a node + * 2) node_start_pfn - the starting page frame number for a node * 3) node_end_pfn - the ending page fram number for a node */ +unsigned long node_start_pfn[MAX_NUMNODES]; +unsigned long node_end_pfn[MAX_NUMNODES]; + +#ifdef CONFIG_DISCONTIGMEM /* + * 4) physnode_map - the mapping between a pfn and owning node * physnode_map keeps track of the physical memory layout of a generic * numa node on a 256Mb break (each element of the array will * represent 256Mb of memory and will be marked by the node id. so, @@ -59,6 +67,7 @@ bootmem_data_t node0_bdata; * physnode_map[8- ] = -1; */ s8 physnode_map[MAX_ELEMENTS] = { [0 ... (MAX_ELEMENTS - 1)] = -1}; +EXPORT_SYMBOL(physnode_map); void memory_present(int nid, unsigned long start, unsigned long end) { @@ -85,9 +94,7 @@ unsigned long node_memmap_size_bytes(int nid, unsigned long start_pfn, return (nr_pages + 1) * sizeof(struct page); } - -unsigned long node_start_pfn[MAX_NUMNODES]; -unsigned long node_end_pfn[MAX_NUMNODES]; +#endif extern unsigned long find_max_low_pfn(void); extern void find_max_pfn(void); @@ -108,6 +115,9 @@ unsigned long node_remap_offset[MAX_NUMNODES]; void *node_remap_start_vaddr[MAX_NUMNODES]; void set_pmd_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags); +void *node_remap_end_vaddr[MAX_NUMNODES]; +void *node_remap_alloc_vaddr[MAX_NUMNODES]; + /* * FLAT - support for basic PC memory model with discontig enabled, essentially * a single node with all available processors in it with a flat @@ -146,6 +156,21 @@ static void __init find_max_pfn_node(int nid) BUG(); } +/* Find the owning node for a pfn. */ +int early_pfn_to_nid(unsigned long pfn) +{ + int nid; + + for_each_node(nid) { + if (node_end_pfn[nid] == 0) + break; + if (node_start_pfn[nid] <= pfn && node_end_pfn[nid] >= pfn) + return nid; + } + + return 0; +} + /* * Allocate memory for the pg_data_t for this node via a crude pre-bootmem * method. For node zero take this from the bottom of memory, for @@ -163,6 +188,21 @@ static void __init allocate_pgdat(int nid) } } +void *alloc_remap(int nid, unsigned long size) +{ + void *allocation = node_remap_alloc_vaddr[nid]; + + size = ALIGN(size, L1_CACHE_BYTES); + + if (!allocation || (allocation + size) >= node_remap_end_vaddr[nid]) + return 0; + + node_remap_alloc_vaddr[nid] += size; + memset(allocation, 0, size); + + return allocation; +} + void __init remap_numa_kva(void) { void *vaddr; @@ -170,8 +210,6 @@ void __init remap_numa_kva(void) int node; for_each_online_node(node) { - if (node == 0) - continue; for (pfn=0; pfn < node_remap_size[node]; pfn += PTRS_PER_PTE) { vaddr = node_remap_start_vaddr[node]+(pfn<<PAGE_SHIFT); set_pmd_pfn((ulong) vaddr, @@ -185,13 +223,9 @@ static unsigned long calculate_numa_remap_pages(void) { int nid; unsigned long size, reserve_pages = 0; + unsigned long pfn; for_each_online_node(nid) { - if (nid == 0) - continue; - if (!node_remap_size[nid]) - continue; - /* * The acpi/srat node info can show hot-add memroy zones * where memory could be added but not currently present. @@ -208,11 +242,24 @@ static unsigned long calculate_numa_remap_pages(void) size = (size + LARGE_PAGE_BYTES - 1) / LARGE_PAGE_BYTES; /* now the roundup is correct, convert to PAGE_SIZE pages */ size = size * PTRS_PER_PTE; + + /* + * Validate the region we are allocating only contains valid + * pages. + */ + for (pfn = node_end_pfn[nid] - size; + pfn < node_end_pfn[nid]; pfn++) + if (!page_is_ram(pfn)) + break; + + if (pfn != node_end_pfn[nid]) + size = 0; + printk("Reserving %ld pages of KVA for lmem_map of node %d\n", size, nid); node_remap_size[nid] = size; - reserve_pages += size; node_remap_offset[nid] = reserve_pages; + reserve_pages += size; printk("Shrinking node %d from %ld pages to %ld pages\n", nid, node_end_pfn[nid], node_end_pfn[nid] - size); node_end_pfn[nid] -= size; @@ -265,12 +312,18 @@ unsigned long __init setup_memory(void) (ulong) pfn_to_kaddr(max_low_pfn)); for_each_online_node(nid) { node_remap_start_vaddr[nid] = pfn_to_kaddr( - (highstart_pfn + reserve_pages) - node_remap_offset[nid]); + highstart_pfn + node_remap_offset[nid]); + /* Init the node remap allocator */ + node_remap_end_vaddr[nid] = node_remap_start_vaddr[nid] + + (node_remap_size[nid] * PAGE_SIZE); + node_remap_alloc_vaddr[nid] = node_remap_start_vaddr[nid] + + ALIGN(sizeof(pg_data_t), PAGE_SIZE); + allocate_pgdat(nid); printk ("node %d will remap to vaddr %08lx - %08lx\n", nid, (ulong) node_remap_start_vaddr[nid], - (ulong) pfn_to_kaddr(highstart_pfn + reserve_pages - - node_remap_offset[nid] + node_remap_size[nid])); + (ulong) pfn_to_kaddr(highstart_pfn + + node_remap_offset[nid] + node_remap_size[nid])); } printk("High memory starts at vaddr %08lx\n", (ulong) pfn_to_kaddr(highstart_pfn)); @@ -333,23 +386,9 @@ void __init zone_sizes_init(void) } zholes_size = get_zholes_size(nid); - /* - * We let the lmem_map for node 0 be allocated from the - * normal bootmem allocator, but other nodes come from the - * remapped KVA area - mbligh - */ - if (!nid) - free_area_init_node(nid, NODE_DATA(nid), - zones_size, start, zholes_size); - else { - unsigned long lmem_map; - lmem_map = (unsigned long)node_remap_start_vaddr[nid]; - lmem_map += sizeof(pg_data_t) + PAGE_SIZE - 1; - lmem_map &= PAGE_MASK; - NODE_DATA(nid)->node_mem_map = (struct page *)lmem_map; - free_area_init_node(nid, NODE_DATA(nid), zones_size, - start, zholes_size); - } + + free_area_init_node(nid, NODE_DATA(nid), zones_size, start, + zholes_size); } return; } @@ -358,24 +397,26 @@ void __init set_highmem_pages_init(int bad_ppro) { #ifdef CONFIG_HIGHMEM struct zone *zone; + struct page *page; for_each_zone(zone) { - unsigned long node_pfn, node_high_size, zone_start_pfn; - struct page * zone_mem_map; - + unsigned long node_pfn, zone_start_pfn, zone_end_pfn; + if (!is_highmem(zone)) continue; - printk("Initializing %s for node %d\n", zone->name, - zone->zone_pgdat->node_id); - - node_high_size = zone->spanned_pages; - zone_mem_map = zone->zone_mem_map; zone_start_pfn = zone->zone_start_pfn; + zone_end_pfn = zone_start_pfn + zone->spanned_pages; + + printk("Initializing %s for node %d (%08lx:%08lx)\n", + zone->name, zone->zone_pgdat->node_id, + zone_start_pfn, zone_end_pfn); - for (node_pfn = 0; node_pfn < node_high_size; node_pfn++) { - one_highpage_init((struct page *)(zone_mem_map + node_pfn), - zone_start_pfn + node_pfn, bad_ppro); + for (node_pfn = zone_start_pfn; node_pfn < zone_end_pfn; node_pfn++) { + if (!pfn_valid(node_pfn)) + continue; + page = pfn_to_page(node_pfn); + one_highpage_init(page, node_pfn, bad_ppro); } } totalram_pages += totalhigh_pages; diff --git a/arch/i386/mm/fault.c b/arch/i386/mm/fault.c index a509237c481..8e90339d6ea 100644 --- a/arch/i386/mm/fault.c +++ b/arch/i386/mm/fault.c @@ -146,7 +146,7 @@ static int __is_prefetch(struct pt_regs *regs, unsigned long addr) if (instr > limit) break; - if (__get_user(opcode, (unsigned char *) instr)) + if (__get_user(opcode, (unsigned char __user *) instr)) break; instr_hi = opcode & 0xf0; @@ -173,7 +173,7 @@ static int __is_prefetch(struct pt_regs *regs, unsigned long addr) scan_more = 0; if (instr > limit) break; - if (__get_user(opcode, (unsigned char *) instr)) + if (__get_user(opcode, (unsigned char __user *) instr)) break; prefetch = (instr_lo == 0xF) && (opcode == 0x0D || opcode == 0x18); @@ -463,6 +463,9 @@ no_context: printk(KERN_ALERT "*pte = %08lx\n", page); } #endif + tsk->thread.cr2 = address; + tsk->thread.trap_no = 14; + tsk->thread.error_code = error_code; die("Oops", regs, error_code); bust_spinlocks(0); do_exit(SIGKILL); diff --git a/arch/i386/mm/highmem.c b/arch/i386/mm/highmem.c index fc4c4cad4e9..b6eb4dcb877 100644 --- a/arch/i386/mm/highmem.c +++ b/arch/i386/mm/highmem.c @@ -1,4 +1,5 @@ #include <linux/highmem.h> +#include <linux/module.h> void *kmap(struct page *page) { @@ -74,6 +75,24 @@ void kunmap_atomic(void *kvaddr, enum km_type type) preempt_check_resched(); } +/* This is the same as kmap_atomic() but can map memory that doesn't + * have a struct page associated with it. + */ +void *kmap_atomic_pfn(unsigned long pfn, enum km_type type) +{ + enum fixed_addresses idx; + unsigned long vaddr; + + inc_preempt_count(); + + idx = type + KM_TYPE_NR*smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot)); + __flush_tlb_one(vaddr); + + return (void*) vaddr; +} + struct page *kmap_atomic_to_page(void *ptr) { unsigned long idx, vaddr = (unsigned long)ptr; @@ -87,3 +106,8 @@ struct page *kmap_atomic_to_page(void *ptr) return pte_page(*pte); } +EXPORT_SYMBOL(kmap); +EXPORT_SYMBOL(kunmap); +EXPORT_SYMBOL(kmap_atomic); +EXPORT_SYMBOL(kunmap_atomic); +EXPORT_SYMBOL(kmap_atomic_to_page); diff --git a/arch/i386/mm/init.c b/arch/i386/mm/init.c index 8766c771bb4..12216b52e28 100644 --- a/arch/i386/mm/init.c +++ b/arch/i386/mm/init.c @@ -191,7 +191,7 @@ static inline int page_kills_ppro(unsigned long pagenr) extern int is_available_memory(efi_memory_desc_t *); -static inline int page_is_ram(unsigned long pagenr) +int page_is_ram(unsigned long pagenr) { int i; unsigned long addr, end; @@ -276,7 +276,9 @@ void __init one_highpage_init(struct page *page, int pfn, int bad_ppro) SetPageReserved(page); } -#ifndef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA +extern void set_highmem_pages_init(int); +#else static void __init set_highmem_pages_init(int bad_ppro) { int pfn; @@ -284,9 +286,7 @@ static void __init set_highmem_pages_init(int bad_ppro) one_highpage_init(pfn_to_page(pfn), pfn, bad_ppro); totalram_pages += totalhigh_pages; } -#else -extern void set_highmem_pages_init(int); -#endif /* !CONFIG_DISCONTIGMEM */ +#endif /* CONFIG_FLATMEM */ #else #define kmap_init() do { } while (0) @@ -295,12 +295,13 @@ extern void set_highmem_pages_init(int); #endif /* CONFIG_HIGHMEM */ unsigned long long __PAGE_KERNEL = _PAGE_KERNEL; +EXPORT_SYMBOL(__PAGE_KERNEL); unsigned long long __PAGE_KERNEL_EXEC = _PAGE_KERNEL_EXEC; -#ifndef CONFIG_DISCONTIGMEM -#define remap_numa_kva() do {} while (0) -#else +#ifdef CONFIG_NUMA extern void __init remap_numa_kva(void); +#else +#define remap_numa_kva() do {} while (0) #endif static void __init pagetable_init (void) @@ -351,7 +352,7 @@ static void __init pagetable_init (void) #endif } -#if defined(CONFIG_PM_DISK) || defined(CONFIG_SOFTWARE_SUSPEND) +#ifdef CONFIG_SOFTWARE_SUSPEND /* * Swap suspend & friends need this for resume because things like the intel-agp * driver might have split up a kernel 4MB mapping. @@ -525,7 +526,7 @@ static void __init set_max_mapnr_init(void) #else num_physpages = max_low_pfn; #endif -#ifndef CONFIG_DISCONTIGMEM +#ifdef CONFIG_FLATMEM max_mapnr = num_physpages; #endif } @@ -539,7 +540,7 @@ void __init mem_init(void) int tmp; int bad_ppro; -#ifndef CONFIG_DISCONTIGMEM +#ifdef CONFIG_FLATMEM if (!mem_map) BUG(); #endif diff --git a/arch/i386/mm/ioremap.c b/arch/i386/mm/ioremap.c index ab542792b27..6b25afc933b 100644 --- a/arch/i386/mm/ioremap.c +++ b/arch/i386/mm/ioremap.c @@ -11,6 +11,7 @@ #include <linux/vmalloc.h> #include <linux/init.h> #include <linux/slab.h> +#include <linux/module.h> #include <asm/io.h> #include <asm/fixmap.h> #include <asm/cacheflush.h> @@ -165,7 +166,7 @@ void __iomem * __ioremap(unsigned long phys_addr, unsigned long size, unsigned l } return (void __iomem *) (offset + (char __iomem *)addr); } - +EXPORT_SYMBOL(__ioremap); /** * ioremap_nocache - map bus memory into CPU space @@ -222,6 +223,7 @@ void __iomem *ioremap_nocache (unsigned long phys_addr, unsigned long size) return p; } +EXPORT_SYMBOL(ioremap_nocache); void iounmap(volatile void __iomem *addr) { @@ -241,7 +243,7 @@ void iounmap(volatile void __iomem *addr) write_lock(&vmlist_lock); p = __remove_vm_area((void *) (PAGE_MASK & (unsigned long __force) addr)); if (!p) { - printk("iounmap: bad address %p\n", addr); + printk(KERN_WARNING "iounmap: bad address %p\n", addr); goto out_unlock; } @@ -255,6 +257,7 @@ out_unlock: write_unlock(&vmlist_lock); kfree(p); } +EXPORT_SYMBOL(iounmap); void __init *bt_ioremap(unsigned long phys_addr, unsigned long size) { diff --git a/arch/i386/mm/pgtable.c b/arch/i386/mm/pgtable.c index dd81479ff88..bd2f7afc7a2 100644 --- a/arch/i386/mm/pgtable.c +++ b/arch/i386/mm/pgtable.c @@ -30,13 +30,14 @@ void show_mem(void) struct page *page; pg_data_t *pgdat; unsigned long i; + struct page_state ps; - printk("Mem-info:\n"); + printk(KERN_INFO "Mem-info:\n"); show_free_areas(); - printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); + printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { for (i = 0; i < pgdat->node_spanned_pages; ++i) { - page = pgdat->node_mem_map + i; + page = pgdat_page_nr(pgdat, i); total++; if (PageHighMem(page)) highmem++; @@ -48,11 +49,18 @@ void show_mem(void) shared += page_count(page) - 1; } } - printk("%d pages of RAM\n", total); - printk("%d pages of HIGHMEM\n",highmem); - printk("%d reserved pages\n",reserved); - printk("%d pages shared\n",shared); - printk("%d pages swap cached\n",cached); + printk(KERN_INFO "%d pages of RAM\n", total); + printk(KERN_INFO "%d pages of HIGHMEM\n", highmem); + printk(KERN_INFO "%d reserved pages\n", reserved); + printk(KERN_INFO "%d pages shared\n", shared); + printk(KERN_INFO "%d pages swap cached\n", cached); + + get_page_state(&ps); + printk(KERN_INFO "%lu pages dirty\n", ps.nr_dirty); + printk(KERN_INFO "%lu pages writeback\n", ps.nr_writeback); + printk(KERN_INFO "%lu pages mapped\n", ps.nr_mapped); + printk(KERN_INFO "%lu pages slab\n", ps.nr_slab); + printk(KERN_INFO "%lu pages pagetables\n", ps.nr_page_table_pages); } /* @@ -105,16 +113,16 @@ void set_pmd_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags) pmd_t *pmd; if (vaddr & (PMD_SIZE-1)) { /* vaddr is misaligned */ - printk ("set_pmd_pfn: vaddr misaligned\n"); + printk(KERN_WARNING "set_pmd_pfn: vaddr misaligned\n"); return; /* BUG(); */ } if (pfn & (PTRS_PER_PTE-1)) { /* pfn is misaligned */ - printk ("set_pmd_pfn: pfn misaligned\n"); + printk(KERN_WARNING "set_pmd_pfn: pfn misaligned\n"); return; /* BUG(); */ } pgd = swapper_pg_dir + pgd_index(vaddr); if (pgd_none(*pgd)) { - printk ("set_pmd_pfn: pgd_none\n"); + printk(KERN_WARNING "set_pmd_pfn: pgd_none\n"); return; /* BUG(); */ } pud = pud_offset(pgd, vaddr); diff --git a/arch/i386/oprofile/backtrace.c b/arch/i386/oprofile/backtrace.c index 52d72e074f7..65dfd2edb67 100644 --- a/arch/i386/oprofile/backtrace.c +++ b/arch/i386/oprofile/backtrace.c @@ -91,7 +91,7 @@ x86_backtrace(struct pt_regs * const regs, unsigned int depth) head = (struct frame_head *)regs->ebp; #endif - if (!user_mode(regs)) { + if (!user_mode_vm(regs)) { while (depth-- && valid_kernel_stack(head, regs)) head = dump_backtrace(head); return; diff --git a/arch/i386/pci/irq.c b/arch/i386/pci/irq.c index da21b1d07c1..83458f81e66 100644 --- a/arch/i386/pci/irq.c +++ b/arch/i386/pci/irq.c @@ -227,6 +227,24 @@ static int pirq_via_set(struct pci_dev *router, struct pci_dev *dev, int pirq, i } /* + * The VIA pirq rules are nibble-based, like ALI, + * but without the ugly irq number munging. + * However, for 82C586, nibble map is different . + */ +static int pirq_via586_get(struct pci_dev *router, struct pci_dev *dev, int pirq) +{ + static unsigned int pirqmap[4] = { 3, 2, 5, 1 }; + return read_config_nybble(router, 0x55, pirqmap[pirq-1]); +} + +static int pirq_via586_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq) +{ + static unsigned int pirqmap[4] = { 3, 2, 5, 1 }; + write_config_nybble(router, 0x55, pirqmap[pirq-1], irq); + return 1; +} + +/* * ITE 8330G pirq rules are nibble-based * FIXME: pirqmap may be { 1, 0, 3, 2 }, * 2+3 are both mapped to irq 9 on my system @@ -512,6 +530,10 @@ static __init int via_router_probe(struct irq_router *r, struct pci_dev *router, switch(device) { case PCI_DEVICE_ID_VIA_82C586_0: + r->name = "VIA"; + r->get = pirq_via586_get; + r->set = pirq_via586_set; + return 1; case PCI_DEVICE_ID_VIA_82C596: case PCI_DEVICE_ID_VIA_82C686: case PCI_DEVICE_ID_VIA_8231: diff --git a/arch/i386/pci/pcbios.c b/arch/i386/pci/pcbios.c index 141421b673b..b9d65f0bc2d 100644 --- a/arch/i386/pci/pcbios.c +++ b/arch/i386/pci/pcbios.c @@ -4,6 +4,7 @@ #include <linux/pci.h> #include <linux/init.h> +#include <linux/module.h> #include "pci.h" #include "pci-functions.h" @@ -456,7 +457,7 @@ struct irq_routing_table * __devinit pcibios_get_irq_routing_table(void) free_page(page); return rt; } - +EXPORT_SYMBOL(pcibios_get_irq_routing_table); int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq) { @@ -473,6 +474,7 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq) "S" (&pci_indirect)); return !(ret & 0xff00); } +EXPORT_SYMBOL(pcibios_set_irq_routing); static int __init pci_pcbios_init(void) { diff --git a/arch/i386/power/cpu.c b/arch/i386/power/cpu.c index cf337c673d9..0e6b45b6125 100644 --- a/arch/i386/power/cpu.c +++ b/arch/i386/power/cpu.c @@ -22,9 +22,11 @@ #include <linux/device.h> #include <linux/suspend.h> #include <linux/acpi.h> + #include <asm/uaccess.h> #include <asm/acpi.h> #include <asm/tlbflush.h> +#include <asm/processor.h> static struct saved_context saved_context; @@ -33,8 +35,6 @@ unsigned long saved_context_esp, saved_context_ebp; unsigned long saved_context_esi, saved_context_edi; unsigned long saved_context_eflags; -extern void enable_sep_cpu(void *); - void __save_processor_state(struct saved_context *ctxt) { kernel_fpu_begin(); @@ -44,7 +44,6 @@ void __save_processor_state(struct saved_context *ctxt) */ asm volatile ("sgdt %0" : "=m" (ctxt->gdt_limit)); asm volatile ("sidt %0" : "=m" (ctxt->idt_limit)); - asm volatile ("sldt %0" : "=m" (ctxt->ldt)); asm volatile ("str %0" : "=m" (ctxt->tr)); /* @@ -94,20 +93,19 @@ static void fix_processor_context(void) * Now maybe reload the debug registers */ if (current->thread.debugreg[7]){ - loaddebug(¤t->thread, 0); - loaddebug(¤t->thread, 1); - loaddebug(¤t->thread, 2); - loaddebug(¤t->thread, 3); - /* no 4 and 5 */ - loaddebug(¤t->thread, 6); - loaddebug(¤t->thread, 7); + set_debugreg(current->thread.debugreg[0], 0); + set_debugreg(current->thread.debugreg[1], 1); + set_debugreg(current->thread.debugreg[2], 2); + set_debugreg(current->thread.debugreg[3], 3); + /* no 4 and 5 */ + set_debugreg(current->thread.debugreg[6], 6); + set_debugreg(current->thread.debugreg[7], 7); } } void __restore_processor_state(struct saved_context *ctxt) { - /* * control registers */ @@ -117,6 +115,13 @@ void __restore_processor_state(struct saved_context *ctxt) asm volatile ("movl %0, %%cr0" :: "r" (ctxt->cr0)); /* + * now restore the descriptor tables to their proper values + * ltr is done i fix_processor_context(). + */ + asm volatile ("lgdt %0" :: "m" (ctxt->gdt_limit)); + asm volatile ("lidt %0" :: "m" (ctxt->idt_limit)); + + /* * segment registers */ asm volatile ("movw %0, %%es" :: "r" (ctxt->es)); @@ -125,18 +130,10 @@ void __restore_processor_state(struct saved_context *ctxt) asm volatile ("movw %0, %%ss" :: "r" (ctxt->ss)); /* - * now restore the descriptor tables to their proper values - * ltr is done i fix_processor_context(). - */ - asm volatile ("lgdt %0" :: "m" (ctxt->gdt_limit)); - asm volatile ("lidt %0" :: "m" (ctxt->idt_limit)); - asm volatile ("lldt %0" :: "m" (ctxt->ldt)); - - /* * sysenter MSRs */ if (boot_cpu_has(X86_FEATURE_SEP)) - enable_sep_cpu(NULL); + enable_sep_cpu(); fix_processor_context(); do_fpu_end(); diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index ce4dfa8b834..01b78e7f992 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -161,6 +161,8 @@ config IA64_PAGE_SIZE_64KB endchoice +source kernel/Kconfig.hz + config IA64_BRL_EMU bool depends on ITANIUM @@ -197,7 +199,7 @@ config HOLES_IN_ZONE bool default y if VIRTUAL_MEM_MAP -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool "Discontiguous memory support" depends on (IA64_DIG || IA64_SGI_SN2 || IA64_GENERIC || IA64_HP_ZX1 || IA64_HP_ZX1_SWIOTLB) && NUMA && VIRTUAL_MEM_MAP default y if (IA64_SGI_SN2 || IA64_GENERIC) && NUMA @@ -300,6 +302,8 @@ config PREEMPT Say Y here if you are building a kernel for a desktop, embedded or real-time system. Say N if you are unsure. +source "mm/Kconfig" + config HAVE_DEC_LOCK bool depends on (SMP || PREEMPT) diff --git a/arch/ia64/Kconfig.debug b/arch/ia64/Kconfig.debug index de9d507ba0f..fda67ac993d 100644 --- a/arch/ia64/Kconfig.debug +++ b/arch/ia64/Kconfig.debug @@ -2,6 +2,17 @@ menu "Kernel hacking" source "lib/Kconfig.debug" +config KPROBES + bool "Kprobes" + depends on DEBUG_KERNEL + help + Kprobes allows you to trap at almost any kernel address and + execute a callback function. register_kprobe() establishes + a probepoint and specifies the callback. Kprobes is useful + for kernel debugging, non-intrusive instrumentation and testing. + If in doubt, say "N". + + choice prompt "Physical memory granularity" default IA64_GRANULE_64MB diff --git a/arch/ia64/configs/sn2_defconfig b/arch/ia64/configs/sn2_defconfig index a01bb02d074..487d2e36b0a 100644 --- a/arch/ia64/configs/sn2_defconfig +++ b/arch/ia64/configs/sn2_defconfig @@ -78,7 +78,7 @@ CONFIG_IA64_L1_CACHE_SHIFT=7 CONFIG_NUMA=y CONFIG_VIRTUAL_MEM_MAP=y CONFIG_HOLES_IN_ZONE=y -CONFIG_DISCONTIGMEM=y +CONFIG_ARCH_DISCONTIGMEM_ENABLE=y # CONFIG_IA64_CYCLONE is not set CONFIG_IOSAPIC=y CONFIG_IA64_SGI_SN_SIM=y diff --git a/arch/ia64/defconfig b/arch/ia64/defconfig index 7be8096e056..8444add7638 100644 --- a/arch/ia64/defconfig +++ b/arch/ia64/defconfig @@ -84,7 +84,7 @@ CONFIG_IA64_L1_CACHE_SHIFT=7 CONFIG_NUMA=y CONFIG_VIRTUAL_MEM_MAP=y CONFIG_HOLES_IN_ZONE=y -CONFIG_DISCONTIGMEM=y +CONFIG_ARCH_DISCONTIGMEM_ENABLE=y CONFIG_IA64_CYCLONE=y CONFIG_IOSAPIC=y CONFIG_FORCE_MAX_ZONEORDER=18 diff --git a/arch/ia64/ia32/ia32priv.h b/arch/ia64/ia32/ia32priv.h index b2de948bdae..e3e9290e3ff 100644 --- a/arch/ia64/ia32/ia32priv.h +++ b/arch/ia64/ia32/ia32priv.h @@ -241,7 +241,7 @@ typedef struct compat_siginfo { /* POSIX.1b timers */ struct { - timer_t _tid; /* timer id */ + compat_timer_t _tid; /* timer id */ int _overrun; /* overrun count */ char _pad[sizeof(unsigned int) - sizeof(int)]; compat_sigval_t _sigval; /* same as below */ diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile index 4c73d8ba2e3..b2e2f6509eb 100644 --- a/arch/ia64/kernel/Makefile +++ b/arch/ia64/kernel/Makefile @@ -20,6 +20,7 @@ obj-$(CONFIG_SMP) += smp.o smpboot.o domain.o obj-$(CONFIG_PERFMON) += perfmon_default_smpl.o obj-$(CONFIG_IA64_CYCLONE) += cyclone.o obj-$(CONFIG_IA64_MCA_RECOVERY) += mca_recovery.o +obj-$(CONFIG_KPROBES) += kprobes.o jprobes.o obj-$(CONFIG_IA64_UNCACHED_ALLOCATOR) += uncached.o mca_recovery-y += mca_drv.o mca_drv_asm.o diff --git a/arch/ia64/kernel/domain.c b/arch/ia64/kernel/domain.c index fe532c97043..d65e87b6394 100644 --- a/arch/ia64/kernel/domain.c +++ b/arch/ia64/kernel/domain.c @@ -14,7 +14,7 @@ #include <linux/topology.h> #include <linux/nodemask.h> -#define SD_NODES_PER_DOMAIN 6 +#define SD_NODES_PER_DOMAIN 16 #ifdef CONFIG_NUMA /** @@ -27,7 +27,7 @@ * * Should use nodemask_t. */ -static int __devinit find_next_best_node(int node, unsigned long *used_nodes) +static int find_next_best_node(int node, unsigned long *used_nodes) { int i, n, val, min_val, best_node = 0; @@ -66,7 +66,7 @@ static int __devinit find_next_best_node(int node, unsigned long *used_nodes) * should be one that prevents unnecessary balancing, but also spreads tasks * out optimally. */ -static cpumask_t __devinit sched_domain_node_span(int node) +static cpumask_t sched_domain_node_span(int node) { int i; cpumask_t span, nodemask; @@ -96,7 +96,7 @@ static cpumask_t __devinit sched_domain_node_span(int node) #ifdef CONFIG_SCHED_SMT static DEFINE_PER_CPU(struct sched_domain, cpu_domains); static struct sched_group sched_group_cpus[NR_CPUS]; -static int __devinit cpu_to_cpu_group(int cpu) +static int cpu_to_cpu_group(int cpu) { return cpu; } @@ -104,7 +104,7 @@ static int __devinit cpu_to_cpu_group(int cpu) static DEFINE_PER_CPU(struct sched_domain, phys_domains); static struct sched_group sched_group_phys[NR_CPUS]; -static int __devinit cpu_to_phys_group(int cpu) +static int cpu_to_phys_group(int cpu) { #ifdef CONFIG_SCHED_SMT return first_cpu(cpu_sibling_map[cpu]); @@ -125,44 +125,36 @@ static struct sched_group *sched_group_nodes[MAX_NUMNODES]; static DEFINE_PER_CPU(struct sched_domain, allnodes_domains); static struct sched_group sched_group_allnodes[MAX_NUMNODES]; -static int __devinit cpu_to_allnodes_group(int cpu) +static int cpu_to_allnodes_group(int cpu) { return cpu_to_node(cpu); } #endif /* - * Set up scheduler domains and groups. Callers must hold the hotplug lock. + * Build sched domains for a given set of cpus and attach the sched domains + * to the individual cpus */ -void __devinit arch_init_sched_domains(void) +void build_sched_domains(const cpumask_t *cpu_map) { int i; - cpumask_t cpu_default_map; /* - * Setup mask for cpus without special case scheduling requirements. - * For now this just excludes isolated cpus, but could be used to - * exclude other special cases in the future. + * Set up domains for cpus specified by the cpu_map. */ - cpus_complement(cpu_default_map, cpu_isolated_map); - cpus_and(cpu_default_map, cpu_default_map, cpu_online_map); - - /* - * Set up domains. Isolated domains just stay on the dummy domain. - */ - for_each_cpu_mask(i, cpu_default_map) { + for_each_cpu_mask(i, *cpu_map) { int group; struct sched_domain *sd = NULL, *p; cpumask_t nodemask = node_to_cpumask(cpu_to_node(i)); - cpus_and(nodemask, nodemask, cpu_default_map); + cpus_and(nodemask, nodemask, *cpu_map); #ifdef CONFIG_NUMA if (num_online_cpus() > SD_NODES_PER_DOMAIN*cpus_weight(nodemask)) { sd = &per_cpu(allnodes_domains, i); *sd = SD_ALLNODES_INIT; - sd->span = cpu_default_map; + sd->span = *cpu_map; group = cpu_to_allnodes_group(i); sd->groups = &sched_group_allnodes[group]; p = sd; @@ -173,7 +165,7 @@ void __devinit arch_init_sched_domains(void) *sd = SD_NODE_INIT; sd->span = sched_domain_node_span(cpu_to_node(i)); sd->parent = p; - cpus_and(sd->span, sd->span, cpu_default_map); + cpus_and(sd->span, sd->span, *cpu_map); #endif p = sd; @@ -190,7 +182,7 @@ void __devinit arch_init_sched_domains(void) group = cpu_to_cpu_group(i); *sd = SD_SIBLING_INIT; sd->span = cpu_sibling_map[i]; - cpus_and(sd->span, sd->span, cpu_default_map); + cpus_and(sd->span, sd->span, *cpu_map); sd->parent = p; sd->groups = &sched_group_cpus[group]; #endif @@ -198,9 +190,9 @@ void __devinit arch_init_sched_domains(void) #ifdef CONFIG_SCHED_SMT /* Set up CPU (sibling) groups */ - for_each_cpu_mask(i, cpu_default_map) { + for_each_cpu_mask(i, *cpu_map) { cpumask_t this_sibling_map = cpu_sibling_map[i]; - cpus_and(this_sibling_map, this_sibling_map, cpu_default_map); + cpus_and(this_sibling_map, this_sibling_map, *cpu_map); if (i != first_cpu(this_sibling_map)) continue; @@ -213,7 +205,7 @@ void __devinit arch_init_sched_domains(void) for (i = 0; i < MAX_NUMNODES; i++) { cpumask_t nodemask = node_to_cpumask(i); - cpus_and(nodemask, nodemask, cpu_default_map); + cpus_and(nodemask, nodemask, *cpu_map); if (cpus_empty(nodemask)) continue; @@ -222,7 +214,7 @@ void __devinit arch_init_sched_domains(void) } #ifdef CONFIG_NUMA - init_sched_build_groups(sched_group_allnodes, cpu_default_map, + init_sched_build_groups(sched_group_allnodes, *cpu_map, &cpu_to_allnodes_group); for (i = 0; i < MAX_NUMNODES; i++) { @@ -233,12 +225,12 @@ void __devinit arch_init_sched_domains(void) cpumask_t covered = CPU_MASK_NONE; int j; - cpus_and(nodemask, nodemask, cpu_default_map); + cpus_and(nodemask, nodemask, *cpu_map); if (cpus_empty(nodemask)) continue; domainspan = sched_domain_node_span(i); - cpus_and(domainspan, domainspan, cpu_default_map); + cpus_and(domainspan, domainspan, *cpu_map); sg = kmalloc(sizeof(struct sched_group), GFP_KERNEL); sched_group_nodes[i] = sg; @@ -266,7 +258,7 @@ void __devinit arch_init_sched_domains(void) int n = (i + j) % MAX_NUMNODES; cpus_complement(notcovered, covered); - cpus_and(tmp, notcovered, cpu_default_map); + cpus_and(tmp, notcovered, *cpu_map); cpus_and(tmp, tmp, domainspan); if (cpus_empty(tmp)) break; @@ -293,7 +285,7 @@ void __devinit arch_init_sched_domains(void) #endif /* Calculate CPU power for physical packages and nodes */ - for_each_cpu_mask(i, cpu_default_map) { + for_each_cpu_mask(i, *cpu_map) { int power; struct sched_domain *sd; #ifdef CONFIG_SCHED_SMT @@ -359,13 +351,35 @@ next_sg: cpu_attach_domain(sd, i); } } +/* + * Set up scheduler domains and groups. Callers must hold the hotplug lock. + */ +void arch_init_sched_domains(const cpumask_t *cpu_map) +{ + cpumask_t cpu_default_map; + + /* + * Setup mask for cpus without special case scheduling requirements. + * For now this just excludes isolated cpus, but could be used to + * exclude other special cases in the future. + */ + cpus_andnot(cpu_default_map, *cpu_map, cpu_isolated_map); + + build_sched_domains(&cpu_default_map); +} -void __devinit arch_destroy_sched_domains(void) +void arch_destroy_sched_domains(const cpumask_t *cpu_map) { #ifdef CONFIG_NUMA int i; for (i = 0; i < MAX_NUMNODES; i++) { + cpumask_t nodemask = node_to_cpumask(i); struct sched_group *oldsg, *sg = sched_group_nodes[i]; + + cpus_and(nodemask, nodemask, *cpu_map); + if (cpus_empty(nodemask)) + continue; + if (sg == NULL) continue; sg = sg->next; diff --git a/arch/ia64/kernel/jprobes.S b/arch/ia64/kernel/jprobes.S new file mode 100644 index 00000000000..b7fa3ccd2b0 --- /dev/null +++ b/arch/ia64/kernel/jprobes.S @@ -0,0 +1,61 @@ +/* + * Jprobe specific operations + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) Intel Corporation, 2005 + * + * 2005-May Rusty Lynch <rusty.lynch@intel.com> and Anil S Keshavamurthy + * <anil.s.keshavamurthy@intel.com> initial implementation + * + * Jprobes (a.k.a. "jump probes" which is built on-top of kprobes) allow a + * probe to be inserted into the beginning of a function call. The fundamental + * difference between a jprobe and a kprobe is the jprobe handler is executed + * in the same context as the target function, while the kprobe handlers + * are executed in interrupt context. + * + * For jprobes we initially gain control by placing a break point in the + * first instruction of the targeted function. When we catch that specific + * break, we: + * * set the return address to our jprobe_inst_return() function + * * jump to the jprobe handler function + * + * Since we fixed up the return address, the jprobe handler will return to our + * jprobe_inst_return() function, giving us control again. At this point we + * are back in the parents frame marker, so we do yet another call to our + * jprobe_break() function to fix up the frame marker as it would normally + * exist in the target function. + * + * Our jprobe_return function then transfers control back to kprobes.c by + * executing a break instruction using one of our reserved numbers. When we + * catch that break in kprobes.c, we continue like we do for a normal kprobe + * by single stepping the emulated instruction, and then returning execution + * to the correct location. + */ +#include <asm/asmmacro.h> + + /* + * void jprobe_break(void) + */ +ENTRY(jprobe_break) + break.m 0x80300 +END(jprobe_break) + + /* + * void jprobe_inst_return(void) + */ +GLOBAL_ENTRY(jprobe_inst_return) + br.call.sptk.many b0=jprobe_break +END(jprobe_inst_return) diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c new file mode 100644 index 00000000000..5978823d5c6 --- /dev/null +++ b/arch/ia64/kernel/kprobes.c @@ -0,0 +1,601 @@ +/* + * Kernel Probes (KProbes) + * arch/ia64/kernel/kprobes.c + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) IBM Corporation, 2002, 2004 + * Copyright (C) Intel Corporation, 2005 + * + * 2005-Apr Rusty Lynch <rusty.lynch@intel.com> and Anil S Keshavamurthy + * <anil.s.keshavamurthy@intel.com> adapted from i386 + */ + +#include <linux/config.h> +#include <linux/kprobes.h> +#include <linux/ptrace.h> +#include <linux/spinlock.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/preempt.h> +#include <linux/moduleloader.h> + +#include <asm/pgtable.h> +#include <asm/kdebug.h> + +extern void jprobe_inst_return(void); + +/* kprobe_status settings */ +#define KPROBE_HIT_ACTIVE 0x00000001 +#define KPROBE_HIT_SS 0x00000002 + +static struct kprobe *current_kprobe, *kprobe_prev; +static unsigned long kprobe_status, kprobe_status_prev; +static struct pt_regs jprobe_saved_regs; + +enum instruction_type {A, I, M, F, B, L, X, u}; +static enum instruction_type bundle_encoding[32][3] = { + { M, I, I }, /* 00 */ + { M, I, I }, /* 01 */ + { M, I, I }, /* 02 */ + { M, I, I }, /* 03 */ + { M, L, X }, /* 04 */ + { M, L, X }, /* 05 */ + { u, u, u }, /* 06 */ + { u, u, u }, /* 07 */ + { M, M, I }, /* 08 */ + { M, M, I }, /* 09 */ + { M, M, I }, /* 0A */ + { M, M, I }, /* 0B */ + { M, F, I }, /* 0C */ + { M, F, I }, /* 0D */ + { M, M, F }, /* 0E */ + { M, M, F }, /* 0F */ + { M, I, B }, /* 10 */ + { M, I, B }, /* 11 */ + { M, B, B }, /* 12 */ + { M, B, B }, /* 13 */ + { u, u, u }, /* 14 */ + { u, u, u }, /* 15 */ + { B, B, B }, /* 16 */ + { B, B, B }, /* 17 */ + { M, M, B }, /* 18 */ + { M, M, B }, /* 19 */ + { u, u, u }, /* 1A */ + { u, u, u }, /* 1B */ + { M, F, B }, /* 1C */ + { M, F, B }, /* 1D */ + { u, u, u }, /* 1E */ + { u, u, u }, /* 1F */ +}; + +/* + * In this function we check to see if the instruction + * is IP relative instruction and update the kprobe + * inst flag accordingly + */ +static void update_kprobe_inst_flag(uint template, uint slot, uint major_opcode, + unsigned long kprobe_inst, struct kprobe *p) +{ + p->ainsn.inst_flag = 0; + p->ainsn.target_br_reg = 0; + + if (bundle_encoding[template][slot] == B) { + switch (major_opcode) { + case INDIRECT_CALL_OPCODE: + p->ainsn.inst_flag |= INST_FLAG_FIX_BRANCH_REG; + p->ainsn.target_br_reg = ((kprobe_inst >> 6) & 0x7); + break; + case IP_RELATIVE_PREDICT_OPCODE: + case IP_RELATIVE_BRANCH_OPCODE: + p->ainsn.inst_flag |= INST_FLAG_FIX_RELATIVE_IP_ADDR; + break; + case IP_RELATIVE_CALL_OPCODE: + p->ainsn.inst_flag |= INST_FLAG_FIX_RELATIVE_IP_ADDR; + p->ainsn.inst_flag |= INST_FLAG_FIX_BRANCH_REG; + p->ainsn.target_br_reg = ((kprobe_inst >> 6) & 0x7); + break; + } + } else if (bundle_encoding[template][slot] == X) { + switch (major_opcode) { + case LONG_CALL_OPCODE: + p->ainsn.inst_flag |= INST_FLAG_FIX_BRANCH_REG; + p->ainsn.target_br_reg = ((kprobe_inst >> 6) & 0x7); + break; + } + } + return; +} + +/* + * In this function we check to see if the instruction + * on which we are inserting kprobe is supported. + * Returns 0 if supported + * Returns -EINVAL if unsupported + */ +static int unsupported_inst(uint template, uint slot, uint major_opcode, + unsigned long kprobe_inst, struct kprobe *p) +{ + unsigned long addr = (unsigned long)p->addr; + + if (bundle_encoding[template][slot] == I) { + switch (major_opcode) { + case 0x0: //I_UNIT_MISC_OPCODE: + /* + * Check for Integer speculation instruction + * - Bit 33-35 to be equal to 0x1 + */ + if (((kprobe_inst >> 33) & 0x7) == 1) { + printk(KERN_WARNING + "Kprobes on speculation inst at <0x%lx> not supported\n", + addr); + return -EINVAL; + } + + /* + * IP relative mov instruction + * - Bit 27-35 to be equal to 0x30 + */ + if (((kprobe_inst >> 27) & 0x1FF) == 0x30) { + printk(KERN_WARNING + "Kprobes on \"mov r1=ip\" at <0x%lx> not supported\n", + addr); + return -EINVAL; + + } + } + } + return 0; +} + + +/* + * In this function we check to see if the instruction + * (qp) cmpx.crel.ctype p1,p2=r2,r3 + * on which we are inserting kprobe is cmp instruction + * with ctype as unc. + */ +static uint is_cmp_ctype_unc_inst(uint template, uint slot, uint major_opcode, +unsigned long kprobe_inst) +{ + cmp_inst_t cmp_inst; + uint ctype_unc = 0; + + if (!((bundle_encoding[template][slot] == I) || + (bundle_encoding[template][slot] == M))) + goto out; + + if (!((major_opcode == 0xC) || (major_opcode == 0xD) || + (major_opcode == 0xE))) + goto out; + + cmp_inst.l = kprobe_inst; + if ((cmp_inst.f.x2 == 0) || (cmp_inst.f.x2 == 1)) { + /* Integere compare - Register Register (A6 type)*/ + if ((cmp_inst.f.tb == 0) && (cmp_inst.f.ta == 0) + &&(cmp_inst.f.c == 1)) + ctype_unc = 1; + } else if ((cmp_inst.f.x2 == 2)||(cmp_inst.f.x2 == 3)) { + /* Integere compare - Immediate Register (A8 type)*/ + if ((cmp_inst.f.ta == 0) &&(cmp_inst.f.c == 1)) + ctype_unc = 1; + } +out: + return ctype_unc; +} + +/* + * In this function we override the bundle with + * the break instruction at the given slot. + */ +static void prepare_break_inst(uint template, uint slot, uint major_opcode, + unsigned long kprobe_inst, struct kprobe *p) +{ + unsigned long break_inst = BREAK_INST; + bundle_t *bundle = &p->ainsn.insn.bundle; + + /* + * Copy the original kprobe_inst qualifying predicate(qp) + * to the break instruction iff !is_cmp_ctype_unc_inst + * because for cmp instruction with ctype equal to unc, + * which is a special instruction always needs to be + * executed regradless of qp + */ + if (!is_cmp_ctype_unc_inst(template, slot, major_opcode, kprobe_inst)) + break_inst |= (0x3f & kprobe_inst); + + switch (slot) { + case 0: + bundle->quad0.slot0 = break_inst; + break; + case 1: + bundle->quad0.slot1_p0 = break_inst; + bundle->quad1.slot1_p1 = break_inst >> (64-46); + break; + case 2: + bundle->quad1.slot2 = break_inst; + break; + } + + /* + * Update the instruction flag, so that we can + * emulate the instruction properly after we + * single step on original instruction + */ + update_kprobe_inst_flag(template, slot, major_opcode, kprobe_inst, p); +} + +static inline void get_kprobe_inst(bundle_t *bundle, uint slot, + unsigned long *kprobe_inst, uint *major_opcode) +{ + unsigned long kprobe_inst_p0, kprobe_inst_p1; + unsigned int template; + + template = bundle->quad0.template; + + switch (slot) { + case 0: + *major_opcode = (bundle->quad0.slot0 >> SLOT0_OPCODE_SHIFT); + *kprobe_inst = bundle->quad0.slot0; + break; + case 1: + *major_opcode = (bundle->quad1.slot1_p1 >> SLOT1_p1_OPCODE_SHIFT); + kprobe_inst_p0 = bundle->quad0.slot1_p0; + kprobe_inst_p1 = bundle->quad1.slot1_p1; + *kprobe_inst = kprobe_inst_p0 | (kprobe_inst_p1 << (64-46)); + break; + case 2: + *major_opcode = (bundle->quad1.slot2 >> SLOT2_OPCODE_SHIFT); + *kprobe_inst = bundle->quad1.slot2; + break; + } +} + +static int valid_kprobe_addr(int template, int slot, unsigned long addr) +{ + if ((slot > 2) || ((bundle_encoding[template][1] == L) && slot > 1)) { + printk(KERN_WARNING "Attempting to insert unaligned kprobe at 0x%lx\n", + addr); + return -EINVAL; + } + return 0; +} + +static inline void save_previous_kprobe(void) +{ + kprobe_prev = current_kprobe; + kprobe_status_prev = kprobe_status; +} + +static inline void restore_previous_kprobe(void) +{ + current_kprobe = kprobe_prev; + kprobe_status = kprobe_status_prev; +} + +static inline void set_current_kprobe(struct kprobe *p) +{ + current_kprobe = p; +} + +int arch_prepare_kprobe(struct kprobe *p) +{ + unsigned long addr = (unsigned long) p->addr; + unsigned long *kprobe_addr = (unsigned long *)(addr & ~0xFULL); + unsigned long kprobe_inst=0; + unsigned int slot = addr & 0xf, template, major_opcode = 0; + bundle_t *bundle = &p->ainsn.insn.bundle; + + memcpy(&p->opcode.bundle, kprobe_addr, sizeof(bundle_t)); + memcpy(&p->ainsn.insn.bundle, kprobe_addr, sizeof(bundle_t)); + + template = bundle->quad0.template; + + if(valid_kprobe_addr(template, slot, addr)) + return -EINVAL; + + /* Move to slot 2, if bundle is MLX type and kprobe slot is 1 */ + if (slot == 1 && bundle_encoding[template][1] == L) + slot++; + + /* Get kprobe_inst and major_opcode from the bundle */ + get_kprobe_inst(bundle, slot, &kprobe_inst, &major_opcode); + + if (unsupported_inst(template, slot, major_opcode, kprobe_inst, p)) + return -EINVAL; + + prepare_break_inst(template, slot, major_opcode, kprobe_inst, p); + + return 0; +} + +void arch_arm_kprobe(struct kprobe *p) +{ + unsigned long addr = (unsigned long)p->addr; + unsigned long arm_addr = addr & ~0xFULL; + + memcpy((char *)arm_addr, &p->ainsn.insn.bundle, sizeof(bundle_t)); + flush_icache_range(arm_addr, arm_addr + sizeof(bundle_t)); +} + +void arch_disarm_kprobe(struct kprobe *p) +{ + unsigned long addr = (unsigned long)p->addr; + unsigned long arm_addr = addr & ~0xFULL; + + /* p->opcode contains the original unaltered bundle */ + memcpy((char *) arm_addr, (char *) &p->opcode.bundle, sizeof(bundle_t)); + flush_icache_range(arm_addr, arm_addr + sizeof(bundle_t)); +} + +void arch_remove_kprobe(struct kprobe *p) +{ +} + +/* + * We are resuming execution after a single step fault, so the pt_regs + * structure reflects the register state after we executed the instruction + * located in the kprobe (p->ainsn.insn.bundle). We still need to adjust + * the ip to point back to the original stack address. To set the IP address + * to original stack address, handle the case where we need to fixup the + * relative IP address and/or fixup branch register. + */ +static void resume_execution(struct kprobe *p, struct pt_regs *regs) +{ + unsigned long bundle_addr = ((unsigned long) (&p->opcode.bundle)) & ~0xFULL; + unsigned long resume_addr = (unsigned long)p->addr & ~0xFULL; + unsigned long template; + int slot = ((unsigned long)p->addr & 0xf); + + template = p->opcode.bundle.quad0.template; + + if (slot == 1 && bundle_encoding[template][1] == L) + slot = 2; + + if (p->ainsn.inst_flag) { + + if (p->ainsn.inst_flag & INST_FLAG_FIX_RELATIVE_IP_ADDR) { + /* Fix relative IP address */ + regs->cr_iip = (regs->cr_iip - bundle_addr) + resume_addr; + } + + if (p->ainsn.inst_flag & INST_FLAG_FIX_BRANCH_REG) { + /* + * Fix target branch register, software convention is + * to use either b0 or b6 or b7, so just checking + * only those registers + */ + switch (p->ainsn.target_br_reg) { + case 0: + if ((regs->b0 == bundle_addr) || + (regs->b0 == bundle_addr + 0x10)) { + regs->b0 = (regs->b0 - bundle_addr) + + resume_addr; + } + break; + case 6: + if ((regs->b6 == bundle_addr) || + (regs->b6 == bundle_addr + 0x10)) { + regs->b6 = (regs->b6 - bundle_addr) + + resume_addr; + } + break; + case 7: + if ((regs->b7 == bundle_addr) || + (regs->b7 == bundle_addr + 0x10)) { + regs->b7 = (regs->b7 - bundle_addr) + + resume_addr; + } + break; + } /* end switch */ + } + goto turn_ss_off; + } + + if (slot == 2) { + if (regs->cr_iip == bundle_addr + 0x10) { + regs->cr_iip = resume_addr + 0x10; + } + } else { + if (regs->cr_iip == bundle_addr) { + regs->cr_iip = resume_addr; + } + } + +turn_ss_off: + /* Turn off Single Step bit */ + ia64_psr(regs)->ss = 0; +} + +static void prepare_ss(struct kprobe *p, struct pt_regs *regs) +{ + unsigned long bundle_addr = (unsigned long) &p->opcode.bundle; + unsigned long slot = (unsigned long)p->addr & 0xf; + + /* Update instruction pointer (IIP) and slot number (IPSR.ri) */ + regs->cr_iip = bundle_addr & ~0xFULL; + + if (slot > 2) + slot = 0; + + ia64_psr(regs)->ri = slot; + + /* turn on single stepping */ + ia64_psr(regs)->ss = 1; +} + +static int pre_kprobes_handler(struct die_args *args) +{ + struct kprobe *p; + int ret = 0; + struct pt_regs *regs = args->regs; + kprobe_opcode_t *addr = (kprobe_opcode_t *)instruction_pointer(regs); + + preempt_disable(); + + /* Handle recursion cases */ + if (kprobe_running()) { + p = get_kprobe(addr); + if (p) { + if (kprobe_status == KPROBE_HIT_SS) { + unlock_kprobes(); + goto no_kprobe; + } + /* We have reentered the pre_kprobe_handler(), since + * another probe was hit while within the handler. + * We here save the original kprobes variables and + * just single step on the instruction of the new probe + * without calling any user handlers. + */ + save_previous_kprobe(); + set_current_kprobe(p); + p->nmissed++; + prepare_ss(p, regs); + kprobe_status = KPROBE_REENTER; + return 1; + } else if (args->err == __IA64_BREAK_JPROBE) { + /* + * jprobe instrumented function just completed + */ + p = current_kprobe; + if (p->break_handler && p->break_handler(p, regs)) { + goto ss_probe; + } + } else { + /* Not our break */ + goto no_kprobe; + } + } + + lock_kprobes(); + p = get_kprobe(addr); + if (!p) { + unlock_kprobes(); + goto no_kprobe; + } + + kprobe_status = KPROBE_HIT_ACTIVE; + set_current_kprobe(p); + + if (p->pre_handler && p->pre_handler(p, regs)) + /* + * Our pre-handler is specifically requesting that we just + * do a return. This is handling the case where the + * pre-handler is really our special jprobe pre-handler. + */ + return 1; + +ss_probe: + prepare_ss(p, regs); + kprobe_status = KPROBE_HIT_SS; + return 1; + +no_kprobe: + preempt_enable_no_resched(); + return ret; +} + +static int post_kprobes_handler(struct pt_regs *regs) +{ + if (!kprobe_running()) + return 0; + + if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) { + kprobe_status = KPROBE_HIT_SSDONE; + current_kprobe->post_handler(current_kprobe, regs, 0); + } + + resume_execution(current_kprobe, regs); + + /*Restore back the original saved kprobes variables and continue. */ + if (kprobe_status == KPROBE_REENTER) { + restore_previous_kprobe(); + goto out; + } + + unlock_kprobes(); + +out: + preempt_enable_no_resched(); + return 1; +} + +static int kprobes_fault_handler(struct pt_regs *regs, int trapnr) +{ + if (!kprobe_running()) + return 0; + + if (current_kprobe->fault_handler && + current_kprobe->fault_handler(current_kprobe, regs, trapnr)) + return 1; + + if (kprobe_status & KPROBE_HIT_SS) { + resume_execution(current_kprobe, regs); + unlock_kprobes(); + preempt_enable_no_resched(); + } + + return 0; +} + +int kprobe_exceptions_notify(struct notifier_block *self, unsigned long val, + void *data) +{ + struct die_args *args = (struct die_args *)data; + switch(val) { + case DIE_BREAK: + if (pre_kprobes_handler(args)) + return NOTIFY_STOP; + break; + case DIE_SS: + if (post_kprobes_handler(args->regs)) + return NOTIFY_STOP; + break; + case DIE_PAGE_FAULT: + if (kprobes_fault_handler(args->regs, args->trapnr)) + return NOTIFY_STOP; + default: + break; + } + return NOTIFY_DONE; +} + +int setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct jprobe *jp = container_of(p, struct jprobe, kp); + unsigned long addr = ((struct fnptr *)(jp->entry))->ip; + + /* save architectural state */ + jprobe_saved_regs = *regs; + + /* after rfi, execute the jprobe instrumented function */ + regs->cr_iip = addr & ~0xFULL; + ia64_psr(regs)->ri = addr & 0xf; + regs->r1 = ((struct fnptr *)(jp->entry))->gp; + + /* + * fix the return address to our jprobe_inst_return() function + * in the jprobes.S file + */ + regs->b0 = ((struct fnptr *)(jprobe_inst_return))->ip; + + return 1; +} + +int longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) +{ + *regs = jprobe_saved_regs; + return 1; +} diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c index 3865f088ffa..623b0a54670 100644 --- a/arch/ia64/kernel/smpboot.c +++ b/arch/ia64/kernel/smpboot.c @@ -346,6 +346,7 @@ smp_callin (void) lock_ipi_calllock(); cpu_set(cpuid, cpu_online_map); unlock_ipi_calllock(); + per_cpu(cpu_state, cpuid) = CPU_ONLINE; smp_setup_percpu_timer(); @@ -611,6 +612,7 @@ void __devinit smp_prepare_boot_cpu(void) { cpu_set(smp_processor_id(), cpu_online_map); cpu_set(smp_processor_id(), cpu_callin_map); + per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; } /* @@ -688,6 +690,7 @@ int __cpu_disable(void) return -EBUSY; remove_siblinginfo(cpu); + cpu_clear(cpu, cpu_online_map); fixup_irqs(); local_flush_tlb_all(); cpu_clear(cpu, cpu_callin_map); @@ -774,6 +777,7 @@ __cpu_up (unsigned int cpu) if (cpu_isset(cpu, cpu_callin_map)) return -EINVAL; + per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; /* Processor goes to start_secondary(), sets online flag */ ret = do_boot_cpu(sapicid, cpu); if (ret < 0) diff --git a/arch/ia64/kernel/traps.c b/arch/ia64/kernel/traps.c index 1861173bd4f..e7e520d90f0 100644 --- a/arch/ia64/kernel/traps.c +++ b/arch/ia64/kernel/traps.c @@ -21,12 +21,26 @@ #include <asm/intrinsics.h> #include <asm/processor.h> #include <asm/uaccess.h> +#include <asm/kdebug.h> extern spinlock_t timerlist_lock; fpswa_interface_t *fpswa_interface; EXPORT_SYMBOL(fpswa_interface); +struct notifier_block *ia64die_chain; +static DEFINE_SPINLOCK(die_notifier_lock); + +int register_die_notifier(struct notifier_block *nb) +{ + int err = 0; + unsigned long flags; + spin_lock_irqsave(&die_notifier_lock, flags); + err = notifier_chain_register(&ia64die_chain, nb); + spin_unlock_irqrestore(&die_notifier_lock, flags); + return err; +} + void __init trap_init (void) { @@ -137,6 +151,10 @@ ia64_bad_break (unsigned long break_num, struct pt_regs *regs) switch (break_num) { case 0: /* unknown error (used by GCC for __builtin_abort()) */ + if (notify_die(DIE_BREAK, "break 0", regs, break_num, TRAP_BRKPT, SIGTRAP) + == NOTIFY_STOP) { + return; + } die_if_kernel("bugcheck!", regs, break_num); sig = SIGILL; code = ILL_ILLOPC; break; @@ -189,6 +207,15 @@ ia64_bad_break (unsigned long break_num, struct pt_regs *regs) sig = SIGILL; code = __ILL_BNDMOD; break; + case 0x80200: + case 0x80300: + if (notify_die(DIE_BREAK, "kprobe", regs, break_num, TRAP_BRKPT, SIGTRAP) + == NOTIFY_STOP) { + return; + } + sig = SIGTRAP; code = TRAP_BRKPT; + break; + default: if (break_num < 0x40000 || break_num > 0x100000) die_if_kernel("Bad break", regs, break_num); @@ -548,7 +575,11 @@ ia64_fault (unsigned long vector, unsigned long isr, unsigned long ifa, #endif break; case 35: siginfo.si_code = TRAP_BRANCH; ifa = 0; break; - case 36: siginfo.si_code = TRAP_TRACE; ifa = 0; break; + case 36: + if (notify_die(DIE_SS, "ss", ®s, vector, + vector, SIGTRAP) == NOTIFY_STOP) + return; + siginfo.si_code = TRAP_TRACE; ifa = 0; break; } siginfo.si_signo = SIGTRAP; siginfo.si_errno = 0; diff --git a/arch/ia64/mm/discontig.c b/arch/ia64/mm/discontig.c index c0071092939..f3fd528ead3 100644 --- a/arch/ia64/mm/discontig.c +++ b/arch/ia64/mm/discontig.c @@ -560,14 +560,15 @@ void show_mem(void) int shared = 0, cached = 0, reserved = 0; printk("Node ID: %d\n", pgdat->node_id); for(i = 0; i < pgdat->node_spanned_pages; i++) { + struct page *page = pgdat_page_nr(pgdat, i); if (!ia64_pfn_valid(pgdat->node_start_pfn+i)) continue; - if (PageReserved(pgdat->node_mem_map+i)) + if (PageReserved(page)) reserved++; - else if (PageSwapCache(pgdat->node_mem_map+i)) + else if (PageSwapCache(page)) cached++; - else if (page_count(pgdat->node_mem_map+i)) - shared += page_count(pgdat->node_mem_map+i)-1; + else if (page_count(page)) + shared += page_count(page)-1; } total_present += present; total_reserved += reserved; diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c index 4174ec999dd..ff62551eb3a 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -14,6 +14,7 @@ #include <asm/processor.h> #include <asm/system.h> #include <asm/uaccess.h> +#include <asm/kdebug.h> extern void die (char *, struct pt_regs *, long); @@ -102,6 +103,13 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re goto bad_area_no_up; #endif + /* + * This is to handle the kprobes on user space access instructions + */ + if (notify_die(DIE_PAGE_FAULT, "page fault", regs, code, TRAP_BRKPT, + SIGSEGV) == NOTIFY_STOP) + return; + down_read(&mm->mmap_sem); vma = find_vma_prev(mm, address, &prev_vma); diff --git a/arch/ia64/sn/kernel/xpc.h b/arch/ia64/sn/kernel/xpc.h index 1a0aed8490d..d0ee635daf2 100644 --- a/arch/ia64/sn/kernel/xpc.h +++ b/arch/ia64/sn/kernel/xpc.h @@ -87,7 +87,7 @@ struct xpc_rsvd_page { u8 partid; /* partition ID from SAL */ u8 version; u8 pad[6]; /* pad to u64 align */ - u64 vars_pa; + volatile u64 vars_pa; u64 part_nasids[XP_NASID_MASK_WORDS] ____cacheline_aligned; u64 mach_nasids[XP_NASID_MASK_WORDS] ____cacheline_aligned; }; @@ -138,7 +138,7 @@ struct xpc_vars { * occupies half a cacheline. */ struct xpc_vars_part { - u64 magic; + volatile u64 magic; u64 openclose_args_pa; /* physical address of open and close args */ u64 GPs_pa; /* physical address of Get/Put values */ @@ -185,8 +185,8 @@ struct xpc_vars_part { * Define a Get/Put value pair (pointers) used with a message queue. */ struct xpc_gp { - s64 get; /* Get value */ - s64 put; /* Put value */ + volatile s64 get; /* Get value */ + volatile s64 put; /* Put value */ }; #define XPC_GP_SIZE \ @@ -231,7 +231,7 @@ struct xpc_openclose_args { */ struct xpc_notify { struct semaphore sema; /* notify semaphore */ - u8 type; /* type of notification */ + volatile u8 type; /* type of notification */ /* the following two fields are only used if type == XPC_N_CALL */ xpc_notify_func func; /* user's notify function */ @@ -439,7 +439,7 @@ struct xpc_partition { /* XPC infrastructure referencing and teardown control */ - u8 setup_state; /* infrastructure setup state */ + volatile u8 setup_state; /* infrastructure setup state */ wait_queue_head_t teardown_wq; /* kthread waiting to teardown infra */ atomic_t references; /* #of references to infrastructure */ diff --git a/arch/ia64/sn/kernel/xpc_channel.c b/arch/ia64/sn/kernel/xpc_channel.c index 0bf6fbcc46d..6d02dac8056 100644 --- a/arch/ia64/sn/kernel/xpc_channel.c +++ b/arch/ia64/sn/kernel/xpc_channel.c @@ -209,7 +209,7 @@ xpc_setup_infrastructure(struct xpc_partition *part) * With the setting of the partition setup_state to XPC_P_SETUP, we're * declaring that this partition is ready to go. */ - (volatile u8) part->setup_state = XPC_P_SETUP; + part->setup_state = XPC_P_SETUP; /* @@ -227,7 +227,7 @@ xpc_setup_infrastructure(struct xpc_partition *part) xpc_vars_part[partid].IPI_phys_cpuid = cpu_physical_id(smp_processor_id()); xpc_vars_part[partid].nchannels = part->nchannels; - (volatile u64) xpc_vars_part[partid].magic = XPC_VP_MAGIC1; + xpc_vars_part[partid].magic = XPC_VP_MAGIC1; return xpcSuccess; } @@ -355,7 +355,7 @@ xpc_pull_remote_vars_part(struct xpc_partition *part) /* let the other side know that we've pulled their variables */ - (volatile u64) xpc_vars_part[partid].magic = XPC_VP_MAGIC2; + xpc_vars_part[partid].magic = XPC_VP_MAGIC2; } if (pulled_entry->magic == XPC_VP_MAGIC1) { @@ -1183,7 +1183,7 @@ xpc_process_msg_IPI(struct xpc_partition *part, int ch_number) */ xpc_clear_local_msgqueue_flags(ch); - (volatile s64) ch->w_remote_GP.get = ch->remote_GP.get; + ch->w_remote_GP.get = ch->remote_GP.get; dev_dbg(xpc_chan, "w_remote_GP.get changed to %ld, partid=%d, " "channel=%d\n", ch->w_remote_GP.get, ch->partid, @@ -1211,7 +1211,7 @@ xpc_process_msg_IPI(struct xpc_partition *part, int ch_number) */ xpc_clear_remote_msgqueue_flags(ch); - (volatile s64) ch->w_remote_GP.put = ch->remote_GP.put; + ch->w_remote_GP.put = ch->remote_GP.put; dev_dbg(xpc_chan, "w_remote_GP.put changed to %ld, partid=%d, " "channel=%d\n", ch->w_remote_GP.put, ch->partid, @@ -1875,7 +1875,7 @@ xpc_send_msg(struct xpc_channel *ch, struct xpc_msg *msg, u8 notify_type, notify = &ch->notify_queue[msg_number % ch->local_nentries]; notify->func = func; notify->key = key; - (volatile u8) notify->type = notify_type; + notify->type = notify_type; // >>> is a mb() needed here? diff --git a/arch/ia64/sn/kernel/xpc_partition.c b/arch/ia64/sn/kernel/xpc_partition.c index cd7ed73f0e7..578265ea9e6 100644 --- a/arch/ia64/sn/kernel/xpc_partition.c +++ b/arch/ia64/sn/kernel/xpc_partition.c @@ -253,7 +253,7 @@ xpc_rsvd_page_init(void) * This signifies to the remote partition that our reserved * page is initialized. */ - (volatile u64) rp->vars_pa = __pa(xpc_vars); + rp->vars_pa = __pa(xpc_vars); return rp; } diff --git a/arch/m32r/Kconfig b/arch/m32r/Kconfig index 64c133344af..42ca8a39798 100644 --- a/arch/m32r/Kconfig +++ b/arch/m32r/Kconfig @@ -172,11 +172,13 @@ config NOHIGHMEM bool default y -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool "Internal RAM Support" depends on CHIP_M32700 || CHIP_M32102 || CHIP_VDEC2 || CHIP_OPSP default y +source "mm/Kconfig" + config IRAM_START hex "Internal memory start address (hex)" default "00f00000" diff --git a/arch/m32r/kernel/signal.c b/arch/m32r/kernel/signal.c index 50311eb07a2..5aef7e406ef 100644 --- a/arch/m32r/kernel/signal.c +++ b/arch/m32r/kernel/signal.c @@ -371,10 +371,8 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset) if (!user_mode(regs)) return 1; - if (current->flags & PF_FREEZE) { - refrigerator(0); + if (try_to_freeze()) goto no_signal; - } if (!oldset) oldset = ¤t->blocked; diff --git a/arch/m32r/mm/init.c b/arch/m32r/mm/init.c index bc423d838fb..d9a40b1fe8b 100644 --- a/arch/m32r/mm/init.c +++ b/arch/m32r/mm/init.c @@ -49,7 +49,7 @@ void show_mem(void) printk("Free swap: %6ldkB\n",nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { for (i = 0; i < pgdat->node_spanned_pages; ++i) { - page = pgdat->node_mem_map + i; + page = pgdat_page_nr(pgdat, i); total++; if (PageHighMem(page)) highmem++; @@ -152,7 +152,7 @@ int __init reservedpages_count(void) reservedpages = 0; for_each_online_node(nid) for (i = 0 ; i < MAX_LOW_PFN(nid) - START_PFN(nid) ; i++) - if (PageReserved(NODE_DATA(nid)->node_mem_map + i)) + if (PageReserved(nid_page_nr(nid, i))) reservedpages++; return reservedpages; diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig index d0713c7d9f0..691a2469ff3 100644 --- a/arch/m68k/Kconfig +++ b/arch/m68k/Kconfig @@ -357,6 +357,8 @@ config 060_WRITETHROUGH is hardwired on. The 53c710 SCSI driver is known to suffer from this problem. +source "mm/Kconfig" + endmenu menu "General setup" diff --git a/arch/m68knommu/Kconfig b/arch/m68knommu/Kconfig index e729bd28062..dbfcdc8e608 100644 --- a/arch/m68knommu/Kconfig +++ b/arch/m68knommu/Kconfig @@ -532,6 +532,8 @@ config ROMKERNEL endchoice +source "mm/Kconfig" + endmenu config ISA_DMA_API diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index ab9944693f1..bd9de7b00c0 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -492,7 +492,7 @@ config SGI_SN0_N_MODE which allows for more memory. Your system is most probably running in M-Mode, so you should say N here. -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool default y if SGI_IP27 help @@ -1416,6 +1416,12 @@ config HIGHMEM bool "High Memory Support" depends on MIPS32 && (CPU_R3000 || CPU_SB1 || CPU_R7000 || CPU_RM9000 || CPU_R10000) && !(MACH_DECSTATION || MOMENCO_JAGUAR_ATX) +config ARCH_FLATMEM_ENABLE + def_bool y + depends on !NUMA + +source "mm/Kconfig" + config SMP bool "Multi-Processing support" depends on CPU_RM9000 || (SIBYTE_SB1250 && !SIBYTE_STANDALONE) || SGI_IP27 diff --git a/arch/mips/configs/ip27_defconfig b/arch/mips/configs/ip27_defconfig index 13472292d0e..b5bab3a42fc 100644 --- a/arch/mips/configs/ip27_defconfig +++ b/arch/mips/configs/ip27_defconfig @@ -82,7 +82,7 @@ CONFIG_STOP_MACHINE=y # CONFIG_SGI_IP22 is not set CONFIG_SGI_IP27=y # CONFIG_SGI_SN0_N_MODE is not set -CONFIG_DISCONTIGMEM=y +CONFIG_ARCH_DISCONTIGMEM_ENABLE=y CONFIG_NUMA=y # CONFIG_MAPPED_KERNEL is not set # CONFIG_REPLICATE_KTEXT is not set diff --git a/arch/mips/kernel/setup.c b/arch/mips/kernel/setup.c index 6018ca25ace..3a240e3e004 100644 --- a/arch/mips/kernel/setup.c +++ b/arch/mips/kernel/setup.c @@ -33,6 +33,7 @@ #include <linux/root_dev.h> #include <linux/highmem.h> #include <linux/console.h> +#include <linux/mmzone.h> #include <asm/addrspace.h> #include <asm/bootinfo.h> @@ -356,6 +357,8 @@ static inline void bootmem_init(void) } #endif + memory_present(0, first_usable_pfn, max_low_pfn); + /* Initialize the boot-time allocator with low memory only. */ bootmap_size = init_bootmem(first_usable_pfn, max_low_pfn); @@ -557,6 +560,7 @@ void __init setup_arch(char **cmdline_p) parse_cmdline_early(); bootmem_init(); + sparse_init(); paging_init(); resource_init(); } diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 73843c52877..9c9a271c8a3 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -128,7 +128,7 @@ static void __init fixrange_init(unsigned long start, unsigned long end, #endif /* CONFIG_MIPS64 */ #endif /* CONFIG_HIGHMEM */ -#ifndef CONFIG_DISCONTIGMEM +#ifndef CONFIG_NEED_MULTIPLE_NODES extern void pagetable_init(void); void __init paging_init(void) @@ -253,7 +253,7 @@ void __init mem_init(void) initsize >> 10, (unsigned long) (totalhigh_pages << (PAGE_SHIFT-10))); } -#endif /* !CONFIG_DISCONTIGMEM */ +#endif /* !CONFIG_NEED_MULTIPLE_NODES */ #ifdef CONFIG_BLK_DEV_INITRD void free_initrd_mem(unsigned long start, unsigned long end) diff --git a/arch/mips/mm/pgtable.c b/arch/mips/mm/pgtable.c index 3b88fdeef32..3fe94202da8 100644 --- a/arch/mips/mm/pgtable.c +++ b/arch/mips/mm/pgtable.c @@ -5,7 +5,7 @@ void show_mem(void) { -#ifndef CONFIG_DISCONTIGMEM /* XXX(hch): later.. */ +#ifndef CONFIG_NEED_MULTIPLE_NODES /* XXX(hch): later.. */ int pfn, total = 0, reserved = 0; int shared = 0, cached = 0; int highmem = 0; diff --git a/arch/mips/sgi-ip27/ip27-memory.c b/arch/mips/sgi-ip27/ip27-memory.c index 0a44a98d7ad..a160d04f7db 100644 --- a/arch/mips/sgi-ip27/ip27-memory.c +++ b/arch/mips/sgi-ip27/ip27-memory.c @@ -549,9 +549,8 @@ void __init mem_init(void) */ numslots = node_getlastslot(node); for (slot = 1; slot <= numslots; slot++) { - p = NODE_DATA(node)->node_mem_map + - (slot_getbasepfn(node, slot) - - slot_getbasepfn(node, 0)); + p = nid_page_nr(node, slot_getbasepfn(node, slot) - + slot_getbasepfn(node, 0)); /* * Free valid memory in current slot. diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig index e7e7c56fc21..ce327c799b4 100644 --- a/arch/parisc/Kconfig +++ b/arch/parisc/Kconfig @@ -148,7 +148,7 @@ config HOTPLUG_CPU default y if SMP select HOTPLUG -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool "Discontiguous memory support (EXPERIMENTAL)" depends on EXPERIMENTAL help @@ -157,6 +157,8 @@ config DISCONTIGMEM or have huge holes in the physical address space for other reasons. See <file:Documentation/vm/numa> for more. +source "mm/Kconfig" + config PREEMPT bool # bool "Preemptible Kernel" diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c index cac37589e35..2886ad70db4 100644 --- a/arch/parisc/mm/init.c +++ b/arch/parisc/mm/init.c @@ -506,7 +506,7 @@ void show_mem(void) for (j = node_start_pfn(i); j < node_end_pfn(i); j++) { struct page *p; - p = node_mem_map(i) + j - node_start_pfn(i); + p = nid_page_nr(i, j) - node_start_pfn(i); total++; if (PageReserved(p)) diff --git a/arch/ppc/Kconfig b/arch/ppc/Kconfig index 10162b187bc..a7835cd3f51 100644 --- a/arch/ppc/Kconfig +++ b/arch/ppc/Kconfig @@ -88,6 +88,9 @@ config 8xx depends on BROKEN bool "8xx" +config E200 + bool "e200" + config E500 bool "e500" @@ -98,12 +101,12 @@ config PPC_FPU config BOOKE bool - depends on E500 + depends on E200 || E500 default y config FSL_BOOKE bool - depends on E500 + depends on E200 || E500 default y config PTE_64BIT @@ -141,16 +144,16 @@ config ALTIVEC config SPE bool "SPE Support" - depends on E500 + depends on E200 || E500 ---help--- This option enables kernel support for the Signal Processing Extensions (SPE) to the PowerPC processor. The kernel currently supports saving and restoring SPE registers, and turning on the 'spe enable' bit so user processes can execute SPE instructions. - This option is only usefully if you have a processor that supports + This option is only useful if you have a processor that supports SPE (e500, otherwise known as 85xx series), but does not have any - affect on a non-spe cpu (it does, however add code to the kernel). + effect on a non-spe cpu (it does, however add code to the kernel). If in doubt, say Y here. @@ -200,7 +203,7 @@ config TAU_AVERAGE config MATH_EMULATION bool "Math emulation" - depends on 4xx || 8xx || E500 + depends on 4xx || 8xx || E200 || E500 ---help--- Some PowerPC chips designed for embedded applications do not have a floating-point unit and therefore do not implement the @@ -214,6 +217,26 @@ config MATH_EMULATION here. Saying Y here will not hurt performance (on any machine) but will increase the size of the kernel. +config KEXEC + bool "kexec system call (EXPERIMENTAL)" + depends on EXPERIMENTAL + help + kexec is a system call that implements the ability to shutdown your + current kernel, and to start another kernel. It is like a reboot + but it is indepedent of the system firmware. And like a reboot + you can start any kernel with it, not just Linux. + + The name comes from the similiarity to the exec system call. + + It is an ongoing process to be certain the hardware in a machine + is properly shutdown, so do not be surprised if this code does not + initially work for you. It may help to enable device hotplugging + support. As of this writing the exact hardware interface is + strongly in flux, so no good recommendation can be made. + + In the GameCube implementation, kexec allows you to load and + run DOL files, including kernel and homebrew DOLs. + source "drivers/cpufreq/Kconfig" config CPU_FREQ_PMAC @@ -254,7 +277,7 @@ config PPC_STD_MMU config NOT_COHERENT_CACHE bool - depends on 4xx || 8xx + depends on 4xx || 8xx || E200 default y endmenu @@ -905,6 +928,8 @@ config PREEMPT config HIGHMEM bool "High memory support" +source "mm/Kconfig" + source "fs/Kconfig.binfmt" config PROC_DEVICETREE diff --git a/arch/ppc/Kconfig.debug b/arch/ppc/Kconfig.debug index d2e1eea8e8e..e16c7710d4b 100644 --- a/arch/ppc/Kconfig.debug +++ b/arch/ppc/Kconfig.debug @@ -66,7 +66,7 @@ config SERIAL_TEXT_DEBUG config PPC_OCP bool - depends on IBM_OCP || FSL_OCP || XILINX_OCP + depends on IBM_OCP || XILINX_OCP default y endmenu diff --git a/arch/ppc/Makefile b/arch/ppc/Makefile index 0432a25b473..f9b0d778dd8 100644 --- a/arch/ppc/Makefile +++ b/arch/ppc/Makefile @@ -29,7 +29,7 @@ CPP = $(CC) -E $(CFLAGS) CHECKFLAGS += -D__powerpc__ -ifndef CONFIG_E500 +ifndef CONFIG_FSL_BOOKE CFLAGS += -mstring endif @@ -38,6 +38,7 @@ cpu-as-$(CONFIG_4xx) += -Wa,-m405 cpu-as-$(CONFIG_6xx) += -Wa,-maltivec cpu-as-$(CONFIG_POWER4) += -Wa,-maltivec cpu-as-$(CONFIG_E500) += -Wa,-me500 +cpu-as-$(CONFIG_E200) += -Wa,-me200 AFLAGS += $(cpu-as-y) CFLAGS += $(cpu-as-y) diff --git a/arch/ppc/boot/openfirmware/chrpmain.c b/arch/ppc/boot/openfirmware/chrpmain.c index 6fb4f738728..effe4a0624b 100644 --- a/arch/ppc/boot/openfirmware/chrpmain.c +++ b/arch/ppc/boot/openfirmware/chrpmain.c @@ -39,7 +39,7 @@ char *avail_high; #define SCRATCH_SIZE (128 << 10) -static char scratch[SCRATCH_SIZE]; /* 1MB of scratch space for gunzip */ +static char scratch[SCRATCH_SIZE]; /* 128k of scratch space for gunzip */ typedef void (*kernel_start_t)(int, int, void *, unsigned int, unsigned int); diff --git a/arch/ppc/boot/simple/misc.c b/arch/ppc/boot/simple/misc.c index ab0f9902cb6..e02de5b467a 100644 --- a/arch/ppc/boot/simple/misc.c +++ b/arch/ppc/boot/simple/misc.c @@ -222,7 +222,7 @@ decompress_kernel(unsigned long load_addr, int num_words, unsigned long cksum) puts("\n"); puts("Uncompressing Linux..."); - gunzip(0x0, 0x400000, zimage_start, &zimage_size); + gunzip(NULL, 0x400000, zimage_start, &zimage_size); puts("done.\n"); /* get the bi_rec address */ diff --git a/arch/ppc/boot/simple/mpc10x_memory.c b/arch/ppc/boot/simple/mpc10x_memory.c index 977daedc14c..20d92a34ceb 100644 --- a/arch/ppc/boot/simple/mpc10x_memory.c +++ b/arch/ppc/boot/simple/mpc10x_memory.c @@ -33,7 +33,7 @@ #define MPC10X_PCI_OP(rw, size, type, op, mask) \ static void \ -mpc10x_##rw##_config_##size(unsigned int *cfg_addr, \ +mpc10x_##rw##_config_##size(unsigned int __iomem *cfg_addr, \ unsigned int *cfg_data, int devfn, int offset, \ type val) \ { \ diff --git a/arch/ppc/kernel/Makefile b/arch/ppc/kernel/Makefile index b284451802c..b1457a8a9c0 100644 --- a/arch/ppc/kernel/Makefile +++ b/arch/ppc/kernel/Makefile @@ -26,7 +26,10 @@ obj-$(CONFIG_KGDB) += ppc-stub.o obj-$(CONFIG_SMP) += smp.o smp-tbsync.o obj-$(CONFIG_TAU) += temp.o obj-$(CONFIG_ALTIVEC) += vecemu.o vector.o +ifndef CONFIG_E200 obj-$(CONFIG_FSL_BOOKE) += perfmon_fsl_booke.o +endif +obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o ifndef CONFIG_MATH_EMULATION obj-$(CONFIG_8xx) += softemu8xx.o diff --git a/arch/ppc/kernel/cputable.c b/arch/ppc/kernel/cputable.c index 01c226008db..50936cda0af 100644 --- a/arch/ppc/kernel/cputable.c +++ b/arch/ppc/kernel/cputable.c @@ -903,7 +903,30 @@ struct cpu_spec cpu_specs[] = { .dcache_bsize = 32, }, #endif /* CONFIG_44x */ -#ifdef CONFIG_E500 +#ifdef CONFIG_FSL_BOOKE + { /* e200z5 */ + .pvr_mask = 0xfff00000, + .pvr_value = 0x81000000, + .cpu_name = "e200z5", + /* xxx - galak: add CPU_FTR_MAYBE_CAN_DOZE */ + .cpu_features = CPU_FTR_USE_TB, + .cpu_user_features = PPC_FEATURE_32 | + PPC_FEATURE_HAS_MMU | PPC_FEATURE_HAS_EFP_SINGLE | + PPC_FEATURE_UNIFIED_CACHE, + .dcache_bsize = 32, + }, + { /* e200z6 */ + .pvr_mask = 0xfff00000, + .pvr_value = 0x81100000, + .cpu_name = "e200z6", + /* xxx - galak: add CPU_FTR_MAYBE_CAN_DOZE */ + .cpu_features = CPU_FTR_USE_TB, + .cpu_user_features = PPC_FEATURE_32 | + PPC_FEATURE_HAS_MMU | PPC_FEATURE_SPE_COMP | + PPC_FEATURE_HAS_EFP_SINGLE | + PPC_FEATURE_UNIFIED_CACHE, + .dcache_bsize = 32, + }, { /* e500 */ .pvr_mask = 0xffff0000, .pvr_value = 0x80200000, diff --git a/arch/ppc/kernel/entry.S b/arch/ppc/kernel/entry.S index 8377b6ca26d..d4df68629cc 100644 --- a/arch/ppc/kernel/entry.S +++ b/arch/ppc/kernel/entry.S @@ -60,6 +60,11 @@ mcheck_transfer_to_handler: TRANSFER_TO_HANDLER_EXC_LEVEL(MCHECK) b transfer_to_handler_full + .globl debug_transfer_to_handler +debug_transfer_to_handler: + TRANSFER_TO_HANDLER_EXC_LEVEL(DEBUG) + b transfer_to_handler_full + .globl crit_transfer_to_handler crit_transfer_to_handler: TRANSFER_TO_HANDLER_EXC_LEVEL(CRIT) @@ -835,6 +840,10 @@ ret_from_crit_exc: RET_FROM_EXC_LEVEL(SPRN_CSRR0, SPRN_CSRR1, RFCI) #ifdef CONFIG_BOOKE + .globl ret_from_debug_exc +ret_from_debug_exc: + RET_FROM_EXC_LEVEL(SPRN_DSRR0, SPRN_DSRR1, RFDI) + .globl ret_from_mcheck_exc ret_from_mcheck_exc: RET_FROM_EXC_LEVEL(SPRN_MCSRR0, SPRN_MCSRR1, RFMCI) diff --git a/arch/ppc/kernel/head_booke.h b/arch/ppc/kernel/head_booke.h index 9c50f9d2657..9342acf12e7 100644 --- a/arch/ppc/kernel/head_booke.h +++ b/arch/ppc/kernel/head_booke.h @@ -49,6 +49,7 @@ * * On 40x critical is the only additional level * On 44x/e500 we have critical and machine check + * On e200 we have critical and debug (machine check occurs via critical) * * Additionally we reserve a SPRG for each priority level so we can free up a * GPR to use as the base for indirect access to the exception stacks. This @@ -60,12 +61,16 @@ /* CRIT_SPRG only used in critical exception handling */ #define CRIT_SPRG SPRN_SPRG2 -/* MCHECK_SPRG only used in critical exception handling */ +/* MCHECK_SPRG only used in machine check exception handling */ #define MCHECK_SPRG SPRN_SPRG6W #define MCHECK_STACK_TOP (exception_stack_top - 4096) #define CRIT_STACK_TOP (exception_stack_top) +/* only on e200 for now */ +#define DEBUG_STACK_TOP (exception_stack_top - 4096) +#define DEBUG_SPRG SPRN_SPRG6W + #ifdef CONFIG_SMP #define BOOKE_LOAD_EXC_LEVEL_STACK(level) \ mfspr r8,SPRN_PIR; \ @@ -124,6 +129,8 @@ #define CRITICAL_EXCEPTION_PROLOG \ EXC_LEVEL_EXCEPTION_PROLOG(CRIT, SPRN_CSRR0, SPRN_CSRR1) +#define DEBUG_EXCEPTION_PROLOG \ + EXC_LEVEL_EXCEPTION_PROLOG(DEBUG, SPRN_DSRR0, SPRN_DSRR1) #define MCHECK_EXCEPTION_PROLOG \ EXC_LEVEL_EXCEPTION_PROLOG(MCHECK, SPRN_MCSRR0, SPRN_MCSRR1) @@ -205,6 +212,60 @@ label: * save (and later restore) the MSR via SPRN_CSRR1, which will still have * the MSR_DE bit set. */ +#ifdef CONFIG_E200 +#define DEBUG_EXCEPTION \ + START_EXCEPTION(Debug); \ + DEBUG_EXCEPTION_PROLOG; \ + \ + /* \ + * If there is a single step or branch-taken exception in an \ + * exception entry sequence, it was probably meant to apply to \ + * the code where the exception occurred (since exception entry \ + * doesn't turn off DE automatically). We simulate the effect \ + * of turning off DE on entry to an exception handler by turning \ + * off DE in the CSRR1 value and clearing the debug status. \ + */ \ + mfspr r10,SPRN_DBSR; /* check single-step/branch taken */ \ + andis. r10,r10,DBSR_IC@h; \ + beq+ 2f; \ + \ + lis r10,KERNELBASE@h; /* check if exception in vectors */ \ + ori r10,r10,KERNELBASE@l; \ + cmplw r12,r10; \ + blt+ 2f; /* addr below exception vectors */ \ + \ + lis r10,Debug@h; \ + ori r10,r10,Debug@l; \ + cmplw r12,r10; \ + bgt+ 2f; /* addr above exception vectors */ \ + \ + /* here it looks like we got an inappropriate debug exception. */ \ +1: rlwinm r9,r9,0,~MSR_DE; /* clear DE in the CDRR1 value */ \ + lis r10,DBSR_IC@h; /* clear the IC event */ \ + mtspr SPRN_DBSR,r10; \ + /* restore state and get out */ \ + lwz r10,_CCR(r11); \ + lwz r0,GPR0(r11); \ + lwz r1,GPR1(r11); \ + mtcrf 0x80,r10; \ + mtspr SPRN_DSRR0,r12; \ + mtspr SPRN_DSRR1,r9; \ + lwz r9,GPR9(r11); \ + lwz r12,GPR12(r11); \ + mtspr DEBUG_SPRG,r8; \ + BOOKE_LOAD_EXC_LEVEL_STACK(DEBUG); /* r8 points to the debug stack */ \ + lwz r10,GPR10-INT_FRAME_SIZE(r8); \ + lwz r11,GPR11-INT_FRAME_SIZE(r8); \ + mfspr r8,DEBUG_SPRG; \ + \ + RFDI; \ + b .; \ + \ + /* continue normal handling for a critical exception... */ \ +2: mfspr r4,SPRN_DBSR; \ + addi r3,r1,STACK_FRAME_OVERHEAD; \ + EXC_XFER_TEMPLATE(DebugException, 0x2002, (MSR_KERNEL & ~(MSR_ME|MSR_DE|MSR_CE)), NOCOPY, debug_transfer_to_handler, ret_from_debug_exc) +#else #define DEBUG_EXCEPTION \ START_EXCEPTION(Debug); \ CRITICAL_EXCEPTION_PROLOG; \ @@ -257,6 +318,7 @@ label: 2: mfspr r4,SPRN_DBSR; \ addi r3,r1,STACK_FRAME_OVERHEAD; \ EXC_XFER_TEMPLATE(DebugException, 0x2002, (MSR_KERNEL & ~(MSR_ME|MSR_DE|MSR_CE)), NOCOPY, crit_transfer_to_handler, ret_from_crit_exc) +#endif #define INSTRUCTION_STORAGE_EXCEPTION \ START_EXCEPTION(InstructionStorage) \ diff --git a/arch/ppc/kernel/head_fsl_booke.S b/arch/ppc/kernel/head_fsl_booke.S index ce36e88ba62..eb804b7a3cb 100644 --- a/arch/ppc/kernel/head_fsl_booke.S +++ b/arch/ppc/kernel/head_fsl_booke.S @@ -102,6 +102,7 @@ invstr: mflr r6 /* Make it accessible */ or r7,r7,r4 mtspr SPRN_MAS6,r7 tlbsx 0,r6 /* search MSR[IS], SPID=PID0 */ +#ifndef CONFIG_E200 mfspr r7,SPRN_MAS1 andis. r7,r7,MAS1_VALID@h bne match_TLB @@ -118,6 +119,7 @@ invstr: mflr r6 /* Make it accessible */ or r7,r7,r4 mtspr SPRN_MAS6,r7 tlbsx 0,r6 /* Fall through, we had to match */ +#endif match_TLB: mfspr r7,SPRN_MAS0 rlwinm r3,r7,16,20,31 /* Extract MAS0(Entry) */ @@ -196,8 +198,10 @@ skpinv: addi r6,r6,1 /* Increment */ /* 4. Clear out PIDs & Search info */ li r6,0 mtspr SPRN_PID0,r6 +#ifndef CONFIG_E200 mtspr SPRN_PID1,r6 mtspr SPRN_PID2,r6 +#endif mtspr SPRN_MAS6,r6 /* 5. Invalidate mapping we started in */ @@ -277,7 +281,9 @@ skpinv: addi r6,r6,1 /* Increment */ SET_IVOR(32, SPEUnavailable); SET_IVOR(33, SPEFloatingPointData); SET_IVOR(34, SPEFloatingPointRound); +#ifndef CONFIG_E200 SET_IVOR(35, PerformanceMonitor); +#endif /* Establish the interrupt vector base */ lis r4,interrupt_base@h /* IVPR only uses the high 16-bits */ @@ -285,6 +291,9 @@ skpinv: addi r6,r6,1 /* Increment */ /* Setup the defaults for TLB entries */ li r2,(MAS4_TSIZED(BOOKE_PAGESZ_4K))@l +#ifdef CONFIG_E200 + oris r2,r2,MAS4_TLBSELD(1)@h +#endif mtspr SPRN_MAS4, r2 #if 0 @@ -293,6 +302,12 @@ skpinv: addi r6,r6,1 /* Increment */ oris r2,r2,HID0_DOZE@h mtspr SPRN_HID0, r2 #endif +#ifdef CONFIG_E200 + /* enable dedicated debug exception handling resources (Debug APU) */ + mfspr r2,SPRN_HID0 + ori r2,r2,HID0_DAPUEN@l + mtspr SPRN_HID0,r2 +#endif #if !defined(CONFIG_BDI_SWITCH) /* @@ -414,7 +429,12 @@ interrupt_base: CRITICAL_EXCEPTION(0x0100, CriticalInput, UnknownException) /* Machine Check Interrupt */ +#ifdef CONFIG_E200 + /* no RFMCI, MCSRRs on E200 */ + CRITICAL_EXCEPTION(0x0200, MachineCheck, MachineCheckException) +#else MCHECK_EXCEPTION(0x0200, MachineCheck, MachineCheckException) +#endif /* Data Storage Interrupt */ START_EXCEPTION(DataStorage) @@ -520,8 +540,13 @@ interrupt_base: #ifdef CONFIG_PPC_FPU FP_UNAVAILABLE_EXCEPTION #else +#ifdef CONFIG_E200 + /* E200 treats 'normal' floating point instructions as FP Unavail exception */ + EXCEPTION(0x0800, FloatingPointUnavailable, ProgramCheckException, EXC_XFER_EE) +#else EXCEPTION(0x0800, FloatingPointUnavailable, UnknownException, EXC_XFER_EE) #endif +#endif /* System Call Interrupt */ START_EXCEPTION(SystemCall) @@ -691,6 +716,7 @@ interrupt_base: /* * Local functions */ + /* * Data TLB exceptions will bail out to this point * if they can't resolve the lightweight TLB fault. @@ -761,6 +787,31 @@ END_FTR_SECTION_IFSET(CPU_FTR_BIG_PHYS) 2: rlwimi r11, r12, 0, 20, 31 /* Extract RPN from PTE and merge with perms */ mtspr SPRN_MAS3, r11 #endif +#ifdef CONFIG_E200 + /* Round robin TLB1 entries assignment */ + mfspr r12, SPRN_MAS0 + + /* Extract TLB1CFG(NENTRY) */ + mfspr r11, SPRN_TLB1CFG + andi. r11, r11, 0xfff + + /* Extract MAS0(NV) */ + andi. r13, r12, 0xfff + addi r13, r13, 1 + cmpw 0, r13, r11 + addi r12, r12, 1 + + /* check if we need to wrap */ + blt 7f + + /* wrap back to first free tlbcam entry */ + lis r13, tlbcam_index@ha + lwz r13, tlbcam_index@l(r13) + rlwimi r12, r13, 0, 20, 31 +7: + mtspr SPRN_MAS0,r12 +#endif /* CONFIG_E200 */ + tlbwe /* Done...restore registers and get out of here. */ diff --git a/arch/ppc/kernel/machine_kexec.c b/arch/ppc/kernel/machine_kexec.c new file mode 100644 index 00000000000..84d65a87191 --- /dev/null +++ b/arch/ppc/kernel/machine_kexec.c @@ -0,0 +1,118 @@ +/* + * machine_kexec.c - handle transition of Linux booting another kernel + * Copyright (C) 2002-2003 Eric Biederman <ebiederm@xmission.com> + * + * GameCube/ppc32 port Copyright (C) 2004 Albert Herranz + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include <linux/mm.h> +#include <linux/kexec.h> +#include <linux/delay.h> +#include <linux/reboot.h> +#include <asm/pgtable.h> +#include <asm/pgalloc.h> +#include <asm/mmu_context.h> +#include <asm/io.h> +#include <asm/hw_irq.h> +#include <asm/cacheflush.h> +#include <asm/machdep.h> + +typedef NORET_TYPE void (*relocate_new_kernel_t)( + unsigned long indirection_page, + unsigned long reboot_code_buffer, + unsigned long start_address) ATTRIB_NORET; + +const extern unsigned char relocate_new_kernel[]; +const extern unsigned int relocate_new_kernel_size; + +void machine_shutdown(void) +{ + if (ppc_md.machine_shutdown) + ppc_md.machine_shutdown(); +} + +void machine_crash_shutdown(struct pt_regs *regs) +{ + if (ppc_md.machine_crash_shutdown) + ppc_md.machine_crash_shutdown(); +} + +/* + * Do what every setup is needed on image and the + * reboot code buffer to allow us to avoid allocations + * later. + */ +int machine_kexec_prepare(struct kimage *image) +{ + if (ppc_md.machine_kexec_prepare) + return ppc_md.machine_kexec_prepare(image); + /* + * Fail if platform doesn't provide its own machine_kexec_prepare + * implementation. + */ + return -ENOSYS; +} + +void machine_kexec_cleanup(struct kimage *image) +{ + if (ppc_md.machine_kexec_cleanup) + ppc_md.machine_kexec_cleanup(image); +} + +/* + * Do not allocate memory (or fail in any way) in machine_kexec(). + * We are past the point of no return, committed to rebooting now. + */ +NORET_TYPE void machine_kexec(struct kimage *image) +{ + if (ppc_md.machine_kexec) + ppc_md.machine_kexec(image); + else { + /* + * Fall back to normal restart if platform doesn't provide + * its own kexec function, and user insist to kexec... + */ + machine_restart(NULL); + } + for(;;); +} + +/* + * This is a generic machine_kexec function suitable at least for + * non-OpenFirmware embedded platforms. + * It merely copies the image relocation code to the control page and + * jumps to it. + * A platform specific function may just call this one. + */ +void machine_kexec_simple(struct kimage *image) +{ + unsigned long page_list; + unsigned long reboot_code_buffer, reboot_code_buffer_phys; + relocate_new_kernel_t rnk; + + /* Interrupts aren't acceptable while we reboot */ + local_irq_disable(); + + page_list = image->head; + + /* we need both effective and real address here */ + reboot_code_buffer = + (unsigned long)page_address(image->control_code_page); + reboot_code_buffer_phys = virt_to_phys((void *)reboot_code_buffer); + + /* copy our kernel relocation code to the control code page */ + memcpy((void *)reboot_code_buffer, relocate_new_kernel, + relocate_new_kernel_size); + + flush_icache_range(reboot_code_buffer, + reboot_code_buffer + KEXEC_CONTROL_CODE_SIZE); + printk(KERN_INFO "Bye!\n"); + + /* now call it */ + rnk = (relocate_new_kernel_t) reboot_code_buffer; + (*rnk)(page_list, reboot_code_buffer_phys, image->start); +} + diff --git a/arch/ppc/kernel/misc.S b/arch/ppc/kernel/misc.S index 7329ef177a1..b6a63a49a23 100644 --- a/arch/ppc/kernel/misc.S +++ b/arch/ppc/kernel/misc.S @@ -593,6 +593,14 @@ _GLOBAL(flush_instruction_cache) iccci 0,r3 #endif #elif CONFIG_FSL_BOOKE +BEGIN_FTR_SECTION + mfspr r3,SPRN_L1CSR0 + ori r3,r3,L1CSR0_CFI|L1CSR0_CLFC + /* msync; isync recommended here */ + mtspr SPRN_L1CSR0,r3 + isync + blr +END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE) mfspr r3,SPRN_L1CSR1 ori r3,r3,L1CSR1_ICFI|L1CSR1_ICLFR mtspr SPRN_L1CSR1,r3 @@ -1436,7 +1444,7 @@ _GLOBAL(sys_call_table) .long sys_mq_timedreceive /* 265 */ .long sys_mq_notify .long sys_mq_getsetattr - .long sys_ni_syscall /* 268 reserved for sys_kexec_load */ + .long sys_kexec_load .long sys_add_key .long sys_request_key /* 270 */ .long sys_keyctl diff --git a/arch/ppc/kernel/perfmon.c b/arch/ppc/kernel/perfmon.c index 918f6b252e4..fa1dad96b83 100644 --- a/arch/ppc/kernel/perfmon.c +++ b/arch/ppc/kernel/perfmon.c @@ -36,7 +36,7 @@ /* A lock to regulate grabbing the interrupt */ DEFINE_SPINLOCK(perfmon_lock); -#ifdef CONFIG_FSL_BOOKE +#if defined (CONFIG_FSL_BOOKE) && !defined (CONFIG_E200) static void dummy_perf(struct pt_regs *regs) { unsigned int pmgc0 = mfpmr(PMRN_PMGC0); diff --git a/arch/ppc/kernel/relocate_kernel.S b/arch/ppc/kernel/relocate_kernel.S new file mode 100644 index 00000000000..7ff69c4af92 --- /dev/null +++ b/arch/ppc/kernel/relocate_kernel.S @@ -0,0 +1,123 @@ +/* + * relocate_kernel.S - put the kernel image in place to boot + * Copyright (C) 2002-2003 Eric Biederman <ebiederm@xmission.com> + * + * GameCube/ppc32 port Copyright (C) 2004 Albert Herranz + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include <asm/reg.h> +#include <asm/ppc_asm.h> +#include <asm/processor.h> + +#include <asm/kexec.h> + +#define PAGE_SIZE 4096 /* must be same value as in <asm/page.h> */ + + /* + * Must be relocatable PIC code callable as a C function. + */ + .globl relocate_new_kernel +relocate_new_kernel: + /* r3 = page_list */ + /* r4 = reboot_code_buffer */ + /* r5 = start_address */ + + li r0, 0 + + /* + * Set Machine Status Register to a known status, + * switch the MMU off and jump to 1: in a single step. + */ + + mr r8, r0 + ori r8, r8, MSR_RI|MSR_ME + mtspr SRR1, r8 + addi r8, r4, 1f - relocate_new_kernel + mtspr SRR0, r8 + sync + rfi + +1: + /* from this point address translation is turned off */ + /* and interrupts are disabled */ + + /* set a new stack at the bottom of our page... */ + /* (not really needed now) */ + addi r1, r4, KEXEC_CONTROL_CODE_SIZE - 8 /* for LR Save+Back Chain */ + stw r0, 0(r1) + + /* Do the copies */ + li r6, 0 /* checksum */ + mr r0, r3 + b 1f + +0: /* top, read another word for the indirection page */ + lwzu r0, 4(r3) + +1: + /* is it a destination page? (r8) */ + rlwinm. r7, r0, 0, 31, 31 /* IND_DESTINATION (1<<0) */ + beq 2f + + rlwinm r8, r0, 0, 0, 19 /* clear kexec flags, page align */ + b 0b + +2: /* is it an indirection page? (r3) */ + rlwinm. r7, r0, 0, 30, 30 /* IND_INDIRECTION (1<<1) */ + beq 2f + + rlwinm r3, r0, 0, 0, 19 /* clear kexec flags, page align */ + subi r3, r3, 4 + b 0b + +2: /* are we done? */ + rlwinm. r7, r0, 0, 29, 29 /* IND_DONE (1<<2) */ + beq 2f + b 3f + +2: /* is it a source page? (r9) */ + rlwinm. r7, r0, 0, 28, 28 /* IND_SOURCE (1<<3) */ + beq 0b + + rlwinm r9, r0, 0, 0, 19 /* clear kexec flags, page align */ + + li r7, PAGE_SIZE / 4 + mtctr r7 + subi r9, r9, 4 + subi r8, r8, 4 +9: + lwzu r0, 4(r9) /* do the copy */ + xor r6, r6, r0 + stwu r0, 4(r8) + dcbst 0, r8 + sync + icbi 0, r8 + bdnz 9b + + addi r9, r9, 4 + addi r8, r8, 4 + b 0b + +3: + + /* To be certain of avoiding problems with self-modifying code + * execute a serializing instruction here. + */ + isync + sync + + /* jump to the entry point, usually the setup routine */ + mtlr r5 + blrl + +1: b 1b + +relocate_new_kernel_end: + + .globl relocate_new_kernel_size +relocate_new_kernel_size: + .long relocate_new_kernel_end - relocate_new_kernel + diff --git a/arch/ppc/kernel/signal.c b/arch/ppc/kernel/signal.c index 7c8437da09d..8aaeb6f4e75 100644 --- a/arch/ppc/kernel/signal.c +++ b/arch/ppc/kernel/signal.c @@ -705,8 +705,7 @@ int do_signal(sigset_t *oldset, struct pt_regs *regs) unsigned long frame, newsp; int signr, ret; - if (current->flags & PF_FREEZE) { - refrigerator(PF_FREEZE); + if (try_to_freeze()) { signr = 0; if (!signal_pending(current)) goto no_signal; diff --git a/arch/ppc/kernel/traps.c b/arch/ppc/kernel/traps.c index 2ca8ecfeefd..9e6ae569665 100644 --- a/arch/ppc/kernel/traps.c +++ b/arch/ppc/kernel/traps.c @@ -173,13 +173,13 @@ static inline int check_io_access(struct pt_regs *regs) /* On 4xx, the reason for the machine check or program exception is in the ESR. */ #define get_reason(regs) ((regs)->dsisr) -#ifndef CONFIG_E500 +#ifndef CONFIG_FSL_BOOKE #define get_mc_reason(regs) ((regs)->dsisr) #else #define get_mc_reason(regs) (mfspr(SPRN_MCSR)) #endif #define REASON_FP ESR_FP -#define REASON_ILLEGAL ESR_PIL +#define REASON_ILLEGAL (ESR_PIL | ESR_PUO) #define REASON_PRIVILEGED ESR_PPR #define REASON_TRAP ESR_PTR @@ -302,7 +302,25 @@ void MachineCheckException(struct pt_regs *regs) printk("Bus - Instruction Parity Error\n"); if (reason & MCSR_BUS_RPERR) printk("Bus - Read Parity Error\n"); -#else /* !CONFIG_4xx && !CONFIG_E500 */ +#elif defined (CONFIG_E200) + printk("Machine check in kernel mode.\n"); + printk("Caused by (from MCSR=%lx): ", reason); + + if (reason & MCSR_MCP) + printk("Machine Check Signal\n"); + if (reason & MCSR_CP_PERR) + printk("Cache Push Parity Error\n"); + if (reason & MCSR_CPERR) + printk("Cache Parity Error\n"); + if (reason & MCSR_EXCP_ERR) + printk("ISI, ITLB, or Bus Error on first instruction fetch for an exception handler\n"); + if (reason & MCSR_BUS_IRERR) + printk("Bus - Read Bus Error on instruction fetch\n"); + if (reason & MCSR_BUS_DRERR) + printk("Bus - Read Bus Error on data load\n"); + if (reason & MCSR_BUS_WRERR) + printk("Bus - Write Bus Error on buffered store or cache line push\n"); +#else /* !CONFIG_4xx && !CONFIG_E500 && !CONFIG_E200 */ printk("Machine check in kernel mode.\n"); printk("Caused by (from SRR1=%lx): ", reason); switch (reason & 0x601F0000) { diff --git a/arch/ppc/mm/44x_mmu.c b/arch/ppc/mm/44x_mmu.c index 72f7c0d1c0e..3d79ce281b6 100644 --- a/arch/ppc/mm/44x_mmu.c +++ b/arch/ppc/mm/44x_mmu.c @@ -39,7 +39,6 @@ #include <linux/vmalloc.h> #include <linux/init.h> #include <linux/delay.h> -#include <linux/bootmem.h> #include <linux/highmem.h> #include <asm/pgalloc.h> diff --git a/arch/ppc/mm/4xx_mmu.c b/arch/ppc/mm/4xx_mmu.c index a7f61614038..b7bcbc232f3 100644 --- a/arch/ppc/mm/4xx_mmu.c +++ b/arch/ppc/mm/4xx_mmu.c @@ -36,7 +36,6 @@ #include <linux/vmalloc.h> #include <linux/init.h> #include <linux/delay.h> -#include <linux/bootmem.h> #include <linux/highmem.h> #include <asm/pgalloc.h> diff --git a/arch/ppc/mm/fsl_booke_mmu.c b/arch/ppc/mm/fsl_booke_mmu.c index e07990efa04..af9ca0eb6d5 100644 --- a/arch/ppc/mm/fsl_booke_mmu.c +++ b/arch/ppc/mm/fsl_booke_mmu.c @@ -41,7 +41,6 @@ #include <linux/vmalloc.h> #include <linux/init.h> #include <linux/delay.h> -#include <linux/bootmem.h> #include <linux/highmem.h> #include <asm/pgalloc.h> @@ -126,7 +125,7 @@ void settlbcam(int index, unsigned long virt, phys_addr_t phys, flags |= _PAGE_COHERENT; #endif - TLBCAM[index].MAS0 = MAS0_TLBSEL(1) | MAS0_ESEL(index); + TLBCAM[index].MAS0 = MAS0_TLBSEL(1) | MAS0_ESEL(index) | MAS0_NV(index+1); TLBCAM[index].MAS1 = MAS1_VALID | MAS1_IPROT | MAS1_TSIZE(tsize) | MAS1_TID(pid); TLBCAM[index].MAS2 = virt & PAGE_MASK; diff --git a/arch/ppc/platforms/83xx/mpc834x_sys.c b/arch/ppc/platforms/83xx/mpc834x_sys.c index 37ece154279..ddd04d4c1ea 100644 --- a/arch/ppc/platforms/83xx/mpc834x_sys.c +++ b/arch/ppc/platforms/83xx/mpc834x_sys.c @@ -94,20 +94,24 @@ mpc834x_sys_setup_arch(void) /* setup the board related information for the enet controllers */ pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC83xx_TSEC1); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC83xx_IRQ_EXT1; - pdata->phyid = 0; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC83xx_IRQ_EXT1; + pdata->phyid = 0; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + } pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC83xx_TSEC2); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC83xx_IRQ_EXT2; - pdata->phyid = 1; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC83xx_IRQ_EXT2; + pdata->phyid = 1; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + } #ifdef CONFIG_BLK_DEV_INITRD if (initrd_start) diff --git a/arch/ppc/platforms/85xx/mpc8540_ads.c b/arch/ppc/platforms/85xx/mpc8540_ads.c index a2ed611cd93..ddd2e9a5bb1 100644 --- a/arch/ppc/platforms/85xx/mpc8540_ads.c +++ b/arch/ppc/platforms/85xx/mpc8540_ads.c @@ -92,28 +92,34 @@ mpc8540ads_setup_arch(void) /* setup the board related information for the enet controllers */ pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC1); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 0; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 0; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + } pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC2); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 1; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); - - pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_FEC); - pdata->board_flags = 0; - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 3; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enet2addr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 1; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + } + + if (pdata) { + pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_FEC); + pdata->board_flags = 0; + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 3; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enet2addr, 6); + } #ifdef CONFIG_BLK_DEV_INITRD if (initrd_start) diff --git a/arch/ppc/platforms/85xx/mpc8560_ads.c b/arch/ppc/platforms/85xx/mpc8560_ads.c index d87dfd5ce0a..e18380258b6 100644 --- a/arch/ppc/platforms/85xx/mpc8560_ads.c +++ b/arch/ppc/platforms/85xx/mpc8560_ads.c @@ -90,20 +90,24 @@ mpc8560ads_setup_arch(void) /* setup the board related information for the enet controllers */ pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC1); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 0; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 0; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + } pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC2); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 1; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 1; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + } #ifdef CONFIG_BLK_DEV_INITRD if (initrd_start) diff --git a/arch/ppc/platforms/85xx/sbc8560.c b/arch/ppc/platforms/85xx/sbc8560.c index 3dbdd73618e..165df94d4aa 100644 --- a/arch/ppc/platforms/85xx/sbc8560.c +++ b/arch/ppc/platforms/85xx/sbc8560.c @@ -129,20 +129,24 @@ sbc8560_setup_arch(void) /* setup the board related information for the enet controllers */ pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC1); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC85xx_IRQ_EXT6; - pdata->phyid = 25; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC85xx_IRQ_EXT6; + pdata->phyid = 25; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + } pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC2); - pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; - pdata->interruptPHY = MPC85xx_IRQ_EXT7; - pdata->phyid = 26; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + if (pdata) { + pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; + pdata->interruptPHY = MPC85xx_IRQ_EXT7; + pdata->phyid = 26; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + } #ifdef CONFIG_BLK_DEV_INITRD if (initrd_start) diff --git a/arch/ppc/platforms/85xx/stx_gp3.c b/arch/ppc/platforms/85xx/stx_gp3.c index 9455bb6b45e..bb41265cfc8 100644 --- a/arch/ppc/platforms/85xx/stx_gp3.c +++ b/arch/ppc/platforms/85xx/stx_gp3.c @@ -122,19 +122,23 @@ gp3_setup_arch(void) /* setup the board related information for the enet controllers */ pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC1); -/* pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; */ - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 2; - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + if (pdata) { + /* pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; */ + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 2; + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enetaddr, 6); + } pdata = (struct gianfar_platform_data *) ppc_sys_get_pdata(MPC85xx_TSEC2); -/* pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; */ - pdata->interruptPHY = MPC85xx_IRQ_EXT5; - pdata->phyid = 4; - /* fixup phy address */ - pdata->phy_reg_addr += binfo->bi_immr_base; - memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + if (pdata) { + /* pdata->board_flags = FSL_GIANFAR_BRD_HAS_PHY_INTR; */ + pdata->interruptPHY = MPC85xx_IRQ_EXT5; + pdata->phyid = 4; + /* fixup phy address */ + pdata->phy_reg_addr += binfo->bi_immr_base; + memcpy(pdata->mac_addr, binfo->bi_enet1addr, 6); + } #ifdef CONFIG_BLK_DEV_INITRD if (initrd_start) diff --git a/arch/ppc/platforms/chrp_pci.c b/arch/ppc/platforms/chrp_pci.c index 7d0ee308f66..7d3fbb5c5db 100644 --- a/arch/ppc/platforms/chrp_pci.c +++ b/arch/ppc/platforms/chrp_pci.c @@ -9,7 +9,6 @@ #include <linux/string.h> #include <linux/init.h> #include <linux/ide.h> -#include <linux/bootmem.h> #include <asm/io.h> #include <asm/pgtable.h> diff --git a/arch/ppc/platforms/katana.c b/arch/ppc/platforms/katana.c index eda922ac316..169dbf6534b 100644 --- a/arch/ppc/platforms/katana.c +++ b/arch/ppc/platforms/katana.c @@ -27,12 +27,12 @@ #include <linux/root_dev.h> #include <linux/delay.h> #include <linux/seq_file.h> -#include <linux/bootmem.h> #include <linux/mtd/physmap.h> #include <linux/mv643xx.h> #ifdef CONFIG_BOOTIMG #include <linux/bootimg.h> #endif +#include <asm/io.h> #include <asm/page.h> #include <asm/time.h> #include <asm/smp.h> diff --git a/arch/ppc/platforms/pmac_pci.c b/arch/ppc/platforms/pmac_pci.c index f6ff5192406..719fb49fe2b 100644 --- a/arch/ppc/platforms/pmac_pci.c +++ b/arch/ppc/platforms/pmac_pci.c @@ -17,7 +17,6 @@ #include <linux/delay.h> #include <linux/string.h> #include <linux/init.h> -#include <linux/bootmem.h> #include <asm/sections.h> #include <asm/io.h> diff --git a/arch/ppc/syslib/cpm2_common.c b/arch/ppc/syslib/cpm2_common.c index ea5e77080e8..4c19a4ac716 100644 --- a/arch/ppc/syslib/cpm2_common.c +++ b/arch/ppc/syslib/cpm2_common.c @@ -21,8 +21,8 @@ #include <linux/string.h> #include <linux/mm.h> #include <linux/interrupt.h> -#include <linux/bootmem.h> #include <linux/module.h> +#include <asm/io.h> #include <asm/irq.h> #include <asm/mpc8260.h> #include <asm/page.h> diff --git a/arch/ppc/syslib/indirect_pci.c b/arch/ppc/syslib/indirect_pci.c index a5a752609e2..e7148846970 100644 --- a/arch/ppc/syslib/indirect_pci.c +++ b/arch/ppc/syslib/indirect_pci.c @@ -14,7 +14,6 @@ #include <linux/delay.h> #include <linux/string.h> #include <linux/init.h> -#include <linux/bootmem.h> #include <asm/io.h> #include <asm/prom.h> diff --git a/arch/ppc/syslib/ipic.c b/arch/ppc/syslib/ipic.c index 580ed658e87..8f01e0f1d84 100644 --- a/arch/ppc/syslib/ipic.c +++ b/arch/ppc/syslib/ipic.c @@ -79,7 +79,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 7, }, [17] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_A, .force = IPIC_SEFCR, @@ -87,7 +87,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 5, }, [18] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_A, .force = IPIC_SEFCR, @@ -95,7 +95,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 6, }, [19] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_A, .force = IPIC_SEFCR, @@ -103,7 +103,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 7, }, [20] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_B, .force = IPIC_SEFCR, @@ -111,7 +111,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 4, }, [21] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_B, .force = IPIC_SEFCR, @@ -119,7 +119,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 5, }, [22] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_B, .force = IPIC_SEFCR, @@ -127,7 +127,7 @@ static struct ipic_info ipic_info[] = { .prio_mask = 6, }, [23] = { - .pend = IPIC_SIPNR_H, + .pend = IPIC_SEPNR, .mask = IPIC_SEMSR, .prio = IPIC_SMPRR_B, .force = IPIC_SEFCR, diff --git a/arch/ppc/syslib/mv64x60.c b/arch/ppc/syslib/mv64x60.c index 7b241e7876b..cc77177fa1c 100644 --- a/arch/ppc/syslib/mv64x60.c +++ b/arch/ppc/syslib/mv64x60.c @@ -17,7 +17,6 @@ #include <linux/slab.h> #include <linux/module.h> #include <linux/string.h> -#include <linux/bootmem.h> #include <linux/spinlock.h> #include <linux/mv643xx.h> diff --git a/arch/ppc/syslib/mv64x60_win.c b/arch/ppc/syslib/mv64x60_win.c index b6f0f5dcf6e..5b827e2bbe2 100644 --- a/arch/ppc/syslib/mv64x60_win.c +++ b/arch/ppc/syslib/mv64x60_win.c @@ -17,7 +17,6 @@ #include <linux/slab.h> #include <linux/module.h> #include <linux/string.h> -#include <linux/bootmem.h> #include <linux/mv643xx.h> #include <asm/byteorder.h> diff --git a/arch/ppc64/Kconfig b/arch/ppc64/Kconfig index 0f1fa289744..f804f25232a 100644 --- a/arch/ppc64/Kconfig +++ b/arch/ppc64/Kconfig @@ -77,6 +77,10 @@ config PPC_PSERIES bool " IBM pSeries & new iSeries" default y +config PPC_BPA + bool " Broadband Processor Architecture" + depends on PPC_MULTIPLATFORM + config PPC_PMAC depends on PPC_MULTIPLATFORM bool " Apple G5 based machines" @@ -106,6 +110,21 @@ config PPC_OF bool default y +config XICS + depends on PPC_PSERIES + bool + default y + +config MPIC + depends on PPC_PSERIES || PPC_PMAC || PPC_MAPLE + bool + default y + +config BPA_IIC + depends on PPC_BPA + bool + default y + # VMX is pSeries only for now until somebody writes the iSeries # exception vectors for it config ALTIVEC @@ -123,6 +142,23 @@ config PPC_SPLPAR processors, that is, which share physical processors between two or more partitions. +config KEXEC + bool "kexec system call (EXPERIMENTAL)" + depends on PPC_MULTIPLATFORM && EXPERIMENTAL + help + kexec is a system call that implements the ability to shutdown your + current kernel, and to start another kernel. It is like a reboot + but it is indepedent of the system firmware. And like a reboot + you can start any kernel with it, not just Linux. + + The name comes from the similiarity to the exec system call. + + It is an ongoing process to be certain the hardware in a machine + is properly shutdown, so do not be surprised if this code does not + initially work for you. It may help to enable device hotplugging + support. As of this writing the exact hardware interface is + strongly in flux, so no good recommendation can be made. + config IBMVIO depends on PPC_PSERIES || PPC_ISERIES bool @@ -198,13 +234,49 @@ config HMT This option enables hardware multithreading on RS64 cpus. pSeries systems p620 and p660 have such a cpu type. -config DISCONTIGMEM - bool "Discontiguous Memory Support" +config ARCH_SELECT_MEMORY_MODEL + def_bool y + +config ARCH_FLATMEM_ENABLE + def_bool y + depends on !NUMA + +config ARCH_DISCONTIGMEM_ENABLE + def_bool y depends on SMP && PPC_PSERIES +config ARCH_DISCONTIGMEM_DEFAULT + def_bool y + depends on ARCH_DISCONTIGMEM_ENABLE + +config ARCH_FLATMEM_ENABLE + def_bool y + +config ARCH_SPARSEMEM_ENABLE + def_bool y + depends on ARCH_DISCONTIGMEM_ENABLE + +source "mm/Kconfig" + +config HAVE_ARCH_EARLY_PFN_TO_NID + def_bool y + depends on NEED_MULTIPLE_NODES + +# Some NUMA nodes have memory ranges that span +# other nodes. Even though a pfn is valid and +# between a node's start and end pfns, it may not +# reside on that node. +# +# This is a relatively temporary hack that should +# be able to go away when sparsemem is fully in +# place +config NODES_SPAN_OTHER_NODES + def_bool y + depends on NEED_MULTIPLE_NODES + config NUMA bool "NUMA support" - depends on DISCONTIGMEM + default y if DISCONTIGMEM || SPARSEMEM config SCHED_SMT bool "SMT (Hyperthreading) scheduler support" @@ -215,26 +287,7 @@ config SCHED_SMT when dealing with POWER5 cpus at a cost of slightly increased overhead in some places. If unsure say N here. -config PREEMPT - bool "Preemptible Kernel" - help - This option reduces the latency of the kernel when reacting to - real-time or interactive events by allowing a low priority process to - be preempted even if it is in kernel mode executing a system call. - - Say Y here if you are building a kernel for a desktop, embedded - or real-time system. Say N if you are unsure. - -config PREEMPT_BKL - bool "Preempt The Big Kernel Lock" - depends on PREEMPT - default y - help - This option reduces the latency of the kernel by making the - big kernel lock preemptible. - - Say Y here if you are building a kernel for a desktop system. - Say N if you are unsure. +source "kernel/Kconfig.preempt" config EEH bool "PCI Extended Error Handling (EEH)" if EMBEDDED @@ -256,7 +309,7 @@ config MSCHUNKS config PPC_RTAS bool - depends on PPC_PSERIES + depends on PPC_PSERIES || PPC_BPA default y config RTAS_PROC diff --git a/arch/ppc64/Makefile b/arch/ppc64/Makefile index 33c752ceca4..731b8475833 100644 --- a/arch/ppc64/Makefile +++ b/arch/ppc64/Makefile @@ -90,12 +90,14 @@ boot := arch/ppc64/boot boottarget-$(CONFIG_PPC_PSERIES) := zImage zImage.initrd boottarget-$(CONFIG_PPC_MAPLE) := zImage zImage.initrd boottarget-$(CONFIG_PPC_ISERIES) := vmlinux.sminitrd vmlinux.initrd vmlinux.sm +boottarget-$(CONFIG_PPC_BPA) := zImage zImage.initrd $(boottarget-y): vmlinux $(Q)$(MAKE) $(build)=$(boot) $(boot)/$@ bootimage-$(CONFIG_PPC_PSERIES) := $(boot)/zImage bootimage-$(CONFIG_PPC_PMAC) := vmlinux bootimage-$(CONFIG_PPC_MAPLE) := $(boot)/zImage +bootimage-$(CONFIG_PPC_BPA) := zImage bootimage-$(CONFIG_PPC_ISERIES) := vmlinux BOOTIMAGE := $(bootimage-y) install: vmlinux diff --git a/arch/ppc64/boot/install.sh b/arch/ppc64/boot/install.sh index 955c5681db6..cb2d6626b55 100644 --- a/arch/ppc64/boot/install.sh +++ b/arch/ppc64/boot/install.sh @@ -22,8 +22,8 @@ # User may have a custom install script -if [ -x ~/bin/installkernel ]; then exec ~/bin/installkernel "$@"; fi -if [ -x /sbin/installkernel ]; then exec /sbin/installkernel "$@"; fi +if [ -x ~/bin/${CROSS_COMPILE}installkernel ]; then exec ~/bin/${CROSS_COMPILE}installkernel "$@"; fi +if [ -x /sbin/${CROSS_COMPILE}installkernel ]; then exec /sbin/${CROSS_COMPILE}installkernel "$@"; fi # Default install diff --git a/arch/ppc64/configs/pSeries_defconfig b/arch/ppc64/configs/pSeries_defconfig index 3eb5ef25d3a..d0db8b5966c 100644 --- a/arch/ppc64/configs/pSeries_defconfig +++ b/arch/ppc64/configs/pSeries_defconfig @@ -88,7 +88,7 @@ CONFIG_IBMVIO=y CONFIG_IOMMU_VMERGE=y CONFIG_SMP=y CONFIG_NR_CPUS=128 -CONFIG_DISCONTIGMEM=y +CONFIG_ARCH_DISCONTIGMEM_ENABLE=y CONFIG_NUMA=y CONFIG_SCHED_SMT=y # CONFIG_PREEMPT is not set diff --git a/arch/ppc64/defconfig b/arch/ppc64/defconfig index 2f31bf3046f..b8e2066dde7 100644 --- a/arch/ppc64/defconfig +++ b/arch/ppc64/defconfig @@ -89,7 +89,7 @@ CONFIG_BOOTX_TEXT=y CONFIG_IOMMU_VMERGE=y CONFIG_SMP=y CONFIG_NR_CPUS=32 -CONFIG_DISCONTIGMEM=y +CONFIG_ARCH_DISCONTIGMEM_ENABLE=y # CONFIG_NUMA is not set # CONFIG_SCHED_SMT is not set # CONFIG_PREEMPT is not set diff --git a/arch/ppc64/kernel/Makefile b/arch/ppc64/kernel/Makefile index b5e167cf1a0..d9b2660ef22 100644 --- a/arch/ppc64/kernel/Makefile +++ b/arch/ppc64/kernel/Makefile @@ -27,17 +27,22 @@ obj-$(CONFIG_PPC_ISERIES) += HvCall.o HvLpConfig.o LparData.o \ mf.o HvLpEvent.o iSeries_proc.o iSeries_htab.o \ iSeries_iommu.o -obj-$(CONFIG_PPC_MULTIPLATFORM) += nvram.o i8259.o prom_init.o prom.o mpic.o +obj-$(CONFIG_PPC_MULTIPLATFORM) += nvram.o i8259.o prom_init.o prom.o obj-$(CONFIG_PPC_PSERIES) += pSeries_pci.o pSeries_lpar.o pSeries_hvCall.o \ pSeries_nvram.o rtasd.o ras.o pSeries_reconfig.o \ - xics.o rtas.o pSeries_setup.o pSeries_iommu.o + pSeries_setup.o pSeries_iommu.o +obj-$(CONFIG_PPC_BPA) += bpa_setup.o bpa_iommu.o bpa_nvram.o \ + bpa_iic.o spider-pic.o + +obj-$(CONFIG_KEXEC) += machine_kexec.o obj-$(CONFIG_EEH) += eeh.o obj-$(CONFIG_PROC_FS) += proc_ppc64.o obj-$(CONFIG_RTAS_FLASH) += rtas_flash.o obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_MODULES) += module.o ppc_ksyms.o +obj-$(CONFIG_PPC_RTAS) += rtas.o rtas_pci.o obj-$(CONFIG_RTAS_PROC) += rtas-proc.o obj-$(CONFIG_SCANLOG) += scanlog.o obj-$(CONFIG_VIOPATH) += viopath.o @@ -46,6 +51,8 @@ obj-$(CONFIG_HVC_CONSOLE) += hvconsole.o obj-$(CONFIG_BOOTX_TEXT) += btext.o obj-$(CONFIG_HVCS) += hvcserver.o obj-$(CONFIG_IBMVIO) += vio.o +obj-$(CONFIG_XICS) += xics.o +obj-$(CONFIG_MPIC) += mpic.o obj-$(CONFIG_PPC_PMAC) += pmac_setup.o pmac_feature.o pmac_pci.o \ pmac_time.o pmac_nvram.o pmac_low_i2c.o @@ -58,6 +65,7 @@ ifdef CONFIG_SMP obj-$(CONFIG_PPC_PMAC) += pmac_smp.o smp-tbsync.o obj-$(CONFIG_PPC_ISERIES) += iSeries_smp.o obj-$(CONFIG_PPC_PSERIES) += pSeries_smp.o +obj-$(CONFIG_PPC_BPA) += pSeries_smp.o obj-$(CONFIG_PPC_MAPLE) += smp-tbsync.o endif diff --git a/arch/ppc64/kernel/bpa_iic.c b/arch/ppc64/kernel/bpa_iic.c new file mode 100644 index 00000000000..c8f3dc3fad7 --- /dev/null +++ b/arch/ppc64/kernel/bpa_iic.c @@ -0,0 +1,270 @@ +/* + * BPA Internal Interrupt Controller + * + * (C) Copyright IBM Deutschland Entwicklung GmbH 2005 + * + * Author: Arnd Bergmann <arndb@de.ibm.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include <linux/config.h> +#include <linux/interrupt.h> +#include <linux/irq.h> +#include <linux/percpu.h> +#include <linux/types.h> + +#include <asm/io.h> +#include <asm/pgtable.h> +#include <asm/prom.h> +#include <asm/ptrace.h> + +#include "bpa_iic.h" + +struct iic_pending_bits { + u32 data; + u8 flags; + u8 class; + u8 source; + u8 prio; +}; + +enum iic_pending_flags { + IIC_VALID = 0x80, + IIC_IPI = 0x40, +}; + +struct iic_regs { + struct iic_pending_bits pending; + struct iic_pending_bits pending_destr; + u64 generate; + u64 prio; +}; + +struct iic { + struct iic_regs __iomem *regs; +}; + +static DEFINE_PER_CPU(struct iic, iic); + +void iic_local_enable(void) +{ + out_be64(&__get_cpu_var(iic).regs->prio, 0xff); +} + +void iic_local_disable(void) +{ + out_be64(&__get_cpu_var(iic).regs->prio, 0x0); +} + +static unsigned int iic_startup(unsigned int irq) +{ + return 0; +} + +static void iic_enable(unsigned int irq) +{ + iic_local_enable(); +} + +static void iic_disable(unsigned int irq) +{ +} + +static void iic_end(unsigned int irq) +{ + iic_local_enable(); +} + +static struct hw_interrupt_type iic_pic = { + .typename = " BPA-IIC ", + .startup = iic_startup, + .enable = iic_enable, + .disable = iic_disable, + .end = iic_end, +}; + +static int iic_external_get_irq(struct iic_pending_bits pending) +{ + int irq; + unsigned char node, unit; + + node = pending.source >> 4; + unit = pending.source & 0xf; + irq = -1; + + /* + * This mapping is specific to the Broadband + * Engine. We might need to get the numbers + * from the device tree to support future CPUs. + */ + switch (unit) { + case 0x00: + case 0x0b: + /* + * One of these units can be connected + * to an external interrupt controller. + */ + if (pending.prio > 0x3f || + pending.class != 2) + break; + irq = IIC_EXT_OFFSET + + spider_get_irq(pending.prio + node * IIC_NODE_STRIDE) + + node * IIC_NODE_STRIDE; + break; + case 0x01 ... 0x04: + case 0x07 ... 0x0a: + /* + * These units are connected to the SPEs + */ + if (pending.class > 2) + break; + irq = IIC_SPE_OFFSET + + pending.class * IIC_CLASS_STRIDE + + node * IIC_NODE_STRIDE + + unit; + break; + } + if (irq == -1) + printk(KERN_WARNING "Unexpected interrupt class %02x, " + "source %02x, prio %02x, cpu %02x\n", pending.class, + pending.source, pending.prio, smp_processor_id()); + return irq; +} + +/* Get an IRQ number from the pending state register of the IIC */ +int iic_get_irq(struct pt_regs *regs) +{ + struct iic *iic; + int irq; + struct iic_pending_bits pending; + + iic = &__get_cpu_var(iic); + *(unsigned long *) &pending = + in_be64((unsigned long __iomem *) &iic->regs->pending_destr); + + irq = -1; + if (pending.flags & IIC_VALID) { + if (pending.flags & IIC_IPI) { + irq = IIC_IPI_OFFSET + (pending.prio >> 4); +/* + if (irq > 0x80) + printk(KERN_WARNING "Unexpected IPI prio %02x" + "on CPU %02x\n", pending.prio, + smp_processor_id()); +*/ + } else { + irq = iic_external_get_irq(pending); + } + } + return irq; +} + +static struct iic_regs __iomem *find_iic(int cpu) +{ + struct device_node *np; + int nodeid = cpu / 2; + unsigned long regs; + struct iic_regs __iomem *iic_regs; + + for (np = of_find_node_by_type(NULL, "cpu"); + np; + np = of_find_node_by_type(np, "cpu")) { + if (nodeid == *(int *)get_property(np, "node-id", NULL)) + break; + } + + if (!np) { + printk(KERN_WARNING "IIC: CPU %d not found\n", cpu); + iic_regs = NULL; + } else { + regs = *(long *)get_property(np, "iic", NULL); + + /* hack until we have decided on the devtree info */ + regs += 0x400; + if (cpu & 1) + regs += 0x20; + + printk(KERN_DEBUG "IIC for CPU %d at %lx\n", cpu, regs); + iic_regs = __ioremap(regs, sizeof(struct iic_regs), + _PAGE_NO_CACHE); + } + return iic_regs; +} + +#ifdef CONFIG_SMP +void iic_setup_cpu(void) +{ + out_be64(&__get_cpu_var(iic).regs->prio, 0xff); +} + +void iic_cause_IPI(int cpu, int mesg) +{ + out_be64(&per_cpu(iic, cpu).regs->generate, mesg); +} + +static irqreturn_t iic_ipi_action(int irq, void *dev_id, struct pt_regs *regs) +{ + + smp_message_recv(irq - IIC_IPI_OFFSET, regs); + return IRQ_HANDLED; +} + +static void iic_request_ipi(int irq, const char *name) +{ + /* IPIs are marked SA_INTERRUPT as they must run with irqs + * disabled */ + get_irq_desc(irq)->handler = &iic_pic; + get_irq_desc(irq)->status |= IRQ_PER_CPU; + request_irq(irq, iic_ipi_action, SA_INTERRUPT, name, NULL); +} + +void iic_request_IPIs(void) +{ + iic_request_ipi(IIC_IPI_OFFSET + PPC_MSG_CALL_FUNCTION, "IPI-call"); + iic_request_ipi(IIC_IPI_OFFSET + PPC_MSG_RESCHEDULE, "IPI-resched"); +#ifdef CONFIG_DEBUGGER + iic_request_ipi(IIC_IPI_OFFSET + PPC_MSG_DEBUGGER_BREAK, "IPI-debug"); +#endif /* CONFIG_DEBUGGER */ +} +#endif /* CONFIG_SMP */ + +static void iic_setup_spe_handlers(void) +{ + int be, isrc; + + /* Assume two threads per BE are present */ + for (be=0; be < num_present_cpus() / 2; be++) { + for (isrc = 0; isrc < IIC_CLASS_STRIDE * 3; isrc++) { + int irq = IIC_NODE_STRIDE * be + IIC_SPE_OFFSET + isrc; + get_irq_desc(irq)->handler = &iic_pic; + } + } +} + +void iic_init_IRQ(void) +{ + int cpu, irq_offset; + struct iic *iic; + + irq_offset = 0; + for_each_cpu(cpu) { + iic = &per_cpu(iic, cpu); + iic->regs = find_iic(cpu); + if (iic->regs) + out_be64(&iic->regs->prio, 0xff); + } + iic_setup_spe_handlers(); +} diff --git a/arch/ppc64/kernel/bpa_iic.h b/arch/ppc64/kernel/bpa_iic.h new file mode 100644 index 00000000000..6833c302216 --- /dev/null +++ b/arch/ppc64/kernel/bpa_iic.h @@ -0,0 +1,62 @@ +#ifndef ASM_BPA_IIC_H +#define ASM_BPA_IIC_H +#ifdef __KERNEL__ +/* + * Mapping of IIC pending bits into per-node + * interrupt numbers. + * + * IRQ FF CC SS PP FF CC SS PP Description + * + * 00-3f 80 02 +0 00 - 80 02 +0 3f South Bridge + * 00-3f 80 02 +b 00 - 80 02 +b 3f South Bridge + * 41-4a 80 00 +1 ** - 80 00 +a ** SPU Class 0 + * 51-5a 80 01 +1 ** - 80 01 +a ** SPU Class 1 + * 61-6a 80 02 +1 ** - 80 02 +a ** SPU Class 2 + * 70-7f C0 ** ** 00 - C0 ** ** 0f IPI + * + * F flags + * C class + * S source + * P Priority + * + node number + * * don't care + * + * A node consists of a Broadband Engine and an optional + * south bridge device providing a maximum of 64 IRQs. + * The south bridge may be connected to either IOIF0 + * or IOIF1. + * Each SPE is represented as three IRQ lines, one per + * interrupt class. + * 16 IRQ numbers are reserved for inter processor + * interruptions, although these are only used in the + * range of the first node. + * + * This scheme needs 128 IRQ numbers per BIF node ID, + * which means that with the total of 512 lines + * available, we can have a maximum of four nodes. + */ + +enum { + IIC_EXT_OFFSET = 0x00, /* Start of south bridge IRQs */ + IIC_NUM_EXT = 0x40, /* Number of south bridge IRQs */ + IIC_SPE_OFFSET = 0x40, /* Start of SPE interrupts */ + IIC_CLASS_STRIDE = 0x10, /* SPE IRQs per class */ + IIC_IPI_OFFSET = 0x70, /* Start of IPI IRQs */ + IIC_NUM_IPIS = 0x10, /* IRQs reserved for IPI */ + IIC_NODE_STRIDE = 0x80, /* Total IRQs per node */ +}; + +extern void iic_init_IRQ(void); +extern int iic_get_irq(struct pt_regs *regs); +extern void iic_cause_IPI(int cpu, int mesg); +extern void iic_request_IPIs(void); +extern void iic_setup_cpu(void); +extern void iic_local_enable(void); +extern void iic_local_disable(void); + + +extern void spider_init_IRQ(void); +extern int spider_get_irq(unsigned long int_pending); + +#endif +#endif /* ASM_BPA_IIC_H */ diff --git a/arch/ppc64/kernel/bpa_iommu.c b/arch/ppc64/kernel/bpa_iommu.c new file mode 100644 index 00000000000..f33a7bccb0d --- /dev/null +++ b/arch/ppc64/kernel/bpa_iommu.c @@ -0,0 +1,377 @@ +/* + * IOMMU implementation for Broadband Processor Architecture + * We just establish a linear mapping at boot by setting all the + * IOPT cache entries in the CPU. + * The mapping functions should be identical to pci_direct_iommu, + * except for the handling of the high order bit that is required + * by the Spider bridge. These should be split into a separate + * file at the point where we get a different bridge chip. + * + * Copyright (C) 2005 IBM Deutschland Entwicklung GmbH, + * Arnd Bergmann <arndb@de.ibm.com> + * + * Based on linear mapping + * Copyright (C) 2003 Benjamin Herrenschmidt (benh@kernel.crashing.org) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#undef DEBUG + +#include <linux/kernel.h> +#include <linux/pci.h> +#include <linux/delay.h> +#include <linux/string.h> +#include <linux/init.h> +#include <linux/bootmem.h> +#include <linux/mm.h> +#include <linux/dma-mapping.h> + +#include <asm/sections.h> +#include <asm/iommu.h> +#include <asm/io.h> +#include <asm/prom.h> +#include <asm/pci-bridge.h> +#include <asm/machdep.h> +#include <asm/pmac_feature.h> +#include <asm/abs_addr.h> +#include <asm/system.h> + +#include "pci.h" +#include "bpa_iommu.h" + +static inline unsigned long +get_iopt_entry(unsigned long real_address, unsigned long ioid, + unsigned long prot) +{ + return (prot & IOPT_PROT_MASK) + | (IOPT_COHERENT) + | (IOPT_ORDER_VC) + | (real_address & IOPT_RPN_MASK) + | (ioid & IOPT_IOID_MASK); +} + +typedef struct { + unsigned long val; +} ioste; + +static inline ioste +mk_ioste(unsigned long val) +{ + ioste ioste = { .val = val, }; + return ioste; +} + +static inline ioste +get_iost_entry(unsigned long iopt_base, unsigned long io_address, unsigned page_size) +{ + unsigned long ps; + unsigned long iostep; + unsigned long nnpt; + unsigned long shift; + + switch (page_size) { + case 0x1000000: + ps = IOST_PS_16M; + nnpt = 0; /* one page per segment */ + shift = 5; /* segment has 16 iopt entries */ + break; + + case 0x100000: + ps = IOST_PS_1M; + nnpt = 0; /* one page per segment */ + shift = 1; /* segment has 256 iopt entries */ + break; + + case 0x10000: + ps = IOST_PS_64K; + nnpt = 0x07; /* 8 pages per io page table */ + shift = 0; /* all entries are used */ + break; + + case 0x1000: + ps = IOST_PS_4K; + nnpt = 0x7f; /* 128 pages per io page table */ + shift = 0; /* all entries are used */ + break; + + default: /* not a known compile time constant */ + BUILD_BUG_ON(1); + break; + } + + iostep = iopt_base + + /* need 8 bytes per iopte */ + (((io_address / page_size * 8) + /* align io page tables on 4k page boundaries */ + << shift) + /* nnpt+1 pages go into each iopt */ + & ~(nnpt << 12)); + + nnpt++; /* this seems to work, but the documentation is not clear + about wether we put nnpt or nnpt-1 into the ioste bits. + In theory, this can't work for 4k pages. */ + return mk_ioste(IOST_VALID_MASK + | (iostep & IOST_PT_BASE_MASK) + | ((nnpt << 5) & IOST_NNPT_MASK) + | (ps & IOST_PS_MASK)); +} + +/* compute the address of an io pte */ +static inline unsigned long +get_ioptep(ioste iost_entry, unsigned long io_address) +{ + unsigned long iopt_base; + unsigned long page_size; + unsigned long page_number; + unsigned long iopt_offset; + + iopt_base = iost_entry.val & IOST_PT_BASE_MASK; + page_size = iost_entry.val & IOST_PS_MASK; + + /* decode page size to compute page number */ + page_number = (io_address & 0x0fffffff) >> (10 + 2 * page_size); + /* page number is an offset into the io page table */ + iopt_offset = (page_number << 3) & 0x7fff8ul; + return iopt_base + iopt_offset; +} + +/* compute the tag field of the iopt cache entry */ +static inline unsigned long +get_ioc_tag(ioste iost_entry, unsigned long io_address) +{ + unsigned long iopte = get_ioptep(iost_entry, io_address); + + return IOPT_VALID_MASK + | ((iopte & 0x00000000000000ff8ul) >> 3) + | ((iopte & 0x0000003fffffc0000ul) >> 9); +} + +/* compute the hashed 6 bit index for the 4-way associative pte cache */ +static inline unsigned long +get_ioc_hash(ioste iost_entry, unsigned long io_address) +{ + unsigned long iopte = get_ioptep(iost_entry, io_address); + + return ((iopte & 0x000000000000001f8ul) >> 3) + ^ ((iopte & 0x00000000000020000ul) >> 17) + ^ ((iopte & 0x00000000000010000ul) >> 15) + ^ ((iopte & 0x00000000000008000ul) >> 13) + ^ ((iopte & 0x00000000000004000ul) >> 11) + ^ ((iopte & 0x00000000000002000ul) >> 9) + ^ ((iopte & 0x00000000000001000ul) >> 7); +} + +/* same as above, but pretend that we have a simpler 1-way associative + pte cache with an 8 bit index */ +static inline unsigned long +get_ioc_hash_1way(ioste iost_entry, unsigned long io_address) +{ + unsigned long iopte = get_ioptep(iost_entry, io_address); + + return ((iopte & 0x000000000000001f8ul) >> 3) + ^ ((iopte & 0x00000000000020000ul) >> 17) + ^ ((iopte & 0x00000000000010000ul) >> 15) + ^ ((iopte & 0x00000000000008000ul) >> 13) + ^ ((iopte & 0x00000000000004000ul) >> 11) + ^ ((iopte & 0x00000000000002000ul) >> 9) + ^ ((iopte & 0x00000000000001000ul) >> 7) + ^ ((iopte & 0x0000000000000c000ul) >> 8); +} + +static inline ioste +get_iost_cache(void __iomem *base, unsigned long index) +{ + unsigned long __iomem *p = (base + IOC_ST_CACHE_DIR); + return mk_ioste(in_be64(&p[index])); +} + +static inline void +set_iost_cache(void __iomem *base, unsigned long index, ioste ste) +{ + unsigned long __iomem *p = (base + IOC_ST_CACHE_DIR); + pr_debug("ioste %02lx was %016lx, store %016lx", index, + get_iost_cache(base, index).val, ste.val); + out_be64(&p[index], ste.val); + pr_debug(" now %016lx\n", get_iost_cache(base, index).val); +} + +static inline unsigned long +get_iopt_cache(void __iomem *base, unsigned long index, unsigned long *tag) +{ + unsigned long __iomem *tags = (void *)(base + IOC_PT_CACHE_DIR); + unsigned long __iomem *p = (void *)(base + IOC_PT_CACHE_REG); + + *tag = tags[index]; + rmb(); + return *p; +} + +static inline void +set_iopt_cache(void __iomem *base, unsigned long index, + unsigned long tag, unsigned long val) +{ + unsigned long __iomem *tags = base + IOC_PT_CACHE_DIR; + unsigned long __iomem *p = base + IOC_PT_CACHE_REG; + pr_debug("iopt %02lx was v%016lx/t%016lx, store v%016lx/t%016lx\n", + index, get_iopt_cache(base, index, &oldtag), oldtag, val, tag); + + out_be64(p, val); + out_be64(&tags[index], tag); +} + +static inline void +set_iost_origin(void __iomem *base) +{ + unsigned long __iomem *p = base + IOC_ST_ORIGIN; + unsigned long origin = IOSTO_ENABLE | IOSTO_SW; + + pr_debug("iost_origin %016lx, now %016lx\n", in_be64(p), origin); + out_be64(p, origin); +} + +static inline void +set_iocmd_config(void __iomem *base) +{ + unsigned long __iomem *p = base + 0xc00; + unsigned long conf; + + conf = in_be64(p); + pr_debug("iost_conf %016lx, now %016lx\n", conf, conf | IOCMD_CONF_TE); + out_be64(p, conf | IOCMD_CONF_TE); +} + +/* FIXME: get these from the device tree */ +#define ioc_base 0x20000511000ull +#define ioc_mmio_base 0x20000510000ull +#define ioid 0x48a +#define iopt_phys_offset (- 0x20000000) /* We have a 512MB offset from the SB */ +#define io_page_size 0x1000000 + +static unsigned long map_iopt_entry(unsigned long address) +{ + switch (address >> 20) { + case 0x600: + address = 0x24020000000ull; /* spider i/o */ + break; + default: + address += iopt_phys_offset; + break; + } + + return get_iopt_entry(address, ioid, IOPT_PROT_RW); +} + +static void iommu_bus_setup_null(struct pci_bus *b) { } +static void iommu_dev_setup_null(struct pci_dev *d) { } + +/* initialize the iommu to support a simple linear mapping + * for each DMA window used by any device. For now, we + * happen to know that there is only one DMA window in use, + * starting at iopt_phys_offset. */ +static void bpa_map_iommu(void) +{ + unsigned long address; + void __iomem *base; + ioste ioste; + unsigned long index; + + base = __ioremap(ioc_base, 0x1000, _PAGE_NO_CACHE); + pr_debug("%lx mapped to %p\n", ioc_base, base); + set_iocmd_config(base); + iounmap(base); + + base = __ioremap(ioc_mmio_base, 0x1000, _PAGE_NO_CACHE); + pr_debug("%lx mapped to %p\n", ioc_mmio_base, base); + + set_iost_origin(base); + + for (address = 0; address < 0x100000000ul; address += io_page_size) { + ioste = get_iost_entry(0x10000000000ul, address, io_page_size); + if ((address & 0xfffffff) == 0) /* segment start */ + set_iost_cache(base, address >> 28, ioste); + index = get_ioc_hash_1way(ioste, address); + pr_debug("addr %08lx, index %02lx, ioste %016lx\n", + address, index, ioste.val); + set_iopt_cache(base, + get_ioc_hash_1way(ioste, address), + get_ioc_tag(ioste, address), + map_iopt_entry(address)); + } + iounmap(base); +} + + +static void *bpa_alloc_coherent(struct device *hwdev, size_t size, + dma_addr_t *dma_handle, unsigned int __nocast flag) +{ + void *ret; + + ret = (void *)__get_free_pages(flag, get_order(size)); + if (ret != NULL) { + memset(ret, 0, size); + *dma_handle = virt_to_abs(ret) | BPA_DMA_VALID; + } + return ret; +} + +static void bpa_free_coherent(struct device *hwdev, size_t size, + void *vaddr, dma_addr_t dma_handle) +{ + free_pages((unsigned long)vaddr, get_order(size)); +} + +static dma_addr_t bpa_map_single(struct device *hwdev, void *ptr, + size_t size, enum dma_data_direction direction) +{ + return virt_to_abs(ptr) | BPA_DMA_VALID; +} + +static void bpa_unmap_single(struct device *hwdev, dma_addr_t dma_addr, + size_t size, enum dma_data_direction direction) +{ +} + +static int bpa_map_sg(struct device *hwdev, struct scatterlist *sg, + int nents, enum dma_data_direction direction) +{ + int i; + + for (i = 0; i < nents; i++, sg++) { + sg->dma_address = (page_to_phys(sg->page) + sg->offset) + | BPA_DMA_VALID; + sg->dma_length = sg->length; + } + + return nents; +} + +static void bpa_unmap_sg(struct device *hwdev, struct scatterlist *sg, + int nents, enum dma_data_direction direction) +{ +} + +static int bpa_dma_supported(struct device *dev, u64 mask) +{ + return mask < 0x100000000ull; +} + +void bpa_init_iommu(void) +{ + bpa_map_iommu(); + + /* Direct I/O, IOMMU off */ + ppc_md.iommu_dev_setup = iommu_dev_setup_null; + ppc_md.iommu_bus_setup = iommu_bus_setup_null; + + pci_dma_ops.alloc_coherent = bpa_alloc_coherent; + pci_dma_ops.free_coherent = bpa_free_coherent; + pci_dma_ops.map_single = bpa_map_single; + pci_dma_ops.unmap_single = bpa_unmap_single; + pci_dma_ops.map_sg = bpa_map_sg; + pci_dma_ops.unmap_sg = bpa_unmap_sg; + pci_dma_ops.dma_supported = bpa_dma_supported; +} diff --git a/arch/ppc64/kernel/bpa_iommu.h b/arch/ppc64/kernel/bpa_iommu.h new file mode 100644 index 00000000000..e547d77dfa0 --- /dev/null +++ b/arch/ppc64/kernel/bpa_iommu.h @@ -0,0 +1,65 @@ +#ifndef BPA_IOMMU_H +#define BPA_IOMMU_H + +/* some constants */ +enum { + /* segment table entries */ + IOST_VALID_MASK = 0x8000000000000000ul, + IOST_TAG_MASK = 0x3000000000000000ul, + IOST_PT_BASE_MASK = 0x000003fffffff000ul, + IOST_NNPT_MASK = 0x0000000000000fe0ul, + IOST_PS_MASK = 0x000000000000000ful, + + IOST_PS_4K = 0x1, + IOST_PS_64K = 0x3, + IOST_PS_1M = 0x5, + IOST_PS_16M = 0x7, + + /* iopt tag register */ + IOPT_VALID_MASK = 0x0000000200000000ul, + IOPT_TAG_MASK = 0x00000001fffffffful, + + /* iopt cache register */ + IOPT_PROT_MASK = 0xc000000000000000ul, + IOPT_PROT_NONE = 0x0000000000000000ul, + IOPT_PROT_READ = 0x4000000000000000ul, + IOPT_PROT_WRITE = 0x8000000000000000ul, + IOPT_PROT_RW = 0xc000000000000000ul, + IOPT_COHERENT = 0x2000000000000000ul, + + IOPT_ORDER_MASK = 0x1800000000000000ul, + /* order access to same IOID/VC on same address */ + IOPT_ORDER_ADDR = 0x0800000000000000ul, + /* similar, but only after a write access */ + IOPT_ORDER_WRITES = 0x1000000000000000ul, + /* Order all accesses to same IOID/VC */ + IOPT_ORDER_VC = 0x1800000000000000ul, + + IOPT_RPN_MASK = 0x000003fffffff000ul, + IOPT_HINT_MASK = 0x0000000000000800ul, + IOPT_IOID_MASK = 0x00000000000007fful, + + IOSTO_ENABLE = 0x8000000000000000ul, + IOSTO_ORIGIN = 0x000003fffffff000ul, + IOSTO_HW = 0x0000000000000800ul, + IOSTO_SW = 0x0000000000000400ul, + + IOCMD_CONF_TE = 0x0000800000000000ul, + + /* memory mapped registers */ + IOC_PT_CACHE_DIR = 0x000, + IOC_ST_CACHE_DIR = 0x800, + IOC_PT_CACHE_REG = 0x910, + IOC_ST_ORIGIN = 0x918, + IOC_CONF = 0x930, + + /* The high bit needs to be set on every DMA address, + only 2GB are addressable */ + BPA_DMA_VALID = 0x80000000, + BPA_DMA_MASK = 0x7fffffff, +}; + + +void bpa_init_iommu(void); + +#endif diff --git a/arch/ppc64/kernel/bpa_nvram.c b/arch/ppc64/kernel/bpa_nvram.c new file mode 100644 index 00000000000..06a119cfceb --- /dev/null +++ b/arch/ppc64/kernel/bpa_nvram.c @@ -0,0 +1,118 @@ +/* + * NVRAM for CPBW + * + * (C) Copyright IBM Corp. 2005 + * + * Authors : Utz Bacher <utz.bacher@de.ibm.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/spinlock.h> +#include <linux/types.h> + +#include <asm/machdep.h> +#include <asm/nvram.h> +#include <asm/prom.h> + +static void __iomem *bpa_nvram_start; +static long bpa_nvram_len; +static spinlock_t bpa_nvram_lock = SPIN_LOCK_UNLOCKED; + +static ssize_t bpa_nvram_read(char *buf, size_t count, loff_t *index) +{ + unsigned long flags; + + if (*index >= bpa_nvram_len) + return 0; + if (*index + count > bpa_nvram_len) + count = bpa_nvram_len - *index; + + spin_lock_irqsave(&bpa_nvram_lock, flags); + + memcpy_fromio(buf, bpa_nvram_start + *index, count); + + spin_unlock_irqrestore(&bpa_nvram_lock, flags); + + *index += count; + return count; +} + +static ssize_t bpa_nvram_write(char *buf, size_t count, loff_t *index) +{ + unsigned long flags; + + if (*index >= bpa_nvram_len) + return 0; + if (*index + count > bpa_nvram_len) + count = bpa_nvram_len - *index; + + spin_lock_irqsave(&bpa_nvram_lock, flags); + + memcpy_toio(bpa_nvram_start + *index, buf, count); + + spin_unlock_irqrestore(&bpa_nvram_lock, flags); + + *index += count; + return count; +} + +static ssize_t bpa_nvram_get_size(void) +{ + return bpa_nvram_len; +} + +int __init bpa_nvram_init(void) +{ + struct device_node *nvram_node; + unsigned long *buffer; + int proplen; + unsigned long nvram_addr; + int ret; + + ret = -ENODEV; + nvram_node = of_find_node_by_type(NULL, "nvram"); + if (!nvram_node) + goto out; + + ret = -EIO; + buffer = (unsigned long *)get_property(nvram_node, "reg", &proplen); + if (proplen != 2*sizeof(unsigned long)) + goto out; + + ret = -ENODEV; + nvram_addr = buffer[0]; + bpa_nvram_len = buffer[1]; + if ( (!bpa_nvram_len) || (!nvram_addr) ) + goto out; + + bpa_nvram_start = ioremap(nvram_addr, bpa_nvram_len); + if (!bpa_nvram_start) + goto out; + + printk(KERN_INFO "BPA NVRAM, %luk mapped to %p\n", + bpa_nvram_len >> 10, bpa_nvram_start); + + ppc_md.nvram_read = bpa_nvram_read; + ppc_md.nvram_write = bpa_nvram_write; + ppc_md.nvram_size = bpa_nvram_get_size; + +out: + of_node_put(nvram_node); + return ret; +} diff --git a/arch/ppc64/kernel/bpa_setup.c b/arch/ppc64/kernel/bpa_setup.c new file mode 100644 index 00000000000..57b3db66f45 --- /dev/null +++ b/arch/ppc64/kernel/bpa_setup.c @@ -0,0 +1,140 @@ +/* + * linux/arch/ppc/kernel/bpa_setup.c + * + * Copyright (C) 1995 Linus Torvalds + * Adapted from 'alpha' version by Gary Thomas + * Modified by Cort Dougan (cort@cs.nmt.edu) + * Modified by PPC64 Team, IBM Corp + * Modified by BPA Team, IBM Deutschland Entwicklung GmbH + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#undef DEBUG + +#include <linux/config.h> +#include <linux/sched.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/stddef.h> +#include <linux/unistd.h> +#include <linux/slab.h> +#include <linux/user.h> +#include <linux/reboot.h> +#include <linux/init.h> +#include <linux/delay.h> +#include <linux/irq.h> +#include <linux/seq_file.h> +#include <linux/root_dev.h> +#include <linux/console.h> + +#include <asm/mmu.h> +#include <asm/processor.h> +#include <asm/io.h> +#include <asm/pgtable.h> +#include <asm/prom.h> +#include <asm/rtas.h> +#include <asm/pci-bridge.h> +#include <asm/iommu.h> +#include <asm/dma.h> +#include <asm/machdep.h> +#include <asm/time.h> +#include <asm/nvram.h> +#include <asm/cputable.h> + +#include "pci.h" +#include "bpa_iic.h" +#include "bpa_iommu.h" + +#ifdef DEBUG +#define DBG(fmt...) udbg_printf(fmt) +#else +#define DBG(fmt...) +#endif + +void bpa_get_cpuinfo(struct seq_file *m) +{ + struct device_node *root; + const char *model = ""; + + root = of_find_node_by_path("/"); + if (root) + model = get_property(root, "model", NULL); + seq_printf(m, "machine\t\t: BPA %s\n", model); + of_node_put(root); +} + +static void bpa_progress(char *s, unsigned short hex) +{ + printk("*** %04x : %s\n", hex, s ? s : ""); +} + +static void __init bpa_setup_arch(void) +{ + ppc_md.init_IRQ = iic_init_IRQ; + ppc_md.get_irq = iic_get_irq; + +#ifdef CONFIG_SMP + smp_init_pSeries(); +#endif + + /* init to some ~sane value until calibrate_delay() runs */ + loops_per_jiffy = 50000000; + + if (ROOT_DEV == 0) { + printk("No ramdisk, default root is /dev/hda2\n"); + ROOT_DEV = Root_HDA2; + } + + /* Find and initialize PCI host bridges */ + init_pci_config_tokens(); + find_and_init_phbs(); + spider_init_IRQ(); +#ifdef CONFIG_DUMMY_CONSOLE + conswitchp = &dummy_con; +#endif + + bpa_nvram_init(); +} + +/* + * Early initialization. Relocation is on but do not reference unbolted pages + */ +static void __init bpa_init_early(void) +{ + DBG(" -> bpa_init_early()\n"); + + hpte_init_native(); + + bpa_init_iommu(); + + ppc64_interrupt_controller = IC_BPA_IIC; + + DBG(" <- bpa_init_early()\n"); +} + + +static int __init bpa_probe(int platform) +{ + if (platform != PLATFORM_BPA) + return 0; + + return 1; +} + +struct machdep_calls __initdata bpa_md = { + .probe = bpa_probe, + .setup_arch = bpa_setup_arch, + .init_early = bpa_init_early, + .get_cpuinfo = bpa_get_cpuinfo, + .restart = rtas_restart, + .power_off = rtas_power_off, + .halt = rtas_halt, + .get_boot_time = rtas_get_boot_time, + .get_rtc_time = rtas_get_rtc_time, + .set_rtc_time = rtas_set_rtc_time, + .calibrate_decr = generic_calibrate_decr, + .progress = bpa_progress, +}; diff --git a/arch/ppc64/kernel/cpu_setup_power4.S b/arch/ppc64/kernel/cpu_setup_power4.S index 3bd95182085..42fc08cf87a 100644 --- a/arch/ppc64/kernel/cpu_setup_power4.S +++ b/arch/ppc64/kernel/cpu_setup_power4.S @@ -73,7 +73,21 @@ _GLOBAL(__970_cpu_preinit) _GLOBAL(__setup_cpu_power4) blr - + +_GLOBAL(__setup_cpu_be) + /* Set large page sizes LP=0: 16MB, LP=1: 64KB */ + addi r3, 0, 0 + ori r3, r3, HID6_LB + sldi r3, r3, 32 + nor r3, r3, r3 + mfspr r4, SPRN_HID6 + and r4, r4, r3 + addi r3, 0, 0x02000 + sldi r3, r3, 32 + or r4, r4, r3 + mtspr SPRN_HID6, r4 + blr + _GLOBAL(__setup_cpu_ppc970) mfspr r0,SPRN_HID0 li r11,5 /* clear DOZE and SLEEP */ diff --git a/arch/ppc64/kernel/cputable.c b/arch/ppc64/kernel/cputable.c index 8644a864805..1d162c7c59d 100644 --- a/arch/ppc64/kernel/cputable.c +++ b/arch/ppc64/kernel/cputable.c @@ -34,6 +34,7 @@ EXPORT_SYMBOL(cur_cpu_spec); extern void __setup_cpu_power3(unsigned long offset, struct cpu_spec* spec); extern void __setup_cpu_power4(unsigned long offset, struct cpu_spec* spec); extern void __setup_cpu_ppc970(unsigned long offset, struct cpu_spec* spec); +extern void __setup_cpu_be(unsigned long offset, struct cpu_spec* spec); /* We only set the altivec features if the kernel was compiled with altivec @@ -162,6 +163,16 @@ struct cpu_spec cpu_specs[] = { __setup_cpu_power4, COMMON_PPC64_FW }, + { /* BE DD1.x */ + 0xffff0000, 0x00700000, "Broadband Engine", + CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | CPU_FTR_HPTE_TABLE | + CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_ALTIVEC_COMP | + CPU_FTR_SMT, + COMMON_USER_PPC64 | PPC_FEATURE_HAS_ALTIVEC_COMP, + 128, 128, + __setup_cpu_be, + COMMON_PPC64_FW + }, { /* default match */ 0x00000000, 0x00000000, "POWER4 (compatible)", CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | CPU_FTR_HPTE_TABLE | diff --git a/arch/ppc64/kernel/head.S b/arch/ppc64/kernel/head.S index 02c8f4e3e4b..675c2708588 100644 --- a/arch/ppc64/kernel/head.S +++ b/arch/ppc64/kernel/head.S @@ -1194,7 +1194,7 @@ _GLOBAL(pSeries_secondary_smp_init) bl .__restore_cpu_setup /* Set up a paca value for this processor. Since we have the - * physical cpu id in r3, we need to search the pacas to find + * physical cpu id in r24, we need to search the pacas to find * which logical id maps to our physical one. */ LOADADDR(r13, paca) /* Get base vaddr of paca array */ @@ -1207,8 +1207,8 @@ _GLOBAL(pSeries_secondary_smp_init) cmpwi r5,NR_CPUS blt 1b -99: HMT_LOW /* Couldn't find our CPU id */ - b 99b + mr r3,r24 /* not found, copy phys to r3 */ + b .kexec_wait /* next kernel might do better */ 2: mtspr SPRG3,r13 /* Save vaddr of paca in SPRG3 */ /* From now on, r24 is expected to be logica cpuid */ diff --git a/arch/ppc64/kernel/iSeries_setup.c b/arch/ppc64/kernel/iSeries_setup.c index b31962436fe..86966ce76b5 100644 --- a/arch/ppc64/kernel/iSeries_setup.c +++ b/arch/ppc64/kernel/iSeries_setup.c @@ -671,9 +671,6 @@ static void __init iSeries_bolt_kernel(unsigned long saddr, unsigned long eaddr) } } -extern unsigned long ppc_proc_freq; -extern unsigned long ppc_tb_freq; - /* * Document me. */ @@ -772,8 +769,6 @@ static void iSeries_halt(void) mf_power_off(); } -extern void setup_default_decr(void); - /* * void __init iSeries_calibrate_decr() * diff --git a/arch/ppc64/kernel/irq.c b/arch/ppc64/kernel/irq.c index d860467b8f0..3defc8c33ad 100644 --- a/arch/ppc64/kernel/irq.c +++ b/arch/ppc64/kernel/irq.c @@ -395,6 +395,9 @@ int virt_irq_create_mapping(unsigned int real_irq) if (ppc64_interrupt_controller == IC_OPEN_PIC) return real_irq; /* no mapping for openpic (for now) */ + if (ppc64_interrupt_controller == IC_BPA_IIC) + return real_irq; /* no mapping for iic either */ + /* don't map interrupts < MIN_VIRT_IRQ */ if (real_irq < MIN_VIRT_IRQ) { virt_irq_to_real_map[real_irq] = real_irq; diff --git a/arch/ppc64/kernel/kprobes.c b/arch/ppc64/kernel/kprobes.c index e950a2058a1..782ce3efa2c 100644 --- a/arch/ppc64/kernel/kprobes.c +++ b/arch/ppc64/kernel/kprobes.c @@ -32,15 +32,14 @@ #include <linux/ptrace.h> #include <linux/spinlock.h> #include <linux/preempt.h> +#include <asm/cacheflush.h> #include <asm/kdebug.h> #include <asm/sstep.h> -/* kprobe_status settings */ -#define KPROBE_HIT_ACTIVE 0x00000001 -#define KPROBE_HIT_SS 0x00000002 - static struct kprobe *current_kprobe; static unsigned long kprobe_status, kprobe_saved_msr; +static struct kprobe *kprobe_prev; +static unsigned long kprobe_status_prev, kprobe_saved_msr_prev; static struct pt_regs jprobe_saved_regs; int arch_prepare_kprobe(struct kprobe *p) @@ -61,16 +60,25 @@ int arch_prepare_kprobe(struct kprobe *p) void arch_copy_kprobe(struct kprobe *p) { memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + p->opcode = *p->addr; } -void arch_remove_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { + *p->addr = BREAKPOINT_INSTRUCTION; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) +void arch_disarm_kprobe(struct kprobe *p) { *p->addr = p->opcode; - regs->nip = (unsigned long)p->addr; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); +} + +void arch_remove_kprobe(struct kprobe *p) +{ } static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -83,6 +91,20 @@ static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) regs->nip = (unsigned long)&p->ainsn.insn; } +static inline void save_previous_kprobe(void) +{ + kprobe_prev = current_kprobe; + kprobe_status_prev = kprobe_status; + kprobe_saved_msr_prev = kprobe_saved_msr; +} + +static inline void restore_previous_kprobe(void) +{ + current_kprobe = kprobe_prev; + kprobe_status = kprobe_status_prev; + kprobe_saved_msr = kprobe_saved_msr_prev; +} + static inline int kprobe_handler(struct pt_regs *regs) { struct kprobe *p; @@ -101,8 +123,19 @@ static inline int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); - ret = 1; + /* We have reentered the kprobe_handler(), since + * another probe was hit while within the handler. + * We here save the original kprobes variables and + * just single step on the instruction of the new probe + * without calling any user handlers. + */ + save_previous_kprobe(); + current_kprobe = p; + kprobe_saved_msr = regs->msr; + p->nmissed++; + prepare_singlestep(p, regs); + kprobe_status = KPROBE_REENTER; + return 1; } else { p = current_kprobe; if (p->break_handler && p->break_handler(p, regs)) { @@ -184,13 +217,21 @@ static inline int post_kprobe_handler(struct pt_regs *regs) if (!kprobe_running()) return 0; - if (current_kprobe->post_handler) + if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) { + kprobe_status = KPROBE_HIT_SSDONE; current_kprobe->post_handler(current_kprobe, regs, 0); + } resume_execution(current_kprobe, regs); regs->msr |= kprobe_saved_msr; + /*Restore back the original saved kprobes variables and continue. */ + if (kprobe_status == KPROBE_REENTER) { + restore_previous_kprobe(); + goto out; + } unlock_kprobes(); +out: preempt_enable_no_resched(); /* diff --git a/arch/ppc64/kernel/lparcfg.c b/arch/ppc64/kernel/lparcfg.c index 387923fcf9b..02e96627fa6 100644 --- a/arch/ppc64/kernel/lparcfg.c +++ b/arch/ppc64/kernel/lparcfg.c @@ -34,6 +34,7 @@ #include <asm/system.h> #include <asm/time.h> #include <asm/iSeries/ItExtVpdPanel.h> +#include <asm/prom.h> #define MODULE_VERS "1.6" #define MODULE_NAME "lparcfg" diff --git a/arch/ppc64/kernel/machine_kexec.c b/arch/ppc64/kernel/machine_kexec.c new file mode 100644 index 00000000000..fdb2fc649d7 --- /dev/null +++ b/arch/ppc64/kernel/machine_kexec.c @@ -0,0 +1,302 @@ +/* + * machine_kexec.c - handle transition of Linux booting another kernel + * + * Copyright (C) 2004-2005, IBM Corp. + * + * Created by: Milton D Miller II + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + + +#include <linux/cpumask.h> +#include <linux/kexec.h> +#include <linux/smp.h> +#include <linux/thread_info.h> +#include <linux/errno.h> + +#include <asm/page.h> +#include <asm/current.h> +#include <asm/machdep.h> +#include <asm/cacheflush.h> +#include <asm/paca.h> +#include <asm/mmu.h> +#include <asm/sections.h> /* _end */ +#include <asm/prom.h> + +#define HASH_GROUP_SIZE 0x80 /* size of each hash group, asm/mmu.h */ + +/* Have this around till we move it into crash specific file */ +note_buf_t crash_notes[NR_CPUS]; + +/* Dummy for now. Not sure if we need to have a crash shutdown in here + * and if what it will achieve. Letting it be now to compile the code + * in generic kexec environment + */ +void machine_crash_shutdown(struct pt_regs *regs) +{ + /* do nothing right now */ + /* smp_relase_cpus() if we want smp on panic kernel */ + /* cpu_irq_down to isolate us until we are ready */ +} + +int machine_kexec_prepare(struct kimage *image) +{ + int i; + unsigned long begin, end; /* limits of segment */ + unsigned long low, high; /* limits of blocked memory range */ + struct device_node *node; + unsigned long *basep; + unsigned int *sizep; + + if (!ppc_md.hpte_clear_all) + return -ENOENT; + + /* + * Since we use the kernel fault handlers and paging code to + * handle the virtual mode, we must make sure no destination + * overlaps kernel static data or bss. + */ + for (i = 0; i < image->nr_segments; i++) + if (image->segment[i].mem < __pa(_end)) + return -ETXTBSY; + + /* + * For non-LPAR, we absolutely can not overwrite the mmu hash + * table, since we are still using the bolted entries in it to + * do the copy. Check that here. + * + * It is safe if the end is below the start of the blocked + * region (end <= low), or if the beginning is after the + * end of the blocked region (begin >= high). Use the + * boolean identity !(a || b) === (!a && !b). + */ + if (htab_address) { + low = __pa(htab_address); + high = low + (htab_hash_mask + 1) * HASH_GROUP_SIZE; + + for (i = 0; i < image->nr_segments; i++) { + begin = image->segment[i].mem; + end = begin + image->segment[i].memsz; + + if ((begin < high) && (end > low)) + return -ETXTBSY; + } + } + + /* We also should not overwrite the tce tables */ + for (node = of_find_node_by_type(NULL, "pci"); node != NULL; + node = of_find_node_by_type(node, "pci")) { + basep = (unsigned long *)get_property(node, "linux,tce-base", + NULL); + sizep = (unsigned int *)get_property(node, "linux,tce-size", + NULL); + if (basep == NULL || sizep == NULL) + continue; + + low = *basep; + high = low + (*sizep); + + for (i = 0; i < image->nr_segments; i++) { + begin = image->segment[i].mem; + end = begin + image->segment[i].memsz; + + if ((begin < high) && (end > low)) + return -ETXTBSY; + } + } + + return 0; +} + +void machine_kexec_cleanup(struct kimage *image) +{ + /* we do nothing in prepare that needs to be undone */ +} + +#define IND_FLAGS (IND_DESTINATION | IND_INDIRECTION | IND_DONE | IND_SOURCE) + +static void copy_segments(unsigned long ind) +{ + unsigned long entry; + unsigned long *ptr; + void *dest; + void *addr; + + /* + * We rely on kexec_load to create a lists that properly + * initializes these pointers before they are used. + * We will still crash if the list is wrong, but at least + * the compiler will be quiet. + */ + ptr = NULL; + dest = NULL; + + for (entry = ind; !(entry & IND_DONE); entry = *ptr++) { + addr = __va(entry & PAGE_MASK); + + switch (entry & IND_FLAGS) { + case IND_DESTINATION: + dest = addr; + break; + case IND_INDIRECTION: + ptr = addr; + break; + case IND_SOURCE: + copy_page(dest, addr); + dest += PAGE_SIZE; + } + } +} + +void kexec_copy_flush(struct kimage *image) +{ + long i, nr_segments = image->nr_segments; + struct kexec_segment ranges[KEXEC_SEGMENT_MAX]; + + /* save the ranges on the stack to efficiently flush the icache */ + memcpy(ranges, image->segment, sizeof(ranges)); + + /* + * After this call we may not use anything allocated in dynamic + * memory, including *image. + * + * Only globals and the stack are allowed. + */ + copy_segments(image->head); + + /* + * we need to clear the icache for all dest pages sometime, + * including ones that were in place on the original copy + */ + for (i = 0; i < nr_segments; i++) + flush_icache_range(ranges[i].mem + KERNELBASE, + ranges[i].mem + KERNELBASE + + ranges[i].memsz); +} + +#ifdef CONFIG_SMP + +/* FIXME: we should schedule this function to be called on all cpus based + * on calling the interrupts, but we would like to call it off irq level + * so that the interrupt controller is clean. + */ +void kexec_smp_down(void *arg) +{ + if (ppc_md.cpu_irq_down) + ppc_md.cpu_irq_down(); + + local_irq_disable(); + kexec_smp_wait(); + /* NOTREACHED */ +} + +static void kexec_prepare_cpus(void) +{ + int my_cpu, i, notified=-1; + + smp_call_function(kexec_smp_down, NULL, 0, /* wait */0); + my_cpu = get_cpu(); + + /* check the others cpus are now down (via paca hw cpu id == -1) */ + for (i=0; i < NR_CPUS; i++) { + if (i == my_cpu) + continue; + + while (paca[i].hw_cpu_id != -1) { + if (!cpu_possible(i)) { + printk("kexec: cpu %d hw_cpu_id %d is not" + " possible, ignoring\n", + i, paca[i].hw_cpu_id); + break; + } + if (!cpu_online(i)) { + /* Fixme: this can be spinning in + * pSeries_secondary_wait with a paca + * waiting for it to go online. + */ + printk("kexec: cpu %d hw_cpu_id %d is not" + " online, ignoring\n", + i, paca[i].hw_cpu_id); + break; + } + if (i != notified) { + printk( "kexec: waiting for cpu %d (physical" + " %d) to go down\n", + i, paca[i].hw_cpu_id); + notified = i; + } + } + } + + /* after we tell the others to go down */ + if (ppc_md.cpu_irq_down) + ppc_md.cpu_irq_down(); + + put_cpu(); + + local_irq_disable(); +} + +#else /* ! SMP */ + +static void kexec_prepare_cpus(void) +{ + /* + * move the secondarys to us so that we can copy + * the new kernel 0-0x100 safely + * + * do this if kexec in setup.c ? + */ + smp_relase_cpus(); + if (ppc_md.cpu_irq_down) + ppc_md.cpu_irq_down(); + local_irq_disable(); +} + +#endif /* SMP */ + +/* + * kexec thread structure and stack. + * + * We need to make sure that this is 16384-byte aligned due to the + * way process stacks are handled. It also must be statically allocated + * or allocated as part of the kimage, because everything else may be + * overwritten when we copy the kexec image. We piggyback on the + * "init_task" linker section here to statically allocate a stack. + * + * We could use a smaller stack if we don't care about anything using + * current, but that audit has not been performed. + */ +union thread_union kexec_stack + __attribute__((__section__(".data.init_task"))) = { }; + +/* Our assembly helper, in kexec_stub.S */ +extern NORET_TYPE void kexec_sequence(void *newstack, unsigned long start, + void *image, void *control, + void (*clear_all)(void)) ATTRIB_NORET; + +/* too late to fail here */ +void machine_kexec(struct kimage *image) +{ + + /* prepare control code if any */ + + /* shutdown other cpus into our wait loop and quiesce interrupts */ + kexec_prepare_cpus(); + + /* switch to a staticly allocated stack. Based on irq stack code. + * XXX: the task struct will likely be invalid once we do the copy! + */ + kexec_stack.thread_info.task = current_thread_info()->task; + kexec_stack.thread_info.flags = 0; + + /* Some things are best done in assembly. Finding globals with + * a toc is easier in C, so pass in what we can. + */ + kexec_sequence(&kexec_stack, image->start, image, + page_address(image->control_code_page), + ppc_md.hpte_clear_all); + /* NOTREACHED */ +} diff --git a/arch/ppc64/kernel/maple_setup.c b/arch/ppc64/kernel/maple_setup.c index 8cf95a27178..da8900b51f4 100644 --- a/arch/ppc64/kernel/maple_setup.c +++ b/arch/ppc64/kernel/maple_setup.c @@ -78,17 +78,77 @@ extern int maple_pci_get_legacy_ide_irq(struct pci_dev *dev, int channel); extern void generic_find_legacy_serial_ports(u64 *physport, unsigned int *default_speed); - static void maple_restart(char *cmd) { + unsigned int maple_nvram_base; + unsigned int maple_nvram_offset; + unsigned int maple_nvram_command; + struct device_node *rtcs; + + /* find NVRAM device */ + rtcs = find_compatible_devices("nvram", "AMD8111"); + if (rtcs && rtcs->addrs) { + maple_nvram_base = rtcs->addrs[0].address; + } else { + printk(KERN_EMERG "Maple: Unable to find NVRAM\n"); + printk(KERN_EMERG "Maple: Manual Restart Required\n"); + return; + } + + /* find service processor device */ + rtcs = find_devices("service-processor"); + if (!rtcs) { + printk(KERN_EMERG "Maple: Unable to find Service Processor\n"); + printk(KERN_EMERG "Maple: Manual Restart Required\n"); + return; + } + maple_nvram_offset = *(unsigned int*) get_property(rtcs, + "restart-addr", NULL); + maple_nvram_command = *(unsigned int*) get_property(rtcs, + "restart-value", NULL); + + /* send command */ + outb_p(maple_nvram_command, maple_nvram_base + maple_nvram_offset); + for (;;) ; } static void maple_power_off(void) { + unsigned int maple_nvram_base; + unsigned int maple_nvram_offset; + unsigned int maple_nvram_command; + struct device_node *rtcs; + + /* find NVRAM device */ + rtcs = find_compatible_devices("nvram", "AMD8111"); + if (rtcs && rtcs->addrs) { + maple_nvram_base = rtcs->addrs[0].address; + } else { + printk(KERN_EMERG "Maple: Unable to find NVRAM\n"); + printk(KERN_EMERG "Maple: Manual Power-Down Required\n"); + return; + } + + /* find service processor device */ + rtcs = find_devices("service-processor"); + if (!rtcs) { + printk(KERN_EMERG "Maple: Unable to find Service Processor\n"); + printk(KERN_EMERG "Maple: Manual Power-Down Required\n"); + return; + } + maple_nvram_offset = *(unsigned int*) get_property(rtcs, + "power-off-addr", NULL); + maple_nvram_command = *(unsigned int*) get_property(rtcs, + "power-off-value", NULL); + + /* send command */ + outb_p(maple_nvram_command, maple_nvram_base + maple_nvram_offset); + for (;;) ; } static void maple_halt(void) { + maple_power_off(); } #ifdef CONFIG_SMP @@ -235,6 +295,6 @@ struct machdep_calls __initdata maple_md = { .get_boot_time = maple_get_boot_time, .set_rtc_time = maple_set_rtc_time, .get_rtc_time = maple_get_rtc_time, - .calibrate_decr = maple_calibrate_decr, + .calibrate_decr = generic_calibrate_decr, .progress = maple_progress, }; diff --git a/arch/ppc64/kernel/maple_time.c b/arch/ppc64/kernel/maple_time.c index 07ce7895b43..d65210abcd0 100644 --- a/arch/ppc64/kernel/maple_time.c +++ b/arch/ppc64/kernel/maple_time.c @@ -42,11 +42,8 @@ #define DBG(x...) #endif -extern void setup_default_decr(void); extern void GregorianDay(struct rtc_time * tm); -extern unsigned long ppc_tb_freq; -extern unsigned long ppc_proc_freq; static int maple_rtc_addr; static int maple_clock_read(int addr) @@ -176,51 +173,3 @@ void __init maple_get_boot_time(struct rtc_time *tm) maple_get_rtc_time(tm); } -/* XXX FIXME: Some sane defaults: 125 MHz timebase, 1GHz processor */ -#define DEFAULT_TB_FREQ 125000000UL -#define DEFAULT_PROC_FREQ (DEFAULT_TB_FREQ * 8) - -void __init maple_calibrate_decr(void) -{ - struct device_node *cpu; - struct div_result divres; - unsigned int *fp = NULL; - - /* - * The cpu node should have a timebase-frequency property - * to tell us the rate at which the decrementer counts. - */ - cpu = of_find_node_by_type(NULL, "cpu"); - - ppc_tb_freq = DEFAULT_TB_FREQ; - if (cpu != 0) - fp = (unsigned int *)get_property(cpu, "timebase-frequency", NULL); - if (fp != NULL) - ppc_tb_freq = *fp; - else - printk(KERN_ERR "WARNING: Estimating decrementer frequency (not found)\n"); - fp = NULL; - ppc_proc_freq = DEFAULT_PROC_FREQ; - if (cpu != 0) - fp = (unsigned int *)get_property(cpu, "clock-frequency", NULL); - if (fp != NULL) - ppc_proc_freq = *fp; - else - printk(KERN_ERR "WARNING: Estimating processor frequency (not found)\n"); - - of_node_put(cpu); - - printk(KERN_INFO "time_init: decrementer frequency = %lu.%.6lu MHz\n", - ppc_tb_freq/1000000, ppc_tb_freq%1000000); - printk(KERN_INFO "time_init: processor frequency = %lu.%.6lu MHz\n", - ppc_proc_freq/1000000, ppc_proc_freq%1000000); - - tb_ticks_per_jiffy = ppc_tb_freq / HZ; - tb_ticks_per_sec = tb_ticks_per_jiffy * HZ; - tb_ticks_per_usec = ppc_tb_freq / 1000000; - tb_to_us = mulhwu_scale_factor(ppc_tb_freq, 1000000); - div128_by_32(1024*1024, 0, tb_ticks_per_sec, &divres); - tb_to_xs = divres.result_low; - - setup_default_decr(); -} diff --git a/arch/ppc64/kernel/misc.S b/arch/ppc64/kernel/misc.S index e3c73b3425d..f3dea0c5a88 100644 --- a/arch/ppc64/kernel/misc.S +++ b/arch/ppc64/kernel/misc.S @@ -680,6 +680,177 @@ _GLOBAL(kernel_thread) ld r30,-16(r1) blr +/* kexec_wait(phys_cpu) + * + * wait for the flag to change, indicating this kernel is going away but + * the slave code for the next one is at addresses 0 to 100. + * + * This is used by all slaves. + * + * Physical (hardware) cpu id should be in r3. + */ +_GLOBAL(kexec_wait) + bl 1f +1: mflr r5 + addi r5,r5,kexec_flag-1b + +99: HMT_LOW +#ifdef CONFIG_KEXEC /* use no memory without kexec */ + lwz r4,0(r5) + cmpwi 0,r4,0 + bnea 0x60 +#endif + b 99b + +/* this can be in text because we won't change it until we are + * running in real anyways + */ +kexec_flag: + .long 0 + + +#ifdef CONFIG_KEXEC + +/* kexec_smp_wait(void) + * + * call with interrupts off + * note: this is a terminal routine, it does not save lr + * + * get phys id from paca + * set paca id to -1 to say we got here + * switch to real mode + * join other cpus in kexec_wait(phys_id) + */ +_GLOBAL(kexec_smp_wait) + lhz r3,PACAHWCPUID(r13) + li r4,-1 + sth r4,PACAHWCPUID(r13) /* let others know we left */ + bl real_mode + b .kexec_wait + +/* + * switch to real mode (turn mmu off) + * we use the early kernel trick that the hardware ignores bits + * 0 and 1 (big endian) of the effective address in real mode + * + * don't overwrite r3 here, it is live for kexec_wait above. + */ +real_mode: /* assume normal blr return */ +1: li r9,MSR_RI + li r10,MSR_DR|MSR_IR + mflr r11 /* return address to SRR0 */ + mfmsr r12 + andc r9,r12,r9 + andc r10,r12,r10 + + mtmsrd r9,1 + mtspr SPRN_SRR1,r10 + mtspr SPRN_SRR0,r11 + rfid + + +/* + * kexec_sequence(newstack, start, image, control, clear_all()) + * + * does the grungy work with stack switching and real mode switches + * also does simple calls to other code + */ + +_GLOBAL(kexec_sequence) + mflr r0 + std r0,16(r1) + + /* switch stacks to newstack -- &kexec_stack.stack */ + stdu r1,THREAD_SIZE-112(r3) + mr r1,r3 + + li r0,0 + std r0,16(r1) + + /* save regs for local vars on new stack. + * yes, we won't go back, but ... + */ + std r31,-8(r1) + std r30,-16(r1) + std r29,-24(r1) + std r28,-32(r1) + std r27,-40(r1) + std r26,-48(r1) + std r25,-56(r1) + + stdu r1,-112-64(r1) + + /* save args into preserved regs */ + mr r31,r3 /* newstack (both) */ + mr r30,r4 /* start (real) */ + mr r29,r5 /* image (virt) */ + mr r28,r6 /* control, unused */ + mr r27,r7 /* clear_all() fn desc */ + mr r26,r8 /* spare */ + lhz r25,PACAHWCPUID(r13) /* get our phys cpu from paca */ + + /* disable interrupts, we are overwriting kernel data next */ + mfmsr r3 + rlwinm r3,r3,0,17,15 + mtmsrd r3,1 + + /* copy dest pages, flush whole dest image */ + mr r3,r29 + bl .kexec_copy_flush /* (image) */ + + /* turn off mmu */ + bl real_mode + + /* clear out hardware hash page table and tlb */ + ld r5,0(r27) /* deref function descriptor */ + mtctr r5 + bctrl /* ppc_md.hash_clear_all(void); */ + +/* + * kexec image calling is: + * the first 0x100 bytes of the entry point are copied to 0 + * + * all slaves branch to slave = 0x60 (absolute) + * slave(phys_cpu_id); + * + * master goes to start = entry point + * start(phys_cpu_id, start, 0); + * + * + * a wrapper is needed to call existing kernels, here is an approximate + * description of one method: + * + * v2: (2.6.10) + * start will be near the boot_block (maybe 0x100 bytes before it?) + * it will have a 0x60, which will b to boot_block, where it will wait + * and 0 will store phys into struct boot-block and load r3 from there, + * copy kernel 0-0x100 and tell slaves to back down to 0x60 again + * + * v1: (2.6.9) + * boot block will have all cpus scanning device tree to see if they + * are the boot cpu ????? + * other device tree differences (prop sizes, va vs pa, etc)... + */ + + /* copy 0x100 bytes starting at start to 0 */ + li r3,0 + mr r4,r30 + li r5,0x100 + li r6,0 + bl .copy_and_flush /* (dest, src, copy limit, start offset) */ +1: /* assume normal blr return */ + + /* release other cpus to the new kernel secondary start at 0x60 */ + mflr r5 + li r6,1 + stw r6,kexec_flag-1b(5) + mr r3,r25 # my phys cpu + mr r4,r30 # start, aka phys mem offset + mtlr 4 + li r5,0 + blr /* image->start(physid, image->start, 0); */ +#endif /* CONFIG_KEXEC */ + /* Why isn't this a) automatic, b) written in 'C'? */ .balign 8 _GLOBAL(sys_call_table32) @@ -951,7 +1122,7 @@ _GLOBAL(sys_call_table32) .llong .compat_sys_mq_timedreceive /* 265 */ .llong .compat_sys_mq_notify .llong .compat_sys_mq_getsetattr - .llong .sys_ni_syscall /* 268 reserved for sys_kexec_load */ + .llong .compat_sys_kexec_load .llong .sys32_add_key .llong .sys32_request_key .llong .compat_sys_keyctl @@ -1227,7 +1398,7 @@ _GLOBAL(sys_call_table) .llong .sys_mq_timedreceive /* 265 */ .llong .sys_mq_notify .llong .sys_mq_getsetattr - .llong .sys_ni_syscall /* 268 reserved for sys_kexec_load */ + .llong .sys_kexec_load .llong .sys_add_key .llong .sys_request_key /* 270 */ .llong .sys_keyctl diff --git a/arch/ppc64/kernel/mpic.c b/arch/ppc64/kernel/mpic.c index 593ea5b82af..e8fbab1df37 100644 --- a/arch/ppc64/kernel/mpic.c +++ b/arch/ppc64/kernel/mpic.c @@ -792,6 +792,35 @@ void mpic_setup_this_cpu(void) #endif /* CONFIG_SMP */ } +/* + * XXX: someone who knows mpic should check this. + * do we need to eoi the ipi here (see xics comments)? + * or can we reset the mpic in the new kernel? + */ +void mpic_teardown_this_cpu(void) +{ + struct mpic *mpic = mpic_primary; + unsigned long flags; + u32 msk = 1 << hard_smp_processor_id(); + unsigned int i; + + BUG_ON(mpic == NULL); + + DBG("%s: teardown_this_cpu(%d)\n", mpic->name, hard_smp_processor_id()); + spin_lock_irqsave(&mpic_lock, flags); + + /* let the mpic know we don't want intrs. */ + for (i = 0; i < mpic->num_sources ; i++) + mpic_irq_write(i, MPIC_IRQ_DESTINATION, + mpic_irq_read(i, MPIC_IRQ_DESTINATION) & ~msk); + + /* Set current processor priority to max */ + mpic_cpu_write(MPIC_CPU_CURRENT_TASK_PRI, 0xf); + + spin_unlock_irqrestore(&mpic_lock, flags); +} + + void mpic_send_ipi(unsigned int ipi_no, unsigned int cpu_mask) { struct mpic *mpic = mpic_primary; diff --git a/arch/ppc64/kernel/mpic.h b/arch/ppc64/kernel/mpic.h index 571b3c99e06..99fbbc9a084 100644 --- a/arch/ppc64/kernel/mpic.h +++ b/arch/ppc64/kernel/mpic.h @@ -255,6 +255,9 @@ extern unsigned int mpic_irq_get_priority(unsigned int irq); /* Setup a non-boot CPU */ extern void mpic_setup_this_cpu(void); +/* Clean up for kexec (or cpu offline or ...) */ +extern void mpic_teardown_this_cpu(void); + /* Request IPIs on primary mpic */ extern void mpic_request_ipis(void); @@ -265,3 +268,6 @@ extern void mpic_send_ipi(unsigned int ipi_no, unsigned int cpu_mask); extern int mpic_get_one_irq(struct mpic *mpic, struct pt_regs *regs); /* This one gets to the primary mpic */ extern int mpic_get_irq(struct pt_regs *regs); + +/* global mpic for pSeries */ +extern struct mpic *pSeries_mpic; diff --git a/arch/ppc64/kernel/pSeries_pci.c b/arch/ppc64/kernel/pSeries_pci.c index 0b1cca28140..1f5f141fb7a 100644 --- a/arch/ppc64/kernel/pSeries_pci.c +++ b/arch/ppc64/kernel/pSeries_pci.c @@ -1,13 +1,11 @@ /* - * pSeries_pci.c + * arch/ppc64/kernel/pSeries_pci.c * * Copyright (C) 2001 Dave Engebretsen, IBM Corporation * Copyright (C) 2003 Anton Blanchard <anton@au.ibm.com>, IBM * * pSeries specific routines for PCI. * - * Based on code from pci.c and chrp_pci.c - * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or @@ -23,430 +21,18 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include <linux/init.h> +#include <linux/ioport.h> #include <linux/kernel.h> -#include <linux/threads.h> #include <linux/pci.h> #include <linux/string.h> -#include <linux/init.h> -#include <linux/bootmem.h> -#include <asm/io.h> -#include <asm/pgtable.h> -#include <asm/irq.h> -#include <asm/prom.h> -#include <asm/machdep.h> #include <asm/pci-bridge.h> -#include <asm/iommu.h> -#include <asm/rtas.h> +#include <asm/prom.h> -#include "mpic.h" #include "pci.h" -/* RTAS tokens */ -static int read_pci_config; -static int write_pci_config; -static int ibm_read_pci_config; -static int ibm_write_pci_config; - -static int s7a_workaround; - -extern struct mpic *pSeries_mpic; - -static int config_access_valid(struct device_node *dn, int where) -{ - if (where < 256) - return 1; - if (where < 4096 && dn->pci_ext_config_space) - return 1; - - return 0; -} - -static int rtas_read_config(struct device_node *dn, int where, int size, u32 *val) -{ - int returnval = -1; - unsigned long buid, addr; - int ret; - - if (!dn) - return PCIBIOS_DEVICE_NOT_FOUND; - if (!config_access_valid(dn, where)) - return PCIBIOS_BAD_REGISTER_NUMBER; - - addr = ((where & 0xf00) << 20) | (dn->busno << 16) | - (dn->devfn << 8) | (where & 0xff); - buid = dn->phb->buid; - if (buid) { - ret = rtas_call(ibm_read_pci_config, 4, 2, &returnval, - addr, buid >> 32, buid & 0xffffffff, size); - } else { - ret = rtas_call(read_pci_config, 2, 2, &returnval, addr, size); - } - *val = returnval; - - if (ret) - return PCIBIOS_DEVICE_NOT_FOUND; - - if (returnval == EEH_IO_ERROR_VALUE(size) - && eeh_dn_check_failure (dn, NULL)) - return PCIBIOS_DEVICE_NOT_FOUND; - - return PCIBIOS_SUCCESSFUL; -} - -static int rtas_pci_read_config(struct pci_bus *bus, - unsigned int devfn, - int where, int size, u32 *val) -{ - struct device_node *busdn, *dn; - - if (bus->self) - busdn = pci_device_to_OF_node(bus->self); - else - busdn = bus->sysdata; /* must be a phb */ - - /* Search only direct children of the bus */ - for (dn = busdn->child; dn; dn = dn->sibling) - if (dn->devfn == devfn) - return rtas_read_config(dn, where, size, val); - return PCIBIOS_DEVICE_NOT_FOUND; -} - -static int rtas_write_config(struct device_node *dn, int where, int size, u32 val) -{ - unsigned long buid, addr; - int ret; - - if (!dn) - return PCIBIOS_DEVICE_NOT_FOUND; - if (!config_access_valid(dn, where)) - return PCIBIOS_BAD_REGISTER_NUMBER; - - addr = ((where & 0xf00) << 20) | (dn->busno << 16) | - (dn->devfn << 8) | (where & 0xff); - buid = dn->phb->buid; - if (buid) { - ret = rtas_call(ibm_write_pci_config, 5, 1, NULL, addr, buid >> 32, buid & 0xffffffff, size, (ulong) val); - } else { - ret = rtas_call(write_pci_config, 3, 1, NULL, addr, size, (ulong)val); - } - - if (ret) - return PCIBIOS_DEVICE_NOT_FOUND; - - return PCIBIOS_SUCCESSFUL; -} - -static int rtas_pci_write_config(struct pci_bus *bus, - unsigned int devfn, - int where, int size, u32 val) -{ - struct device_node *busdn, *dn; - - if (bus->self) - busdn = pci_device_to_OF_node(bus->self); - else - busdn = bus->sysdata; /* must be a phb */ - - /* Search only direct children of the bus */ - for (dn = busdn->child; dn; dn = dn->sibling) - if (dn->devfn == devfn) - return rtas_write_config(dn, where, size, val); - return PCIBIOS_DEVICE_NOT_FOUND; -} - -struct pci_ops rtas_pci_ops = { - rtas_pci_read_config, - rtas_pci_write_config -}; - -int is_python(struct device_node *dev) -{ - char *model = (char *)get_property(dev, "model", NULL); - - if (model && strstr(model, "Python")) - return 1; - - return 0; -} - -static int get_phb_reg_prop(struct device_node *dev, - unsigned int addr_size_words, - struct reg_property64 *reg) -{ - unsigned int *ui_ptr = NULL, len; - - /* Found a PHB, now figure out where his registers are mapped. */ - ui_ptr = (unsigned int *)get_property(dev, "reg", &len); - if (ui_ptr == NULL) - return 1; - - if (addr_size_words == 1) { - reg->address = ((struct reg_property32 *)ui_ptr)->address; - reg->size = ((struct reg_property32 *)ui_ptr)->size; - } else { - *reg = *((struct reg_property64 *)ui_ptr); - } - - return 0; -} - -static void python_countermeasures(struct device_node *dev, - unsigned int addr_size_words) -{ - struct reg_property64 reg_struct; - void __iomem *chip_regs; - volatile u32 val; - - if (get_phb_reg_prop(dev, addr_size_words, ®_struct)) - return; - - /* Python's register file is 1 MB in size. */ - chip_regs = ioremap(reg_struct.address & ~(0xfffffUL), 0x100000); - - /* - * Firmware doesn't always clear this bit which is critical - * for good performance - Anton - */ - -#define PRG_CL_RESET_VALID 0x00010000 - - val = in_be32(chip_regs + 0xf6030); - if (val & PRG_CL_RESET_VALID) { - printk(KERN_INFO "Python workaround: "); - val &= ~PRG_CL_RESET_VALID; - out_be32(chip_regs + 0xf6030, val); - /* - * We must read it back for changes to - * take effect - */ - val = in_be32(chip_regs + 0xf6030); - printk("reg0: %x\n", val); - } - - iounmap(chip_regs); -} - -void __init init_pci_config_tokens (void) -{ - read_pci_config = rtas_token("read-pci-config"); - write_pci_config = rtas_token("write-pci-config"); - ibm_read_pci_config = rtas_token("ibm,read-pci-config"); - ibm_write_pci_config = rtas_token("ibm,write-pci-config"); -} - -unsigned long __devinit get_phb_buid (struct device_node *phb) -{ - int addr_cells; - unsigned int *buid_vals; - unsigned int len; - unsigned long buid; - - if (ibm_read_pci_config == -1) return 0; - - /* PHB's will always be children of the root node, - * or so it is promised by the current firmware. */ - if (phb->parent == NULL) - return 0; - if (phb->parent->parent) - return 0; - - buid_vals = (unsigned int *) get_property(phb, "reg", &len); - if (buid_vals == NULL) - return 0; - - addr_cells = prom_n_addr_cells(phb); - if (addr_cells == 1) { - buid = (unsigned long) buid_vals[0]; - } else { - buid = (((unsigned long)buid_vals[0]) << 32UL) | - (((unsigned long)buid_vals[1]) & 0xffffffff); - } - return buid; -} - -static int phb_set_bus_ranges(struct device_node *dev, - struct pci_controller *phb) -{ - int *bus_range; - unsigned int len; - - bus_range = (int *) get_property(dev, "bus-range", &len); - if (bus_range == NULL || len < 2 * sizeof(int)) { - return 1; - } - - phb->first_busno = bus_range[0]; - phb->last_busno = bus_range[1]; - - return 0; -} - -static int __devinit setup_phb(struct device_node *dev, - struct pci_controller *phb, - unsigned int addr_size_words) -{ - pci_setup_pci_controller(phb); - - if (is_python(dev)) - python_countermeasures(dev, addr_size_words); - - if (phb_set_bus_ranges(dev, phb)) - return 1; - - phb->arch_data = dev; - phb->ops = &rtas_pci_ops; - phb->buid = get_phb_buid(dev); - - return 0; -} - -static void __devinit add_linux_pci_domain(struct device_node *dev, - struct pci_controller *phb, - struct property *of_prop) -{ - memset(of_prop, 0, sizeof(struct property)); - of_prop->name = "linux,pci-domain"; - of_prop->length = sizeof(phb->global_number); - of_prop->value = (unsigned char *)&of_prop[1]; - memcpy(of_prop->value, &phb->global_number, sizeof(phb->global_number)); - prom_add_property(dev, of_prop); -} - -static struct pci_controller * __init alloc_phb(struct device_node *dev, - unsigned int addr_size_words) -{ - struct pci_controller *phb; - struct property *of_prop; - - phb = alloc_bootmem(sizeof(struct pci_controller)); - if (phb == NULL) - return NULL; - - of_prop = alloc_bootmem(sizeof(struct property) + - sizeof(phb->global_number)); - if (!of_prop) - return NULL; - - if (setup_phb(dev, phb, addr_size_words)) - return NULL; - - add_linux_pci_domain(dev, phb, of_prop); - - return phb; -} - -static struct pci_controller * __devinit alloc_phb_dynamic(struct device_node *dev, unsigned int addr_size_words) -{ - struct pci_controller *phb; - - phb = (struct pci_controller *)kmalloc(sizeof(struct pci_controller), - GFP_KERNEL); - if (phb == NULL) - return NULL; - - if (setup_phb(dev, phb, addr_size_words)) - return NULL; - - phb->is_dynamic = 1; - - /* TODO: linux,pci-domain? */ - - return phb; -} - -unsigned long __init find_and_init_phbs(void) -{ - struct device_node *node; - struct pci_controller *phb; - unsigned int root_size_cells = 0; - unsigned int index; - unsigned int *opprop = NULL; - struct device_node *root = of_find_node_by_path("/"); - - if (ppc64_interrupt_controller == IC_OPEN_PIC) { - opprop = (unsigned int *)get_property(root, - "platform-open-pic", NULL); - } - - root_size_cells = prom_n_size_cells(root); - - index = 0; - - for (node = of_get_next_child(root, NULL); - node != NULL; - node = of_get_next_child(root, node)) { - if (node->type == NULL || strcmp(node->type, "pci") != 0) - continue; - - phb = alloc_phb(node, root_size_cells); - if (!phb) - continue; - - pci_process_bridge_OF_ranges(phb, node); - pci_setup_phb_io(phb, index == 0); - - if (ppc64_interrupt_controller == IC_OPEN_PIC && pSeries_mpic) { - int addr = root_size_cells * (index + 2) - 1; - mpic_assign_isu(pSeries_mpic, index, opprop[addr]); - } - - index++; - } - - of_node_put(root); - pci_devs_phb_init(); - - /* - * pci_probe_only and pci_assign_all_buses can be set via properties - * in chosen. - */ - if (of_chosen) { - int *prop; - - prop = (int *)get_property(of_chosen, "linux,pci-probe-only", - NULL); - if (prop) - pci_probe_only = *prop; - - prop = (int *)get_property(of_chosen, - "linux,pci-assign-all-buses", NULL); - if (prop) - pci_assign_all_buses = *prop; - } - - return 0; -} - -struct pci_controller * __devinit init_phb_dynamic(struct device_node *dn) -{ - struct device_node *root = of_find_node_by_path("/"); - unsigned int root_size_cells = 0; - struct pci_controller *phb; - struct pci_bus *bus; - int primary; - - root_size_cells = prom_n_size_cells(root); - - primary = list_empty(&hose_list); - phb = alloc_phb_dynamic(dn, root_size_cells); - if (!phb) - return NULL; - - pci_process_bridge_OF_ranges(phb, dn); - - pci_setup_phb_io_dynamic(phb, primary); - of_node_put(root); - - pci_devs_phb_init_dynamic(phb); - phb->last_busno = 0xff; - bus = pci_scan_bus(phb->first_busno, phb->ops, phb->arch_data); - phb->bus = bus; - phb->last_busno = bus->subordinate; - - return phb; -} -EXPORT_SYMBOL(init_phb_dynamic); +static int __initdata s7a_workaround = -1; #if 0 void pcibios_name_device(struct pci_dev *dev) @@ -474,11 +60,12 @@ void pcibios_name_device(struct pci_dev *dev) DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, pcibios_name_device); #endif -static void check_s7a(void) +static void __init check_s7a(void) { struct device_node *root; char *model; + s7a_workaround = 0; root = of_find_node_by_path("/"); if (root) { model = get_property(root, "model", NULL); @@ -488,55 +75,23 @@ static void check_s7a(void) } } -/* RPA-specific bits for removing PHBs */ -int pcibios_remove_root_bus(struct pci_controller *phb) +void __devinit pSeries_irq_bus_setup(struct pci_bus *bus) { - struct pci_bus *b = phb->bus; - struct resource *res; - int rc, i; - - res = b->resource[0]; - if (!res->flags) { - printk(KERN_ERR "%s: no IO resource for PHB %s\n", __FUNCTION__, - b->name); - return 1; - } - - rc = unmap_bus_range(b); - if (rc) { - printk(KERN_ERR "%s: failed to unmap IO on bus %s\n", - __FUNCTION__, b->name); - return 1; - } + struct pci_dev *dev; - if (release_resource(res)) { - printk(KERN_ERR "%s: failed to release IO on bus %s\n", - __FUNCTION__, b->name); - return 1; - } - - for (i = 1; i < 3; ++i) { - res = b->resource[i]; - if (!res->flags && i == 0) { - printk(KERN_ERR "%s: no MEM resource for PHB %s\n", - __FUNCTION__, b->name); - return 1; - } - if (res->flags && release_resource(res)) { - printk(KERN_ERR - "%s: failed to release IO %d on bus %s\n", - __FUNCTION__, i, b->name); - return 1; + if (s7a_workaround < 0) + check_s7a(); + list_for_each_entry(dev, &bus->devices, bus_list) { + pci_read_irq_line(dev); + if (s7a_workaround) { + if (dev->irq > 16) { + dev->irq -= 3; + pci_write_config_byte(dev, PCI_INTERRUPT_LINE, + dev->irq); + } } } - - list_del(&phb->list_node); - if (phb->is_dynamic) - kfree(phb); - - return 0; } -EXPORT_SYMBOL(pcibios_remove_root_bus); static void __init pSeries_request_regions(void) { @@ -553,20 +108,6 @@ static void __init pSeries_request_regions(void) void __init pSeries_final_fixup(void) { - struct pci_dev *dev = NULL; - - check_s7a(); - - for_each_pci_dev(dev) { - pci_read_irq_line(dev); - if (s7a_workaround) { - if (dev->irq > 16) { - dev->irq -= 3; - pci_write_config_byte(dev, PCI_INTERRUPT_LINE, dev->irq); - } - } - } - phbs_remap_io(); pSeries_request_regions(); diff --git a/arch/ppc64/kernel/pSeries_setup.c b/arch/ppc64/kernel/pSeries_setup.c index 6c0d1d58a55..44d9af72d22 100644 --- a/arch/ppc64/kernel/pSeries_setup.c +++ b/arch/ppc64/kernel/pSeries_setup.c @@ -71,11 +71,6 @@ #define DBG(fmt...) #endif -extern void pSeries_final_fixup(void); - -extern void pSeries_get_boot_time(struct rtc_time *rtc_time); -extern void pSeries_get_rtc_time(struct rtc_time *rtc_time); -extern int pSeries_set_rtc_time(struct rtc_time *rtc_time); extern void find_udbg_vterm(void); extern void system_reset_fwnmi(void); /* from head.S */ extern void machine_check_fwnmi(void); /* from head.S */ @@ -84,9 +79,6 @@ extern void generic_find_legacy_serial_ports(u64 *physport, int fwnmi_active; /* TRUE if an FWNMI handler is present */ -extern unsigned long ppc_proc_freq; -extern unsigned long ppc_tb_freq; - extern void pSeries_system_reset_exception(struct pt_regs *regs); extern int pSeries_machine_check_exception(struct pt_regs *regs); @@ -195,14 +187,16 @@ static void __init pSeries_setup_arch(void) { /* Fixup ppc_md depending on the type of interrupt controller */ if (ppc64_interrupt_controller == IC_OPEN_PIC) { - ppc_md.init_IRQ = pSeries_init_mpic; + ppc_md.init_IRQ = pSeries_init_mpic; ppc_md.get_irq = mpic_get_irq; + ppc_md.cpu_irq_down = mpic_teardown_this_cpu; /* Allocate the mpic now, so that find_and_init_phbs() can * fill the ISUs */ pSeries_setup_mpic(); } else { ppc_md.init_IRQ = xics_init_IRQ; ppc_md.get_irq = xics_get_irq; + ppc_md.cpu_irq_down = xics_teardown_cpu; } #ifdef CONFIG_SMP @@ -381,171 +375,6 @@ static void __init pSeries_init_early(void) } -static void pSeries_progress(char *s, unsigned short hex) -{ - struct device_node *root; - int width, *p; - char *os; - static int display_character, set_indicator; - static int max_width; - static DEFINE_SPINLOCK(progress_lock); - static int pending_newline = 0; /* did last write end with unprinted newline? */ - - if (!rtas.base) - return; - - if (max_width == 0) { - if ((root = find_path_device("/rtas")) && - (p = (unsigned int *)get_property(root, - "ibm,display-line-length", - NULL))) - max_width = *p; - else - max_width = 0x10; - display_character = rtas_token("display-character"); - set_indicator = rtas_token("set-indicator"); - } - - if (display_character == RTAS_UNKNOWN_SERVICE) { - /* use hex display if available */ - if (set_indicator != RTAS_UNKNOWN_SERVICE) - rtas_call(set_indicator, 3, 1, NULL, 6, 0, hex); - return; - } - - spin_lock(&progress_lock); - - /* - * Last write ended with newline, but we didn't print it since - * it would just clear the bottom line of output. Print it now - * instead. - * - * If no newline is pending, print a CR to start output at the - * beginning of the line. - */ - if (pending_newline) { - rtas_call(display_character, 1, 1, NULL, '\r'); - rtas_call(display_character, 1, 1, NULL, '\n'); - pending_newline = 0; - } else { - rtas_call(display_character, 1, 1, NULL, '\r'); - } - - width = max_width; - os = s; - while (*os) { - if (*os == '\n' || *os == '\r') { - /* Blank to end of line. */ - while (width-- > 0) - rtas_call(display_character, 1, 1, NULL, ' '); - - /* If newline is the last character, save it - * until next call to avoid bumping up the - * display output. - */ - if (*os == '\n' && !os[1]) { - pending_newline = 1; - spin_unlock(&progress_lock); - return; - } - - /* RTAS wants CR-LF, not just LF */ - - if (*os == '\n') { - rtas_call(display_character, 1, 1, NULL, '\r'); - rtas_call(display_character, 1, 1, NULL, '\n'); - } else { - /* CR might be used to re-draw a line, so we'll - * leave it alone and not add LF. - */ - rtas_call(display_character, 1, 1, NULL, *os); - } - - width = max_width; - } else { - width--; - rtas_call(display_character, 1, 1, NULL, *os); - } - - os++; - - /* if we overwrite the screen length */ - if (width <= 0) - while ((*os != 0) && (*os != '\n') && (*os != '\r')) - os++; - } - - /* Blank to end of line. */ - while (width-- > 0) - rtas_call(display_character, 1, 1, NULL, ' '); - - spin_unlock(&progress_lock); -} - -extern void setup_default_decr(void); - -/* Some sane defaults: 125 MHz timebase, 1GHz processor */ -#define DEFAULT_TB_FREQ 125000000UL -#define DEFAULT_PROC_FREQ (DEFAULT_TB_FREQ * 8) - -static void __init pSeries_calibrate_decr(void) -{ - struct device_node *cpu; - struct div_result divres; - unsigned int *fp; - int node_found; - - /* - * The cpu node should have a timebase-frequency property - * to tell us the rate at which the decrementer counts. - */ - cpu = of_find_node_by_type(NULL, "cpu"); - - ppc_tb_freq = DEFAULT_TB_FREQ; /* hardcoded default */ - node_found = 0; - if (cpu != 0) { - fp = (unsigned int *)get_property(cpu, "timebase-frequency", - NULL); - if (fp != 0) { - node_found = 1; - ppc_tb_freq = *fp; - } - } - if (!node_found) - printk(KERN_ERR "WARNING: Estimating decrementer frequency " - "(not found)\n"); - - ppc_proc_freq = DEFAULT_PROC_FREQ; - node_found = 0; - if (cpu != 0) { - fp = (unsigned int *)get_property(cpu, "clock-frequency", - NULL); - if (fp != 0) { - node_found = 1; - ppc_proc_freq = *fp; - } - } - if (!node_found) - printk(KERN_ERR "WARNING: Estimating processor frequency " - "(not found)\n"); - - of_node_put(cpu); - - printk(KERN_INFO "time_init: decrementer frequency = %lu.%.6lu MHz\n", - ppc_tb_freq/1000000, ppc_tb_freq%1000000); - printk(KERN_INFO "time_init: processor frequency = %lu.%.6lu MHz\n", - ppc_proc_freq/1000000, ppc_proc_freq%1000000); - - tb_ticks_per_jiffy = ppc_tb_freq / HZ; - tb_ticks_per_sec = tb_ticks_per_jiffy * HZ; - tb_ticks_per_usec = ppc_tb_freq / 1000000; - tb_to_us = mulhwu_scale_factor(ppc_tb_freq, 1000000); - div128_by_32(1024*1024, 0, tb_ticks_per_sec, &divres); - tb_to_xs = divres.result_low; - - setup_default_decr(); -} - static int pSeries_check_legacy_ioport(unsigned int baseport) { struct device_node *np; @@ -596,16 +425,17 @@ struct machdep_calls __initdata pSeries_md = { .get_cpuinfo = pSeries_get_cpuinfo, .log_error = pSeries_log_error, .pcibios_fixup = pSeries_final_fixup, + .irq_bus_setup = pSeries_irq_bus_setup, .restart = rtas_restart, .power_off = rtas_power_off, .halt = rtas_halt, .panic = rtas_os_term, .cpu_die = pSeries_mach_cpu_die, - .get_boot_time = pSeries_get_boot_time, - .get_rtc_time = pSeries_get_rtc_time, - .set_rtc_time = pSeries_set_rtc_time, - .calibrate_decr = pSeries_calibrate_decr, - .progress = pSeries_progress, + .get_boot_time = rtas_get_boot_time, + .get_rtc_time = rtas_get_rtc_time, + .set_rtc_time = rtas_set_rtc_time, + .calibrate_decr = generic_calibrate_decr, + .progress = rtas_progress, .check_legacy_ioport = pSeries_check_legacy_ioport, .system_reset_exception = pSeries_system_reset_exception, .machine_check_exception = pSeries_machine_check_exception, diff --git a/arch/ppc64/kernel/pSeries_smp.c b/arch/ppc64/kernel/pSeries_smp.c index 4203bd020c8..62c55a12356 100644 --- a/arch/ppc64/kernel/pSeries_smp.c +++ b/arch/ppc64/kernel/pSeries_smp.c @@ -1,5 +1,5 @@ /* - * SMP support for pSeries machines. + * SMP support for pSeries and BPA machines. * * Dave Engebretsen, Peter Bergner, and * Mike Corrigan {engebret|bergner|mikec}@us.ibm.com @@ -47,6 +47,7 @@ #include <asm/pSeries_reconfig.h> #include "mpic.h" +#include "bpa_iic.h" #ifdef DEBUG #define DBG(fmt...) udbg_printf(fmt) @@ -92,10 +93,13 @@ static int query_cpu_stopped(unsigned int pcpu) int pSeries_cpu_disable(void) { + int cpu = smp_processor_id(); + + cpu_clear(cpu, cpu_online_map); systemcfg->processorCount--; /*fix boot_cpuid here*/ - if (smp_processor_id() == boot_cpuid) + if (cpu == boot_cpuid) boot_cpuid = any_online_cpu(cpu_online_map); /* FIXME: abstract this to not be platform specific later on */ @@ -286,6 +290,7 @@ static inline int __devinit smp_startup_cpu(unsigned int lcpu) return 1; } +#ifdef CONFIG_XICS static inline void smp_xics_do_message(int cpu, int msg) { set_bit(msg, &xics_ipi_message[cpu].value); @@ -327,6 +332,37 @@ static void __devinit smp_xics_setup_cpu(int cpu) cpu_clear(cpu, of_spin_map); } +#endif /* CONFIG_XICS */ +#ifdef CONFIG_BPA_IIC +static void smp_iic_message_pass(int target, int msg) +{ + unsigned int i; + + if (target < NR_CPUS) { + iic_cause_IPI(target, msg); + } else { + for_each_online_cpu(i) { + if (target == MSG_ALL_BUT_SELF + && i == smp_processor_id()) + continue; + iic_cause_IPI(i, msg); + } + } +} + +static int __init smp_iic_probe(void) +{ + iic_request_IPIs(); + + return cpus_weight(cpu_possible_map); +} + +static void __devinit smp_iic_setup_cpu(int cpu) +{ + if (cpu != boot_cpuid) + iic_setup_cpu(); +} +#endif /* CONFIG_BPA_IIC */ static DEFINE_SPINLOCK(timebase_lock); static unsigned long timebase = 0; @@ -381,14 +417,15 @@ static int smp_pSeries_cpu_bootable(unsigned int nr) return 1; } - +#ifdef CONFIG_MPIC static struct smp_ops_t pSeries_mpic_smp_ops = { .message_pass = smp_mpic_message_pass, .probe = smp_mpic_probe, .kick_cpu = smp_pSeries_kick_cpu, .setup_cpu = smp_mpic_setup_cpu, }; - +#endif +#ifdef CONFIG_XICS static struct smp_ops_t pSeries_xics_smp_ops = { .message_pass = smp_xics_message_pass, .probe = smp_xics_probe, @@ -396,6 +433,16 @@ static struct smp_ops_t pSeries_xics_smp_ops = { .setup_cpu = smp_xics_setup_cpu, .cpu_bootable = smp_pSeries_cpu_bootable, }; +#endif +#ifdef CONFIG_BPA_IIC +static struct smp_ops_t bpa_iic_smp_ops = { + .message_pass = smp_iic_message_pass, + .probe = smp_iic_probe, + .kick_cpu = smp_pSeries_kick_cpu, + .setup_cpu = smp_iic_setup_cpu, + .cpu_bootable = smp_pSeries_cpu_bootable, +}; +#endif /* This is called very early */ void __init smp_init_pSeries(void) @@ -404,10 +451,25 @@ void __init smp_init_pSeries(void) DBG(" -> smp_init_pSeries()\n"); - if (ppc64_interrupt_controller == IC_OPEN_PIC) + switch (ppc64_interrupt_controller) { +#ifdef CONFIG_MPIC + case IC_OPEN_PIC: smp_ops = &pSeries_mpic_smp_ops; - else + break; +#endif +#ifdef CONFIG_XICS + case IC_PPC_XIC: smp_ops = &pSeries_xics_smp_ops; + break; +#endif +#ifdef CONFIG_BPA_IIC + case IC_BPA_IIC: + smp_ops = &bpa_iic_smp_ops; + break; +#endif + default: + panic("Invalid interrupt controller"); + } #ifdef CONFIG_HOTPLUG_CPU smp_ops->cpu_disable = pSeries_cpu_disable; diff --git a/arch/ppc64/kernel/pci.c b/arch/ppc64/kernel/pci.c index 2bf0513f3ec..580676f87d2 100644 --- a/arch/ppc64/kernel/pci.c +++ b/arch/ppc64/kernel/pci.c @@ -902,6 +902,9 @@ void __devinit pcibios_fixup_bus(struct pci_bus *bus) list_for_each_entry(dev, &bus->devices, bus_list) ppc_md.iommu_dev_setup(dev); + if (ppc_md.irq_bus_setup) + ppc_md.irq_bus_setup(bus); + if (!pci_probe_only) return; diff --git a/arch/ppc64/kernel/pci.h b/arch/ppc64/kernel/pci.h index 0fd7d849aa7..26be78b13af 100644 --- a/arch/ppc64/kernel/pci.h +++ b/arch/ppc64/kernel/pci.h @@ -40,10 +40,14 @@ struct device_node *fetch_dev_dn(struct pci_dev *dev); void pci_addr_cache_insert_device(struct pci_dev *dev); void pci_addr_cache_remove_device(struct pci_dev *dev); -/* From pSeries_pci.h */ +/* From rtas_pci.h */ void init_pci_config_tokens (void); unsigned long get_phb_buid (struct device_node *); +/* From pSeries_pci.h */ +extern void pSeries_final_fixup(void); +extern void pSeries_irq_bus_setup(struct pci_bus *bus); + extern unsigned long pci_probe_only; extern unsigned long pci_assign_all_buses; extern int pci_read_irq_line(struct pci_dev *pci_dev); diff --git a/arch/ppc64/kernel/pmac_time.c b/arch/ppc64/kernel/pmac_time.c index f24827581dd..3059edb09cc 100644 --- a/arch/ppc64/kernel/pmac_time.c +++ b/arch/ppc64/kernel/pmac_time.c @@ -40,11 +40,6 @@ #define DBG(x...) #endif -extern void setup_default_decr(void); - -extern unsigned long ppc_tb_freq; -extern unsigned long ppc_proc_freq; - /* Apparently the RTC stores seconds since 1 Jan 1904 */ #define RTC_OFFSET 2082844800 @@ -161,8 +156,7 @@ void __init pmac_get_boot_time(struct rtc_time *tm) /* * Query the OF and get the decr frequency. - * This was taken from the pmac time_init() when merging the prep/pmac - * time functions. + * FIXME: merge this with generic_calibrate_decr */ void __init pmac_calibrate_decr(void) { diff --git a/arch/ppc64/kernel/proc_ppc64.c b/arch/ppc64/kernel/proc_ppc64.c index 0914b0669b0..a87c66a9652 100644 --- a/arch/ppc64/kernel/proc_ppc64.c +++ b/arch/ppc64/kernel/proc_ppc64.c @@ -53,7 +53,7 @@ static int __init proc_ppc64_create(void) if (!root) return 1; - if (!(systemcfg->platform & PLATFORM_PSERIES)) + if (!(systemcfg->platform & (PLATFORM_PSERIES | PLATFORM_BPA))) return 0; if (!proc_mkdir("rtas", root)) diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c index b7683abfbe6..e248a7950ae 100644 --- a/arch/ppc64/kernel/prom_init.c +++ b/arch/ppc64/kernel/prom_init.c @@ -1915,9 +1915,9 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, unsigned long prom_send_capabilities(); /* - * On pSeries, copy the CPU hold code + * On pSeries and BPA, copy the CPU hold code */ - if (RELOC(of_platform) & PLATFORM_PSERIES) + if (RELOC(of_platform) & (PLATFORM_PSERIES | PLATFORM_BPA)) copy_and_flush(0, KERNELBASE - offset, 0x100, 0); /* diff --git a/arch/ppc64/kernel/ptrace.c b/arch/ppc64/kernel/ptrace.c index 9f8c6087ae5..2993f108d96 100644 --- a/arch/ppc64/kernel/ptrace.c +++ b/arch/ppc64/kernel/ptrace.c @@ -305,6 +305,8 @@ static void do_syscall_trace(void) void do_syscall_trace_enter(struct pt_regs *regs) { + secure_computing(regs->gpr[0]); + if (test_thread_flag(TIF_SYSCALL_TRACE) && (current->ptrace & PT_PTRACED)) do_syscall_trace(); @@ -320,8 +322,6 @@ void do_syscall_trace_enter(struct pt_regs *regs) void do_syscall_trace_leave(struct pt_regs *regs) { - secure_computing(regs->gpr[0]); - if (unlikely(current->audit_context)) audit_syscall_exit(current, (regs->ccr&0x1000)?AUDITSC_FAILURE:AUDITSC_SUCCESS, diff --git a/arch/ppc64/kernel/rtas-proc.c b/arch/ppc64/kernel/rtas-proc.c index 28b1f1521f2..1f3ff860fdf 100644 --- a/arch/ppc64/kernel/rtas-proc.c +++ b/arch/ppc64/kernel/rtas-proc.c @@ -371,11 +371,11 @@ static ssize_t ppc_rtas_progress_write(struct file *file, /* Lets see if the user passed hexdigits */ hex = simple_strtoul(progress_led, NULL, 10); - ppc_md.progress ((char *)progress_led, hex); + rtas_progress ((char *)progress_led, hex); return count; /* clear the line */ - /* ppc_md.progress(" ", 0xffff);*/ + /* rtas_progress(" ", 0xffff);*/ } /* ****************************************************************** */ static int ppc_rtas_progress_show(struct seq_file *m, void *v) diff --git a/arch/ppc64/kernel/rtas.c b/arch/ppc64/kernel/rtas.c index 5575603def2..5e8eb33b8e5 100644 --- a/arch/ppc64/kernel/rtas.c +++ b/arch/ppc64/kernel/rtas.c @@ -91,6 +91,123 @@ call_rtas_display_status_delay(unsigned char c) } } +void +rtas_progress(char *s, unsigned short hex) +{ + struct device_node *root; + int width, *p; + char *os; + static int display_character, set_indicator; + static int display_width, display_lines, *row_width, form_feed; + static DEFINE_SPINLOCK(progress_lock); + static int current_line; + static int pending_newline = 0; /* did last write end with unprinted newline? */ + + if (!rtas.base) + return; + + if (display_width == 0) { + display_width = 0x10; + if ((root = find_path_device("/rtas"))) { + if ((p = (unsigned int *)get_property(root, + "ibm,display-line-length", NULL))) + display_width = *p; + if ((p = (unsigned int *)get_property(root, + "ibm,form-feed", NULL))) + form_feed = *p; + if ((p = (unsigned int *)get_property(root, + "ibm,display-number-of-lines", NULL))) + display_lines = *p; + row_width = (unsigned int *)get_property(root, + "ibm,display-truncation-length", NULL); + } + display_character = rtas_token("display-character"); + set_indicator = rtas_token("set-indicator"); + } + + if (display_character == RTAS_UNKNOWN_SERVICE) { + /* use hex display if available */ + if (set_indicator != RTAS_UNKNOWN_SERVICE) + rtas_call(set_indicator, 3, 1, NULL, 6, 0, hex); + return; + } + + spin_lock(&progress_lock); + + /* + * Last write ended with newline, but we didn't print it since + * it would just clear the bottom line of output. Print it now + * instead. + * + * If no newline is pending and form feed is supported, clear the + * display with a form feed; otherwise, print a CR to start output + * at the beginning of the line. + */ + if (pending_newline) { + rtas_call(display_character, 1, 1, NULL, '\r'); + rtas_call(display_character, 1, 1, NULL, '\n'); + pending_newline = 0; + } else { + current_line = 0; + if (form_feed) + rtas_call(display_character, 1, 1, NULL, + (char)form_feed); + else + rtas_call(display_character, 1, 1, NULL, '\r'); + } + + if (row_width) + width = row_width[current_line]; + else + width = display_width; + os = s; + while (*os) { + if (*os == '\n' || *os == '\r') { + /* If newline is the last character, save it + * until next call to avoid bumping up the + * display output. + */ + if (*os == '\n' && !os[1]) { + pending_newline = 1; + current_line++; + if (current_line > display_lines-1) + current_line = display_lines-1; + spin_unlock(&progress_lock); + return; + } + + /* RTAS wants CR-LF, not just LF */ + + if (*os == '\n') { + rtas_call(display_character, 1, 1, NULL, '\r'); + rtas_call(display_character, 1, 1, NULL, '\n'); + } else { + /* CR might be used to re-draw a line, so we'll + * leave it alone and not add LF. + */ + rtas_call(display_character, 1, 1, NULL, *os); + } + + if (row_width) + width = row_width[current_line]; + else + width = display_width; + } else { + width--; + rtas_call(display_character, 1, 1, NULL, *os); + } + + os++; + + /* if we overwrite the screen length */ + if (width <= 0) + while ((*os != 0) && (*os != '\n') && (*os != '\r')) + os++; + } + + spin_unlock(&progress_lock); +} + int rtas_token(const char *service) { @@ -425,8 +542,8 @@ rtas_flash_firmware(void) printk(KERN_ALERT "FLASH: flash image is %ld bytes\n", image_size); printk(KERN_ALERT "FLASH: performing flash and reboot\n"); - ppc_md.progress("Flashing \n", 0x0); - ppc_md.progress("Please Wait... ", 0x0); + rtas_progress("Flashing \n", 0x0); + rtas_progress("Please Wait... ", 0x0); printk(KERN_ALERT "FLASH: this will take several minutes. Do not power off!\n"); status = rtas_call(update_token, 1, 1, NULL, rtas_block_list); switch (status) { /* should only get "bad" status */ diff --git a/arch/ppc64/kernel/rtas_pci.c b/arch/ppc64/kernel/rtas_pci.c new file mode 100644 index 00000000000..1048817befb --- /dev/null +++ b/arch/ppc64/kernel/rtas_pci.c @@ -0,0 +1,495 @@ +/* + * arch/ppc64/kernel/rtas_pci.c + * + * Copyright (C) 2001 Dave Engebretsen, IBM Corporation + * Copyright (C) 2003 Anton Blanchard <anton@au.ibm.com>, IBM + * + * RTAS specific routines for PCI. + * + * Based on code from pci.c, chrp_pci.c and pSeries_pci.c + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include <linux/kernel.h> +#include <linux/threads.h> +#include <linux/pci.h> +#include <linux/string.h> +#include <linux/init.h> +#include <linux/bootmem.h> + +#include <asm/io.h> +#include <asm/pgtable.h> +#include <asm/irq.h> +#include <asm/prom.h> +#include <asm/machdep.h> +#include <asm/pci-bridge.h> +#include <asm/iommu.h> +#include <asm/rtas.h> + +#include "mpic.h" +#include "pci.h" + +/* RTAS tokens */ +static int read_pci_config; +static int write_pci_config; +static int ibm_read_pci_config; +static int ibm_write_pci_config; + +static int config_access_valid(struct device_node *dn, int where) +{ + if (where < 256) + return 1; + if (where < 4096 && dn->pci_ext_config_space) + return 1; + + return 0; +} + +static int rtas_read_config(struct device_node *dn, int where, int size, u32 *val) +{ + int returnval = -1; + unsigned long buid, addr; + int ret; + + if (!dn) + return PCIBIOS_DEVICE_NOT_FOUND; + if (!config_access_valid(dn, where)) + return PCIBIOS_BAD_REGISTER_NUMBER; + + addr = ((where & 0xf00) << 20) | (dn->busno << 16) | + (dn->devfn << 8) | (where & 0xff); + buid = dn->phb->buid; + if (buid) { + ret = rtas_call(ibm_read_pci_config, 4, 2, &returnval, + addr, buid >> 32, buid & 0xffffffff, size); + } else { + ret = rtas_call(read_pci_config, 2, 2, &returnval, addr, size); + } + *val = returnval; + + if (ret) + return PCIBIOS_DEVICE_NOT_FOUND; + + if (returnval == EEH_IO_ERROR_VALUE(size) + && eeh_dn_check_failure (dn, NULL)) + return PCIBIOS_DEVICE_NOT_FOUND; + + return PCIBIOS_SUCCESSFUL; +} + +static int rtas_pci_read_config(struct pci_bus *bus, + unsigned int devfn, + int where, int size, u32 *val) +{ + struct device_node *busdn, *dn; + + if (bus->self) + busdn = pci_device_to_OF_node(bus->self); + else + busdn = bus->sysdata; /* must be a phb */ + + /* Search only direct children of the bus */ + for (dn = busdn->child; dn; dn = dn->sibling) + if (dn->devfn == devfn) + return rtas_read_config(dn, where, size, val); + return PCIBIOS_DEVICE_NOT_FOUND; +} + +static int rtas_write_config(struct device_node *dn, int where, int size, u32 val) +{ + unsigned long buid, addr; + int ret; + + if (!dn) + return PCIBIOS_DEVICE_NOT_FOUND; + if (!config_access_valid(dn, where)) + return PCIBIOS_BAD_REGISTER_NUMBER; + + addr = ((where & 0xf00) << 20) | (dn->busno << 16) | + (dn->devfn << 8) | (where & 0xff); + buid = dn->phb->buid; + if (buid) { + ret = rtas_call(ibm_write_pci_config, 5, 1, NULL, addr, buid >> 32, buid & 0xffffffff, size, (ulong) val); + } else { + ret = rtas_call(write_pci_config, 3, 1, NULL, addr, size, (ulong)val); + } + + if (ret) + return PCIBIOS_DEVICE_NOT_FOUND; + + return PCIBIOS_SUCCESSFUL; +} + +static int rtas_pci_write_config(struct pci_bus *bus, + unsigned int devfn, + int where, int size, u32 val) +{ + struct device_node *busdn, *dn; + + if (bus->self) + busdn = pci_device_to_OF_node(bus->self); + else + busdn = bus->sysdata; /* must be a phb */ + + /* Search only direct children of the bus */ + for (dn = busdn->child; dn; dn = dn->sibling) + if (dn->devfn == devfn) + return rtas_write_config(dn, where, size, val); + return PCIBIOS_DEVICE_NOT_FOUND; +} + +struct pci_ops rtas_pci_ops = { + rtas_pci_read_config, + rtas_pci_write_config +}; + +int is_python(struct device_node *dev) +{ + char *model = (char *)get_property(dev, "model", NULL); + + if (model && strstr(model, "Python")) + return 1; + + return 0; +} + +static int get_phb_reg_prop(struct device_node *dev, + unsigned int addr_size_words, + struct reg_property64 *reg) +{ + unsigned int *ui_ptr = NULL, len; + + /* Found a PHB, now figure out where his registers are mapped. */ + ui_ptr = (unsigned int *)get_property(dev, "reg", &len); + if (ui_ptr == NULL) + return 1; + + if (addr_size_words == 1) { + reg->address = ((struct reg_property32 *)ui_ptr)->address; + reg->size = ((struct reg_property32 *)ui_ptr)->size; + } else { + *reg = *((struct reg_property64 *)ui_ptr); + } + + return 0; +} + +static void python_countermeasures(struct device_node *dev, + unsigned int addr_size_words) +{ + struct reg_property64 reg_struct; + void __iomem *chip_regs; + volatile u32 val; + + if (get_phb_reg_prop(dev, addr_size_words, ®_struct)) + return; + + /* Python's register file is 1 MB in size. */ + chip_regs = ioremap(reg_struct.address & ~(0xfffffUL), 0x100000); + + /* + * Firmware doesn't always clear this bit which is critical + * for good performance - Anton + */ + +#define PRG_CL_RESET_VALID 0x00010000 + + val = in_be32(chip_regs + 0xf6030); + if (val & PRG_CL_RESET_VALID) { + printk(KERN_INFO "Python workaround: "); + val &= ~PRG_CL_RESET_VALID; + out_be32(chip_regs + 0xf6030, val); + /* + * We must read it back for changes to + * take effect + */ + val = in_be32(chip_regs + 0xf6030); + printk("reg0: %x\n", val); + } + + iounmap(chip_regs); +} + +void __init init_pci_config_tokens (void) +{ + read_pci_config = rtas_token("read-pci-config"); + write_pci_config = rtas_token("write-pci-config"); + ibm_read_pci_config = rtas_token("ibm,read-pci-config"); + ibm_write_pci_config = rtas_token("ibm,write-pci-config"); +} + +unsigned long __devinit get_phb_buid (struct device_node *phb) +{ + int addr_cells; + unsigned int *buid_vals; + unsigned int len; + unsigned long buid; + + if (ibm_read_pci_config == -1) return 0; + + /* PHB's will always be children of the root node, + * or so it is promised by the current firmware. */ + if (phb->parent == NULL) + return 0; + if (phb->parent->parent) + return 0; + + buid_vals = (unsigned int *) get_property(phb, "reg", &len); + if (buid_vals == NULL) + return 0; + + addr_cells = prom_n_addr_cells(phb); + if (addr_cells == 1) { + buid = (unsigned long) buid_vals[0]; + } else { + buid = (((unsigned long)buid_vals[0]) << 32UL) | + (((unsigned long)buid_vals[1]) & 0xffffffff); + } + return buid; +} + +static int phb_set_bus_ranges(struct device_node *dev, + struct pci_controller *phb) +{ + int *bus_range; + unsigned int len; + + bus_range = (int *) get_property(dev, "bus-range", &len); + if (bus_range == NULL || len < 2 * sizeof(int)) { + return 1; + } + + phb->first_busno = bus_range[0]; + phb->last_busno = bus_range[1]; + + return 0; +} + +static int __devinit setup_phb(struct device_node *dev, + struct pci_controller *phb, + unsigned int addr_size_words) +{ + pci_setup_pci_controller(phb); + + if (is_python(dev)) + python_countermeasures(dev, addr_size_words); + + if (phb_set_bus_ranges(dev, phb)) + return 1; + + phb->arch_data = dev; + phb->ops = &rtas_pci_ops; + phb->buid = get_phb_buid(dev); + + return 0; +} + +static void __devinit add_linux_pci_domain(struct device_node *dev, + struct pci_controller *phb, + struct property *of_prop) +{ + memset(of_prop, 0, sizeof(struct property)); + of_prop->name = "linux,pci-domain"; + of_prop->length = sizeof(phb->global_number); + of_prop->value = (unsigned char *)&of_prop[1]; + memcpy(of_prop->value, &phb->global_number, sizeof(phb->global_number)); + prom_add_property(dev, of_prop); +} + +static struct pci_controller * __init alloc_phb(struct device_node *dev, + unsigned int addr_size_words) +{ + struct pci_controller *phb; + struct property *of_prop; + + phb = alloc_bootmem(sizeof(struct pci_controller)); + if (phb == NULL) + return NULL; + + of_prop = alloc_bootmem(sizeof(struct property) + + sizeof(phb->global_number)); + if (!of_prop) + return NULL; + + if (setup_phb(dev, phb, addr_size_words)) + return NULL; + + add_linux_pci_domain(dev, phb, of_prop); + + return phb; +} + +static struct pci_controller * __devinit alloc_phb_dynamic(struct device_node *dev, unsigned int addr_size_words) +{ + struct pci_controller *phb; + + phb = (struct pci_controller *)kmalloc(sizeof(struct pci_controller), + GFP_KERNEL); + if (phb == NULL) + return NULL; + + if (setup_phb(dev, phb, addr_size_words)) + return NULL; + + phb->is_dynamic = 1; + + /* TODO: linux,pci-domain? */ + + return phb; +} + +unsigned long __init find_and_init_phbs(void) +{ + struct device_node *node; + struct pci_controller *phb; + unsigned int root_size_cells = 0; + unsigned int index; + unsigned int *opprop = NULL; + struct device_node *root = of_find_node_by_path("/"); + + if (ppc64_interrupt_controller == IC_OPEN_PIC) { + opprop = (unsigned int *)get_property(root, + "platform-open-pic", NULL); + } + + root_size_cells = prom_n_size_cells(root); + + index = 0; + + for (node = of_get_next_child(root, NULL); + node != NULL; + node = of_get_next_child(root, node)) { + if (node->type == NULL || strcmp(node->type, "pci") != 0) + continue; + + phb = alloc_phb(node, root_size_cells); + if (!phb) + continue; + + pci_process_bridge_OF_ranges(phb, node); + pci_setup_phb_io(phb, index == 0); +#ifdef CONFIG_PPC_PSERIES + if (ppc64_interrupt_controller == IC_OPEN_PIC && pSeries_mpic) { + int addr = root_size_cells * (index + 2) - 1; + mpic_assign_isu(pSeries_mpic, index, opprop[addr]); + } +#endif + index++; + } + + of_node_put(root); + pci_devs_phb_init(); + + /* + * pci_probe_only and pci_assign_all_buses can be set via properties + * in chosen. + */ + if (of_chosen) { + int *prop; + + prop = (int *)get_property(of_chosen, "linux,pci-probe-only", + NULL); + if (prop) + pci_probe_only = *prop; + + prop = (int *)get_property(of_chosen, + "linux,pci-assign-all-buses", NULL); + if (prop) + pci_assign_all_buses = *prop; + } + + return 0; +} + +struct pci_controller * __devinit init_phb_dynamic(struct device_node *dn) +{ + struct device_node *root = of_find_node_by_path("/"); + unsigned int root_size_cells = 0; + struct pci_controller *phb; + struct pci_bus *bus; + int primary; + + root_size_cells = prom_n_size_cells(root); + + primary = list_empty(&hose_list); + phb = alloc_phb_dynamic(dn, root_size_cells); + if (!phb) + return NULL; + + pci_process_bridge_OF_ranges(phb, dn); + + pci_setup_phb_io_dynamic(phb, primary); + of_node_put(root); + + pci_devs_phb_init_dynamic(phb); + phb->last_busno = 0xff; + bus = pci_scan_bus(phb->first_busno, phb->ops, phb->arch_data); + phb->bus = bus; + phb->last_busno = bus->subordinate; + + return phb; +} +EXPORT_SYMBOL(init_phb_dynamic); + +/* RPA-specific bits for removing PHBs */ +int pcibios_remove_root_bus(struct pci_controller *phb) +{ + struct pci_bus *b = phb->bus; + struct resource *res; + int rc, i; + + res = b->resource[0]; + if (!res->flags) { + printk(KERN_ERR "%s: no IO resource for PHB %s\n", __FUNCTION__, + b->name); + return 1; + } + + rc = unmap_bus_range(b); + if (rc) { + printk(KERN_ERR "%s: failed to unmap IO on bus %s\n", + __FUNCTION__, b->name); + return 1; + } + + if (release_resource(res)) { + printk(KERN_ERR "%s: failed to release IO on bus %s\n", + __FUNCTION__, b->name); + return 1; + } + + for (i = 1; i < 3; ++i) { + res = b->resource[i]; + if (!res->flags && i == 0) { + printk(KERN_ERR "%s: no MEM resource for PHB %s\n", + __FUNCTION__, b->name); + return 1; + } + if (res->flags && release_resource(res)) { + printk(KERN_ERR + "%s: failed to release IO %d on bus %s\n", + __FUNCTION__, i, b->name); + return 1; + } + } + + list_del(&phb->list_node); + if (phb->is_dynamic) + kfree(phb); + + return 0; +} +EXPORT_SYMBOL(pcibios_remove_root_bus); diff --git a/arch/ppc64/kernel/rtc.c b/arch/ppc64/kernel/rtc.c index de02aedbe08..d729fefa0df 100644 --- a/arch/ppc64/kernel/rtc.c +++ b/arch/ppc64/kernel/rtc.c @@ -301,7 +301,7 @@ void iSeries_get_boot_time(struct rtc_time *tm) #ifdef CONFIG_PPC_RTAS #define MAX_RTC_WAIT 5000 /* 5 sec */ #define RTAS_CLOCK_BUSY (-2) -void pSeries_get_boot_time(struct rtc_time *rtc_tm) +void rtas_get_boot_time(struct rtc_time *rtc_tm) { int ret[8]; int error, wait_time; @@ -336,7 +336,7 @@ void pSeries_get_boot_time(struct rtc_time *rtc_tm) * and if a delay is needed to read the clock. In this case we just * silently return without updating rtc_tm. */ -void pSeries_get_rtc_time(struct rtc_time *rtc_tm) +void rtas_get_rtc_time(struct rtc_time *rtc_tm) { int ret[8]; int error, wait_time; @@ -371,7 +371,7 @@ void pSeries_get_rtc_time(struct rtc_time *rtc_tm) rtc_tm->tm_year = ret[0] - 1900; } -int pSeries_set_rtc_time(struct rtc_time *tm) +int rtas_set_rtc_time(struct rtc_time *tm) { int error, wait_time; unsigned long max_wait_tb; diff --git a/arch/ppc64/kernel/setup.c b/arch/ppc64/kernel/setup.c index 8e439a81764..d5e4866e9ac 100644 --- a/arch/ppc64/kernel/setup.c +++ b/arch/ppc64/kernel/setup.c @@ -344,6 +344,7 @@ static void __init setup_cpu_maps(void) extern struct machdep_calls pSeries_md; extern struct machdep_calls pmac_md; extern struct machdep_calls maple_md; +extern struct machdep_calls bpa_md; /* Ultimately, stuff them in an elf section like initcalls... */ static struct machdep_calls __initdata *machines[] = { @@ -356,6 +357,9 @@ static struct machdep_calls __initdata *machines[] = { #ifdef CONFIG_PPC_MAPLE &maple_md, #endif /* CONFIG_PPC_MAPLE */ +#ifdef CONFIG_PPC_BPA + &bpa_md, +#endif NULL }; @@ -673,37 +677,52 @@ void __init setup_system(void) DBG(" <- setup_system()\n"); } - -void machine_restart(char *cmd) +/* also used by kexec */ +void machine_shutdown(void) { if (ppc_md.nvram_sync) ppc_md.nvram_sync(); - ppc_md.restart(cmd); } +void machine_restart(char *cmd) +{ + machine_shutdown(); + ppc_md.restart(cmd); +#ifdef CONFIG_SMP + smp_send_stop(); +#endif + printk(KERN_EMERG "System Halted, OK to turn off power\n"); + local_irq_disable(); + while (1) ; +} EXPORT_SYMBOL(machine_restart); - + void machine_power_off(void) { - if (ppc_md.nvram_sync) - ppc_md.nvram_sync(); + machine_shutdown(); ppc_md.power_off(); +#ifdef CONFIG_SMP + smp_send_stop(); +#endif + printk(KERN_EMERG "System Halted, OK to turn off power\n"); + local_irq_disable(); + while (1) ; } - EXPORT_SYMBOL(machine_power_off); - + void machine_halt(void) { - if (ppc_md.nvram_sync) - ppc_md.nvram_sync(); + machine_shutdown(); ppc_md.halt(); +#ifdef CONFIG_SMP + smp_send_stop(); +#endif + printk(KERN_EMERG "System Halted, OK to turn off power\n"); + local_irq_disable(); + while (1) ; } - EXPORT_SYMBOL(machine_halt); -unsigned long ppc_proc_freq; -unsigned long ppc_tb_freq; - static int ppc64_panic_event(struct notifier_block *this, unsigned long event, void *ptr) { @@ -1055,6 +1074,7 @@ void __init setup_arch(char **cmdline_p) /* set up the bootmem stuff with available memory */ do_init_bootmem(); + sparse_init(); /* initialize the syscall map in systemcfg */ setup_syscall_map(); @@ -1079,11 +1099,11 @@ void __init setup_arch(char **cmdline_p) static void ppc64_do_msg(unsigned int src, const char *msg) { if (ppc_md.progress) { - char buf[32]; + char buf[128]; - sprintf(buf, "%08x \n", src); + sprintf(buf, "%08X\n", src); ppc_md.progress(buf, 0); - sprintf(buf, "%-16s", msg); + snprintf(buf, 128, "%s", msg); ppc_md.progress(buf, 0); } } @@ -1117,7 +1137,7 @@ void ppc64_dump_msg(unsigned int src, const char *msg) } /* This should only be called on processor 0 during calibrate decr */ -void setup_default_decr(void) +void __init setup_default_decr(void) { struct paca_struct *lpaca = get_paca(); diff --git a/arch/ppc64/kernel/smp.c b/arch/ppc64/kernel/smp.c index 9ef5d36d6b2..2fcddfcb594 100644 --- a/arch/ppc64/kernel/smp.c +++ b/arch/ppc64/kernel/smp.c @@ -71,7 +71,7 @@ void smp_call_function_interrupt(void); int smt_enabled_at_boot = 1; -#ifdef CONFIG_PPC_MULTIPLATFORM +#ifdef CONFIG_MPIC void smp_mpic_message_pass(int target, int msg) { /* make sure we're sending something that translates to an IPI */ @@ -128,7 +128,7 @@ void __devinit smp_generic_kick_cpu(int nr) smp_mb(); } -#endif /* CONFIG_PPC_MULTIPLATFORM */ +#endif /* CONFIG_MPIC */ static void __init smp_space_timers(unsigned int max_cpus) { diff --git a/arch/ppc64/kernel/spider-pic.c b/arch/ppc64/kernel/spider-pic.c new file mode 100644 index 00000000000..d5c9a02fb11 --- /dev/null +++ b/arch/ppc64/kernel/spider-pic.c @@ -0,0 +1,191 @@ +/* + * External Interrupt Controller on Spider South Bridge + * + * (C) Copyright IBM Deutschland Entwicklung GmbH 2005 + * + * Author: Arnd Bergmann <arndb@de.ibm.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include <linux/interrupt.h> +#include <linux/irq.h> + +#include <asm/pgtable.h> +#include <asm/prom.h> +#include <asm/io.h> + +#include "bpa_iic.h" + +/* register layout taken from Spider spec, table 7.4-4 */ +enum { + TIR_DEN = 0x004, /* Detection Enable Register */ + TIR_MSK = 0x084, /* Mask Level Register */ + TIR_EDC = 0x0c0, /* Edge Detection Clear Register */ + TIR_PNDA = 0x100, /* Pending Register A */ + TIR_PNDB = 0x104, /* Pending Register B */ + TIR_CS = 0x144, /* Current Status Register */ + TIR_LCSA = 0x150, /* Level Current Status Register A */ + TIR_LCSB = 0x154, /* Level Current Status Register B */ + TIR_LCSC = 0x158, /* Level Current Status Register C */ + TIR_LCSD = 0x15c, /* Level Current Status Register D */ + TIR_CFGA = 0x200, /* Setting Register A0 */ + TIR_CFGB = 0x204, /* Setting Register B0 */ + /* 0x208 ... 0x3ff Setting Register An/Bn */ + TIR_PPNDA = 0x400, /* Packet Pending Register A */ + TIR_PPNDB = 0x404, /* Packet Pending Register B */ + TIR_PIERA = 0x408, /* Packet Output Error Register A */ + TIR_PIERB = 0x40c, /* Packet Output Error Register B */ + TIR_PIEN = 0x444, /* Packet Output Enable Register */ + TIR_PIPND = 0x454, /* Packet Output Pending Register */ + TIRDID = 0x484, /* Spider Device ID Register */ + REISTIM = 0x500, /* Reissue Command Timeout Time Setting */ + REISTIMEN = 0x504, /* Reissue Command Timeout Setting */ + REISWAITEN = 0x508, /* Reissue Wait Control*/ +}; + +static void __iomem *spider_pics[4]; + +static void __iomem *spider_get_pic(int irq) +{ + int node = irq / IIC_NODE_STRIDE; + irq %= IIC_NODE_STRIDE; + + if (irq >= IIC_EXT_OFFSET && + irq < IIC_EXT_OFFSET + IIC_NUM_EXT && + spider_pics) + return spider_pics[node]; + return NULL; +} + +static int spider_get_nr(unsigned int irq) +{ + return (irq % IIC_NODE_STRIDE) - IIC_EXT_OFFSET; +} + +static void __iomem *spider_get_irq_config(int irq) +{ + void __iomem *pic; + pic = spider_get_pic(irq); + return pic + TIR_CFGA + 8 * spider_get_nr(irq); +} + +static void spider_enable_irq(unsigned int irq) +{ + void __iomem *cfg = spider_get_irq_config(irq); + irq = spider_get_nr(irq); + + out_be32(cfg, in_be32(cfg) | 0x3107000eu); + out_be32(cfg + 4, in_be32(cfg + 4) | 0x00020000u | irq); +} + +static void spider_disable_irq(unsigned int irq) +{ + void __iomem *cfg = spider_get_irq_config(irq); + irq = spider_get_nr(irq); + + out_be32(cfg, in_be32(cfg) & ~0x30000000u); +} + +static unsigned int spider_startup_irq(unsigned int irq) +{ + spider_enable_irq(irq); + return 0; +} + +static void spider_shutdown_irq(unsigned int irq) +{ + spider_disable_irq(irq); +} + +static void spider_end_irq(unsigned int irq) +{ + spider_enable_irq(irq); +} + +static void spider_ack_irq(unsigned int irq) +{ + spider_disable_irq(irq); + iic_local_enable(); +} + +static struct hw_interrupt_type spider_pic = { + .typename = " SPIDER ", + .startup = spider_startup_irq, + .shutdown = spider_shutdown_irq, + .enable = spider_enable_irq, + .disable = spider_disable_irq, + .ack = spider_ack_irq, + .end = spider_end_irq, +}; + + +int spider_get_irq(unsigned long int_pending) +{ + void __iomem *regs = spider_get_pic(int_pending); + unsigned long cs; + int irq; + + cs = in_be32(regs + TIR_CS); + + irq = cs >> 24; + if (irq != 63) + return irq; + + return -1; +} + +void spider_init_IRQ(void) +{ + int node; + struct device_node *dn; + unsigned int *property; + long spiderpic; + int n; + +/* FIXME: detect multiple PICs as soon as the device tree has them */ + for (node = 0; node < 1; node++) { + dn = of_find_node_by_path("/"); + n = prom_n_addr_cells(dn); + property = (unsigned int *) get_property(dn, + "platform-spider-pic", NULL); + + if (!property) + continue; + for (spiderpic = 0; n > 0; --n) + spiderpic = (spiderpic << 32) + *property++; + printk(KERN_DEBUG "SPIDER addr: %lx\n", spiderpic); + spider_pics[node] = __ioremap(spiderpic, 0x800, _PAGE_NO_CACHE); + for (n = 0; n < IIC_NUM_EXT; n++) { + int irq = n + IIC_EXT_OFFSET + node * IIC_NODE_STRIDE; + get_irq_desc(irq)->handler = &spider_pic; + + /* do not mask any interrupts because of level */ + out_be32(spider_pics[node] + TIR_MSK, 0x0); + + /* disable edge detection clear */ + /* out_be32(spider_pics[node] + TIR_EDC, 0x0); */ + + /* enable interrupt packets to be output */ + out_be32(spider_pics[node] + TIR_PIEN, + in_be32(spider_pics[node] + TIR_PIEN) | 0x1); + + /* Enable the interrupt detection enable bit. Do this last! */ + out_be32(spider_pics[node] + TIR_DEN, + in_be32(spider_pics[node] +TIR_DEN) | 0x1); + + } + } +} diff --git a/arch/ppc64/kernel/time.c b/arch/ppc64/kernel/time.c index 33364a7d2cd..2348a75e050 100644 --- a/arch/ppc64/kernel/time.c +++ b/arch/ppc64/kernel/time.c @@ -107,6 +107,9 @@ void ppc_adjtimex(void); static unsigned adjusting_time = 0; +unsigned long ppc_proc_freq; +unsigned long ppc_tb_freq; + static __inline__ void timer_check_rtc(void) { /* @@ -472,6 +475,66 @@ int do_settimeofday(struct timespec *tv) EXPORT_SYMBOL(do_settimeofday); +#if defined(CONFIG_PPC_PSERIES) || defined(CONFIG_PPC_MAPLE) || defined(CONFIG_PPC_BPA) +void __init generic_calibrate_decr(void) +{ + struct device_node *cpu; + struct div_result divres; + unsigned int *fp; + int node_found; + + /* + * The cpu node should have a timebase-frequency property + * to tell us the rate at which the decrementer counts. + */ + cpu = of_find_node_by_type(NULL, "cpu"); + + ppc_tb_freq = DEFAULT_TB_FREQ; /* hardcoded default */ + node_found = 0; + if (cpu != 0) { + fp = (unsigned int *)get_property(cpu, "timebase-frequency", + NULL); + if (fp != 0) { + node_found = 1; + ppc_tb_freq = *fp; + } + } + if (!node_found) + printk(KERN_ERR "WARNING: Estimating decrementer frequency " + "(not found)\n"); + + ppc_proc_freq = DEFAULT_PROC_FREQ; + node_found = 0; + if (cpu != 0) { + fp = (unsigned int *)get_property(cpu, "clock-frequency", + NULL); + if (fp != 0) { + node_found = 1; + ppc_proc_freq = *fp; + } + } + if (!node_found) + printk(KERN_ERR "WARNING: Estimating processor frequency " + "(not found)\n"); + + of_node_put(cpu); + + printk(KERN_INFO "time_init: decrementer frequency = %lu.%.6lu MHz\n", + ppc_tb_freq/1000000, ppc_tb_freq%1000000); + printk(KERN_INFO "time_init: processor frequency = %lu.%.6lu MHz\n", + ppc_proc_freq/1000000, ppc_proc_freq%1000000); + + tb_ticks_per_jiffy = ppc_tb_freq / HZ; + tb_ticks_per_sec = tb_ticks_per_jiffy * HZ; + tb_ticks_per_usec = ppc_tb_freq / 1000000; + tb_to_us = mulhwu_scale_factor(ppc_tb_freq, 1000000); + div128_by_32(1024*1024, 0, tb_ticks_per_sec, &divres); + tb_to_xs = divres.result_low; + + setup_default_decr(); +} +#endif + void __init time_init(void) { /* This function is only called on the boot processor */ diff --git a/arch/ppc64/kernel/traps.c b/arch/ppc64/kernel/traps.c index 7e52cb2605e..a8d5e83ee89 100644 --- a/arch/ppc64/kernel/traps.c +++ b/arch/ppc64/kernel/traps.c @@ -126,6 +126,10 @@ int die(const char *str, struct pt_regs *regs, long err) printk("POWERMAC "); nl = 1; break; + case PLATFORM_BPA: + printk("BPA "); + nl = 1; + break; } if (nl) printk("\n"); diff --git a/arch/ppc64/kernel/xics.c b/arch/ppc64/kernel/xics.c index 879f39b90a3..677c4450984 100644 --- a/arch/ppc64/kernel/xics.c +++ b/arch/ppc64/kernel/xics.c @@ -647,6 +647,31 @@ static void xics_set_affinity(unsigned int virq, cpumask_t cpumask) } } +void xics_teardown_cpu(void) +{ + int cpu = smp_processor_id(); + int status; + + ops->cppr_info(cpu, 0x00); + iosync(); + + /* + * we need to EOI the IPI if we got here from kexec down IPI + * + * xics doesn't care if we duplicate an EOI as long as we + * don't EOI and raise priority. + * + * probably need to check all the other interrupts too + * should we be flagging idle loop instead? + * or creating some task to be scheduled? + */ + ops->xirr_info_set(cpu, XICS_IPI); + + status = rtas_set_indicator(GLOBAL_INTERRUPT_QUEUE, + (1UL << interrupt_server_size) - 1 - default_distrib_server, 0); + WARN_ON(status != 0); +} + #ifdef CONFIG_HOTPLUG_CPU /* Interrupts are disabled. */ diff --git a/arch/ppc64/mm/Makefile b/arch/ppc64/mm/Makefile index ac522d57b2a..3695d00d347 100644 --- a/arch/ppc64/mm/Makefile +++ b/arch/ppc64/mm/Makefile @@ -6,6 +6,6 @@ EXTRA_CFLAGS += -mno-minimal-toc obj-y := fault.o init.o imalloc.o hash_utils.o hash_low.o tlb.o \ slb_low.o slb.o stab.o mmap.o -obj-$(CONFIG_DISCONTIGMEM) += numa.o +obj-$(CONFIG_NEED_MULTIPLE_NODES) += numa.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o obj-$(CONFIG_PPC_MULTIPLATFORM) += hash_native.o diff --git a/arch/ppc64/mm/hash_native.c b/arch/ppc64/mm/hash_native.c index 52b6b930534..4fec05817d6 100644 --- a/arch/ppc64/mm/hash_native.c +++ b/arch/ppc64/mm/hash_native.c @@ -304,6 +304,50 @@ static void native_hpte_invalidate(unsigned long slot, unsigned long va, local_irq_restore(flags); } +/* + * clear all mappings on kexec. All cpus are in real mode (or they will + * be when they isi), and we are the only one left. We rely on our kernel + * mapping being 0xC0's and the hardware ignoring those two real bits. + * + * TODO: add batching support when enabled. remember, no dynamic memory here, + * athough there is the control page available... + */ +static void native_hpte_clear(void) +{ + unsigned long slot, slots, flags; + HPTE *hptep = htab_address; + Hpte_dword0 dw0; + unsigned long pteg_count; + + pteg_count = htab_hash_mask + 1; + + local_irq_save(flags); + + /* we take the tlbie lock and hold it. Some hardware will + * deadlock if we try to tlbie from two processors at once. + */ + spin_lock(&native_tlbie_lock); + + slots = pteg_count * HPTES_PER_GROUP; + + for (slot = 0; slot < slots; slot++, hptep++) { + /* + * we could lock the pte here, but we are the only cpu + * running, right? and for crash dump, we probably + * don't want to wait for a maybe bad cpu. + */ + dw0 = hptep->dw0.dw0; + + if (dw0.v) { + hptep->dw0.dword0 = 0; + tlbie(slot2va(dw0.avpn, dw0.l, dw0.h, slot), dw0.l); + } + } + + spin_unlock(&native_tlbie_lock); + local_irq_restore(flags); +} + static void native_flush_hash_range(unsigned long context, unsigned long number, int local) { @@ -415,7 +459,8 @@ void hpte_init_native(void) ppc_md.hpte_updatepp = native_hpte_updatepp; ppc_md.hpte_updateboltedpp = native_hpte_updateboltedpp; ppc_md.hpte_insert = native_hpte_insert; - ppc_md.hpte_remove = native_hpte_remove; + ppc_md.hpte_remove = native_hpte_remove; + ppc_md.hpte_clear_all = native_hpte_clear; if (tlb_batching_enabled()) ppc_md.flush_hash_range = native_flush_hash_range; htab_finish_init(); diff --git a/arch/ppc64/mm/init.c b/arch/ppc64/mm/init.c index 6fa1e6490b5..b50b3a446db 100644 --- a/arch/ppc64/mm/init.c +++ b/arch/ppc64/mm/init.c @@ -98,7 +98,7 @@ void show_mem(void) printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { for (i = 0; i < pgdat->node_spanned_pages; i++) { - page = pgdat->node_mem_map + i; + page = pgdat_page_nr(pgdat, i); total++; if (PageReserved(page)) reserved++; @@ -531,7 +531,7 @@ EXPORT_SYMBOL(page_is_ram); * Initialize the bootmem system and give it all the memory we * have available. */ -#ifndef CONFIG_DISCONTIGMEM +#ifndef CONFIG_NEED_MULTIPLE_NODES void __init do_init_bootmem(void) { unsigned long i; @@ -553,12 +553,20 @@ void __init do_init_bootmem(void) max_pfn = max_low_pfn; - /* add all physical memory to the bootmem map. Also find the first */ + /* Add all physical memory to the bootmem map, mark each area + * present. + */ for (i=0; i < lmb.memory.cnt; i++) { unsigned long physbase, size; + unsigned long start_pfn, end_pfn; physbase = lmb.memory.region[i].physbase; size = lmb.memory.region[i].size; + + start_pfn = physbase >> PAGE_SHIFT; + end_pfn = start_pfn + (size >> PAGE_SHIFT); + memory_present(0, start_pfn, end_pfn); + free_bootmem(physbase, size); } @@ -597,7 +605,7 @@ void __init paging_init(void) free_area_init_node(0, NODE_DATA(0), zones_size, __pa(PAGE_OFFSET) >> PAGE_SHIFT, zholes_size); } -#endif /* CONFIG_DISCONTIGMEM */ +#endif /* ! CONFIG_NEED_MULTIPLE_NODES */ static struct kcore_list kcore_vmem; @@ -628,7 +636,7 @@ module_init(setup_kcore); void __init mem_init(void) { -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NEED_MULTIPLE_NODES int nid; #endif pg_data_t *pgdat; @@ -639,7 +647,7 @@ void __init mem_init(void) num_physpages = max_low_pfn; /* RAM is assumed contiguous */ high_memory = (void *) __va(max_low_pfn * PAGE_SIZE); -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NEED_MULTIPLE_NODES for_each_online_node(nid) { if (NODE_DATA(nid)->node_spanned_pages != 0) { printk("freeing bootmem node %x\n", nid); @@ -654,7 +662,7 @@ void __init mem_init(void) for_each_pgdat(pgdat) { for (i = 0; i < pgdat->node_spanned_pages; i++) { - page = pgdat->node_mem_map + i; + page = pgdat_page_nr(pgdat, i); if (PageReserved(page)) reservedpages++; } diff --git a/arch/ppc64/mm/numa.c b/arch/ppc64/mm/numa.c index ea862ec643d..cafd91aef28 100644 --- a/arch/ppc64/mm/numa.c +++ b/arch/ppc64/mm/numa.c @@ -440,6 +440,8 @@ new_range: for (i = start ; i < (start+size); i += MEMORY_INCREMENT) numa_memory_lookup_table[i >> MEMORY_INCREMENT_SHIFT] = numa_domain; + memory_present(numa_domain, start >> PAGE_SHIFT, + (start + size) >> PAGE_SHIFT); if (--ranges) goto new_range; @@ -481,6 +483,7 @@ static void __init setup_nonnuma(void) for (i = 0 ; i < top_of_ram; i += MEMORY_INCREMENT) numa_memory_lookup_table[i >> MEMORY_INCREMENT_SHIFT] = 0; + memory_present(0, 0, init_node_data[0].node_end_pfn); } static void __init dump_numa_topology(void) diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index ab79af84699..6600ee87f89 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -226,6 +226,8 @@ config WARN_STACK_SIZE This allows you to specify the maximum frame size a function may have without the compiler complaining about it. +source "mm/Kconfig" + comment "I/O subsystem configuration" config MACHCHK_WARNING @@ -453,6 +455,14 @@ config NO_IDLE_HZ_INIT The HZ timer is switched off in idle by default. That means the HZ timer is already disabled at boot time. +config KEXEC + bool "kexec system call (EXPERIMENTAL)" + depends on EXPERIMENTAL + help + kexec is a system call that implements the ability to shutdown your + current kernel, and to start another kernel. It is like a reboot + but is independent of hardware/microcode support. + endmenu config PCMCIA diff --git a/arch/s390/boot/install.sh b/arch/s390/boot/install.sh index 278a8139cb1..d4026f62cb0 100644 --- a/arch/s390/boot/install.sh +++ b/arch/s390/boot/install.sh @@ -21,8 +21,8 @@ # User may have a custom install script -if [ -x ~/bin/installkernel ]; then exec ~/bin/installkernel "$@"; fi -if [ -x /sbin/installkernel ]; then exec /sbin/installkernel "$@"; fi +if [ -x ~/bin/${CROSS_COMPILE}installkernel ]; then exec ~/bin/${CROSS_COMPILE}installkernel "$@"; fi +if [ -x /sbin/${CROSS_COMPILE}installkernel ]; then exec /sbin/${CROSS_COMPILE}installkernel "$@"; fi # Default install - same as make zlilo diff --git a/arch/s390/defconfig b/arch/s390/defconfig index 07fd0414a4b..89850b2c27e 100644 --- a/arch/s390/defconfig +++ b/arch/s390/defconfig @@ -245,6 +245,7 @@ CONFIG_S390_TAPE_BLOCK=y # CONFIG_S390_TAPE_34XX=m # CONFIG_VMLOGRDR is not set +# CONFIG_VMCP is not set # CONFIG_MONREADER is not set # CONFIG_DCSS_SHM is not set diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index b41e0e199a7..ab1e49d2e51 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -25,6 +25,16 @@ obj-$(CONFIG_ARCH_S390X) += entry64.o reipl64.o obj-$(CONFIG_VIRT_TIMER) += vtime.o +# Kexec part +S390_KEXEC_OBJS := machine_kexec.o crash.o +ifeq ($(CONFIG_ARCH_S390X),y) +S390_KEXEC_OBJS += relocate_kernel64.o +else +S390_KEXEC_OBJS += relocate_kernel.o +endif +obj-$(CONFIG_KEXEC) += $(S390_KEXEC_OBJS) + + # # This is just to get the dependencies... # diff --git a/arch/s390/kernel/compat_linux.h b/arch/s390/kernel/compat_linux.h index bf33dcfec7d..3898f66d0b2 100644 --- a/arch/s390/kernel/compat_linux.h +++ b/arch/s390/kernel/compat_linux.h @@ -45,7 +45,7 @@ typedef struct compat_siginfo { /* POSIX.1b timers */ struct { - timer_t _tid; /* timer id */ + compat_timer_t _tid; /* timer id */ int _overrun; /* overrun count */ compat_sigval_t _sigval; /* same as below */ int _sys_private; /* not to be passed to user */ diff --git a/arch/s390/kernel/compat_wrapper.S b/arch/s390/kernel/compat_wrapper.S index 7a607b1d038..bf529739c8a 100644 --- a/arch/s390/kernel/compat_wrapper.S +++ b/arch/s390/kernel/compat_wrapper.S @@ -1441,3 +1441,11 @@ compat_sys_waitid_wrapper: lgfr %r5,%r5 # int llgtr %r6,%r6 # struct rusage_emu31 * jg compat_sys_waitid + + .globl compat_sys_kexec_load_wrapper +compat_sys_kexec_load_wrapper: + llgfr %r2,%r2 # unsigned long + llgfr %r3,%r3 # unsigned long + llgtr %r4,%r4 # struct kexec_segment * + llgfr %r5,%r5 # unsigned long + jg compat_sys_kexec_load diff --git a/arch/s390/kernel/cpcmd.c b/arch/s390/kernel/cpcmd.c index 44df8dc07c5..20062145e84 100644 --- a/arch/s390/kernel/cpcmd.c +++ b/arch/s390/kernel/cpcmd.c @@ -2,7 +2,7 @@ * arch/s390/kernel/cpcmd.c * * S390 version - * Copyright (C) 1999,2000 IBM Deutschland Entwicklung GmbH, IBM Corporation + * Copyright (C) 1999,2005 IBM Deutschland Entwicklung GmbH, IBM Corporation * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com), * Christian Borntraeger (cborntra@de.ibm.com), */ @@ -18,93 +18,114 @@ #include <asm/system.h> static DEFINE_SPINLOCK(cpcmd_lock); -static char cpcmd_buf[240]; +static char cpcmd_buf[241]; /* * the caller of __cpcmd has to ensure that the response buffer is below 2 GB */ -void __cpcmd(char *cmd, char *response, int rlen) +int __cpcmd(const char *cmd, char *response, int rlen, int *response_code) { const int mask = 0x40000000L; unsigned long flags; + int return_code; + int return_len; int cmdlen; spin_lock_irqsave(&cpcmd_lock, flags); cmdlen = strlen(cmd); BUG_ON(cmdlen > 240); - strcpy(cpcmd_buf, cmd); + memcpy(cpcmd_buf, cmd, cmdlen); ASCEBC(cpcmd_buf, cmdlen); if (response != NULL && rlen > 0) { memset(response, 0, rlen); #ifndef CONFIG_ARCH_S390X - asm volatile ("LRA 2,0(%0)\n\t" - "LR 4,%1\n\t" - "O 4,%4\n\t" - "LRA 3,0(%2)\n\t" - "LR 5,%3\n\t" - ".long 0x83240008 # Diagnose X'08'\n\t" - : /* no output */ - : "a" (cpcmd_buf), "d" (cmdlen), - "a" (response), "d" (rlen), "m" (mask) - : "cc", "2", "3", "4", "5" ); + asm volatile ( "lra 2,0(%2)\n" + "lr 4,%3\n" + "o 4,%6\n" + "lra 3,0(%4)\n" + "lr 5,%5\n" + "diag 2,4,0x8\n" + "brc 8, .Litfits\n" + "ar 5, %5\n" + ".Litfits: \n" + "lr %0,4\n" + "lr %1,5\n" + : "=d" (return_code), "=d" (return_len) + : "a" (cpcmd_buf), "d" (cmdlen), + "a" (response), "d" (rlen), "m" (mask) + : "cc", "2", "3", "4", "5" ); #else /* CONFIG_ARCH_S390X */ - asm volatile (" lrag 2,0(%0)\n" - " lgr 4,%1\n" - " o 4,%4\n" - " lrag 3,0(%2)\n" - " lgr 5,%3\n" - " sam31\n" - " .long 0x83240008 # Diagnose X'08'\n" - " sam64" - : /* no output */ - : "a" (cpcmd_buf), "d" (cmdlen), - "a" (response), "d" (rlen), "m" (mask) - : "cc", "2", "3", "4", "5" ); + asm volatile ( "lrag 2,0(%2)\n" + "lgr 4,%3\n" + "o 4,%6\n" + "lrag 3,0(%4)\n" + "lgr 5,%5\n" + "sam31\n" + "diag 2,4,0x8\n" + "sam64\n" + "brc 8, .Litfits\n" + "agr 5, %5\n" + ".Litfits: \n" + "lgr %0,4\n" + "lgr %1,5\n" + : "=d" (return_code), "=d" (return_len) + : "a" (cpcmd_buf), "d" (cmdlen), + "a" (response), "d" (rlen), "m" (mask) + : "cc", "2", "3", "4", "5" ); #endif /* CONFIG_ARCH_S390X */ EBCASC(response, rlen); } else { + return_len = 0; #ifndef CONFIG_ARCH_S390X - asm volatile ("LRA 2,0(%0)\n\t" - "LR 3,%1\n\t" - ".long 0x83230008 # Diagnose X'08'\n\t" - : /* no output */ - : "a" (cpcmd_buf), "d" (cmdlen) - : "2", "3" ); + asm volatile ( "lra 2,0(%1)\n" + "lr 3,%2\n" + "diag 2,3,0x8\n" + "lr %0,3\n" + : "=d" (return_code) + : "a" (cpcmd_buf), "d" (cmdlen) + : "2", "3" ); #else /* CONFIG_ARCH_S390X */ - asm volatile (" lrag 2,0(%0)\n" - " lgr 3,%1\n" - " sam31\n" - " .long 0x83230008 # Diagnose X'08'\n" - " sam64" - : /* no output */ - : "a" (cpcmd_buf), "d" (cmdlen) - : "2", "3" ); + asm volatile ( "lrag 2,0(%1)\n" + "lgr 3,%2\n" + "sam31\n" + "diag 2,3,0x8\n" + "sam64\n" + "lgr %0,3\n" + : "=d" (return_code) + : "a" (cpcmd_buf), "d" (cmdlen) + : "2", "3" ); #endif /* CONFIG_ARCH_S390X */ } spin_unlock_irqrestore(&cpcmd_lock, flags); + if (response_code != NULL) + *response_code = return_code; + return return_len; } EXPORT_SYMBOL(__cpcmd); #ifdef CONFIG_ARCH_S390X -void cpcmd(char *cmd, char *response, int rlen) +int cpcmd(const char *cmd, char *response, int rlen, int *response_code) { char *lowbuf; + int len; + if ((rlen == 0) || (response == NULL) || !((unsigned long)response >> 31)) - __cpcmd(cmd, response, rlen); + len = __cpcmd(cmd, response, rlen, response_code); else { lowbuf = kmalloc(rlen, GFP_KERNEL | GFP_DMA); if (!lowbuf) { printk(KERN_WARNING "cpcmd: could not allocate response buffer\n"); - return; + return -ENOMEM; } - __cpcmd(cmd, lowbuf, rlen); + len = __cpcmd(cmd, lowbuf, rlen, response_code); memcpy(response, lowbuf, rlen); kfree(lowbuf); } + return len; } EXPORT_SYMBOL(cpcmd); diff --git a/arch/s390/kernel/crash.c b/arch/s390/kernel/crash.c new file mode 100644 index 00000000000..7bd169c58b0 --- /dev/null +++ b/arch/s390/kernel/crash.c @@ -0,0 +1,17 @@ +/* + * arch/s390/kernel/crash.c + * + * (C) Copyright IBM Corp. 2005 + * + * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com> + * + */ + +#include <linux/threads.h> +#include <linux/kexec.h> + +note_buf_t crash_notes[NR_CPUS]; + +void machine_crash_shutdown(struct pt_regs *regs) +{ +} diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index 91f8ce5543d..960ba6029c3 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -19,22 +19,27 @@ #include <linux/sysctl.h> #include <asm/uaccess.h> #include <asm/semaphore.h> - #include <linux/module.h> #include <linux/init.h> +#include <linux/fs.h> +#include <linux/debugfs.h> #include <asm/debug.h> #define DEBUG_PROLOG_ENTRY -1 +#define ALL_AREAS 0 /* copy all debug areas */ +#define NO_AREAS 1 /* copy no debug areas */ + /* typedefs */ typedef struct file_private_info { loff_t offset; /* offset of last read in file */ int act_area; /* number of last formated area */ + int act_page; /* act page in given area */ int act_entry; /* last formated entry (offset */ /* relative to beginning of last */ - /* formated area) */ + /* formated page) */ size_t act_entry_offset; /* up to this offset we copied */ /* in last read the last formated */ /* entry to userland */ @@ -51,8 +56,8 @@ typedef struct * This assumes that all args are converted into longs * on L/390 this is the case for all types of parameter * except of floats, and long long (32 bit) - * - */ + * + */ long args[0]; } debug_sprintf_entry_t; @@ -63,32 +68,38 @@ extern void tod_to_timeval(uint64_t todval, struct timeval *xtime); static int debug_init(void); static ssize_t debug_output(struct file *file, char __user *user_buf, - size_t user_len, loff_t * offset); + size_t user_len, loff_t * offset); static ssize_t debug_input(struct file *file, const char __user *user_buf, - size_t user_len, loff_t * offset); + size_t user_len, loff_t * offset); static int debug_open(struct inode *inode, struct file *file); static int debug_close(struct inode *inode, struct file *file); -static debug_info_t* debug_info_create(char *name, int page_order, int nr_areas, int buf_size); +static debug_info_t* debug_info_create(char *name, int pages_per_area, + int nr_areas, int buf_size); static void debug_info_get(debug_info_t *); static void debug_info_put(debug_info_t *); static int debug_prolog_level_fn(debug_info_t * id, - struct debug_view *view, char *out_buf); + struct debug_view *view, char *out_buf); static int debug_input_level_fn(debug_info_t * id, struct debug_view *view, - struct file *file, const char __user *user_buf, - size_t user_buf_size, loff_t * offset); + struct file *file, const char __user *user_buf, + size_t user_buf_size, loff_t * offset); +static int debug_prolog_pages_fn(debug_info_t * id, + struct debug_view *view, char *out_buf); +static int debug_input_pages_fn(debug_info_t * id, struct debug_view *view, + struct file *file, const char __user *user_buf, + size_t user_buf_size, loff_t * offset); static int debug_input_flush_fn(debug_info_t * id, struct debug_view *view, - struct file *file, const char __user *user_buf, - size_t user_buf_size, loff_t * offset); + struct file *file, const char __user *user_buf, + size_t user_buf_size, loff_t * offset); static int debug_hex_ascii_format_fn(debug_info_t * id, struct debug_view *view, - char *out_buf, const char *in_buf); + char *out_buf, const char *in_buf); static int debug_raw_format_fn(debug_info_t * id, - struct debug_view *view, char *out_buf, - const char *in_buf); + struct debug_view *view, char *out_buf, + const char *in_buf); static int debug_raw_header_fn(debug_info_t * id, struct debug_view *view, - int area, debug_entry_t * entry, char *out_buf); + int area, debug_entry_t * entry, char *out_buf); static int debug_sprintf_format_fn(debug_info_t * id, struct debug_view *view, - char *out_buf, debug_sprintf_entry_t *curr_event); + char *out_buf, debug_sprintf_entry_t *curr_event); /* globals */ @@ -119,6 +130,15 @@ struct debug_view debug_level_view = { NULL }; +struct debug_view debug_pages_view = { + "pages", + &debug_prolog_pages_fn, + NULL, + NULL, + &debug_input_pages_fn, + NULL +}; + struct debug_view debug_flush_view = { "flush", NULL, @@ -149,98 +169,161 @@ DECLARE_MUTEX(debug_lock); static int initialized; static struct file_operations debug_file_ops = { - .owner = THIS_MODULE, + .owner = THIS_MODULE, .read = debug_output, - .write = debug_input, + .write = debug_input, .open = debug_open, .release = debug_close, }; -static struct proc_dir_entry *debug_proc_root_entry; +static struct dentry *debug_debugfs_root_entry; /* functions */ /* + * debug_areas_alloc + * - Debug areas are implemented as a threedimensonal array: + * areas[areanumber][pagenumber][pageoffset] + */ + +static debug_entry_t*** +debug_areas_alloc(int pages_per_area, int nr_areas) +{ + debug_entry_t*** areas; + int i,j; + + areas = (debug_entry_t ***) kmalloc(nr_areas * + sizeof(debug_entry_t**), + GFP_KERNEL); + if (!areas) + goto fail_malloc_areas; + for (i = 0; i < nr_areas; i++) { + areas[i] = (debug_entry_t**) kmalloc(pages_per_area * + sizeof(debug_entry_t*),GFP_KERNEL); + if (!areas[i]) { + goto fail_malloc_areas2; + } + for(j = 0; j < pages_per_area; j++) { + areas[i][j] = (debug_entry_t*)kmalloc(PAGE_SIZE, + GFP_KERNEL); + if(!areas[i][j]) { + for(j--; j >=0 ; j--) { + kfree(areas[i][j]); + } + kfree(areas[i]); + goto fail_malloc_areas2; + } else { + memset(areas[i][j],0,PAGE_SIZE); + } + } + } + return areas; + +fail_malloc_areas2: + for(i--; i >= 0; i--){ + for(j=0; j < pages_per_area;j++){ + kfree(areas[i][j]); + } + kfree(areas[i]); + } + kfree(areas); +fail_malloc_areas: + return NULL; + +} + + +/* * debug_info_alloc * - alloc new debug-info */ -static debug_info_t* debug_info_alloc(char *name, int page_order, - int nr_areas, int buf_size) +static debug_info_t* +debug_info_alloc(char *name, int pages_per_area, int nr_areas, int buf_size, + int level, int mode) { debug_info_t* rc; - int i; /* alloc everything */ - rc = (debug_info_t*) kmalloc(sizeof(debug_info_t), GFP_ATOMIC); + rc = (debug_info_t*) kmalloc(sizeof(debug_info_t), GFP_KERNEL); if(!rc) goto fail_malloc_rc; - rc->active_entry = (int*)kmalloc(nr_areas * sizeof(int), GFP_ATOMIC); - if(!rc->active_entry) - goto fail_malloc_active_entry; - memset(rc->active_entry, 0, nr_areas * sizeof(int)); - rc->areas = (debug_entry_t **) kmalloc(nr_areas * - sizeof(debug_entry_t *), - GFP_ATOMIC); - if (!rc->areas) - goto fail_malloc_areas; - for (i = 0; i < nr_areas; i++) { - rc->areas[i] = (debug_entry_t *) __get_free_pages(GFP_ATOMIC, - page_order); - if (!rc->areas[i]) { - for (i--; i >= 0; i--) { - free_pages((unsigned long) rc->areas[i], - page_order); - } - goto fail_malloc_areas2; - } else { - memset(rc->areas[i], 0, PAGE_SIZE << page_order); - } + rc->active_entries = (int*)kmalloc(nr_areas * sizeof(int), GFP_KERNEL); + if(!rc->active_entries) + goto fail_malloc_active_entries; + memset(rc->active_entries, 0, nr_areas * sizeof(int)); + rc->active_pages = (int*)kmalloc(nr_areas * sizeof(int), GFP_KERNEL); + if(!rc->active_pages) + goto fail_malloc_active_pages; + memset(rc->active_pages, 0, nr_areas * sizeof(int)); + if((mode == ALL_AREAS) && (pages_per_area != 0)){ + rc->areas = debug_areas_alloc(pages_per_area, nr_areas); + if(!rc->areas) + goto fail_malloc_areas; + } else { + rc->areas = NULL; } /* initialize members */ spin_lock_init(&rc->lock); - rc->page_order = page_order; - rc->nr_areas = nr_areas; - rc->active_area = 0; - rc->level = DEBUG_DEFAULT_LEVEL; - rc->buf_size = buf_size; - rc->entry_size = sizeof(debug_entry_t) + buf_size; - strlcpy(rc->name, name, sizeof(rc->name)); + rc->pages_per_area = pages_per_area; + rc->nr_areas = nr_areas; + rc->active_area = 0; + rc->level = level; + rc->buf_size = buf_size; + rc->entry_size = sizeof(debug_entry_t) + buf_size; + strlcpy(rc->name, name, sizeof(rc->name)-1); memset(rc->views, 0, DEBUG_MAX_VIEWS * sizeof(struct debug_view *)); -#ifdef CONFIG_PROC_FS - memset(rc->proc_entries, 0 ,DEBUG_MAX_VIEWS * - sizeof(struct proc_dir_entry*)); -#endif /* CONFIG_PROC_FS */ + memset(rc->debugfs_entries, 0 ,DEBUG_MAX_VIEWS * + sizeof(struct dentry*)); atomic_set(&(rc->ref_count), 0); return rc; -fail_malloc_areas2: - kfree(rc->areas); fail_malloc_areas: - kfree(rc->active_entry); -fail_malloc_active_entry: + kfree(rc->active_pages); +fail_malloc_active_pages: + kfree(rc->active_entries); +fail_malloc_active_entries: kfree(rc); fail_malloc_rc: return NULL; } /* - * debug_info_free - * - free memory debug-info + * debug_areas_free + * - free all debug areas */ -static void debug_info_free(debug_info_t* db_info){ - int i; +static void +debug_areas_free(debug_info_t* db_info) +{ + int i,j; + + if(!db_info->areas) + return; for (i = 0; i < db_info->nr_areas; i++) { - free_pages((unsigned long) db_info->areas[i], - db_info->page_order); + for(j = 0; j < db_info->pages_per_area; j++) { + kfree(db_info->areas[i][j]); + } + kfree(db_info->areas[i]); } kfree(db_info->areas); - kfree(db_info->active_entry); + db_info->areas = NULL; +} + +/* + * debug_info_free + * - free memory debug-info + */ + +static void +debug_info_free(debug_info_t* db_info){ + debug_areas_free(db_info); + kfree(db_info->active_entries); + kfree(db_info->active_pages); kfree(db_info); } @@ -249,21 +332,22 @@ static void debug_info_free(debug_info_t* db_info){ * - create new debug-info */ -static debug_info_t* debug_info_create(char *name, int page_order, - int nr_areas, int buf_size) +static debug_info_t* +debug_info_create(char *name, int pages_per_area, int nr_areas, int buf_size) { debug_info_t* rc; - rc = debug_info_alloc(name, page_order, nr_areas, buf_size); + rc = debug_info_alloc(name, pages_per_area, nr_areas, buf_size, + DEBUG_DEFAULT_LEVEL, ALL_AREAS); if(!rc) goto out; - - /* create proc rood directory */ - rc->proc_root_entry = proc_mkdir(rc->name, debug_proc_root_entry); + /* create root directory */ + rc->debugfs_root_entry = debugfs_create_dir(rc->name, + debug_debugfs_root_entry); /* append new element to linked list */ - if (debug_area_first == NULL) { + if (!debug_area_first) { /* first element in list */ debug_area_first = rc; rc->prev = NULL; @@ -285,17 +369,21 @@ out: * - copy debug-info */ -static debug_info_t* debug_info_copy(debug_info_t* in) +static debug_info_t* +debug_info_copy(debug_info_t* in, int mode) { - int i; + int i,j; debug_info_t* rc; - rc = debug_info_alloc(in->name, in->page_order, - in->nr_areas, in->buf_size); - if(!rc) + + rc = debug_info_alloc(in->name, in->pages_per_area, in->nr_areas, + in->buf_size, in->level, mode); + if(!rc || (mode == NO_AREAS)) goto out; for(i = 0; i < in->nr_areas; i++){ - memcpy(rc->areas[i],in->areas[i], PAGE_SIZE << in->page_order); + for(j = 0; j < in->pages_per_area; j++) { + memcpy(rc->areas[i][j], in->areas[i][j],PAGE_SIZE); + } } out: return rc; @@ -306,7 +394,8 @@ out: * - increments reference count for debug-info */ -static void debug_info_get(debug_info_t * db_info) +static void +debug_info_get(debug_info_t * db_info) { if (db_info) atomic_inc(&db_info->ref_count); @@ -317,29 +406,20 @@ static void debug_info_get(debug_info_t * db_info) * - decreases reference count for debug-info and frees it if necessary */ -static void debug_info_put(debug_info_t *db_info) +static void +debug_info_put(debug_info_t *db_info) { int i; if (!db_info) return; if (atomic_dec_and_test(&db_info->ref_count)) { -#ifdef DEBUG - printk(KERN_INFO "debug: freeing debug area %p (%s)\n", - db_info, db_info->name); -#endif for (i = 0; i < DEBUG_MAX_VIEWS; i++) { - if (db_info->views[i] == NULL) + if (!db_info->views[i]) continue; -#ifdef CONFIG_PROC_FS - remove_proc_entry(db_info->proc_entries[i]->name, - db_info->proc_root_entry); -#endif + debugfs_remove(db_info->debugfs_entries[i]); } -#ifdef CONFIG_PROC_FS - remove_proc_entry(db_info->proc_root_entry->name, - debug_proc_root_entry); -#endif + debugfs_remove(db_info->debugfs_root_entry); if(db_info == debug_area_first) debug_area_first = db_info->next; if(db_info == debug_area_last) @@ -355,9 +435,9 @@ static void debug_info_put(debug_info_t *db_info) * - format one debug entry and return size of formated data */ -static int debug_format_entry(file_private_info_t *p_info) +static int +debug_format_entry(file_private_info_t *p_info) { - debug_info_t *id_org = p_info->debug_info_org; debug_info_t *id_snap = p_info->debug_info_snap; struct debug_view *view = p_info->view; debug_entry_t *act_entry; @@ -365,22 +445,23 @@ static int debug_format_entry(file_private_info_t *p_info) if(p_info->act_entry == DEBUG_PROLOG_ENTRY){ /* print prolog */ if (view->prolog_proc) - len += view->prolog_proc(id_org, view,p_info->temp_buf); + len += view->prolog_proc(id_snap,view,p_info->temp_buf); goto out; } - - act_entry = (debug_entry_t *) ((char*)id_snap->areas[p_info->act_area] + - p_info->act_entry); + if (!id_snap->areas) /* this is true, if we have a prolog only view */ + goto out; /* or if 'pages_per_area' is 0 */ + act_entry = (debug_entry_t *) ((char*)id_snap->areas[p_info->act_area] + [p_info->act_page] + p_info->act_entry); if (act_entry->id.stck == 0LL) goto out; /* empty entry */ if (view->header_proc) - len += view->header_proc(id_org, view, p_info->act_area, + len += view->header_proc(id_snap, view, p_info->act_area, act_entry, p_info->temp_buf + len); if (view->format_proc) - len += view->format_proc(id_org, view, p_info->temp_buf + len, + len += view->format_proc(id_snap, view, p_info->temp_buf + len, DEBUG_DATA(act_entry)); - out: +out: return len; } @@ -389,20 +470,30 @@ static int debug_format_entry(file_private_info_t *p_info) * - goto next entry in p_info */ -extern inline int debug_next_entry(file_private_info_t *p_info) +extern inline int +debug_next_entry(file_private_info_t *p_info) { - debug_info_t *id = p_info->debug_info_snap; + debug_info_t *id; + + id = p_info->debug_info_snap; if(p_info->act_entry == DEBUG_PROLOG_ENTRY){ p_info->act_entry = 0; + p_info->act_page = 0; goto out; } - if ((p_info->act_entry += id->entry_size) - > ((PAGE_SIZE << (id->page_order)) - - id->entry_size)){ - - /* next area */ + if(!id->areas) + return 1; + p_info->act_entry += id->entry_size; + /* switch to next page, if we reached the end of the page */ + if (p_info->act_entry > (PAGE_SIZE - id->entry_size)){ + /* next page */ p_info->act_entry = 0; - p_info->act_area++; + p_info->act_page += 1; + if((p_info->act_page % id->pages_per_area) == 0) { + /* next area */ + p_info->act_area++; + p_info->act_page=0; + } if(p_info->act_area >= id->nr_areas) return 1; } @@ -416,13 +507,14 @@ out: * - copies formated debug entries to the user buffer */ -static ssize_t debug_output(struct file *file, /* file descriptor */ - char __user *user_buf, /* user buffer */ - size_t len, /* length of buffer */ - loff_t *offset) /* offset in the file */ +static ssize_t +debug_output(struct file *file, /* file descriptor */ + char __user *user_buf, /* user buffer */ + size_t len, /* length of buffer */ + loff_t *offset) /* offset in the file */ { size_t count = 0; - size_t entry_offset, size = 0; + size_t entry_offset; file_private_info_t *p_info; p_info = ((file_private_info_t *) file->private_data); @@ -430,27 +522,33 @@ static ssize_t debug_output(struct file *file, /* file descriptor */ return -EPIPE; if(p_info->act_area >= p_info->debug_info_snap->nr_areas) return 0; - entry_offset = p_info->act_entry_offset; - while(count < len){ - size = debug_format_entry(p_info); - size = min((len - count), (size - entry_offset)); - - if(size){ - if (copy_to_user(user_buf + count, - p_info->temp_buf + entry_offset, size)) - return -EFAULT; + int formatted_line_size; + int formatted_line_residue; + int user_buf_residue; + size_t copy_size; + + formatted_line_size = debug_format_entry(p_info); + formatted_line_residue = formatted_line_size - entry_offset; + user_buf_residue = len-count; + copy_size = min(user_buf_residue, formatted_line_residue); + if(copy_size){ + if (copy_to_user(user_buf + count, p_info->temp_buf + + entry_offset, copy_size)) + return -EFAULT; + count += copy_size; + entry_offset += copy_size; } - count += size; - entry_offset = 0; - if(count != len) - if(debug_next_entry(p_info)) + if(copy_size == formatted_line_residue){ + entry_offset = 0; + if(debug_next_entry(p_info)) goto out; + } } out: p_info->offset = *offset + count; - p_info->act_entry_offset = size; + p_info->act_entry_offset = entry_offset; *offset = p_info->offset; return count; } @@ -461,9 +559,9 @@ out: * - calls input function of view */ -static ssize_t debug_input(struct file *file, - const char __user *user_buf, size_t length, - loff_t *offset) +static ssize_t +debug_input(struct file *file, const char __user *user_buf, size_t length, + loff_t *offset) { int rc = 0; file_private_info_t *p_info; @@ -487,26 +585,23 @@ static ssize_t debug_input(struct file *file, * handle */ -static int debug_open(struct inode *inode, struct file *file) +static int +debug_open(struct inode *inode, struct file *file) { int i = 0, rc = 0; file_private_info_t *p_info; debug_info_t *debug_info, *debug_info_snapshot; -#ifdef DEBUG - printk("debug_open\n"); -#endif down(&debug_lock); /* find debug log and view */ - debug_info = debug_area_first; while(debug_info != NULL){ for (i = 0; i < DEBUG_MAX_VIEWS; i++) { - if (debug_info->views[i] == NULL) + if (!debug_info->views[i]) continue; - else if (debug_info->proc_entries[i] == - PDE(file->f_dentry->d_inode)) { + else if (debug_info->debugfs_entries[i] == + file->f_dentry) { goto found; /* found view ! */ } } @@ -516,41 +611,42 @@ static int debug_open(struct inode *inode, struct file *file) rc = -EINVAL; goto out; - found: +found: - /* make snapshot of current debug areas to get it consistent */ + /* Make snapshot of current debug areas to get it consistent. */ + /* To copy all the areas is only needed, if we have a view which */ + /* formats the debug areas. */ - debug_info_snapshot = debug_info_copy(debug_info); + if(!debug_info->views[i]->format_proc && + !debug_info->views[i]->header_proc){ + debug_info_snapshot = debug_info_copy(debug_info, NO_AREAS); + } else { + debug_info_snapshot = debug_info_copy(debug_info, ALL_AREAS); + } if(!debug_info_snapshot){ -#ifdef DEBUG - printk(KERN_ERR "debug_open: debug_info_copy failed (out of mem)\n"); -#endif rc = -ENOMEM; goto out; } - - if ((file->private_data = - kmalloc(sizeof(file_private_info_t), GFP_ATOMIC)) == 0) { -#ifdef DEBUG - printk(KERN_ERR "debug_open: kmalloc failed\n"); -#endif - debug_info_free(debug_info_snapshot); + p_info = (file_private_info_t *) kmalloc(sizeof(file_private_info_t), + GFP_KERNEL); + if(!p_info){ + if(debug_info_snapshot) + debug_info_free(debug_info_snapshot); rc = -ENOMEM; goto out; } - p_info = (file_private_info_t *) file->private_data; p_info->offset = 0; p_info->debug_info_snap = debug_info_snapshot; p_info->debug_info_org = debug_info; p_info->view = debug_info->views[i]; p_info->act_area = 0; + p_info->act_page = 0; p_info->act_entry = DEBUG_PROLOG_ENTRY; p_info->act_entry_offset = 0; - + file->private_data = p_info; debug_info_get(debug_info); - - out: +out: up(&debug_lock); return rc; } @@ -561,14 +657,13 @@ static int debug_open(struct inode *inode, struct file *file) * - deletes private_data area of the file handle */ -static int debug_close(struct inode *inode, struct file *file) +static int +debug_close(struct inode *inode, struct file *file) { file_private_info_t *p_info; -#ifdef DEBUG - printk("debug_close\n"); -#endif p_info = (file_private_info_t *) file->private_data; - debug_info_free(p_info->debug_info_snap); + if(p_info->debug_info_snap) + debug_info_free(p_info->debug_info_snap); debug_info_put(p_info->debug_info_org); kfree(file->private_data); return 0; /* success */ @@ -580,8 +675,8 @@ static int debug_close(struct inode *inode, struct file *file) * - returns handle for debug area */ -debug_info_t *debug_register - (char *name, int page_order, int nr_areas, int buf_size) +debug_info_t* +debug_register (char *name, int pages_per_area, int nr_areas, int buf_size) { debug_info_t *rc = NULL; @@ -591,18 +686,14 @@ debug_info_t *debug_register /* create new debug_info */ - rc = debug_info_create(name, page_order, nr_areas, buf_size); + rc = debug_info_create(name, pages_per_area, nr_areas, buf_size); if(!rc) goto out; debug_register_view(rc, &debug_level_view); debug_register_view(rc, &debug_flush_view); -#ifdef DEBUG - printk(KERN_INFO - "debug: reserved %d areas of %d pages for debugging %s\n", - nr_areas, 1 << page_order, rc->name); -#endif - out: - if (rc == NULL){ + debug_register_view(rc, &debug_pages_view); +out: + if (!rc){ printk(KERN_ERR "debug: debug_register failed for %s\n",name); } up(&debug_lock); @@ -614,27 +705,65 @@ debug_info_t *debug_register * - give back debug area */ -void debug_unregister(debug_info_t * id) +void +debug_unregister(debug_info_t * id) { if (!id) goto out; down(&debug_lock); -#ifdef DEBUG - printk(KERN_INFO "debug: unregistering %s\n", id->name); -#endif debug_info_put(id); up(&debug_lock); - out: +out: return; } /* + * debug_set_size: + * - set area size (number of pages) and number of areas + */ +static int +debug_set_size(debug_info_t* id, int nr_areas, int pages_per_area) +{ + unsigned long flags; + debug_entry_t *** new_areas; + int rc=0; + + if(!id || (nr_areas <= 0) || (pages_per_area < 0)) + return -EINVAL; + if(pages_per_area > 0){ + new_areas = debug_areas_alloc(pages_per_area, nr_areas); + if(!new_areas) { + printk(KERN_WARNING "debug: could not allocate memory "\ + "for pagenumber: %i\n",pages_per_area); + rc = -ENOMEM; + goto out; + } + } else { + new_areas = NULL; + } + spin_lock_irqsave(&id->lock,flags); + debug_areas_free(id); + id->areas = new_areas; + id->nr_areas = nr_areas; + id->pages_per_area = pages_per_area; + id->active_area = 0; + memset(id->active_entries,0,sizeof(int)*id->nr_areas); + memset(id->active_pages, 0, sizeof(int)*id->nr_areas); + spin_unlock_irqrestore(&id->lock,flags); + printk(KERN_INFO "debug: %s: set new size (%i pages)\n"\ + ,id->name, pages_per_area); +out: + return rc; +} + +/* * debug_set_level: * - set actual debug level */ -void debug_set_level(debug_info_t* id, int new_level) +void +debug_set_level(debug_info_t* id, int new_level) { unsigned long flags; if(!id) @@ -649,10 +778,6 @@ void debug_set_level(debug_info_t* id, int new_level) id->name, new_level, 0, DEBUG_MAX_LEVEL); } else { id->level = new_level; -#ifdef DEBUG - printk(KERN_INFO - "debug: %s: new level %i\n",id->name,id->level); -#endif } spin_unlock_irqrestore(&id->lock,flags); } @@ -663,11 +788,16 @@ void debug_set_level(debug_info_t* id, int new_level) * - set active entry to next in the ring buffer */ -extern inline void proceed_active_entry(debug_info_t * id) +extern inline void +proceed_active_entry(debug_info_t * id) { - if ((id->active_entry[id->active_area] += id->entry_size) - > ((PAGE_SIZE << (id->page_order)) - id->entry_size)) - id->active_entry[id->active_area] = 0; + if ((id->active_entries[id->active_area] += id->entry_size) + > (PAGE_SIZE - id->entry_size)){ + id->active_entries[id->active_area] = 0; + id->active_pages[id->active_area] = + (id->active_pages[id->active_area] + 1) % + id->pages_per_area; + } } /* @@ -675,7 +805,8 @@ extern inline void proceed_active_entry(debug_info_t * id) * - set active area to next in the ring buffer */ -extern inline void proceed_active_area(debug_info_t * id) +extern inline void +proceed_active_area(debug_info_t * id) { id->active_area++; id->active_area = id->active_area % id->nr_areas; @@ -685,10 +816,12 @@ extern inline void proceed_active_area(debug_info_t * id) * get_active_entry: */ -extern inline debug_entry_t *get_active_entry(debug_info_t * id) +extern inline debug_entry_t* +get_active_entry(debug_info_t * id) { - return (debug_entry_t *) ((char *) id->areas[id->active_area] + - id->active_entry[id->active_area]); + return (debug_entry_t *) (((char *) id->areas[id->active_area] + [id->active_pages[id->active_area]]) + + id->active_entries[id->active_area]); } /* @@ -696,8 +829,9 @@ extern inline debug_entry_t *get_active_entry(debug_info_t * id) * - set timestamp, caller address, cpu number etc. */ -extern inline void debug_finish_entry(debug_info_t * id, debug_entry_t* active, - int level, int exception) +extern inline void +debug_finish_entry(debug_info_t * id, debug_entry_t* active, int level, + int exception) { STCK(active->id.stck); active->id.fields.cpuid = smp_processor_id(); @@ -721,7 +855,8 @@ static int debug_active=1; * always allow read, allow write only if debug_stoppable is set or * if debug_active is already off */ -static int s390dbf_procactive(ctl_table *table, int write, struct file *filp, +static int +s390dbf_procactive(ctl_table *table, int write, struct file *filp, void __user *buffer, size_t *lenp, loff_t *ppos) { if (!write || debug_stoppable || !debug_active) @@ -766,7 +901,8 @@ static struct ctl_table s390dbf_dir_table[] = { struct ctl_table_header *s390dbf_sysctl_header; -void debug_stop_all(void) +void +debug_stop_all(void) { if (debug_stoppable) debug_active = 0; @@ -778,13 +914,13 @@ void debug_stop_all(void) * - write debug entry with given size */ -debug_entry_t *debug_event_common(debug_info_t * id, int level, const void *buf, - int len) +debug_entry_t* +debug_event_common(debug_info_t * id, int level, const void *buf, int len) { unsigned long flags; debug_entry_t *active; - if (!debug_active) + if (!debug_active || !id->areas) return NULL; spin_lock_irqsave(&id->lock, flags); active = get_active_entry(id); @@ -801,13 +937,13 @@ debug_entry_t *debug_event_common(debug_info_t * id, int level, const void *buf, * - write debug entry with given size and switch to next debug area */ -debug_entry_t *debug_exception_common(debug_info_t * id, int level, - const void *buf, int len) +debug_entry_t +*debug_exception_common(debug_info_t * id, int level, const void *buf, int len) { unsigned long flags; debug_entry_t *active; - if (!debug_active) + if (!debug_active || !id->areas) return NULL; spin_lock_irqsave(&id->lock, flags); active = get_active_entry(id); @@ -823,7 +959,8 @@ debug_entry_t *debug_exception_common(debug_info_t * id, int level, * counts arguments in format string for sprintf view */ -extern inline int debug_count_numargs(char *string) +extern inline int +debug_count_numargs(char *string) { int numargs=0; @@ -838,8 +975,8 @@ extern inline int debug_count_numargs(char *string) * debug_sprintf_event: */ -debug_entry_t *debug_sprintf_event(debug_info_t* id, - int level,char *string,...) +debug_entry_t* +debug_sprintf_event(debug_info_t* id, int level,char *string,...) { va_list ap; int numargs,idx; @@ -849,7 +986,7 @@ debug_entry_t *debug_sprintf_event(debug_info_t* id, if((!id) || (level > id->level)) return NULL; - if (!debug_active) + if (!debug_active || !id->areas) return NULL; numargs=debug_count_numargs(string); @@ -871,8 +1008,8 @@ debug_entry_t *debug_sprintf_event(debug_info_t* id, * debug_sprintf_exception: */ -debug_entry_t *debug_sprintf_exception(debug_info_t* id, - int level,char *string,...) +debug_entry_t* +debug_sprintf_exception(debug_info_t* id, int level,char *string,...) { va_list ap; int numargs,idx; @@ -882,7 +1019,7 @@ debug_entry_t *debug_sprintf_exception(debug_info_t* id, if((!id) || (level > id->level)) return NULL; - if (!debug_active) + if (!debug_active || !id->areas) return NULL; numargs=debug_count_numargs(string); @@ -906,15 +1043,14 @@ debug_entry_t *debug_sprintf_exception(debug_info_t* id, * - is called exactly once to initialize the debug feature */ -static int __init debug_init(void) +static int +__init debug_init(void) { int rc = 0; s390dbf_sysctl_header = register_sysctl_table(s390dbf_dir_table, 1); down(&debug_lock); -#ifdef CONFIG_PROC_FS - debug_proc_root_entry = proc_mkdir(DEBUG_DIR_ROOT, NULL); -#endif /* CONFIG_PROC_FS */ + debug_debugfs_root_entry = debugfs_create_dir(DEBUG_DIR_ROOT,NULL); printk(KERN_INFO "debug: Initialization complete\n"); initialized = 1; up(&debug_lock); @@ -926,13 +1062,14 @@ static int __init debug_init(void) * debug_register_view: */ -int debug_register_view(debug_info_t * id, struct debug_view *view) +int +debug_register_view(debug_info_t * id, struct debug_view *view) { int rc = 0; int i; unsigned long flags; mode_t mode = S_IFREG; - struct proc_dir_entry *pde; + struct dentry *pde; if (!id) goto out; @@ -940,16 +1077,17 @@ int debug_register_view(debug_info_t * id, struct debug_view *view) mode |= S_IRUSR; if (view->input_proc) mode |= S_IWUSR; - pde = create_proc_entry(view->name, mode, id->proc_root_entry); + pde = debugfs_create_file(view->name, mode, id->debugfs_root_entry, + NULL, &debug_file_ops); if (!pde){ - printk(KERN_WARNING "debug: create_proc_entry() failed! Cannot register view %s/%s\n", id->name,view->name); + printk(KERN_WARNING "debug: debugfs_create_file() failed!"\ + " Cannot register view %s/%s\n", id->name,view->name); rc = -1; goto out; } - spin_lock_irqsave(&id->lock, flags); for (i = 0; i < DEBUG_MAX_VIEWS; i++) { - if (id->views[i] == NULL) + if (!id->views[i]) break; } if (i == DEBUG_MAX_VIEWS) { @@ -957,16 +1095,14 @@ int debug_register_view(debug_info_t * id, struct debug_view *view) id->name,view->name); printk(KERN_WARNING "debug: maximum number of views reached (%i)!\n", i); - remove_proc_entry(pde->name, id->proc_root_entry); + debugfs_remove(pde); rc = -1; - } - else { + } else { id->views[i] = view; - pde->proc_fops = &debug_file_ops; - id->proc_entries[i] = pde; + id->debugfs_entries[i] = pde; } spin_unlock_irqrestore(&id->lock, flags); - out: +out: return rc; } @@ -974,7 +1110,8 @@ int debug_register_view(debug_info_t * id, struct debug_view *view) * debug_unregister_view: */ -int debug_unregister_view(debug_info_t * id, struct debug_view *view) +int +debug_unregister_view(debug_info_t * id, struct debug_view *view) { int rc = 0; int i; @@ -990,15 +1127,46 @@ int debug_unregister_view(debug_info_t * id, struct debug_view *view) if (i == DEBUG_MAX_VIEWS) rc = -1; else { -#ifdef CONFIG_PROC_FS - remove_proc_entry(id->proc_entries[i]->name, - id->proc_root_entry); -#endif + debugfs_remove(id->debugfs_entries[i]); id->views[i] = NULL; rc = 0; } spin_unlock_irqrestore(&id->lock, flags); - out: +out: + return rc; +} + +static inline char * +debug_get_user_string(const char __user *user_buf, size_t user_len) +{ + char* buffer; + + buffer = kmalloc(user_len + 1, GFP_KERNEL); + if (!buffer) + return ERR_PTR(-ENOMEM); + if (copy_from_user(buffer, user_buf, user_len) != 0) { + kfree(buffer); + return ERR_PTR(-EFAULT); + } + /* got the string, now strip linefeed. */ + if (buffer[user_len - 1] == '\n') + buffer[user_len - 1] = 0; + else + buffer[user_len] = 0; + return buffer; +} + +static inline int +debug_get_uint(char *buf) +{ + int rc; + + for(; isspace(*buf); buf++); + rc = simple_strtoul(buf, &buf, 10); + if(*buf){ + printk("debug: no integer specified!\n"); + rc = -EINVAL; + } return rc; } @@ -1011,13 +1179,69 @@ int debug_unregister_view(debug_info_t * id, struct debug_view *view) * prints out actual debug level */ -static int debug_prolog_level_fn(debug_info_t * id, +static int +debug_prolog_pages_fn(debug_info_t * id, struct debug_view *view, char *out_buf) { + return sprintf(out_buf, "%i\n", id->pages_per_area); +} + +/* + * reads new size (number of pages per debug area) + */ + +static int +debug_input_pages_fn(debug_info_t * id, struct debug_view *view, + struct file *file, const char __user *user_buf, + size_t user_len, loff_t * offset) +{ + char *str; + int rc,new_pages; + + if (user_len > 0x10000) + user_len = 0x10000; + if (*offset != 0){ + rc = -EPIPE; + goto out; + } + str = debug_get_user_string(user_buf,user_len); + if(IS_ERR(str)){ + rc = PTR_ERR(str); + goto out; + } + new_pages = debug_get_uint(str); + if(new_pages < 0){ + rc = -EINVAL; + goto free_str; + } + rc = debug_set_size(id,id->nr_areas, new_pages); + if(rc != 0){ + rc = -EINVAL; + goto free_str; + } + rc = user_len; +free_str: + kfree(str); +out: + *offset += user_len; + return rc; /* number of input characters */ +} + +/* + * prints out actual debug level + */ + +static int +debug_prolog_level_fn(debug_info_t * id, struct debug_view *view, char *out_buf) +{ int rc = 0; - if(id->level == -1) rc = sprintf(out_buf,"-\n"); - else rc = sprintf(out_buf, "%i\n", id->level); + if(id->level == DEBUG_OFF_LEVEL) { + rc = sprintf(out_buf,"-\n"); + } + else { + rc = sprintf(out_buf, "%i\n", id->level); + } return rc; } @@ -1025,30 +1249,43 @@ static int debug_prolog_level_fn(debug_info_t * id, * reads new debug level */ -static int debug_input_level_fn(debug_info_t * id, struct debug_view *view, - struct file *file, const char __user *user_buf, - size_t in_buf_size, loff_t * offset) +static int +debug_input_level_fn(debug_info_t * id, struct debug_view *view, + struct file *file, const char __user *user_buf, + size_t user_len, loff_t * offset) { - char input_buf[1]; - int rc = in_buf_size; + char *str; + int rc,new_level; - if (*offset != 0) + if (user_len > 0x10000) + user_len = 0x10000; + if (*offset != 0){ + rc = -EPIPE; goto out; - if (copy_from_user(input_buf, user_buf, 1)){ - rc = -EFAULT; + } + str = debug_get_user_string(user_buf,user_len); + if(IS_ERR(str)){ + rc = PTR_ERR(str); goto out; } - if (isdigit(input_buf[0])) { - int new_level = ((int) input_buf[0] - (int) '0'); - debug_set_level(id, new_level); - } else if(input_buf[0] == '-') { + if(str[0] == '-'){ debug_set_level(id, DEBUG_OFF_LEVEL); + rc = user_len; + goto free_str; } else { - printk(KERN_INFO "debug: level `%c` is not valid\n", - input_buf[0]); + new_level = debug_get_uint(str); } - out: - *offset += in_buf_size; + if(new_level < 0) { + printk(KERN_INFO "debug: level `%s` is not valid\n", str); + rc = -EINVAL; + } else { + debug_set_level(id, new_level); + rc = user_len; + } +free_str: + kfree(str); +out: + *offset += user_len; return rc; /* number of input characters */ } @@ -1057,29 +1294,36 @@ static int debug_input_level_fn(debug_info_t * id, struct debug_view *view, * flushes debug areas */ -void debug_flush(debug_info_t* id, int area) +void +debug_flush(debug_info_t* id, int area) { unsigned long flags; - int i; + int i,j; - if(!id) + if(!id || !id->areas) return; spin_lock_irqsave(&id->lock,flags); if(area == DEBUG_FLUSH_ALL){ id->active_area = 0; - memset(id->active_entry, 0, id->nr_areas * sizeof(int)); - for (i = 0; i < id->nr_areas; i++) - memset(id->areas[i], 0, PAGE_SIZE << id->page_order); + memset(id->active_entries, 0, id->nr_areas * sizeof(int)); + for (i = 0; i < id->nr_areas; i++) { + id->active_pages[i] = 0; + for(j = 0; j < id->pages_per_area; j++) { + memset(id->areas[i][j], 0, PAGE_SIZE); + } + } printk(KERN_INFO "debug: %s: all areas flushed\n",id->name); } else if(area >= 0 && area < id->nr_areas) { - id->active_entry[area] = 0; - memset(id->areas[area], 0, PAGE_SIZE << id->page_order); - printk(KERN_INFO - "debug: %s: area %i has been flushed\n", + id->active_entries[area] = 0; + id->active_pages[area] = 0; + for(i = 0; i < id->pages_per_area; i++) { + memset(id->areas[area][i],0,PAGE_SIZE); + } + printk(KERN_INFO "debug: %s: area %i has been flushed\n", id->name, area); } else { printk(KERN_INFO - "debug: %s: area %i cannot be flushed (range: %i - %i)\n", + "debug: %s: area %i cannot be flushed (range: %i - %i)\n", id->name, area, 0, id->nr_areas-1); } spin_unlock_irqrestore(&id->lock,flags); @@ -1089,15 +1333,20 @@ void debug_flush(debug_info_t* id, int area) * view function: flushes debug areas */ -static int debug_input_flush_fn(debug_info_t * id, struct debug_view *view, - struct file *file, const char __user *user_buf, - size_t in_buf_size, loff_t * offset) +static int +debug_input_flush_fn(debug_info_t * id, struct debug_view *view, + struct file *file, const char __user *user_buf, + size_t user_len, loff_t * offset) { char input_buf[1]; - int rc = in_buf_size; - - if (*offset != 0) + int rc = user_len; + + if (user_len > 0x10000) + user_len = 0x10000; + if (*offset != 0){ + rc = -EPIPE; goto out; + } if (copy_from_user(input_buf, user_buf, 1)){ rc = -EFAULT; goto out; @@ -1114,8 +1363,8 @@ static int debug_input_flush_fn(debug_info_t * id, struct debug_view *view, printk(KERN_INFO "debug: area `%c` is not valid\n", input_buf[0]); - out: - *offset += in_buf_size; +out: + *offset += user_len; return rc; /* number of input characters */ } @@ -1123,8 +1372,9 @@ static int debug_input_flush_fn(debug_info_t * id, struct debug_view *view, * prints debug header in raw format */ -int debug_raw_header_fn(debug_info_t * id, struct debug_view *view, - int area, debug_entry_t * entry, char *out_buf) +static int +debug_raw_header_fn(debug_info_t * id, struct debug_view *view, + int area, debug_entry_t * entry, char *out_buf) { int rc; @@ -1137,7 +1387,8 @@ int debug_raw_header_fn(debug_info_t * id, struct debug_view *view, * prints debug data in raw format */ -static int debug_raw_format_fn(debug_info_t * id, struct debug_view *view, +static int +debug_raw_format_fn(debug_info_t * id, struct debug_view *view, char *out_buf, const char *in_buf) { int rc; @@ -1151,8 +1402,9 @@ static int debug_raw_format_fn(debug_info_t * id, struct debug_view *view, * prints debug data in hex/ascii format */ -static int debug_hex_ascii_format_fn(debug_info_t * id, struct debug_view *view, - char *out_buf, const char *in_buf) +static int +debug_hex_ascii_format_fn(debug_info_t * id, struct debug_view *view, + char *out_buf, const char *in_buf) { int i, rc = 0; @@ -1176,7 +1428,8 @@ static int debug_hex_ascii_format_fn(debug_info_t * id, struct debug_view *view, * prints header for debug entry */ -int debug_dflt_header_fn(debug_info_t * id, struct debug_view *view, +int +debug_dflt_header_fn(debug_info_t * id, struct debug_view *view, int area, debug_entry_t * entry, char *out_buf) { struct timeval time_val; @@ -1210,8 +1463,9 @@ int debug_dflt_header_fn(debug_info_t * id, struct debug_view *view, #define DEBUG_SPRINTF_MAX_ARGS 10 -int debug_sprintf_format_fn(debug_info_t * id, struct debug_view *view, - char *out_buf, debug_sprintf_entry_t *curr_event) +static int +debug_sprintf_format_fn(debug_info_t * id, struct debug_view *view, + char *out_buf, debug_sprintf_entry_t *curr_event) { int num_longs, num_used_args = 0,i, rc = 0; int index[DEBUG_SPRINTF_MAX_ARGS]; @@ -1251,14 +1505,10 @@ out: /* * clean up module */ -void __exit debug_exit(void) +void +__exit debug_exit(void) { -#ifdef DEBUG - printk("debug_cleanup_module: \n"); -#endif -#ifdef CONFIG_PROC_FS - remove_proc_entry(debug_proc_root_entry->name, NULL); -#endif /* CONFIG_PROC_FS */ + debugfs_remove(debug_debugfs_root_entry); unregister_sysctl_table(s390dbf_sysctl_header); return; } @@ -1266,7 +1516,7 @@ void __exit debug_exit(void) /* * module definitions */ -core_initcall(debug_init); +postcore_initcall(debug_init); module_exit(debug_exit); MODULE_LICENSE("GPL"); diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S index c0e09b33feb..5b262b5d869 100644 --- a/arch/s390/kernel/entry.S +++ b/arch/s390/kernel/entry.S @@ -7,6 +7,7 @@ * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com), * Hartmut Penner (hp@de.ibm.com), * Denis Joseph Barrow (djbarrow@de.ibm.com,barrow_dj@yahoo.com), + * Heiko Carstens <heiko.carstens@de.ibm.com> */ #include <linux/sys.h> @@ -49,9 +50,9 @@ SP_ILC = STACK_FRAME_OVERHEAD + __PT_ILC SP_TRAP = STACK_FRAME_OVERHEAD + __PT_TRAP SP_SIZE = STACK_FRAME_OVERHEAD + __PT_SIZE -_TIF_WORK_SVC = (_TIF_SIGPENDING | _TIF_NEED_RESCHED | \ +_TIF_WORK_SVC = (_TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_MCCK_PENDING | \ _TIF_RESTART_SVC | _TIF_SINGLE_STEP ) -_TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NEED_RESCHED) +_TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_MCCK_PENDING) STACK_SHIFT = PAGE_SHIFT + THREAD_ORDER STACK_SIZE = 1 << STACK_SHIFT @@ -121,7 +122,11 @@ STACK_SIZE = 1 << STACK_SHIFT bz BASED(stack_overflow) 3: #endif -2: s %r15,BASED(.Lc_spsize) # make room for registers & psw +2: + .endm + + .macro CREATE_STACK_FRAME psworg,savearea + s %r15,BASED(.Lc_spsize) # make room for registers & psw mvc SP_PSW(8,%r15),0(%r12) # move user PSW to stack la %r12,\psworg st %r2,SP_ORIG_R2(%r15) # store original content of gpr 2 @@ -161,6 +166,13 @@ __switch_to_base: be __switch_to_noper-__switch_to_base(%r1) # we got away w/o bashing TLB's lctl %c9,%c11,__THREAD_per(%r3) # Nope we didn't __switch_to_noper: + l %r4,__THREAD_info(%r2) # get thread_info of prev + tm __TI_flags+3(%r4),_TIF_MCCK_PENDING # machine check pending? + bz __switch_to_no_mcck-__switch_to_base(%r1) + ni __TI_flags+3(%r4),255-_TIF_MCCK_PENDING # clear flag in prev + l %r4,__THREAD_info(%r3) # get thread_info of next + oi __TI_flags+3(%r4),_TIF_MCCK_PENDING # set it in next +__switch_to_no_mcck: stm %r6,%r15,__SF_GPRS(%r15)# store __switch_to registers of prev task st %r15,__THREAD_ksp(%r2) # store kernel stack to prev->tss.ksp l %r15,__THREAD_ksp(%r3) # load kernel stack from next->tss.ksp @@ -185,6 +197,7 @@ system_call: sysc_saveall: SAVE_ALL_BASE __LC_SAVE_AREA SAVE_ALL __LC_SVC_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA lh %r7,0x8a # get svc number from lowcore #ifdef CONFIG_VIRT_CPU_ACCOUNTING sysc_vtime: @@ -234,6 +247,8 @@ sysc_work_loop: # One of the work bits is on. Find out which one. # sysc_work: + tm __TI_flags+3(%r9),_TIF_MCCK_PENDING + bo BASED(sysc_mcck_pending) tm __TI_flags+3(%r9),_TIF_NEED_RESCHED bo BASED(sysc_reschedule) tm __TI_flags+3(%r9),_TIF_SIGPENDING @@ -253,6 +268,14 @@ sysc_reschedule: br %r1 # call scheduler # +# _TIF_MCCK_PENDING is set, call handler +# +sysc_mcck_pending: + l %r1,BASED(.Ls390_handle_mcck) + la %r14,BASED(sysc_work_loop) + br %r1 # TIF bit will be cleared by handler + +# # _TIF_SIGPENDING is set, call do_signal # sysc_sigpending: @@ -430,6 +453,7 @@ pgm_check_handler: tm __LC_PGM_INT_CODE+1,0x80 # check whether we got a per exception bnz BASED(pgm_per) # got per exception -> special case SAVE_ALL __LC_PGM_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? bz BASED(pgm_no_vtime) @@ -468,6 +492,7 @@ pgm_per: # pgm_per_std: SAVE_ALL __LC_PGM_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? bz BASED(pgm_no_vtime2) @@ -493,6 +518,7 @@ pgm_no_vtime2: # pgm_svcper: SAVE_ALL __LC_SVC_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? bz BASED(pgm_no_vtime3) @@ -521,6 +547,7 @@ io_int_handler: stck __LC_INT_CLOCK SAVE_ALL_BASE __LC_SAVE_AREA+16 SAVE_ALL __LC_IO_OLD_PSW,__LC_SAVE_AREA+16,0 + CREATE_STACK_FRAME __LC_IO_OLD_PSW,__LC_SAVE_AREA+16 #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? bz BASED(io_no_vtime) @@ -578,9 +605,11 @@ io_work: lr %r15,%r1 # # One of the work bits is on. Find out which one. -# Checked are: _TIF_SIGPENDING and _TIF_NEED_RESCHED +# Checked are: _TIF_SIGPENDING, _TIF_NEED_RESCHED and _TIF_MCCK_PENDING # io_work_loop: + tm __TI_flags+3(%r9),_TIF_MCCK_PENDING + bo BASED(io_mcck_pending) tm __TI_flags+3(%r9),_TIF_NEED_RESCHED bo BASED(io_reschedule) tm __TI_flags+3(%r9),_TIF_SIGPENDING @@ -588,6 +617,14 @@ io_work_loop: b BASED(io_leave) # +# _TIF_MCCK_PENDING is set, call handler +# +io_mcck_pending: + l %r1,BASED(.Ls390_handle_mcck) + l %r14,BASED(io_work_loop) + br %r1 # TIF bit will be cleared by handler + +# # _TIF_NEED_RESCHED is set, call schedule # io_reschedule: @@ -621,6 +658,7 @@ ext_int_handler: stck __LC_INT_CLOCK SAVE_ALL_BASE __LC_SAVE_AREA+16 SAVE_ALL __LC_EXT_OLD_PSW,__LC_SAVE_AREA+16,0 + CREATE_STACK_FRAME __LC_EXT_OLD_PSW,__LC_SAVE_AREA+16 #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? bz BASED(ext_no_vtime) @@ -642,19 +680,62 @@ ext_no_vtime: .globl mcck_int_handler mcck_int_handler: - STORE_TIMER __LC_ASYNC_ENTER_TIMER + spt __LC_CPU_TIMER_SAVE_AREA # revalidate cpu timer + lm %r0,%r15,__LC_GPREGS_SAVE_AREA # revalidate gprs SAVE_ALL_BASE __LC_SAVE_AREA+32 - SAVE_ALL __LC_MCK_OLD_PSW,__LC_SAVE_AREA+32,0 + la %r12,__LC_MCK_OLD_PSW + tm __LC_MCCK_CODE,0x80 # system damage? + bo BASED(mcck_int_main) # yes -> rest of mcck code invalid + tm __LC_MCCK_CODE+5,0x02 # stored cpu timer value valid? + bo BASED(0f) + spt __LC_LAST_UPDATE_TIMER # revalidate cpu timer #ifdef CONFIG_VIRT_CPU_ACCOUNTING - tm SP_PSW+1(%r15),0x01 # interrupting from user ? + mvc __LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER + mvc __LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER + mvc __LC_LAST_UPDATE_TIMER(8),__LC_EXIT_TIMER +0: tm __LC_MCCK_CODE+2,0x08 # mwp of old psw valid? + bno BASED(mcck_no_vtime) # no -> skip cleanup critical + tm __LC_MCK_OLD_PSW+1,0x01 # interrupting from user ? bz BASED(mcck_no_vtime) UPDATE_VTIME __LC_EXIT_TIMER,__LC_ASYNC_ENTER_TIMER,__LC_USER_TIMER UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER mvc __LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER mcck_no_vtime: #endif +0: + tm __LC_MCCK_CODE+2,0x09 # mwp + ia of old psw valid? + bno BASED(mcck_int_main) # no -> skip cleanup critical + tm __LC_MCK_OLD_PSW+1,0x01 # test problem state bit + bnz BASED(mcck_int_main) # from user -> load async stack + clc __LC_MCK_OLD_PSW+4(4),BASED(.Lcritical_end) + bhe BASED(mcck_int_main) + clc __LC_MCK_OLD_PSW+4(4),BASED(.Lcritical_start) + bl BASED(mcck_int_main) + l %r14,BASED(.Lcleanup_critical) + basr %r14,%r14 +mcck_int_main: + l %r14,__LC_PANIC_STACK # are we already on the panic stack? + slr %r14,%r15 + sra %r14,PAGE_SHIFT + be BASED(0f) + l %r15,__LC_PANIC_STACK # load panic stack +0: CREATE_STACK_FRAME __LC_MCK_OLD_PSW,__LC_SAVE_AREA+32 + l %r9,__LC_THREAD_INFO # load pointer to thread_info struct + la %r2,SP_PTREGS(%r15) # load pt_regs l %r1,BASED(.Ls390_mcck) - basr %r14,%r1 # call machine check handler + basr %r14,%r1 # call machine check handler + tm SP_PSW+1(%r15),0x01 # returning to user ? + bno BASED(mcck_return) + l %r1,__LC_KERNEL_STACK # switch to kernel stack + s %r1,BASED(.Lc_spsize) + mvc SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15) + xc __SF_BACKCHAIN(4,%r1),__SF_BACKCHAIN(%r1) # clear back chain + lr %r15,%r1 + stosm __SF_EMPTY(%r15),0x04 # turn dat on + tm __TI_flags+3(%r9),_TIF_MCCK_PENDING + bno BASED(mcck_return) + l %r1,BASED(.Ls390_handle_mcck) + basr %r14,%r1 # call machine check handler mcck_return: RESTORE_ALL 0 @@ -742,7 +823,7 @@ cleanup_critical: clc 4(4,%r12),BASED(cleanup_table_sysc_work_loop) bl BASED(0f) clc 4(4,%r12),BASED(cleanup_table_sysc_work_loop+4) - bl BASED(cleanup_sysc_leave) + bl BASED(cleanup_sysc_return) 0: br %r14 @@ -760,6 +841,7 @@ cleanup_system_call: mvc __LC_SAVE_AREA(16),__LC_SAVE_AREA+16 0: st %r13,__LC_SAVE_AREA+20 SAVE_ALL __LC_SVC_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA st %r15,__LC_SAVE_AREA+28 lh %r7,0x8a #ifdef CONFIG_VIRT_CPU_ACCOUNTING @@ -834,6 +916,8 @@ cleanup_sysc_leave_insn: * Symbol constants */ .Ls390_mcck: .long s390_do_machine_check +.Ls390_handle_mcck: + .long s390_handle_mcck .Ldo_IRQ: .long do_IRQ .Ldo_extint: .long do_extint .Ldo_signal: .long do_signal diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index 51527ab8c8f..57ca75d0ad7 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -7,6 +7,7 @@ * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com), * Hartmut Penner (hp@de.ibm.com), * Denis Joseph Barrow (djbarrow@de.ibm.com,barrow_dj@yahoo.com), + * Heiko Carstens <heiko.carstens@de.ibm.com> */ #include <linux/sys.h> @@ -52,9 +53,9 @@ SP_SIZE = STACK_FRAME_OVERHEAD + __PT_SIZE STACK_SHIFT = PAGE_SHIFT + THREAD_ORDER STACK_SIZE = 1 << STACK_SHIFT -_TIF_WORK_SVC = (_TIF_SIGPENDING | _TIF_NEED_RESCHED | \ +_TIF_WORK_SVC = (_TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_MCCK_PENDING | \ _TIF_RESTART_SVC | _TIF_SINGLE_STEP ) -_TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NEED_RESCHED) +_TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_MCCK_PENDING) #define BASED(name) name-system_call(%r13) @@ -114,7 +115,11 @@ _TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NEED_RESCHED) jz stack_overflow 3: #endif -2: aghi %r15,-SP_SIZE # make room for registers & psw +2: + .endm + + .macro CREATE_STACK_FRAME psworg,savearea + aghi %r15,-SP_SIZE # make room for registers & psw mvc SP_PSW(16,%r15),0(%r12) # move user PSW to stack la %r12,\psworg stg %r2,SP_ORIG_R2(%r15) # store original content of gpr 2 @@ -152,6 +157,13 @@ __switch_to: je __switch_to_noper # we got away without bashing TLB's lctlg %c9,%c11,__THREAD_per(%r3) # Nope we didn't __switch_to_noper: + lg %r4,__THREAD_info(%r2) # get thread_info of prev + tm __TI_flags+7(%r4),_TIF_MCCK_PENDING # machine check pending? + jz __switch_to_no_mcck + ni __TI_flags+7(%r4),255-_TIF_MCCK_PENDING # clear flag in prev + lg %r4,__THREAD_info(%r3) # get thread_info of next + oi __TI_flags+7(%r4),_TIF_MCCK_PENDING # set it in next +__switch_to_no_mcck: stmg %r6,%r15,__SF_GPRS(%r15)# store __switch_to registers of prev task stg %r15,__THREAD_ksp(%r2) # store kernel stack to prev->tss.ksp lg %r15,__THREAD_ksp(%r3) # load kernel stack from next->tss.ksp @@ -176,6 +188,7 @@ system_call: sysc_saveall: SAVE_ALL_BASE __LC_SAVE_AREA SAVE_ALL __LC_SVC_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA llgh %r7,__LC_SVC_INT_CODE # get svc number from lowcore #ifdef CONFIG_VIRT_CPU_ACCOUNTING sysc_vtime: @@ -232,6 +245,8 @@ sysc_work_loop: # One of the work bits is on. Find out which one. # sysc_work: + tm __TI_flags+7(%r9),_TIF_MCCK_PENDING + jo sysc_mcck_pending tm __TI_flags+7(%r9),_TIF_NEED_RESCHED jo sysc_reschedule tm __TI_flags+7(%r9),_TIF_SIGPENDING @@ -250,6 +265,13 @@ sysc_reschedule: jg schedule # return point is sysc_return # +# _TIF_MCCK_PENDING is set, call handler +# +sysc_mcck_pending: + larl %r14,sysc_work_loop + jg s390_handle_mcck # TIF bit will be cleared by handler + +# # _TIF_SIGPENDING is set, call do_signal # sysc_sigpending: @@ -474,6 +496,7 @@ pgm_check_handler: tm __LC_PGM_INT_CODE+1,0x80 # check whether we got a per exception jnz pgm_per # got per exception -> special case SAVE_ALL __LC_PGM_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? jz pgm_no_vtime @@ -512,6 +535,7 @@ pgm_per: # pgm_per_std: SAVE_ALL __LC_PGM_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? jz pgm_no_vtime2 @@ -537,6 +561,7 @@ pgm_no_vtime2: # pgm_svcper: SAVE_ALL __LC_SVC_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? jz pgm_no_vtime3 @@ -564,6 +589,7 @@ io_int_handler: stck __LC_INT_CLOCK SAVE_ALL_BASE __LC_SAVE_AREA+32 SAVE_ALL __LC_IO_OLD_PSW,__LC_SAVE_AREA+32,0 + CREATE_STACK_FRAME __LC_IO_OLD_PSW,__LC_SAVE_AREA+32 #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? jz io_no_vtime @@ -621,9 +647,11 @@ io_work: lgr %r15,%r1 # # One of the work bits is on. Find out which one. -# Checked are: _TIF_SIGPENDING and _TIF_NEED_RESCHED +# Checked are: _TIF_SIGPENDING, _TIF_NEED_RESCHED and _TIF_MCCK_PENDING # io_work_loop: + tm __TI_flags+7(%r9),_TIF_MCCK_PENDING + jo io_mcck_pending tm __TI_flags+7(%r9),_TIF_NEED_RESCHED jo io_reschedule tm __TI_flags+7(%r9),_TIF_SIGPENDING @@ -631,6 +659,13 @@ io_work_loop: j io_leave # +# _TIF_MCCK_PENDING is set, call handler +# +io_mcck_pending: + larl %r14,io_work_loop + jg s390_handle_mcck # TIF bit will be cleared by handler + +# # _TIF_NEED_RESCHED is set, call schedule # io_reschedule: @@ -661,6 +696,7 @@ ext_int_handler: stck __LC_INT_CLOCK SAVE_ALL_BASE __LC_SAVE_AREA+32 SAVE_ALL __LC_EXT_OLD_PSW,__LC_SAVE_AREA+32,0 + CREATE_STACK_FRAME __LC_EXT_OLD_PSW,__LC_SAVE_AREA+32 #ifdef CONFIG_VIRT_CPU_ACCOUNTING tm SP_PSW+1(%r15),0x01 # interrupting from user ? jz ext_no_vtime @@ -680,18 +716,60 @@ ext_no_vtime: */ .globl mcck_int_handler mcck_int_handler: - STORE_TIMER __LC_ASYNC_ENTER_TIMER + la %r1,4095 # revalidate r1 + spt __LC_CPU_TIMER_SAVE_AREA-4095(%r1) # revalidate cpu timer + lmg %r0,%r15,__LC_GPREGS_SAVE_AREA-4095(%r1)# revalidate gprs SAVE_ALL_BASE __LC_SAVE_AREA+64 - SAVE_ALL __LC_MCK_OLD_PSW,__LC_SAVE_AREA+64,0 + la %r12,__LC_MCK_OLD_PSW + tm __LC_MCCK_CODE,0x80 # system damage? + jo mcck_int_main # yes -> rest of mcck code invalid + tm __LC_MCCK_CODE+5,0x02 # stored cpu timer value valid? + jo 0f + spt __LC_LAST_UPDATE_TIMER #ifdef CONFIG_VIRT_CPU_ACCOUNTING - tm SP_PSW+1(%r15),0x01 # interrupting from user ? + mvc __LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER + mvc __LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER + mvc __LC_LAST_UPDATE_TIMER(8),__LC_EXIT_TIMER +0: tm __LC_MCCK_CODE+2,0x08 # mwp of old psw valid? + jno mcck_no_vtime # no -> no timer update + tm __LC_MCK_OLD_PSW+1,0x01 # interrupting from user ? jz mcck_no_vtime UPDATE_VTIME __LC_EXIT_TIMER,__LC_ASYNC_ENTER_TIMER,__LC_USER_TIMER UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER mvc __LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER mcck_no_vtime: #endif - brasl %r14,s390_do_machine_check +0: + tm __LC_MCCK_CODE+2,0x09 # mwp + ia of old psw valid? + jno mcck_int_main # no -> skip cleanup critical + tm __LC_MCK_OLD_PSW+1,0x01 # test problem state bit + jnz mcck_int_main # from user -> load kernel stack + clc __LC_MCK_OLD_PSW+8(8),BASED(.Lcritical_end) + jhe mcck_int_main + clc __LC_MCK_OLD_PSW+8(8),BASED(.Lcritical_start) + jl mcck_int_main + brasl %r14,cleanup_critical +mcck_int_main: + lg %r14,__LC_PANIC_STACK # are we already on the panic stack? + slgr %r14,%r15 + srag %r14,%r14,PAGE_SHIFT + jz 0f + lg %r15,__LC_PANIC_STACK # load panic stack +0: CREATE_STACK_FRAME __LC_MCK_OLD_PSW,__LC_SAVE_AREA+64 + lg %r9,__LC_THREAD_INFO # load pointer to thread_info struct + la %r2,SP_PTREGS(%r15) # load pt_regs + brasl %r14,s390_do_machine_check + tm SP_PSW+1(%r15),0x01 # returning to user ? + jno mcck_return + lg %r1,__LC_KERNEL_STACK # switch to kernel stack + aghi %r1,-SP_SIZE + mvc SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15) + xc __SF_BACKCHAIN(8,%r1),__SF_BACKCHAIN(%r1) # clear back chain + lgr %r15,%r1 + stosm __SF_EMPTY(%r15),0x04 # turn dat on + tm __TI_flags+7(%r9),_TIF_MCCK_PENDING + jno mcck_return + brasl %r14,s390_handle_mcck mcck_return: RESTORE_ALL 0 @@ -775,7 +853,7 @@ cleanup_critical: clc 8(8,%r12),BASED(cleanup_table_sysc_work_loop) jl 0f clc 8(8,%r12),BASED(cleanup_table_sysc_work_loop+8) - jl cleanup_sysc_leave + jl cleanup_sysc_return 0: br %r14 @@ -793,6 +871,7 @@ cleanup_system_call: mvc __LC_SAVE_AREA(32),__LC_SAVE_AREA+32 0: stg %r13,__LC_SAVE_AREA+40 SAVE_ALL __LC_SVC_OLD_PSW,__LC_SAVE_AREA,1 + CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA stg %r15,__LC_SAVE_AREA+56 llgh %r7,__LC_SVC_INT_CODE #ifdef CONFIG_VIRT_CPU_ACCOUNTING diff --git a/arch/s390/kernel/machine_kexec.c b/arch/s390/kernel/machine_kexec.c new file mode 100644 index 00000000000..2721c3a32b8 --- /dev/null +++ b/arch/s390/kernel/machine_kexec.c @@ -0,0 +1,98 @@ +/* + * arch/s390/kernel/machine_kexec.c + * + * (C) Copyright IBM Corp. 2005 + * + * Author(s): Rolf Adelsberger <adelsberger@de.ibm.com> + * + */ + +/* + * s390_machine_kexec.c - handle the transition of Linux booting another kernel + * on the S390 architecture. + */ + +#include <asm/cio.h> +#include <asm/setup.h> +#include <linux/device.h> +#include <linux/mm.h> +#include <linux/kexec.h> +#include <linux/delay.h> +#include <asm/pgtable.h> +#include <asm/pgalloc.h> +#include <asm/system.h> + +static void kexec_halt_all_cpus(void *); + +typedef void (*relocate_kernel_t) (kimage_entry_t *, unsigned long); + +const extern unsigned char relocate_kernel[]; +const extern unsigned long long relocate_kernel_len; + +int +machine_kexec_prepare(struct kimage *image) +{ + unsigned long reboot_code_buffer; + + /* We don't support anything but the default image type for now. */ + if (image->type != KEXEC_TYPE_DEFAULT) + return -EINVAL; + + /* Get the destination where the assembler code should be copied to.*/ + reboot_code_buffer = page_to_pfn(image->control_code_page)<<PAGE_SHIFT; + + /* Then copy it */ + memcpy((void *) reboot_code_buffer, relocate_kernel, + relocate_kernel_len); + return 0; +} + +void +machine_kexec_cleanup(struct kimage *image) +{ +} + +void +machine_shutdown(void) +{ + printk(KERN_INFO "kexec: machine_shutdown called\n"); +} + +NORET_TYPE void +machine_kexec(struct kimage *image) +{ + clear_all_subchannels(); + + /* Disable lowcore protection */ + ctl_clear_bit(0,28); + + on_each_cpu(kexec_halt_all_cpus, image, 0, 0); + for (;;); +} + +static void +kexec_halt_all_cpus(void *kernel_image) +{ + static atomic_t cpuid = ATOMIC_INIT(-1); + int cpu; + struct kimage *image; + relocate_kernel_t data_mover; + + if (atomic_compare_and_swap(-1, smp_processor_id(), &cpuid)) + signal_processor(smp_processor_id(), sigp_stop); + + /* Wait for all other cpus to enter stopped state */ + for_each_online_cpu(cpu) { + if (cpu == smp_processor_id()) + continue; + while (!smp_cpu_not_running(cpu)) + cpu_relax(); + } + + image = (struct kimage *) kernel_image; + data_mover = (relocate_kernel_t) + (page_to_pfn(image->control_code_page) << PAGE_SHIFT); + + /* Call the moving routine */ + (*data_mover) (&image->head, image->start); +} diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index 7aea25d6e30..9f3dff6c0b7 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -91,13 +91,12 @@ void do_monitor_call(struct pt_regs *regs, long interruption_code) (void *)(long) smp_processor_id()); } +extern void s390_handle_mcck(void); /* * The idle loop on a S390... */ void default_idle(void) { - psw_t wait_psw; - unsigned long reg; int cpu, rc; local_irq_disable(); @@ -125,38 +124,17 @@ void default_idle(void) cpu_die(); #endif - /* - * Wait for external, I/O or machine check interrupt and - * switch off machine check bit after the wait has ended. - */ - wait_psw.mask = PSW_KERNEL_BITS | PSW_MASK_MCHECK | PSW_MASK_WAIT | - PSW_MASK_IO | PSW_MASK_EXT; -#ifndef CONFIG_ARCH_S390X - asm volatile ( - " basr %0,0\n" - "0: la %0,1f-0b(%0)\n" - " st %0,4(%1)\n" - " oi 4(%1),0x80\n" - " lpsw 0(%1)\n" - "1: la %0,2f-1b(%0)\n" - " st %0,4(%1)\n" - " oi 4(%1),0x80\n" - " ni 1(%1),0xf9\n" - " lpsw 0(%1)\n" - "2:" - : "=&a" (reg) : "a" (&wait_psw) : "memory", "cc" ); -#else /* CONFIG_ARCH_S390X */ - asm volatile ( - " larl %0,0f\n" - " stg %0,8(%1)\n" - " lpswe 0(%1)\n" - "0: larl %0,1f\n" - " stg %0,8(%1)\n" - " ni 1(%1),0xf9\n" - " lpswe 0(%1)\n" - "1:" - : "=&a" (reg) : "a" (&wait_psw) : "memory", "cc" ); -#endif /* CONFIG_ARCH_S390X */ + local_mcck_disable(); + if (test_thread_flag(TIF_MCCK_PENDING)) { + local_mcck_enable(); + local_irq_enable(); + s390_handle_mcck(); + return; + } + + /* Wait for external, I/O or machine check interrupt. */ + __load_psw_mask(PSW_KERNEL_BITS | PSW_MASK_WAIT | + PSW_MASK_IO | PSW_MASK_EXT); } void cpu_idle(void) diff --git a/arch/s390/kernel/relocate_kernel.S b/arch/s390/kernel/relocate_kernel.S new file mode 100644 index 00000000000..d5e4a62fbb7 --- /dev/null +++ b/arch/s390/kernel/relocate_kernel.S @@ -0,0 +1,81 @@ +/* + * arch/s390/kernel/relocate_kernel.S + * + * (C) Copyright IBM Corp. 2005 + * + * Author(s): Rolf Adelsberger <adelsberger@de.ibm.com> + * + */ + +/* + * moves the new kernel to its destination... + * %r2 = pointer to first kimage_entry_t + * %r3 = start address - where to jump to after the job is done... + * + * %r5 will be used as temp. storage + * %r6 holds the destination address + * %r7 = PAGE_SIZE + * %r8 holds the source address + * %r9 = PAGE_SIZE + * %r10 is a page mask + */ + + .text + .globl relocate_kernel + relocate_kernel: + basr %r13,0 #base address + .base: + spx zero64-.base(%r13) #absolute addressing mode + stnsm sys_msk-.base(%r13),0xf8 #disable DAT and IRQ (external) + lhi %r10,-1 #preparing the mask + sll %r10,12 #shift it such that it becomes 0xf000 + .top: + lhi %r7,4096 #load PAGE_SIZE in r7 + lhi %r9,4096 #load PAGE_SIZE in r9 + l %r5,0(%r2) #read another word for indirection page + ahi %r2,4 #increment pointer + tml %r5,0x1 #is it a destination page? + je .indir_check #NO, goto "indir_check" + lr %r6,%r5 #r6 = r5 + nr %r6,%r10 #mask it out and... + j .top #...next iteration + .indir_check: + tml %r5,0x2 #is it a indirection page? + je .done_test #NO, goto "done_test" + nr %r5,%r10 #YES, mask out, + lr %r2,%r5 #move it into the right register, + j .top #and read next... + .done_test: + tml %r5,0x4 #is it the done indicator? + je .source_test #NO! Well, then it should be the source indicator... + j .done #ok, lets finish it here... + .source_test: + tml %r5,0x8 #it should be a source indicator... + je .top #NO, ignore it... + lr %r8,%r5 #r8 = r5 + nr %r8,%r10 #masking + 0: mvcle %r6,%r8,0x0 #copy PAGE_SIZE bytes from r8 to r6 - pad with 0 + jo 0b + j .top + .done: + sr %r0,%r0 #clear register r0 + la %r4,load_psw-.base(%r13) #load psw-address into the register + o %r3,4(%r4) #or load address into psw + st %r3,4(%r4) + mvc 0(8,%r0),0(%r4) #copy psw to absolute address 0 + sr %r1,%r1 #clear %r1 + sr %r2,%r2 #clear %r2 + sigp %r1,%r2,0x12 #set cpuid to zero + lpsw 0 #hopefully start new kernel... + + .align 8 + zero64: + .quad 0 + load_psw: + .long 0x00080000,0x80000000 + sys_msk: + .quad 0 + relocate_kernel_end: + .globl relocate_kernel_len + relocate_kernel_len: + .quad relocate_kernel_end - relocate_kernel diff --git a/arch/s390/kernel/relocate_kernel64.S b/arch/s390/kernel/relocate_kernel64.S new file mode 100644 index 00000000000..96290cc4eb3 --- /dev/null +++ b/arch/s390/kernel/relocate_kernel64.S @@ -0,0 +1,82 @@ +/* + * arch/s390/kernel/relocate_kernel64.S + * + * (C) Copyright IBM Corp. 2005 + * + * Author(s): Rolf Adelsberger <adelsberger@de.ibm.com> + * + */ + +/* + * moves the new kernel to its destination... + * %r2 = pointer to first kimage_entry_t + * %r3 = start address - where to jump to after the job is done... + * + * %r5 will be used as temp. storage + * %r6 holds the destination address + * %r7 = PAGE_SIZE + * %r8 holds the source address + * %r9 = PAGE_SIZE + * + * 0xf000 is a page_mask + */ + + .text + .globl relocate_kernel + relocate_kernel: + basr %r13,0 #base address + .base: + spx zero64-.base(%r13) #absolute addressing mode + stnsm sys_msk-.base(%r13),0xf8 #disable DAT and IRQ (external) + .top: + lghi %r7,4096 #load PAGE_SIZE in r7 + lghi %r9,4096 #load PAGE_SIZE in r9 + lg %r5,0(%r2) #read another word for indirection page + aghi %r2,8 #increment pointer + tml %r5,0x1 #is it a destination page? + je .indir_check #NO, goto "indir_check" + lgr %r6,%r5 #r6 = r5 + nill %r6,0xf000 #mask it out and... + j .top #...next iteration + .indir_check: + tml %r5,0x2 #is it a indirection page? + je .done_test #NO, goto "done_test" + nill %r5,0xf000 #YES, mask out, + lgr %r2,%r5 #move it into the right register, + j .top #and read next... + .done_test: + tml %r5,0x4 #is it the done indicator? + je .source_test #NO! Well, then it should be the source indicator... + j .done #ok, lets finish it here... + .source_test: + tml %r5,0x8 #it should be a source indicator... + je .top #NO, ignore it... + lgr %r8,%r5 #r8 = r5 + nill %r8,0xf000 #masking + 0: mvcle %r6,%r8,0x0 #copy PAGE_SIZE bytes from r8 to r6 - pad with 0 + jo 0b + j .top + .done: + sgr %r0,%r0 #clear register r0 + la %r4,load_psw-.base(%r13) #load psw-address into the register + o %r3,4(%r4) #or load address into psw + st %r3,4(%r4) + mvc 0(8,%r0),0(%r4) #copy psw to absolute address 0 + sam31 #31 bit mode + sr %r1,%r1 #erase register r1 + sr %r2,%r2 #erase register r2 + sigp %r1,%r2,0x12 #set cpuid to zero + lpsw 0 #hopefully start new kernel... + + .align 8 + zero64: + .quad 0 + load_psw: + .long 0x00080000,0x80000000 + sys_msk: + .quad 0 + relocate_kernel_end: + .globl relocate_kernel_len + relocate_kernel_len: + .quad relocate_kernel_end - relocate_kernel + diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index df83215beac..b6d740ac0e6 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -198,11 +198,11 @@ static void __init conmode_default(void) char *ptr; if (MACHINE_IS_VM) { - __cpcmd("QUERY CONSOLE", query_buffer, 1024); + __cpcmd("QUERY CONSOLE", query_buffer, 1024, NULL); console_devno = simple_strtoul(query_buffer + 5, NULL, 16); ptr = strstr(query_buffer, "SUBCHANNEL ="); console_irq = simple_strtoul(ptr + 13, NULL, 16); - __cpcmd("QUERY TERM", query_buffer, 1024); + __cpcmd("QUERY TERM", query_buffer, 1024, NULL); ptr = strstr(query_buffer, "CONMODE"); /* * Set the conmode to 3215 so that the device recognition @@ -211,7 +211,7 @@ static void __init conmode_default(void) * 3215 and the 3270 driver will try to access the console * device (3215 as console and 3270 as normal tty). */ - __cpcmd("TERM CONMODE 3215", NULL, 0); + __cpcmd("TERM CONMODE 3215", NULL, 0, NULL); if (ptr == NULL) { #if defined(CONFIG_SCLP_CONSOLE) SET_CONSOLE_SCLP; @@ -414,7 +414,8 @@ setup_lowcore(void) lc->program_new_psw.mask = PSW_KERNEL_BITS; lc->program_new_psw.addr = PSW_ADDR_AMODE | (unsigned long)pgm_check_handler; - lc->mcck_new_psw.mask = PSW_KERNEL_BITS; + lc->mcck_new_psw.mask = + PSW_KERNEL_BITS & ~PSW_MASK_MCHECK & ~PSW_MASK_DAT; lc->mcck_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) mcck_int_handler; lc->io_new_psw.mask = PSW_KERNEL_BITS; @@ -424,12 +425,18 @@ setup_lowcore(void) lc->kernel_stack = ((unsigned long) &init_thread_union) + THREAD_SIZE; lc->async_stack = (unsigned long) __alloc_bootmem(ASYNC_SIZE, ASYNC_SIZE, 0) + ASYNC_SIZE; -#ifdef CONFIG_CHECK_STACK lc->panic_stack = (unsigned long) __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0) + PAGE_SIZE; -#endif lc->current_task = (unsigned long) init_thread_union.thread_info.task; lc->thread_info = (unsigned long) &init_thread_union; +#ifndef CONFIG_ARCH_S390X + if (MACHINE_HAS_IEEE) { + lc->extended_save_area_addr = (__u32) + __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0); + /* enable extended save area */ + ctl_set_bit(14, 29); + } +#endif #ifdef CONFIG_ARCH_S390X if (MACHINE_HAS_DIAG44) lc->diag44_opcode = 0x83000044; diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index fdfcf0488b4..642572a8e33 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -284,7 +284,7 @@ static void do_machine_restart(void * __unused) * locks are always held disabled). */ if (MACHINE_IS_VM) - cpcmd ("IPL", NULL, 0); + cpcmd ("IPL", NULL, 0, NULL); else reipl (0x10000 | S390_lowcore.ipl_device); } @@ -313,7 +313,7 @@ static void do_machine_halt(void * __unused) if (atomic_compare_and_swap(-1, smp_processor_id(), &cpuid) == 0) { smp_send_stop(); if (MACHINE_IS_VM && strlen(vmhalt_cmd) > 0) - cpcmd(vmhalt_cmd, NULL, 0); + cpcmd(vmhalt_cmd, NULL, 0, NULL); signal_processor(smp_processor_id(), sigp_stop_and_store_status); } @@ -332,7 +332,7 @@ static void do_machine_power_off(void * __unused) if (atomic_compare_and_swap(-1, smp_processor_id(), &cpuid) == 0) { smp_send_stop(); if (MACHINE_IS_VM && strlen(vmpoff_cmd) > 0) - cpcmd(vmpoff_cmd, NULL, 0); + cpcmd(vmpoff_cmd, NULL, 0, NULL); signal_processor(smp_processor_id(), sigp_stop_and_store_status); } @@ -679,12 +679,14 @@ __cpu_disable(void) { unsigned long flags; ec_creg_mask_parms cr_parms; + int cpu = smp_processor_id(); spin_lock_irqsave(&smp_reserve_lock, flags); - if (smp_cpu_reserved[smp_processor_id()] != 0) { + if (smp_cpu_reserved[cpu] != 0) { spin_unlock_irqrestore(&smp_reserve_lock, flags); return -EBUSY; } + cpu_clear(cpu, cpu_online_map); #ifdef CONFIG_PFAULT /* Disable pfault pseudo page faults on this cpu. */ @@ -771,13 +773,24 @@ void __init smp_prepare_cpus(unsigned int max_cpus) *(lowcore_ptr[i]) = S390_lowcore; lowcore_ptr[i]->async_stack = stack + (ASYNC_SIZE); -#ifdef CONFIG_CHECK_STACK stack = __get_free_pages(GFP_KERNEL,0); if (stack == 0ULL) panic("smp_boot_cpus failed to allocate memory\n"); lowcore_ptr[i]->panic_stack = stack + (PAGE_SIZE); +#ifndef __s390x__ + if (MACHINE_HAS_IEEE) { + lowcore_ptr[i]->extended_save_area_addr = + (__u32) __get_free_pages(GFP_KERNEL,0); + if (lowcore_ptr[i]->extended_save_area_addr == 0) + panic("smp_boot_cpus failed to " + "allocate memory\n"); + } #endif } +#ifndef __s390x__ + if (MACHINE_HAS_IEEE) + ctl_set_bit(14, 29); /* enable extended save area */ +#endif set_prefix((u32)(unsigned long) lowcore_ptr[smp_processor_id()]); for_each_cpu(cpu) diff --git a/arch/s390/kernel/syscalls.S b/arch/s390/kernel/syscalls.S index 515938628f8..a8668afb5f8 100644 --- a/arch/s390/kernel/syscalls.S +++ b/arch/s390/kernel/syscalls.S @@ -285,7 +285,7 @@ SYSCALL(sys_mq_timedsend,sys_mq_timedsend,compat_sys_mq_timedsend_wrapper) SYSCALL(sys_mq_timedreceive,sys_mq_timedreceive,compat_sys_mq_timedreceive_wrapper) SYSCALL(sys_mq_notify,sys_mq_notify,compat_sys_mq_notify_wrapper) /* 275 */ SYSCALL(sys_mq_getsetattr,sys_mq_getsetattr,compat_sys_mq_getsetattr_wrapper) -NI_SYSCALL /* reserved for kexec */ +SYSCALL(sys_kexec_load,sys_kexec_load,compat_sys_kexec_load_wrapper) SYSCALL(sys_add_key,sys_add_key,compat_sys_add_key_wrapper) SYSCALL(sys_request_key,sys_request_key,compat_sys_request_key_wrapper) SYSCALL(sys_keyctl,sys_keyctl,compat_sys_keyctl) /* 280 */ diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c index ca34b6f34b3..bc7b7be7acb 100644 --- a/arch/s390/kernel/traps.c +++ b/arch/s390/kernel/traps.c @@ -735,7 +735,7 @@ void __init trap_init(void) &ext_int_pfault); #endif #ifndef CONFIG_ARCH_S390X - cpcmd("SET PAGEX ON", NULL, 0); + cpcmd("SET PAGEX ON", NULL, 0, NULL); #endif } } diff --git a/arch/s390/mm/extmem.c b/arch/s390/mm/extmem.c index 648deed17e2..c5348108ca3 100644 --- a/arch/s390/mm/extmem.c +++ b/arch/s390/mm/extmem.c @@ -576,8 +576,8 @@ segment_save(char *name) segtype_string[seg->range[i].start & 0xff]); } sprintf(cmd2, "SAVESEG %s", name); - cpcmd(cmd1, NULL, 0); - cpcmd(cmd2, NULL, 0); + cpcmd(cmd1, NULL, 0, NULL); + cpcmd(cmd2, NULL, 0, NULL); spin_unlock(&dcss_lock); } diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig index 3468d512722..a7c8bfc1160 100644 --- a/arch/sh/Kconfig +++ b/arch/sh/Kconfig @@ -486,7 +486,7 @@ config CPU_SUBTYPE_ST40 depends on CPU_SUBTYPE_ST40STB1 || CPU_SUBTYPE_ST40GX1 default y -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool depends on SH_HP690 default y @@ -496,6 +496,8 @@ config DISCONTIGMEM or have huge holes in the physical address space for other reasons. See <file:Documentation/vm/numa> for more. +source "mm/Kconfig" + config ZERO_PAGE_OFFSET hex "Zero page offset" default "0x00001000" if !(SH_MPC1211 || SH_SH03) diff --git a/arch/sh64/Kconfig b/arch/sh64/Kconfig index 76eb81fba45..708e59736a4 100644 --- a/arch/sh64/Kconfig +++ b/arch/sh64/Kconfig @@ -217,6 +217,8 @@ config PREEMPT bool "Preemptible Kernel (EXPERIMENTAL)" depends on EXPERIMENTAL +source "mm/Kconfig" + endmenu menu "Bus options (PCI, PCMCIA, EISA, MCA, ISA)" diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 237f922520f..262e13d086f 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -264,7 +264,11 @@ config SUNOS_EMUL want to run SunOS binaries on an Ultra you must also say Y to "Kernel support for 32-bit a.out binaries" above. -source "drivers/parport/Kconfig" +source "mm/Kconfig" + +endmenu + +source "drivers/Kconfig" config PRINTER tristate "Parallel printer support" @@ -291,6 +295,8 @@ config PRINTER If you have more than 8 printers, you need to increase the LP_NO macro in lp.c and the PARPORT_MAX macro in parport.h. +source "mm/Kconfig" + endmenu source "drivers/base/Kconfig" @@ -372,18 +378,8 @@ config UNIX98_PTY_COUNT endmenu -source "drivers/input/Kconfig" - source "fs/Kconfig" -source "sound/Kconfig" - -source "drivers/usb/Kconfig" - -source "drivers/infiniband/Kconfig" - -source "drivers/char/watchdog/Kconfig" - source "arch/sparc/Kconfig.debug" source "security/Kconfig" diff --git a/arch/sparc64/Kconfig b/arch/sparc64/Kconfig index a72fd15d5ea..e2b050eb3b9 100644 --- a/arch/sparc64/Kconfig +++ b/arch/sparc64/Kconfig @@ -484,6 +484,8 @@ config CMDLINE NOTE: This option WILL override the PROM bootargs setting! +source "mm/Kconfig" + endmenu source "drivers/base/Kconfig" diff --git a/arch/sparc64/kernel/kprobes.c b/arch/sparc64/kernel/kprobes.c index 7066d7ba667..bdac631cf01 100644 --- a/arch/sparc64/kernel/kprobes.c +++ b/arch/sparc64/kernel/kprobes.c @@ -6,7 +6,6 @@ #include <linux/config.h> #include <linux/kernel.h> #include <linux/kprobes.h> - #include <asm/kdebug.h> #include <asm/signal.h> @@ -47,25 +46,59 @@ void arch_copy_kprobe(struct kprobe *p) { p->ainsn.insn[0] = *p->addr; p->ainsn.insn[1] = BREAKPOINT_INSTRUCTION_2; + p->opcode = *p->addr; } -void arch_remove_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { + *p->addr = BREAKPOINT_INSTRUCTION; + flushi(p->addr); } -/* kprobe_status settings */ -#define KPROBE_HIT_ACTIVE 0x00000001 -#define KPROBE_HIT_SS 0x00000002 +void arch_disarm_kprobe(struct kprobe *p) +{ + *p->addr = p->opcode; + flushi(p->addr); +} + +void arch_remove_kprobe(struct kprobe *p) +{ +} static struct kprobe *current_kprobe; static unsigned long current_kprobe_orig_tnpc; static unsigned long current_kprobe_orig_tstate_pil; static unsigned int kprobe_status; +static struct kprobe *kprobe_prev; +static unsigned long kprobe_orig_tnpc_prev; +static unsigned long kprobe_orig_tstate_pil_prev; +static unsigned int kprobe_status_prev; -static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) +static inline void save_previous_kprobe(void) +{ + kprobe_status_prev = kprobe_status; + kprobe_orig_tnpc_prev = current_kprobe_orig_tnpc; + kprobe_orig_tstate_pil_prev = current_kprobe_orig_tstate_pil; + kprobe_prev = current_kprobe; +} + +static inline void restore_previous_kprobe(void) +{ + kprobe_status = kprobe_status_prev; + current_kprobe_orig_tnpc = kprobe_orig_tnpc_prev; + current_kprobe_orig_tstate_pil = kprobe_orig_tstate_pil_prev; + current_kprobe = kprobe_prev; +} + +static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs) { current_kprobe_orig_tnpc = regs->tnpc; current_kprobe_orig_tstate_pil = (regs->tstate & TSTATE_PIL); + current_kprobe = p; +} + +static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) +{ regs->tstate |= TSTATE_PIL; /*single step inline, if it a breakpoint instruction*/ @@ -78,17 +111,6 @@ static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) } } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) -{ - *p->addr = p->opcode; - flushi(p->addr); - - regs->tpc = (unsigned long) p->addr; - regs->tnpc = current_kprobe_orig_tnpc; - regs->tstate = ((regs->tstate & ~TSTATE_PIL) | - current_kprobe_orig_tstate_pil); -} - static int kprobe_handler(struct pt_regs *regs) { struct kprobe *p; @@ -109,8 +131,18 @@ static int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); - ret = 1; + /* We have reentered the kprobe_handler(), since + * another probe was hit while within the handler. + * We here save the original kprobes variables and + * just single step on the instruction of the new probe + * without calling any user handlers. + */ + save_previous_kprobe(); + set_current_kprobe(p, regs); + p->nmissed++; + kprobe_status = KPROBE_REENTER; + prepare_singlestep(p, regs); + return 1; } else { p = current_kprobe; if (p->break_handler && p->break_handler(p, regs)) @@ -138,8 +170,8 @@ static int kprobe_handler(struct pt_regs *regs) goto no_kprobe; } + set_current_kprobe(p, regs); kprobe_status = KPROBE_HIT_ACTIVE; - current_kprobe = p; if (p->pre_handler && p->pre_handler(p, regs)) return 1; @@ -245,12 +277,20 @@ static inline int post_kprobe_handler(struct pt_regs *regs) if (!kprobe_running()) return 0; - if (current_kprobe->post_handler) + if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) { + kprobe_status = KPROBE_HIT_SSDONE; current_kprobe->post_handler(current_kprobe, regs, 0); + } resume_execution(current_kprobe, regs); + /*Restore back the original saved kprobes variables and continue. */ + if (kprobe_status == KPROBE_REENTER) { + restore_previous_kprobe(); + goto out; + } unlock_kprobes(); +out: preempt_enable_no_resched(); return 1; @@ -392,3 +432,4 @@ int longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) } return 0; } + diff --git a/arch/sparc64/kernel/signal32.c b/arch/sparc64/kernel/signal32.c index 9a375e975cf..f28428f4170 100644 --- a/arch/sparc64/kernel/signal32.c +++ b/arch/sparc64/kernel/signal32.c @@ -102,7 +102,7 @@ typedef struct compat_siginfo{ /* POSIX.1b timers */ struct { - timer_t _tid; /* timer id */ + compat_timer_t _tid; /* timer id */ int _overrun; /* overrun count */ compat_sigval_t _sigval; /* same as below */ int _sys_private; /* not to be passed to user */ diff --git a/arch/um/Kconfig b/arch/um/Kconfig index b8e952c88fd..9469e77303e 100644 --- a/arch/um/Kconfig +++ b/arch/um/Kconfig @@ -74,6 +74,7 @@ config MODE_SKAS option will shrink the UML binary slightly. source "arch/um/Kconfig_arch" +source "mm/Kconfig" config LD_SCRIPT_STATIC bool diff --git a/arch/um/drivers/daemon_user.c b/arch/um/drivers/daemon_user.c index cf15b4a8b51..c1b03f7c1da 100644 --- a/arch/um/drivers/daemon_user.c +++ b/arch/um/drivers/daemon_user.c @@ -157,9 +157,9 @@ static void daemon_remove(void *data) os_close_file(pri->fd); os_close_file(pri->control); - if(pri->data_addr != NULL) kfree(pri->data_addr); - if(pri->ctl_addr != NULL) kfree(pri->ctl_addr); - if(pri->local_addr != NULL) kfree(pri->local_addr); + kfree(pri->data_addr); + kfree(pri->ctl_addr); + kfree(pri->local_addr); } int daemon_user_write(int fd, void *buf, int len, struct daemon_data *pri) diff --git a/arch/um/drivers/line.c b/arch/um/drivers/line.c index 0f59736db32..2bb4c4f5dec 100644 --- a/arch/um/drivers/line.c +++ b/arch/um/drivers/line.c @@ -602,11 +602,26 @@ int line_get_config(char *name, struct line *lines, unsigned int num, char *str, return n; } -int line_remove(struct line *lines, unsigned int num, char *str) +int line_id(char **str, int *start_out, int *end_out) +{ + char *end; + int n; + + n = simple_strtoul(*str, &end, 0); + if((*end != '\0') || (end == *str)) + return -1; + + *str = end; + *start_out = n; + *end_out = n; + return n; +} + +int line_remove(struct line *lines, unsigned int num, int n) { char config[sizeof("conxxxx=none\0")]; - sprintf(config, "%s=none", str); + sprintf(config, "%d=none", n); return !line_setup(lines, num, config, 0); } diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_kern.c index d7c7adcc0a6..404de41a4f6 100644 --- a/arch/um/drivers/mconsole_kern.c +++ b/arch/um/drivers/mconsole_kern.c @@ -419,8 +419,9 @@ void mconsole_config(struct mc_request *req) void mconsole_remove(struct mc_request *req) { struct mc_device *dev; - char *ptr = req->request.data; - int err; + char *ptr = req->request.data, *err_msg = ""; + char error[256]; + int err, start, end, n; ptr += strlen("remove"); while(isspace(*ptr)) ptr++; @@ -429,8 +430,35 @@ void mconsole_remove(struct mc_request *req) mconsole_reply(req, "Bad remove option", 1, 0); return; } - err = (*dev->remove)(&ptr[strlen(dev->name)]); - mconsole_reply(req, "", err, 0); + + ptr = &ptr[strlen(dev->name)]; + + err = 1; + n = (*dev->id)(&ptr, &start, &end); + if(n < 0){ + err_msg = "Couldn't parse device number"; + goto out; + } + else if((n < start) || (n > end)){ + sprintf(error, "Invalid device number - must be between " + "%d and %d", start, end); + err_msg = error; + goto out; + } + + err = (*dev->remove)(n); + switch(err){ + case -ENODEV: + err_msg = "Device doesn't exist"; + break; + case -EBUSY: + err_msg = "Device is currently open"; + break; + default: + break; + } + out: + mconsole_reply(req, err_msg, err, 0); } #ifdef CONFIG_MAGIC_SYSRQ diff --git a/arch/um/drivers/net_kern.c b/arch/um/drivers/net_kern.c index 5388a742869..1495007bf6c 100644 --- a/arch/um/drivers/net_kern.c +++ b/arch/um/drivers/net_kern.c @@ -612,25 +612,35 @@ static int net_config(char *str) return(err); } -static int net_remove(char *str) +static int net_id(char **str, int *start_out, int *end_out) +{ + char *end; + int n; + + n = simple_strtoul(*str, &end, 0); + if((*end != '\0') || (end == *str)) + return -1; + + *start_out = n; + *end_out = n; + *str = end; + return n; +} + +static int net_remove(int n) { struct uml_net *device; struct net_device *dev; struct uml_net_private *lp; - char *end; - int n; - - n = simple_strtoul(str, &end, 0); - if((*end != '\0') || (end == str)) - return(-1); device = find_device(n); if(device == NULL) - return(0); + return -ENODEV; dev = device->dev; lp = dev->priv; - if(lp->fd > 0) return(-1); + if(lp->fd > 0) + return -EBUSY; if(lp->remove != NULL) (*lp->remove)(&lp->user); unregister_netdev(dev); platform_device_unregister(&device->pdev); @@ -638,13 +648,14 @@ static int net_remove(char *str) list_del(&device->list); kfree(device); free_netdev(dev); - return(0); + return 0; } static struct mc_device net_mc = { .name = "eth", .config = net_config, .get_config = NULL, + .id = net_id, .remove = net_remove, }; diff --git a/arch/um/drivers/ssl.c b/arch/um/drivers/ssl.c index b32a77010fb..62e04ecfada 100644 --- a/arch/um/drivers/ssl.c +++ b/arch/um/drivers/ssl.c @@ -49,7 +49,7 @@ static struct chan_opts opts = { static int ssl_config(char *str); static int ssl_get_config(char *dev, char *str, int size, char **error_out); -static int ssl_remove(char *str); +static int ssl_remove(int n); static struct line_driver driver = { .name = "UML serial line", @@ -69,6 +69,7 @@ static struct line_driver driver = { .name = "ssl", .config = ssl_config, .get_config = ssl_get_config, + .id = line_id, .remove = ssl_remove, }, }; @@ -94,10 +95,10 @@ static int ssl_get_config(char *dev, char *str, int size, char **error_out) str, size, error_out)); } -static int ssl_remove(char *str) +static int ssl_remove(int n) { - return(line_remove(serial_lines, - sizeof(serial_lines)/sizeof(serial_lines[0]), str)); + return line_remove(serial_lines, + sizeof(serial_lines)/sizeof(serial_lines[0]), n); } int ssl_open(struct tty_struct *tty, struct file *filp) diff --git a/arch/um/drivers/stdio_console.c b/arch/um/drivers/stdio_console.c index afbe1e71ed8..005aa6333b6 100644 --- a/arch/um/drivers/stdio_console.c +++ b/arch/um/drivers/stdio_console.c @@ -55,7 +55,7 @@ static struct chan_opts opts = { static int con_config(char *str); static int con_get_config(char *dev, char *str, int size, char **error_out); -static int con_remove(char *str); +static int con_remove(int n); static struct line_driver driver = { .name = "UML console", @@ -75,6 +75,7 @@ static struct line_driver driver = { .name = "con", .config = con_config, .get_config = con_get_config, + .id = line_id, .remove = con_remove, }, }; @@ -99,9 +100,9 @@ static int con_get_config(char *dev, char *str, int size, char **error_out) size, error_out)); } -static int con_remove(char *str) +static int con_remove(int n) { - return(line_remove(vts, sizeof(vts)/sizeof(vts[0]), str)); + return line_remove(vts, sizeof(vts)/sizeof(vts[0]), n); } static int con_open(struct tty_struct *tty, struct file *filp) diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c index 2a7f6892c55..344b24d09a7 100644 --- a/arch/um/drivers/ubd_kern.c +++ b/arch/um/drivers/ubd_kern.c @@ -754,24 +754,34 @@ static int ubd_get_config(char *name, char *str, int size, char **error_out) return(len); } -static int ubd_remove(char *str) +static int ubd_id(char **str, int *start_out, int *end_out) +{ + int n; + + n = parse_unit(str); + *start_out = 0; + *end_out = MAX_DEV - 1; + return n; +} + +static int ubd_remove(int n) { struct ubd *dev; - int n, err = -ENODEV; + int err = -ENODEV; - n = parse_unit(&str); + spin_lock(&ubd_lock); - if((n < 0) || (n >= MAX_DEV)) - return(err); + if(ubd_gendisk[n] == NULL) + goto out; dev = &ubd_dev[n]; - if(dev->count > 0) - return(-EBUSY); /* you cannot remove a open disk */ - err = 0; - spin_lock(&ubd_lock); + if(dev->file == NULL) + goto out; - if(ubd_gendisk[n] == NULL) + /* you cannot remove a open disk */ + err = -EBUSY; + if(dev->count > 0) goto out; del_gendisk(ubd_gendisk[n]); @@ -787,15 +797,16 @@ static int ubd_remove(char *str) platform_device_unregister(&dev->pdev); *dev = ((struct ubd) DEFAULT_UBD); err = 0; - out: - spin_unlock(&ubd_lock); - return(err); +out: + spin_unlock(&ubd_lock); + return err; } static struct mc_device ubd_mc = { .name = "ubd", .config = ubd_config, .get_config = ubd_get_config, + .id = ubd_id, .remove = ubd_remove, }; diff --git a/arch/um/include/line.h b/arch/um/include/line.h index 4c5e92c04cc..5323d22a6ca 100644 --- a/arch/um/include/line.h +++ b/arch/um/include/line.h @@ -101,7 +101,8 @@ extern void lines_init(struct line *lines, int nlines); extern void close_lines(struct line *lines, int nlines); extern int line_config(struct line *lines, unsigned int sizeof_lines, char *str); -extern int line_remove(struct line *lines, unsigned int sizeof_lines, char *str); +extern int line_id(char **str, int *start_out, int *end_out); +extern int line_remove(struct line *lines, unsigned int sizeof_lines, int n); extern int line_get_config(char *dev, struct line *lines, unsigned int sizeof_lines, char *str, int size, char **error_out); diff --git a/arch/um/include/mconsole_kern.h b/arch/um/include/mconsole_kern.h index 61c274fcee5..d86ee14260c 100644 --- a/arch/um/include/mconsole_kern.h +++ b/arch/um/include/mconsole_kern.h @@ -20,7 +20,8 @@ struct mc_device { char *name; int (*config)(char *); int (*get_config)(char *, char *, int, char **); - int (*remove)(char *); + int (*id)(char **, int *, int *); + int (*remove)(int); }; #define CONFIG_CHUNK(str, size, current, chunk, end) \ diff --git a/arch/um/include/time_user.h b/arch/um/include/time_user.h index 6793a2fcd0a..f64ef77019a 100644 --- a/arch/um/include/time_user.h +++ b/arch/um/include/time_user.h @@ -8,11 +8,11 @@ extern void timer(void); extern void switch_timers(int to_real); -extern void set_interval(int timer_type); extern void idle_sleep(int secs); extern void enable_timer(void); extern void disable_timer(void); extern unsigned long time_lock(void); extern void time_unlock(unsigned long); +extern void user_time_init(void); #endif diff --git a/arch/um/kernel/main.c b/arch/um/kernel/main.c index e59f5815267..1e1a87f1c51 100644 --- a/arch/um/kernel/main.c +++ b/arch/um/kernel/main.c @@ -69,7 +69,6 @@ static __init void do_uml_initcalls(void) static void last_ditch_exit(int sig) { - kmalloc_ok = 0; signal(SIGINT, SIG_DFL); signal(SIGTERM, SIG_DFL); signal(SIGHUP, SIG_DFL); diff --git a/arch/um/kernel/process_kern.c b/arch/um/kernel/process_kern.c index 804c6bbdf67..d4036ed680b 100644 --- a/arch/um/kernel/process_kern.c +++ b/arch/um/kernel/process_kern.c @@ -8,6 +8,7 @@ #include "linux/kernel.h" #include "linux/sched.h" #include "linux/interrupt.h" +#include "linux/string.h" #include "linux/mm.h" #include "linux/slab.h" #include "linux/utsname.h" @@ -95,8 +96,8 @@ int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags) current->thread.request.u.thread.proc = fn; current->thread.request.u.thread.arg = arg; - pid = do_fork(CLONE_VM | CLONE_UNTRACED | flags, 0, NULL, 0, NULL, - NULL); + pid = do_fork(CLONE_VM | CLONE_UNTRACED | flags, 0, + ¤t->thread.regs, 0, NULL, NULL); if(pid < 0) panic("do_fork failed in kernel_thread, errno = %d", pid); return(pid); @@ -168,7 +169,7 @@ int current_pid(void) void default_idle(void) { - uml_idle_timer(); + CHOOSE_MODE(uml_idle_timer(), (void) 0); atomic_inc(&init_mm.mm_count); current->mm = &init_mm; @@ -322,12 +323,7 @@ void do_uml_exitcalls(void) char *uml_strdup(char *string) { - char *new; - - new = kmalloc(strlen(string) + 1, GFP_KERNEL); - if(new == NULL) return(NULL); - strcpy(new, string); - return(new); + return kstrdup(string, GFP_KERNEL); } int copy_to_user_proc(void __user *to, void *from, int size) diff --git a/arch/um/kernel/reboot.c b/arch/um/kernel/reboot.c index 207f89d7490..fcec51da1d3 100644 --- a/arch/um/kernel/reboot.c +++ b/arch/um/kernel/reboot.c @@ -38,14 +38,14 @@ static void kill_off_processes(void) void uml_cleanup(void) { - kill_off_processes(); + kmalloc_ok = 0; do_uml_exitcalls(); + kill_off_processes(); } void machine_restart(char * __unused) { - do_uml_exitcalls(); - kill_off_processes(); + uml_cleanup(); CHOOSE_MODE(reboot_tt(), reboot_skas()); } @@ -53,8 +53,7 @@ EXPORT_SYMBOL(machine_restart); void machine_power_off(void) { - do_uml_exitcalls(); - kill_off_processes(); + uml_cleanup(); CHOOSE_MODE(halt_tt(), halt_skas()); } diff --git a/arch/um/kernel/skas/Makefile b/arch/um/kernel/skas/Makefile index d37d1bfcd6f..ff69c4b312c 100644 --- a/arch/um/kernel/skas/Makefile +++ b/arch/um/kernel/skas/Makefile @@ -4,10 +4,10 @@ # obj-y := exec_kern.o mem.o mem_user.o mmu.o process.o process_kern.o \ - syscall_kern.o syscall_user.o time.o tlb.o trap_user.o uaccess.o \ + syscall_kern.o syscall_user.o tlb.o trap_user.o uaccess.o \ subdir- := util -USER_OBJS := process.o time.o +USER_OBJS := process.o include arch/um/scripts/Makefile.rules diff --git a/arch/um/kernel/skas/include/mode-skas.h b/arch/um/kernel/skas/include/mode-skas.h index c1e33bd788d..bcd26a6a388 100644 --- a/arch/um/kernel/skas/include/mode-skas.h +++ b/arch/um/kernel/skas/include/mode-skas.h @@ -13,7 +13,6 @@ extern unsigned long exec_fp_regs[]; extern unsigned long exec_fpx_regs[]; extern int have_fpx_regs; -extern void user_time_init_skas(void); extern void sig_handler_common_skas(int sig, void *sc_ptr); extern void halt_skas(void); extern void reboot_skas(void); diff --git a/arch/um/kernel/skas/process_kern.c b/arch/um/kernel/skas/process_kern.c index fc71ef29578..0a7b8aa55db 100644 --- a/arch/um/kernel/skas/process_kern.c +++ b/arch/um/kernel/skas/process_kern.c @@ -111,8 +111,7 @@ int copy_thread_skas(int nr, unsigned long clone_flags, unsigned long sp, void (*handler)(int); if(current->thread.forking){ - memcpy(&p->thread.regs.regs.skas, - ¤t->thread.regs.regs.skas, + memcpy(&p->thread.regs.regs.skas, ®s->regs.skas, sizeof(p->thread.regs.regs.skas)); REGS_SET_SYSCALL_RETURN(p->thread.regs.regs.skas.regs, 0); if(sp != 0) REGS_SP(p->thread.regs.regs.skas.regs) = sp; @@ -181,7 +180,6 @@ int start_uml_skas(void) start_userspace(0); init_new_thread_signals(1); - uml_idle_timer(); init_task.thread.request.u.thread.proc = start_kernel_proc; init_task.thread.request.u.thread.arg = NULL; @@ -201,14 +199,3 @@ int thread_pid_skas(struct task_struct *task) #warning Need to look up userspace_pid by cpu return(userspace_pid[0]); } - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/kernel/skas/time.c b/arch/um/kernel/skas/time.c deleted file mode 100644 index 98091494b89..00000000000 --- a/arch/um/kernel/skas/time.c +++ /dev/null @@ -1,30 +0,0 @@ -/* - * Copyright (C) 2002 Jeff Dike (jdike@karaya.com) - * Licensed under the GPL - */ - -#include <sys/signal.h> -#include <sys/time.h> -#include "time_user.h" -#include "process.h" -#include "user.h" - -void user_time_init_skas(void) -{ - if(signal(SIGALRM, (__sighandler_t) alarm_handler) == SIG_ERR) - panic("Couldn't set SIGALRM handler"); - if(signal(SIGVTALRM, (__sighandler_t) alarm_handler) == SIG_ERR) - panic("Couldn't set SIGVTALRM handler"); - set_interval(ITIMER_VIRTUAL); -} - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/kernel/syscall_kern.c b/arch/um/kernel/syscall_kern.c index b7a55251e89..8e1a3501ff4 100644 --- a/arch/um/kernel/syscall_kern.c +++ b/arch/um/kernel/syscall_kern.c @@ -31,7 +31,8 @@ long sys_fork(void) long ret; current->thread.forking = 1; - ret = do_fork(SIGCHLD, 0, NULL, 0, NULL, NULL); + ret = do_fork(SIGCHLD, UPT_SP(¤t->thread.regs.regs), + ¤t->thread.regs, 0, NULL, NULL); current->thread.forking = 0; return(ret); } @@ -41,8 +42,9 @@ long sys_vfork(void) long ret; current->thread.forking = 1; - ret = do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, 0, NULL, 0, NULL, - NULL); + ret = do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, + UPT_SP(¤t->thread.regs.regs), + ¤t->thread.regs, 0, NULL, NULL); current->thread.forking = 0; return(ret); } @@ -162,14 +164,3 @@ int next_syscall_index(int limit) spin_unlock(&syscall_lock); return(ret); } - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/kernel/time.c b/arch/um/kernel/time.c index c40c86a3f91..f829b309b63 100644 --- a/arch/um/kernel/time.c +++ b/arch/um/kernel/time.c @@ -33,7 +33,7 @@ void timer(void) timeradd(&xtime, &local_offset, &xtime); } -void set_interval(int timer_type) +static void set_interval(int timer_type) { int usec = 1000000/hz(); struct itimerval interval = ((struct itimerval) { { 0, usec }, @@ -45,12 +45,7 @@ void set_interval(int timer_type) void enable_timer(void) { - int usec = 1000000/hz(); - struct itimerval enable = ((struct itimerval) { { 0, usec }, - { 0, usec }}); - if(setitimer(ITIMER_VIRTUAL, &enable, NULL)) - printk("enable_timer - setitimer failed, errno = %d\n", - errno); + set_interval(ITIMER_VIRTUAL); } void disable_timer(void) @@ -155,13 +150,15 @@ void idle_sleep(int secs) nanosleep(&ts, NULL); } -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ +/* XXX This partly duplicates init_irq_signals */ + +void user_time_init(void) +{ + set_handler(SIGVTALRM, (__sighandler_t) alarm_handler, + SA_ONSTACK | SA_RESTART, SIGUSR1, SIGIO, SIGWINCH, + SIGALRM, SIGUSR2, -1); + set_handler(SIGALRM, (__sighandler_t) alarm_handler, + SA_ONSTACK | SA_RESTART, SIGUSR1, SIGIO, SIGWINCH, + SIGVTALRM, SIGUSR2, -1); + set_interval(ITIMER_VIRTUAL); +} diff --git a/arch/um/kernel/time_kern.c b/arch/um/kernel/time_kern.c index 6516fc52afe..a8b4ef601f5 100644 --- a/arch/um/kernel/time_kern.c +++ b/arch/um/kernel/time_kern.c @@ -162,7 +162,7 @@ int __init timer_init(void) { int err; - CHOOSE_MODE(user_time_init_tt(), user_time_init_skas()); + user_time_init(); err = request_irq(TIMER_IRQ, um_timer, SA_INTERRUPT, "timer", NULL); if(err != 0) printk(KERN_ERR "timer_init : request_irq failed - " diff --git a/arch/um/kernel/tt/Makefile b/arch/um/kernel/tt/Makefile index 3fd2554e60b..6939e5af847 100644 --- a/arch/um/kernel/tt/Makefile +++ b/arch/um/kernel/tt/Makefile @@ -4,11 +4,11 @@ # obj-y = exec_kern.o exec_user.o gdb.o ksyms.o mem.o mem_user.o process_kern.o \ - syscall_kern.o syscall_user.o time.o tlb.o tracer.o trap_user.o \ + syscall_kern.o syscall_user.o tlb.o tracer.o trap_user.o \ uaccess.o uaccess_user.o obj-$(CONFIG_PT_PROXY) += gdb_kern.o ptproxy/ -USER_OBJS := gdb.o time.o tracer.o +USER_OBJS := gdb.o tracer.o include arch/um/scripts/Makefile.rules diff --git a/arch/um/kernel/tt/gdb.c b/arch/um/kernel/tt/gdb.c index 19a0ad7b35b..37e22d71a0d 100644 --- a/arch/um/kernel/tt/gdb.c +++ b/arch/um/kernel/tt/gdb.c @@ -153,10 +153,10 @@ void remove_gdb_cb(void *unused) exit_debugger_cb(NULL); } -int gdb_remove(char *unused) +int gdb_remove(int unused) { initial_thread_cb(remove_gdb_cb, NULL); - return(0); + return 0; } void signal_usr1(int sig) diff --git a/arch/um/kernel/tt/gdb_kern.c b/arch/um/kernel/tt/gdb_kern.c index 93fb121f86a..26506388a6a 100644 --- a/arch/um/kernel/tt/gdb_kern.c +++ b/arch/um/kernel/tt/gdb_kern.c @@ -10,7 +10,7 @@ #ifdef CONFIG_MCONSOLE extern int gdb_config(char *str); -extern int gdb_remove(char *unused); +extern int gdb_remove(int n); static struct mc_device gdb_mc = { .name = "gdb", diff --git a/arch/um/kernel/tt/include/debug.h b/arch/um/kernel/tt/include/debug.h index 8eff674107c..738435461e1 100644 --- a/arch/um/kernel/tt/include/debug.h +++ b/arch/um/kernel/tt/include/debug.h @@ -4,8 +4,8 @@ * Licensed under the GPL */ -#ifndef __DEBUG_H -#define __DEBUG_H +#ifndef __UML_TT_DEBUG_H +#define __UML_TT_DEBUG_H extern int debugger_proxy(int status, pid_t pid); extern void child_proxy(pid_t pid, int status); @@ -13,17 +13,6 @@ extern void init_proxy (pid_t pid, int waiting, int status); extern int start_debugger(char *prog, int startup, int stop, int *debugger_fd); extern void fake_child_exit(void); extern int gdb_config(char *str); -extern int gdb_remove(char *unused); +extern int gdb_remove(int unused); #endif - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/kernel/tt/include/mode-tt.h b/arch/um/kernel/tt/include/mode-tt.h index efe46201906..e171e15fead 100644 --- a/arch/um/kernel/tt/include/mode-tt.h +++ b/arch/um/kernel/tt/include/mode-tt.h @@ -13,7 +13,6 @@ enum { OP_NONE, OP_EXEC, OP_FORK, OP_TRACE_ON, OP_REBOOT, OP_HALT, OP_CB }; extern int tracing_pid; extern int tracer(int (*init_proc)(void *), void *sp); -extern void user_time_init_tt(void); extern void sig_handler_common_tt(int sig, void *sc); extern void syscall_handler_tt(int sig, union uml_pt_regs *regs); extern void reboot_tt(void); diff --git a/arch/um/kernel/tt/process_kern.c b/arch/um/kernel/tt/process_kern.c index 776310fd5b8..a189a2b9293 100644 --- a/arch/um/kernel/tt/process_kern.c +++ b/arch/um/kernel/tt/process_kern.c @@ -266,10 +266,10 @@ int copy_thread_tt(int nr, unsigned long clone_flags, unsigned long sp, } if(current->thread.forking){ - sc_to_sc(UPT_SC(&p->thread.regs.regs), - UPT_SC(¤t->thread.regs.regs)); + sc_to_sc(UPT_SC(&p->thread.regs.regs), UPT_SC(®s->regs)); SC_SET_SYSCALL_RETURN(UPT_SC(&p->thread.regs.regs), 0); - if(sp != 0) SC_SP(UPT_SC(&p->thread.regs.regs)) = sp; + if(sp != 0) + SC_SP(UPT_SC(&p->thread.regs.regs)) = sp; } p->thread.mode.tt.extern_pid = new_pid; @@ -459,14 +459,3 @@ int is_valid_pid(int pid) read_unlock(&tasklist_lock); return(0); } - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/kernel/tt/time.c b/arch/um/kernel/tt/time.c deleted file mode 100644 index 8565b71b07c..00000000000 --- a/arch/um/kernel/tt/time.c +++ /dev/null @@ -1,28 +0,0 @@ -/* - * Copyright (C) 2000, 2001, 2002 Jeff Dike (jdike@karaya.com) - * Licensed under the GPL - */ - -#include <signal.h> -#include <sys/time.h> -#include <time_user.h> -#include "process.h" -#include "user.h" - -void user_time_init_tt(void) -{ - if(signal(SIGVTALRM, (__sighandler_t) alarm_handler) == SIG_ERR) - panic("Couldn't set SIGVTALRM handler"); - set_interval(ITIMER_VIRTUAL); -} - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/sys-i386/signal.c b/arch/um/sys-i386/signal.c index 03913ca5d25..4efc69a039d 100644 --- a/arch/um/sys-i386/signal.c +++ b/arch/um/sys-i386/signal.c @@ -312,7 +312,7 @@ long sys_sigreturn(struct pt_regs regs) unsigned long __user *extramask = frame->extramask; int sig_size = (_NSIG_WORDS - 1) * sizeof(unsigned long); - if(copy_from_user(&set.sig[0], oldmask, sizeof(&set.sig[0])) || + if(copy_from_user(&set.sig[0], oldmask, sizeof(set.sig[0])) || copy_from_user(&set.sig[1], extramask, sig_size)) goto segfault; diff --git a/arch/um/sys-i386/syscalls.c b/arch/um/sys-i386/syscalls.c index 335e2d89504..83e9be820a8 100644 --- a/arch/um/sys-i386/syscalls.c +++ b/arch/um/sys-i386/syscalls.c @@ -69,15 +69,11 @@ long sys_clone(unsigned long clone_flags, unsigned long newsp, { long ret; - /* XXX: normal arch do here this pass, and also pass the regs to - * do_fork, instead of NULL. Currently the arch-independent code - * ignores these values, while the UML code (actually it's - * copy_thread) does the right thing. But this should change, - probably. */ - /*if (!newsp) - newsp = UPT_SP(current->thread.regs);*/ + if (!newsp) + newsp = UPT_SP(¤t->thread.regs.regs); current->thread.forking = 1; - ret = do_fork(clone_flags, newsp, NULL, 0, parent_tid, child_tid); + ret = do_fork(clone_flags, newsp, ¤t->thread.regs, 0, parent_tid, + child_tid); current->thread.forking = 0; return(ret); } @@ -197,14 +193,3 @@ long sys_sigaction(int sig, const struct old_sigaction __user *act, return ret; } - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/sys-x86_64/syscalls.c b/arch/um/sys-x86_64/syscalls.c index 6f44f40204e..3259a4db453 100644 --- a/arch/um/sys-x86_64/syscalls.c +++ b/arch/um/sys-x86_64/syscalls.c @@ -174,26 +174,11 @@ long sys_clone(unsigned long clone_flags, unsigned long newsp, { long ret; - /* XXX: normal arch do here this pass, and also pass the regs to - * do_fork, instead of NULL. Currently the arch-independent code - * ignores these values, while the UML code (actually it's - * copy_thread) does the right thing. But this should change, - probably. */ - /*if (!newsp) - newsp = UPT_SP(current->thread.regs);*/ + if (!newsp) + newsp = UPT_SP(¤t->thread.regs.regs); current->thread.forking = 1; - ret = do_fork(clone_flags, newsp, NULL, 0, parent_tid, child_tid); + ret = do_fork(clone_flags, newsp, ¤t->thread.regs, 0, parent_tid, + child_tid); current->thread.forking = 0; return(ret); } - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/v850/Kconfig b/arch/v850/Kconfig index 90cd4baa75e..27febd6ffa8 100644 --- a/arch/v850/Kconfig +++ b/arch/v850/Kconfig @@ -218,6 +218,8 @@ menu "Processor type and features" a lot of RAM, and you need to able to allocate very large contiguous chunks. If unsure, say N. +source "mm/Kconfig" + endmenu diff --git a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig index 289f448ac89..d09437b5c48 100644 --- a/arch/x86_64/Kconfig +++ b/arch/x86_64/Kconfig @@ -207,33 +207,6 @@ config SMP If you don't know what to do here, say N. -config PREEMPT - bool "Preemptible Kernel" - ---help--- - This option reduces the latency of the kernel when reacting to - real-time or interactive events by allowing a low priority process to - be preempted even if it is in kernel mode executing a system call. - This allows applications to run more reliably even when the system is - under load. On contrary it may also break your drivers and add - priority inheritance problems to your system. Don't select it if - you rely on a stable system or have slightly obscure hardware. - It's also not very well tested on x86-64 currently. - You have been warned. - - Say Y here if you are feeling brave and building a kernel for a - desktop, embedded or real-time system. Say N if you are unsure. - -config PREEMPT_BKL - bool "Preempt The Big Kernel Lock" - depends on PREEMPT - default y - help - This option reduces the latency of the kernel by making the - big kernel lock preemptible. - - Say Y here if you are building a kernel for a desktop system. - Say N if you are unsure. - config SCHED_SMT bool "SMT (Hyperthreading) scheduler support" depends on SMP @@ -244,6 +217,8 @@ config SCHED_SMT cost of slightly increased overhead in some places. If unsure say N here. +source "kernel/Kconfig.preempt" + config K8_NUMA bool "K8 NUMA support" select NUMA @@ -265,7 +240,7 @@ config NUMA_EMU into virtual nodes when booted with "numa=fake=N", where N is the number of nodes. This is only useful for debugging. -config DISCONTIGMEM +config ARCH_DISCONTIGMEM_ENABLE bool depends on NUMA default y @@ -274,6 +249,27 @@ config NUMA bool default n +config ARCH_DISCONTIGMEM_ENABLE + def_bool y + depends on NUMA + +config ARCH_DISCONTIGMEM_DEFAULT + def_bool y + depends on NUMA + +config ARCH_SPARSEMEM_ENABLE + def_bool y + depends on NUMA + +config ARCH_FLATMEM_ENABLE + def_bool y + depends on !NUMA + +source "mm/Kconfig" + +config HAVE_ARCH_EARLY_PFN_TO_NID + def_bool y + config HAVE_DEC_LOCK bool depends on SMP @@ -292,6 +288,15 @@ config NR_CPUS This is purely to save memory - each supported CPU requires memory in the static kernel configuration. +config HOTPLUG_CPU + bool "Support for hot-pluggable CPUs (EXPERIMENTAL)" + depends on SMP && HOTPLUG && EXPERIMENTAL + help + Say Y here to experiment with turning CPUs off and on. CPUs + can be controlled through /sys/devices/system/cpu/cpu#. + Say N if you want to disable CPU hotplug. + + config HPET_TIMER bool default y @@ -364,6 +369,34 @@ config X86_MCE_INTEL Additional support for intel specific MCE features such as the thermal monitor. +config PHYSICAL_START + hex "Physical address where the kernel is loaded" if EMBEDDED + default "0x100000" + help + This gives the physical address where the kernel is loaded. + Primarily used in the case of kexec on panic where the + fail safe kernel needs to run at a different address than + the panic-ed kernel. + + Don't change this unless you know what you are doing. + +config KEXEC + bool "kexec system call (EXPERIMENTAL)" + depends on EXPERIMENTAL + help + kexec is a system call that implements the ability to shutdown your + current kernel, and to start another kernel. It is like a reboot + but it is indepedent of the system firmware. And like a reboot + you can start any kernel with it, not just Linux. + + The name comes from the similiarity to the exec system call. + + It is an ongoing process to be certain the hardware in a machine + is properly shutdown, so do not be surprised if this code does not + initially work for you. It may help to enable device hotplugging + support. As of this writing the exact hardware interface is + strongly in flux, so no good recommendation can be made. + config SECCOMP bool "Enable seccomp to safely compute untrusted bytecode" depends on PROC_FS @@ -381,6 +414,8 @@ config SECCOMP If unsure, say Y. Only embedded should say N here. +source kernel/Kconfig.hz + endmenu # diff --git a/arch/x86_64/Makefile b/arch/x86_64/Makefile index 6f90c246c41..8a73794f9b9 100644 --- a/arch/x86_64/Makefile +++ b/arch/x86_64/Makefile @@ -35,7 +35,7 @@ export IA32_CC IA32_LD IA32_AS IA32_OBJCOPY IA32_CPP LDFLAGS := -m elf_x86_64 OBJCOPYFLAGS := -O binary -R .note -R .comment -S -LDFLAGS_vmlinux := -e stext +LDFLAGS_vmlinux := CHECKFLAGS += -D__x86_64__ -m64 diff --git a/arch/x86_64/boot/compressed/head.S b/arch/x86_64/boot/compressed/head.S index 27264dbd575..6f55565e4d4 100644 --- a/arch/x86_64/boot/compressed/head.S +++ b/arch/x86_64/boot/compressed/head.S @@ -2,8 +2,6 @@ * linux/boot/head.S * * Copyright (C) 1991, 1992, 1993 Linus Torvalds - * - * $Id: head.S,v 1.3 2001/04/20 00:59:28 ak Exp $ */ /* @@ -21,13 +19,14 @@ */ /* - * High loaded stuff by Hans Lermen & Werner Almesberger, Feb. 1996 + * High loaded stuff by Hans Lermen & Werner Almesberger, Feb. 1996 */ .code32 .text #include <linux/linkage.h> #include <asm/segment.h> +#include <asm/page.h> .code32 .globl startup_32 @@ -77,7 +76,7 @@ startup_32: jnz 3f addl $8,%esp xorl %ebx,%ebx - ljmp $(__KERNEL_CS), $0x100000 + ljmp $(__KERNEL_CS), $__PHYSICAL_START /* * We come here, if we were loaded high. @@ -103,7 +102,7 @@ startup_32: popl %ecx # lcount popl %edx # high_buffer_start popl %eax # hcount - movl $0x100000,%edi + movl $__PHYSICAL_START,%edi cli # make sure we don't get interrupted ljmp $(__KERNEL_CS), $0x1000 # and jump to the move routine @@ -128,7 +127,7 @@ move_routine_start: movsl movl %ebx,%esi # Restore setup pointer xorl %ebx,%ebx - ljmp $(__KERNEL_CS), $0x100000 + ljmp $(__KERNEL_CS), $__PHYSICAL_START move_routine_end: diff --git a/arch/x86_64/boot/compressed/misc.c b/arch/x86_64/boot/compressed/misc.c index c8b9216f9e6..b38d5b8b5fb 100644 --- a/arch/x86_64/boot/compressed/misc.c +++ b/arch/x86_64/boot/compressed/misc.c @@ -11,6 +11,7 @@ #include "miscsetup.h" #include <asm/io.h> +#include <asm/page.h> /* * gzip declarations @@ -92,8 +93,11 @@ static unsigned long output_ptr = 0; static void *malloc(int size); static void free(void *where); +void* memset(void* s, int c, unsigned n); +void* memcpy(void* dest, const void* src, unsigned n); + static void putstr(const char *); - + extern int end; static long free_mem_ptr = (long)&end; static long free_mem_end_ptr; @@ -284,7 +288,7 @@ void setup_normal_output_buffer(void) #else if ((ALT_MEM_K > EXT_MEM_K ? ALT_MEM_K : EXT_MEM_K) < 1024) error("Less than 2MB of memory"); #endif - output_data = (char *)0x100000; /* Points to 1M */ + output_data = (char *)__PHYSICAL_START; /* Normally Points to 1M */ free_mem_end_ptr = (long)real_mode; } @@ -307,8 +311,8 @@ void setup_output_buffer_if_we_run_high(struct moveparams *mv) low_buffer_size = low_buffer_end - LOW_BUFFER_START; high_loaded = 1; free_mem_end_ptr = (long)high_buffer_start; - if ( (0x100000 + low_buffer_size) > ((ulg)high_buffer_start)) { - high_buffer_start = (uch *)(0x100000 + low_buffer_size); + if ( (__PHYSICAL_START + low_buffer_size) > ((ulg)high_buffer_start)) { + high_buffer_start = (uch *)(__PHYSICAL_START + low_buffer_size); mv->hcount = 0; /* say: we need not to move high_buffer */ } else mv->hcount = -1; diff --git a/arch/x86_64/boot/install.sh b/arch/x86_64/boot/install.sh index 90f2452b3b9..198af15a775 100644 --- a/arch/x86_64/boot/install.sh +++ b/arch/x86_64/boot/install.sh @@ -1,6 +1,6 @@ #!/bin/sh # -# arch/i386/boot/install.sh +# arch/x86_64/boot/install.sh # # This file is subject to the terms and conditions of the GNU General Public # License. See the file "COPYING" in the main directory of this archive @@ -21,8 +21,8 @@ # User may have a custom install script -if [ -x ~/bin/installkernel ]; then exec ~/bin/installkernel "$@"; fi -if [ -x /sbin/installkernel ]; then exec /sbin/installkernel "$@"; fi +if [ -x ~/bin/${CROSS_COMPILE}installkernel ]; then exec ~/bin/${CROSS_COMPILE}installkernel "$@"; fi +if [ -x /sbin/${CROSS_COMPILE}installkernel ]; then exec /sbin/${CROSS_COMPILE}installkernel "$@"; fi # Default install - same as make zlilo diff --git a/arch/x86_64/boot/setup.S b/arch/x86_64/boot/setup.S index 75d4d2ad93b..ff58b2832b7 100644 --- a/arch/x86_64/boot/setup.S +++ b/arch/x86_64/boot/setup.S @@ -33,7 +33,7 @@ * Transcribed from Intel (as86) -> AT&T (gas) by Chris Noe, May 1999. * <stiker@northlink.com> * - * Fix to work around buggy BIOSes which dont use carry bit correctly + * Fix to work around buggy BIOSes which don't use carry bit correctly * and/or report extended memory in CX/DX for e801h memory size detection * call. As a result the kernel got wrong figures. The int15/e801h docs * from Ralf Brown interrupt list seem to indicate AX/BX should be used @@ -383,7 +383,7 @@ sse_ok: # a whole bunch of different types, and allows memory holes and # everything. We scan through this memory map and build a list # of the first 32 memory areas, which we return at [E820MAP]. -# This is documented at http://www.teleport.com/~acpi/acpihtml/topic245.htm +# This is documented at http://www.acpi.info/, in the ACPI 2.0 specification. #define SMAP 0x534d4150 @@ -436,7 +436,7 @@ bail820: meme801: stc # fix to work around buggy - xorw %cx,%cx # BIOSes which dont clear/set + xorw %cx,%cx # BIOSes which don't clear/set xorw %dx,%dx # carry on pass/error of # e801h memory size call # or merely pass cx,dx though @@ -733,7 +733,7 @@ flush_instr: # # but we yet haven't reloaded the CS register, so the default size # of the target offset still is 16 bit. -# However, using an operant prefix (0x66), the CPU will properly +# However, using an operand prefix (0x66), the CPU will properly # take our 48 bit far pointer. (INTeL 80386 Programmer's Reference # Manual, Mixing 16-bit and 32-bit code, page 16-6) diff --git a/arch/x86_64/boot/tools/build.c b/arch/x86_64/boot/tools/build.c index c2fa6631317..18b5bac1c42 100644 --- a/arch/x86_64/boot/tools/build.c +++ b/arch/x86_64/boot/tools/build.c @@ -1,6 +1,4 @@ /* - * $Id: build.c,v 1.3 2001/06/26 15:14:50 pavel Exp $ - * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright (C) 1997 Martin Mares */ @@ -8,7 +6,8 @@ /* * This file builds a disk-image from three different files: * - * - bootsect: exactly 512 bytes of 8086 machine code, loads the rest + * - bootsect: compatibility mbr which prints an error message if + * someone tries to boot the kernel directly. * - setup: 8086 machine code, sets up system parm * - system: 80386 code for actual system * diff --git a/arch/x86_64/ia32/ia32_signal.c b/arch/x86_64/ia32/ia32_signal.c index fbd09b5126c..66e2821533d 100644 --- a/arch/x86_64/ia32/ia32_signal.c +++ b/arch/x86_64/ia32/ia32_signal.c @@ -428,8 +428,8 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) return (void __user *)((rsp - frame_size) & -8UL); } -void ia32_setup_frame(int sig, struct k_sigaction *ka, - compat_sigset_t *set, struct pt_regs * regs) +int ia32_setup_frame(int sig, struct k_sigaction *ka, + compat_sigset_t *set, struct pt_regs * regs) { struct sigframe __user *frame; int err = 0; @@ -514,14 +514,15 @@ void ia32_setup_frame(int sig, struct k_sigaction *ka, current->comm, current->pid, frame, regs->rip, frame->pretcode); #endif - return; + return 1; give_sigsegv: force_sigsegv(sig, current); + return 0; } -void ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, - compat_sigset_t *set, struct pt_regs * regs) +int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, + compat_sigset_t *set, struct pt_regs * regs) { struct rt_sigframe __user *frame; int err = 0; @@ -613,9 +614,9 @@ void ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, current->comm, current->pid, frame, regs->rip, frame->pretcode); #endif - return; + return 1; give_sigsegv: force_sigsegv(sig, current); + return 0; } - diff --git a/arch/x86_64/ia32/ia32entry.S b/arch/x86_64/ia32/ia32entry.S index f3ca0db85b5..cc935427d53 100644 --- a/arch/x86_64/ia32/ia32entry.S +++ b/arch/x86_64/ia32/ia32entry.S @@ -589,7 +589,7 @@ ia32_sys_call_table: .quad compat_sys_mq_timedreceive /* 280 */ .quad compat_sys_mq_notify .quad compat_sys_mq_getsetattr - .quad quiet_ni_syscall /* reserved for kexec */ + .quad compat_sys_kexec_load /* reserved for kexec */ .quad compat_sys_waitid .quad quiet_ni_syscall /* sys_altroot */ .quad sys_add_key diff --git a/arch/x86_64/kernel/Makefile b/arch/x86_64/kernel/Makefile index 5ca4a4598fd..48f9e2c19cd 100644 --- a/arch/x86_64/kernel/Makefile +++ b/arch/x86_64/kernel/Makefile @@ -20,6 +20,7 @@ obj-$(CONFIG_SMP) += smp.o smpboot.o trampoline.o obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o obj-$(CONFIG_X86_IO_APIC) += io_apic.o mpparse.o \ genapic.o genapic_cluster.o genapic_flat.o +obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o obj-$(CONFIG_PM) += suspend.o obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend_asm.o obj-$(CONFIG_CPU_FREQ) += cpufreq/ diff --git a/arch/x86_64/kernel/acpi/wakeup.S b/arch/x86_64/kernel/acpi/wakeup.S index a4c630034cd..185faa911db 100644 --- a/arch/x86_64/kernel/acpi/wakeup.S +++ b/arch/x86_64/kernel/acpi/wakeup.S @@ -67,7 +67,7 @@ wakeup_code: shll $4, %eax addl $(gdta - wakeup_code), %eax movl %eax, gdt_48a +2 - wakeup_code - lgdt %ds:gdt_48a - wakeup_code # load gdt with whatever is + lgdtl %ds:gdt_48a - wakeup_code # load gdt with whatever is # appropriate movl $1, %eax # protected mode (PE) bit diff --git a/arch/x86_64/kernel/aperture.c b/arch/x86_64/kernel/aperture.c index 504e6347499..c9a6b812e92 100644 --- a/arch/x86_64/kernel/aperture.c +++ b/arch/x86_64/kernel/aperture.c @@ -40,11 +40,7 @@ int fix_aperture __initdata = 1; static u32 __init allocate_aperture(void) { -#ifdef CONFIG_DISCONTIGMEM pg_data_t *nd0 = NODE_DATA(0); -#else - pg_data_t *nd0 = &contig_page_data; -#endif u32 aper_size; void *p; diff --git a/arch/x86_64/kernel/apic.c b/arch/x86_64/kernel/apic.c index f8e6cc4fecd..375d369570c 100644 --- a/arch/x86_64/kernel/apic.c +++ b/arch/x86_64/kernel/apic.c @@ -133,7 +133,7 @@ void __init connect_bsp_APIC(void) } } -void disconnect_bsp_APIC(void) +void disconnect_bsp_APIC(int virt_wire_setup) { if (pic_mode) { /* @@ -146,6 +146,42 @@ void disconnect_bsp_APIC(void) outb(0x70, 0x22); outb(0x00, 0x23); } + else { + /* Go back to Virtual Wire compatibility mode */ + unsigned long value; + + /* For the spurious interrupt use vector F, and enable it */ + value = apic_read(APIC_SPIV); + value &= ~APIC_VECTOR_MASK; + value |= APIC_SPIV_APIC_ENABLED; + value |= 0xf; + apic_write_around(APIC_SPIV, value); + + if (!virt_wire_setup) { + /* For LVT0 make it edge triggered, active high, external and enabled */ + value = apic_read(APIC_LVT0); + value &= ~(APIC_MODE_MASK | APIC_SEND_PENDING | + APIC_INPUT_POLARITY | APIC_LVT_REMOTE_IRR | + APIC_LVT_LEVEL_TRIGGER | APIC_LVT_MASKED ); + value |= APIC_LVT_REMOTE_IRR | APIC_SEND_PENDING; + value = SET_APIC_DELIVERY_MODE(value, APIC_MODE_EXTINT); + apic_write_around(APIC_LVT0, value); + } + else { + /* Disable LVT0 */ + apic_write_around(APIC_LVT0, APIC_LVT_MASKED); + } + + /* For LVT1 make it edge triggered, active high, nmi and enabled */ + value = apic_read(APIC_LVT1); + value &= ~( + APIC_MODE_MASK | APIC_SEND_PENDING | + APIC_INPUT_POLARITY | APIC_LVT_REMOTE_IRR | + APIC_LVT_LEVEL_TRIGGER | APIC_LVT_MASKED); + value |= APIC_LVT_REMOTE_IRR | APIC_SEND_PENDING; + value = SET_APIC_DELIVERY_MODE(value, APIC_MODE_NMI); + apic_write_around(APIC_LVT1, value); + } } void disable_local_APIC(void) @@ -285,7 +321,7 @@ void __init init_bsp_APIC(void) apic_write_around(APIC_LVT1, value); } -void __init setup_local_APIC (void) +void __cpuinit setup_local_APIC (void) { unsigned int value, ver, maxlvt; @@ -534,7 +570,7 @@ static struct sys_device device_lapic = { .cls = &lapic_sysclass, }; -static void __init apic_pm_activate(void) +static void __cpuinit apic_pm_activate(void) { apic_pm_state.active = 1; } @@ -774,14 +810,14 @@ void __init setup_boot_APIC_clock (void) local_irq_enable(); } -void __init setup_secondary_APIC_clock(void) +void __cpuinit setup_secondary_APIC_clock(void) { local_irq_disable(); /* FIXME: Do we need this? --RR */ setup_APIC_timer(calibration_result); local_irq_enable(); } -void __init disable_APIC_timer(void) +void __cpuinit disable_APIC_timer(void) { if (using_apic_timer) { unsigned long v; diff --git a/arch/x86_64/kernel/crash.c b/arch/x86_64/kernel/crash.c new file mode 100644 index 00000000000..d7fa4248501 --- /dev/null +++ b/arch/x86_64/kernel/crash.c @@ -0,0 +1,35 @@ +/* + * Architecture specific (x86_64) functions for kexec based crash dumps. + * + * Created by: Hariprasad Nellitheertha (hari@in.ibm.com) + * + * Copyright (C) IBM Corporation, 2004. All rights reserved. + * + */ + +#include <linux/init.h> +#include <linux/types.h> +#include <linux/kernel.h> +#include <linux/smp.h> +#include <linux/irq.h> +#include <linux/reboot.h> +#include <linux/kexec.h> + +#include <asm/processor.h> +#include <asm/hardirq.h> +#include <asm/nmi.h> +#include <asm/hw_irq.h> + +note_buf_t crash_notes[NR_CPUS]; + +void machine_crash_shutdown(struct pt_regs *regs) +{ + /* This function is only called after the system + * has paniced or is otherwise in a critical state. + * The minimum amount of code to allow a kexec'd kernel + * to run successfully needs to happen here. + * + * In practice this means shooting down the other cpus in + * an SMP system. + */ +} diff --git a/arch/x86_64/kernel/e820.c b/arch/x86_64/kernel/e820.c index 7c154dfff64..6ded3a50dfe 100644 --- a/arch/x86_64/kernel/e820.c +++ b/arch/x86_64/kernel/e820.c @@ -16,6 +16,7 @@ #include <linux/bootmem.h> #include <linux/ioport.h> #include <linux/string.h> +#include <linux/kexec.h> #include <asm/page.h> #include <asm/e820.h> #include <asm/proto.h> @@ -191,8 +192,6 @@ void __init e820_reserve_resources(void) int i; for (i = 0; i < e820.nr_map; i++) { struct resource *res; - if (e820.map[i].addr + e820.map[i].size > 0x100000000ULL) - continue; res = alloc_bootmem_low(sizeof(struct resource)); switch (e820.map[i].type) { case E820_RAM: res->name = "System RAM"; break; @@ -212,6 +211,9 @@ void __init e820_reserve_resources(void) */ request_resource(res, &code_resource); request_resource(res, &data_resource); +#ifdef CONFIG_KEXEC + request_resource(res, &crashk_res); +#endif } } } diff --git a/arch/x86_64/kernel/early_printk.c b/arch/x86_64/kernel/early_printk.c index e3a19e8ebbf..9631c747c5e 100644 --- a/arch/x86_64/kernel/early_printk.c +++ b/arch/x86_64/kernel/early_printk.c @@ -2,20 +2,24 @@ #include <linux/kernel.h> #include <linux/init.h> #include <linux/string.h> +#include <linux/tty.h> #include <asm/io.h> #include <asm/processor.h> /* Simple VGA output */ #ifdef __i386__ +#include <asm/setup.h> #define VGABASE (__ISA_IO_base + 0xb8000) #else +#include <asm/bootsetup.h> #define VGABASE ((void __iomem *)0xffffffff800b8000UL) #endif -#define MAX_YPOS 25 -#define MAX_XPOS 80 +#define MAX_YPOS max_ypos +#define MAX_XPOS max_xpos +static int max_ypos = 25, max_xpos = 80; static int current_ypos = 1, current_xpos = 0; static void early_vga_write(struct console *con, const char *str, unsigned n) @@ -196,7 +200,10 @@ int __init setup_early_printk(char *opt) } else if (!strncmp(buf, "ttyS", 4)) { early_serial_init(buf); early_console = &early_serial_console; - } else if (!strncmp(buf, "vga", 3)) { + } else if (!strncmp(buf, "vga", 3) + && SCREEN_INFO.orig_video_isVGA == 1) { + max_xpos = SCREEN_INFO.orig_video_cols; + max_ypos = SCREEN_INFO.orig_video_lines; early_console = &early_vga_console; } early_console_initialized = 1; diff --git a/arch/x86_64/kernel/genapic_flat.c b/arch/x86_64/kernel/genapic_flat.c index b4cbbad0422..28284696508 100644 --- a/arch/x86_64/kernel/genapic_flat.c +++ b/arch/x86_64/kernel/genapic_flat.c @@ -7,6 +7,8 @@ * Hacked for x86-64 by James Cleverdon from i386 architecture code by * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and * James Cleverdon. + * Ashok Raj <ashok.raj@intel.com> + * Removed IPI broadcast shortcut to support CPU hotplug */ #include <linux/config.h> #include <linux/threads.h> @@ -18,6 +20,46 @@ #include <asm/smp.h> #include <asm/ipi.h> +/* + * The following permit choosing broadcast IPI shortcut v.s sending IPI only + * to online cpus via the send_IPI_mask varient. + * The mask version is my preferred option, since it eliminates a lot of + * other extra code that would need to be written to cleanup intrs sent + * to a CPU while offline. + * + * Sending broadcast introduces lots of trouble in CPU hotplug situations. + * These IPI's are delivered to cpu's irrespective of their offline status + * and could pickup stale intr data when these CPUS are turned online. + * + * Not using broadcast is a cleaner approach IMO, but Andi Kleen disagrees with + * the idea of not using broadcast IPI's anymore. Hence the run time check + * is introduced, on his request so we can choose an alternate mechanism. + * + * Initial wacky performance tests that collect cycle counts show + * no increase in using mask v.s broadcast version. In fact they seem + * identical in terms of cycle counts. + * + * if we need to use broadcast, we need to do the following. + * + * cli; + * hold call_lock; + * clear any pending IPI, just ack and clear all pending intr + * set cpu_online_map; + * release call_lock; + * sti; + * + * The complicated dummy irq processing shown above is not required if + * we didnt sent IPI's to wrong CPU's in the first place. + * + * - Ashok Raj <ashok.raj@intel.com> + */ +#ifdef CONFIG_HOTPLUG_CPU +#define DEFAULT_SEND_IPI (1) +#else +#define DEFAULT_SEND_IPI (0) +#endif + +static int no_broadcast=DEFAULT_SEND_IPI; static cpumask_t flat_target_cpus(void) { @@ -45,22 +87,6 @@ static void flat_init_apic_ldr(void) apic_write_around(APIC_LDR, val); } -static void flat_send_IPI_allbutself(int vector) -{ - /* - * if there are no other CPUs in the system then - * we get an APIC send error if we try to broadcast. - * thus we have to avoid sending IPIs in this case. - */ - if (num_online_cpus() > 1) - __send_IPI_shortcut(APIC_DEST_ALLBUT, vector, APIC_DEST_LOGICAL); -} - -static void flat_send_IPI_all(int vector) -{ - __send_IPI_shortcut(APIC_DEST_ALLINC, vector, APIC_DEST_LOGICAL); -} - static void flat_send_IPI_mask(cpumask_t cpumask, int vector) { unsigned long mask = cpus_addr(cpumask)[0]; @@ -93,6 +119,39 @@ static void flat_send_IPI_mask(cpumask_t cpumask, int vector) local_irq_restore(flags); } +static inline void __local_flat_send_IPI_allbutself(int vector) +{ + if (no_broadcast) { + cpumask_t mask = cpu_online_map; + int this_cpu = get_cpu(); + + cpu_clear(this_cpu, mask); + flat_send_IPI_mask(mask, vector); + put_cpu(); + } + else + __send_IPI_shortcut(APIC_DEST_ALLBUT, vector, APIC_DEST_LOGICAL); +} + +static inline void __local_flat_send_IPI_all(int vector) +{ + if (no_broadcast) + flat_send_IPI_mask(cpu_online_map, vector); + else + __send_IPI_shortcut(APIC_DEST_ALLINC, vector, APIC_DEST_LOGICAL); +} + +static void flat_send_IPI_allbutself(int vector) +{ + if (((num_online_cpus()) - 1) >= 1) + __local_flat_send_IPI_allbutself(vector); +} + +static void flat_send_IPI_all(int vector) +{ + __local_flat_send_IPI_all(vector); +} + static int flat_apic_id_registered(void) { return physid_isset(GET_APIC_ID(apic_read(APIC_ID)), phys_cpu_present_map); @@ -111,6 +170,16 @@ static unsigned int phys_pkg_id(int index_msb) return ((ebx >> 24) & 0xFF) >> index_msb; } +static __init int no_ipi_broadcast(char *str) +{ + get_option(&str, &no_broadcast); + printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" : + "IPI Broadcast"); + return 1; +} + +__setup("no_ipi_broadcast", no_ipi_broadcast); + struct genapic apic_flat = { .name = "flat", .int_delivery_mode = dest_LowestPrio, @@ -125,3 +194,12 @@ struct genapic apic_flat = { .cpu_mask_to_apicid = flat_cpu_mask_to_apicid, .phys_pkg_id = phys_pkg_id, }; + +static int __init print_ipi_mode(void) +{ + printk ("Using IPI %s mode\n", no_broadcast ? "No-Shortcut" : + "Shortcut"); + return 0; +} + +late_initcall(print_ipi_mode); diff --git a/arch/x86_64/kernel/head.S b/arch/x86_64/kernel/head.S index 9bd2e7a4b81..8d765aa77a2 100644 --- a/arch/x86_64/kernel/head.S +++ b/arch/x86_64/kernel/head.S @@ -248,23 +248,23 @@ ENTRY(_stext) */ .org 0x1000 ENTRY(init_level4_pgt) - .quad 0x0000000000102007 /* -> level3_ident_pgt */ + .quad 0x0000000000002007 + __PHYSICAL_START /* -> level3_ident_pgt */ .fill 255,8,0 - .quad 0x000000000010a007 + .quad 0x000000000000a007 + __PHYSICAL_START .fill 254,8,0 /* (2^48-(2*1024*1024*1024))/(2^39) = 511 */ - .quad 0x0000000000103007 /* -> level3_kernel_pgt */ + .quad 0x0000000000003007 + __PHYSICAL_START /* -> level3_kernel_pgt */ .org 0x2000 ENTRY(level3_ident_pgt) - .quad 0x0000000000104007 + .quad 0x0000000000004007 + __PHYSICAL_START .fill 511,8,0 .org 0x3000 ENTRY(level3_kernel_pgt) .fill 510,8,0 /* (2^48-(2*1024*1024*1024)-((2^39)*511))/(2^30) = 510 */ - .quad 0x0000000000105007 /* -> level2_kernel_pgt */ + .quad 0x0000000000005007 + __PHYSICAL_START /* -> level2_kernel_pgt */ .fill 1,8,0 .org 0x4000 @@ -337,17 +337,17 @@ ENTRY(empty_bad_pmd_table) .org 0xa000 ENTRY(level3_physmem_pgt) - .quad 0x0000000000105007 /* -> level2_kernel_pgt (so that __va works even before pagetable_init) */ + .quad 0x0000000000005007 + __PHYSICAL_START /* -> level2_kernel_pgt (so that __va works even before pagetable_init) */ .org 0xb000 #ifdef CONFIG_ACPI_SLEEP ENTRY(wakeup_level4_pgt) - .quad 0x0000000000102007 /* -> level3_ident_pgt */ + .quad 0x0000000000002007 + __PHYSICAL_START /* -> level3_ident_pgt */ .fill 255,8,0 - .quad 0x000000000010a007 + .quad 0x000000000000a007 + __PHYSICAL_START .fill 254,8,0 /* (2^48-(2*1024*1024*1024))/(2^39) = 511 */ - .quad 0x0000000000103007 /* -> level3_kernel_pgt */ + .quad 0x0000000000003007 + __PHYSICAL_START /* -> level3_kernel_pgt */ #endif .data diff --git a/arch/x86_64/kernel/head64.c b/arch/x86_64/kernel/head64.c index 0f8c78dcd38..cf6ab147a2a 100644 --- a/arch/x86_64/kernel/head64.c +++ b/arch/x86_64/kernel/head64.c @@ -94,7 +94,7 @@ void __init x86_64_start_kernel(char * real_mode_data) s = strstr(saved_command_line, "earlyprintk="); if (s != NULL) setup_early_printk(s); -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA s = strstr(saved_command_line, "numa="); if (s != NULL) numa_setup(s+5); diff --git a/arch/x86_64/kernel/i387.c b/arch/x86_64/kernel/i387.c index ba139cac57c..d9b22b633e3 100644 --- a/arch/x86_64/kernel/i387.c +++ b/arch/x86_64/kernel/i387.c @@ -42,7 +42,7 @@ void mxcsr_feature_mask_init(void) * Called at bootup to set up the initial FPU state that is later cloned * into all processes. */ -void __init fpu_init(void) +void __cpuinit fpu_init(void) { unsigned long oldcr0 = read_cr0(); extern void __bad_fxsave_alignment(void); diff --git a/arch/x86_64/kernel/i8259.c b/arch/x86_64/kernel/i8259.c index 7873d9ba881..a8916909512 100644 --- a/arch/x86_64/kernel/i8259.c +++ b/arch/x86_64/kernel/i8259.c @@ -157,14 +157,13 @@ static unsigned int startup_8259A_irq(unsigned int irq) } static struct hw_interrupt_type i8259A_irq_type = { - "XT-PIC", - startup_8259A_irq, - shutdown_8259A_irq, - enable_8259A_irq, - disable_8259A_irq, - mask_and_ack_8259A, - end_8259A_irq, - NULL + .typename = "XT-PIC", + .startup = startup_8259A_irq, + .shutdown = shutdown_8259A_irq, + .enable = enable_8259A_irq, + .disable = disable_8259A_irq, + .ack = mask_and_ack_8259A, + .end = end_8259A_irq, }; /* @@ -415,10 +414,22 @@ static int i8259A_suspend(struct sys_device *dev, pm_message_t state) return 0; } +static int i8259A_shutdown(struct sys_device *dev) +{ + /* Put the i8259A into a quiescent state that + * the kernel initialization code can get it + * out of. + */ + outb(0xff, 0x21); /* mask all of 8259A-1 */ + outb(0xff, 0xA1); /* mask all of 8259A-1 */ + return 0; +} + static struct sysdev_class i8259_sysdev_class = { set_kset_name("i8259"), .suspend = i8259A_suspend, .resume = i8259A_resume, + .shutdown = i8259A_shutdown, }; static struct sys_device device_i8259A = { diff --git a/arch/x86_64/kernel/io_apic.c b/arch/x86_64/kernel/io_apic.c index afd87e64d0a..157190d986b 100644 --- a/arch/x86_64/kernel/io_apic.c +++ b/arch/x86_64/kernel/io_apic.c @@ -330,7 +330,7 @@ static int find_irq_entry(int apic, int pin, int type) /* * Find the pin to which IRQ[irq] (ISA) is connected */ -static int __init find_isa_irq_pin(int irq, int type) +static int find_isa_irq_pin(int irq, int type) { int i; @@ -1132,12 +1132,44 @@ static void __init enable_IO_APIC(void) */ void disable_IO_APIC(void) { + int pin; /* * Clear the IO-APIC before rebooting: */ clear_IO_APIC(); - disconnect_bsp_APIC(); + /* + * If the i82559 is routed through an IOAPIC + * Put that IOAPIC in virtual wire mode + * so legacy interrups can be delivered. + */ + pin = find_isa_irq_pin(0, mp_ExtINT); + if (pin != -1) { + struct IO_APIC_route_entry entry; + unsigned long flags; + + memset(&entry, 0, sizeof(entry)); + entry.mask = 0; /* Enabled */ + entry.trigger = 0; /* Edge */ + entry.irr = 0; + entry.polarity = 0; /* High */ + entry.delivery_status = 0; + entry.dest_mode = 0; /* Physical */ + entry.delivery_mode = 7; /* ExtInt */ + entry.vector = 0; + entry.dest.physical.physical_dest = 0; + + + /* + * Add it to the IO-APIC irq-routing table: + */ + spin_lock_irqsave(&ioapic_lock, flags); + io_apic_write(0, 0x11+2*pin, *(((int *)&entry)+1)); + io_apic_write(0, 0x10+2*pin, *(((int *)&entry)+0)); + spin_unlock_irqrestore(&ioapic_lock, flags); + } + + disconnect_bsp_APIC(pin != -1); } /* diff --git a/arch/x86_64/kernel/irq.c b/arch/x86_64/kernel/irq.c index 62b112e4deb..cc3fb85f514 100644 --- a/arch/x86_64/kernel/irq.c +++ b/arch/x86_64/kernel/irq.c @@ -14,6 +14,7 @@ #include <linux/interrupt.h> #include <linux/seq_file.h> #include <linux/module.h> +#include <linux/delay.h> #include <asm/uaccess.h> #include <asm/io_apic.h> @@ -106,3 +107,31 @@ asmlinkage unsigned int do_IRQ(struct pt_regs *regs) return 1; } +#ifdef CONFIG_HOTPLUG_CPU +void fixup_irqs(cpumask_t map) +{ + unsigned int irq; + static int warned; + + for (irq = 0; irq < NR_IRQS; irq++) { + cpumask_t mask; + if (irq == 2) + continue; + + cpus_and(mask, irq_affinity[irq], map); + if (any_online_cpu(mask) == NR_CPUS) { + printk("Breaking affinity for irq %i\n", irq); + mask = map; + } + if (irq_desc[irq].handler->set_affinity) + irq_desc[irq].handler->set_affinity(irq, mask); + else if (irq_desc[irq].action && !(warned++)) + printk("Cannot set affinity for irq %i\n", irq); + } + + /* That doesn't seem sufficient. Give it 1ms. */ + local_irq_enable(); + mdelay(1); + local_irq_disable(); +} +#endif diff --git a/arch/x86_64/kernel/kprobes.c b/arch/x86_64/kernel/kprobes.c index f77f8a0ff18..4e680f87a75 100644 --- a/arch/x86_64/kernel/kprobes.c +++ b/arch/x86_64/kernel/kprobes.c @@ -27,6 +27,8 @@ * <prasanna@in.ibm.com> adapted for x86_64 * 2005-Mar Roland McGrath <roland@redhat.com> * Fixed to handle %rip-relative addressing mode correctly. + * 2005-May Rusty Lynch <rusty.lynch@intel.com> + * Added function return probes functionality */ #include <linux/config.h> @@ -37,18 +39,16 @@ #include <linux/slab.h> #include <linux/preempt.h> #include <linux/moduleloader.h> - +#include <asm/cacheflush.h> #include <asm/pgtable.h> #include <asm/kdebug.h> static DECLARE_MUTEX(kprobe_mutex); -/* kprobe_status settings */ -#define KPROBE_HIT_ACTIVE 0x00000001 -#define KPROBE_HIT_SS 0x00000002 - static struct kprobe *current_kprobe; static unsigned long kprobe_status, kprobe_old_rflags, kprobe_saved_rflags; +static struct kprobe *kprobe_prev; +static unsigned long kprobe_status_prev, kprobe_old_rflags_prev, kprobe_saved_rflags_prev; static struct pt_regs jprobe_saved_regs; static long *jprobe_saved_rsp; static kprobe_opcode_t *get_insn_slot(void); @@ -214,6 +214,21 @@ void arch_copy_kprobe(struct kprobe *p) BUG_ON((s64) (s32) disp != disp); /* Sanity check. */ *ripdisp = disp; } + p->opcode = *p->addr; +} + +void arch_arm_kprobe(struct kprobe *p) +{ + *p->addr = BREAKPOINT_INSTRUCTION; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); +} + +void arch_disarm_kprobe(struct kprobe *p) +{ + *p->addr = p->opcode; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); } void arch_remove_kprobe(struct kprobe *p) @@ -223,10 +238,29 @@ void arch_remove_kprobe(struct kprobe *p) down(&kprobe_mutex); } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) +static inline void save_previous_kprobe(void) { - *p->addr = p->opcode; - regs->rip = (unsigned long)p->addr; + kprobe_prev = current_kprobe; + kprobe_status_prev = kprobe_status; + kprobe_old_rflags_prev = kprobe_old_rflags; + kprobe_saved_rflags_prev = kprobe_saved_rflags; +} + +static inline void restore_previous_kprobe(void) +{ + current_kprobe = kprobe_prev; + kprobe_status = kprobe_status_prev; + kprobe_old_rflags = kprobe_old_rflags_prev; + kprobe_saved_rflags = kprobe_saved_rflags_prev; +} + +static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs) +{ + current_kprobe = p; + kprobe_saved_rflags = kprobe_old_rflags + = (regs->eflags & (TF_MASK | IF_MASK)); + if (is_IF_modifier(p->ainsn.insn)) + kprobe_saved_rflags &= ~IF_MASK; } static void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -240,6 +274,50 @@ static void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) regs->rip = (unsigned long)p->ainsn.insn; } +struct task_struct *arch_get_kprobe_task(void *ptr) +{ + return ((struct thread_info *) (((unsigned long) ptr) & + (~(THREAD_SIZE -1))))->task; +} + +void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs) +{ + unsigned long *sara = (unsigned long *)regs->rsp; + struct kretprobe_instance *ri; + static void *orig_ret_addr; + + /* + * Save the return address when the return probe hits + * the first time, and use it to populate the (krprobe + * instance)->ret_addr for subsequent return probes at + * the same addrress since stack address would have + * the kretprobe_trampoline by then. + */ + if (((void*) *sara) != kretprobe_trampoline) + orig_ret_addr = (void*) *sara; + + if ((ri = get_free_rp_inst(rp)) != NULL) { + ri->rp = rp; + ri->stack_addr = sara; + ri->ret_addr = orig_ret_addr; + add_rp_inst(ri); + /* Replace the return addr with trampoline addr */ + *sara = (unsigned long) &kretprobe_trampoline; + } else { + rp->nmissed++; + } +} + +void arch_kprobe_flush_task(struct task_struct *tk) +{ + struct kretprobe_instance *ri; + while ((ri = get_rp_inst_tsk(tk)) != NULL) { + *((unsigned long *)(ri->stack_addr)) = + (unsigned long) ri->ret_addr; + recycle_rp_inst(ri); + } +} + /* * Interrupts are disabled on entry as trap3 is an interrupt gate and they * remain disabled thorough out this function. @@ -264,9 +342,30 @@ int kprobe_handler(struct pt_regs *regs) regs->eflags |= kprobe_saved_rflags; unlock_kprobes(); goto no_kprobe; + } else if (kprobe_status == KPROBE_HIT_SSDONE) { + /* TODO: Provide re-entrancy from + * post_kprobes_handler() and avoid exception + * stack corruption while single-stepping on + * the instruction of the new probe. + */ + arch_disarm_kprobe(p); + regs->rip = (unsigned long)p->addr; + ret = 1; + } else { + /* We have reentered the kprobe_handler(), since + * another probe was hit while within the + * handler. We here save the original kprobe + * variables and just single step on instruction + * of the new probe without calling any user + * handlers. + */ + save_previous_kprobe(); + set_current_kprobe(p, regs); + p->nmissed++; + prepare_singlestep(p, regs); + kprobe_status = KPROBE_REENTER; + return 1; } - disarm_kprobe(p, regs); - ret = 1; } else { p = current_kprobe; if (p->break_handler && p->break_handler(p, regs)) { @@ -296,11 +395,7 @@ int kprobe_handler(struct pt_regs *regs) } kprobe_status = KPROBE_HIT_ACTIVE; - current_kprobe = p; - kprobe_saved_rflags = kprobe_old_rflags - = (regs->eflags & (TF_MASK | IF_MASK)); - if (is_IF_modifier(p->ainsn.insn)) - kprobe_saved_rflags &= ~IF_MASK; + set_current_kprobe(p, regs); if (p->pre_handler && p->pre_handler(p, regs)) /* handler has already set things up, so skip ss setup */ @@ -317,6 +412,55 @@ no_kprobe: } /* + * For function-return probes, init_kprobes() establishes a probepoint + * here. When a retprobed function returns, this probe is hit and + * trampoline_probe_handler() runs, calling the kretprobe's handler. + */ + void kretprobe_trampoline_holder(void) + { + asm volatile ( ".global kretprobe_trampoline\n" + "kretprobe_trampoline: \n" + "nop\n"); + } + +/* + * Called when we hit the probe point at kretprobe_trampoline + */ +int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct task_struct *tsk; + struct kretprobe_instance *ri; + struct hlist_head *head; + struct hlist_node *node; + unsigned long *sara = (unsigned long *)regs->rsp - 1; + + tsk = arch_get_kprobe_task(sara); + head = kretprobe_inst_table_head(tsk); + + hlist_for_each_entry(ri, node, head, hlist) { + if (ri->stack_addr == sara && ri->rp) { + if (ri->rp->handler) + ri->rp->handler(ri, regs); + } + } + return 0; +} + +void trampoline_post_handler(struct kprobe *p, struct pt_regs *regs, + unsigned long flags) +{ + struct kretprobe_instance *ri; + /* RA already popped */ + unsigned long *sara = ((unsigned long *)regs->rsp) - 1; + + while ((ri = get_rp_inst(sara))) { + regs->rip = (unsigned long)ri->ret_addr; + recycle_rp_inst(ri); + } + regs->eflags &= ~TF_MASK; +} + +/* * Called after single-stepping. p->addr is the address of the * instruction whose first byte has been replaced by the "int 3" * instruction. To avoid the SMP problems that can occur when we @@ -401,13 +545,23 @@ int post_kprobe_handler(struct pt_regs *regs) if (!kprobe_running()) return 0; - if (current_kprobe->post_handler) + if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) { + kprobe_status = KPROBE_HIT_SSDONE; current_kprobe->post_handler(current_kprobe, regs, 0); + } - resume_execution(current_kprobe, regs); + if (current_kprobe->post_handler != trampoline_post_handler) + resume_execution(current_kprobe, regs); regs->eflags |= kprobe_saved_rflags; - unlock_kprobes(); + /* Restore the original saved kprobes variables and continue. */ + if (kprobe_status == KPROBE_REENTER) { + restore_previous_kprobe(); + goto out; + } else { + unlock_kprobes(); + } +out: preempt_enable_no_resched(); /* diff --git a/arch/x86_64/kernel/machine_kexec.c b/arch/x86_64/kernel/machine_kexec.c new file mode 100644 index 00000000000..60d1eff4156 --- /dev/null +++ b/arch/x86_64/kernel/machine_kexec.c @@ -0,0 +1,250 @@ +/* + * machine_kexec.c - handle transition of Linux booting another kernel + * Copyright (C) 2002-2005 Eric Biederman <ebiederm@xmission.com> + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include <linux/mm.h> +#include <linux/kexec.h> +#include <linux/delay.h> +#include <linux/string.h> +#include <linux/reboot.h> +#include <asm/pda.h> +#include <asm/pgtable.h> +#include <asm/pgalloc.h> +#include <asm/tlbflush.h> +#include <asm/mmu_context.h> +#include <asm/io.h> +#include <asm/apic.h> +#include <asm/cpufeature.h> +#include <asm/hw_irq.h> + +#define LEVEL0_SIZE (1UL << 12UL) +#define LEVEL1_SIZE (1UL << 21UL) +#define LEVEL2_SIZE (1UL << 30UL) +#define LEVEL3_SIZE (1UL << 39UL) +#define LEVEL4_SIZE (1UL << 48UL) + +#define L0_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define L1_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_PSE) +#define L2_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define L3_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) + +static void init_level2_page(u64 *level2p, unsigned long addr) +{ + unsigned long end_addr; + + addr &= PAGE_MASK; + end_addr = addr + LEVEL2_SIZE; + while (addr < end_addr) { + *(level2p++) = addr | L1_ATTR; + addr += LEVEL1_SIZE; + } +} + +static int init_level3_page(struct kimage *image, u64 *level3p, + unsigned long addr, unsigned long last_addr) +{ + unsigned long end_addr; + int result; + + result = 0; + addr &= PAGE_MASK; + end_addr = addr + LEVEL3_SIZE; + while ((addr < last_addr) && (addr < end_addr)) { + struct page *page; + u64 *level2p; + + page = kimage_alloc_control_pages(image, 0); + if (!page) { + result = -ENOMEM; + goto out; + } + level2p = (u64 *)page_address(page); + init_level2_page(level2p, addr); + *(level3p++) = __pa(level2p) | L2_ATTR; + addr += LEVEL2_SIZE; + } + /* clear the unused entries */ + while (addr < end_addr) { + *(level3p++) = 0; + addr += LEVEL2_SIZE; + } +out: + return result; +} + + +static int init_level4_page(struct kimage *image, u64 *level4p, + unsigned long addr, unsigned long last_addr) +{ + unsigned long end_addr; + int result; + + result = 0; + addr &= PAGE_MASK; + end_addr = addr + LEVEL4_SIZE; + while ((addr < last_addr) && (addr < end_addr)) { + struct page *page; + u64 *level3p; + + page = kimage_alloc_control_pages(image, 0); + if (!page) { + result = -ENOMEM; + goto out; + } + level3p = (u64 *)page_address(page); + result = init_level3_page(image, level3p, addr, last_addr); + if (result) { + goto out; + } + *(level4p++) = __pa(level3p) | L3_ATTR; + addr += LEVEL3_SIZE; + } + /* clear the unused entries */ + while (addr < end_addr) { + *(level4p++) = 0; + addr += LEVEL3_SIZE; + } +out: + return result; +} + + +static int init_pgtable(struct kimage *image, unsigned long start_pgtable) +{ + u64 *level4p; + level4p = (u64 *)__va(start_pgtable); + return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT); +} + +static void set_idt(void *newidt, u16 limit) +{ + unsigned char curidt[10]; + + /* x86-64 supports unaliged loads & stores */ + (*(u16 *)(curidt)) = limit; + (*(u64 *)(curidt +2)) = (unsigned long)(newidt); + + __asm__ __volatile__ ( + "lidt %0\n" + : "=m" (curidt) + ); +}; + + +static void set_gdt(void *newgdt, u16 limit) +{ + unsigned char curgdt[10]; + + /* x86-64 supports unaligned loads & stores */ + (*(u16 *)(curgdt)) = limit; + (*(u64 *)(curgdt +2)) = (unsigned long)(newgdt); + + __asm__ __volatile__ ( + "lgdt %0\n" + : "=m" (curgdt) + ); +}; + +static void load_segments(void) +{ + __asm__ __volatile__ ( + "\tmovl $"STR(__KERNEL_DS)",%eax\n" + "\tmovl %eax,%ds\n" + "\tmovl %eax,%es\n" + "\tmovl %eax,%ss\n" + "\tmovl %eax,%fs\n" + "\tmovl %eax,%gs\n" + ); +#undef STR +#undef __STR +} + +typedef NORET_TYPE void (*relocate_new_kernel_t)(unsigned long indirection_page, + unsigned long control_code_buffer, + unsigned long start_address, + unsigned long pgtable) ATTRIB_NORET; + +const extern unsigned char relocate_new_kernel[]; +const extern unsigned long relocate_new_kernel_size; + +int machine_kexec_prepare(struct kimage *image) +{ + unsigned long start_pgtable, control_code_buffer; + int result; + + /* Calculate the offsets */ + start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT; + control_code_buffer = start_pgtable + 4096UL; + + /* Setup the identity mapped 64bit page table */ + result = init_pgtable(image, start_pgtable); + if (result) + return result; + + /* Place the code in the reboot code buffer */ + memcpy(__va(control_code_buffer), relocate_new_kernel, + relocate_new_kernel_size); + + return 0; +} + +void machine_kexec_cleanup(struct kimage *image) +{ + return; +} + +/* + * Do not allocate memory (or fail in any way) in machine_kexec(). + * We are past the point of no return, committed to rebooting now. + */ +NORET_TYPE void machine_kexec(struct kimage *image) +{ + unsigned long page_list; + unsigned long control_code_buffer; + unsigned long start_pgtable; + relocate_new_kernel_t rnk; + + /* Interrupts aren't acceptable while we reboot */ + local_irq_disable(); + + /* Calculate the offsets */ + page_list = image->head; + start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT; + control_code_buffer = start_pgtable + 4096UL; + + /* Set the low half of the page table to my identity mapped + * page table for kexec. Leave the high half pointing at the + * kernel pages. Don't bother to flush the global pages + * as that will happen when I fully switch to my identity mapped + * page table anyway. + */ + memcpy(__va(read_cr3()), __va(start_pgtable), PAGE_SIZE/2); + __flush_tlb(); + + + /* The segment registers are funny things, they are + * automatically loaded from a table, in memory wherever you + * set them to a specific selector, but this table is never + * accessed again unless you set the segment to a different selector. + * + * The more common model are caches where the behide + * the scenes work is done, but is also dropped at arbitrary + * times. + * + * I take advantage of this here by force loading the + * segments, before I zap the gdt with an invalid value. + */ + load_segments(); + /* The gdt & idt are now invalid. + * If you want to load them you must set up your own idt & gdt. + */ + set_gdt(phys_to_virt(0),0); + set_idt(phys_to_virt(0),0); + /* now call it */ + rnk = (relocate_new_kernel_t) control_code_buffer; + (*rnk)(page_list, control_code_buffer, image->start, start_pgtable); +} diff --git a/arch/x86_64/kernel/mce.c b/arch/x86_64/kernel/mce.c index 3a89d735a4f..21e70625a49 100644 --- a/arch/x86_64/kernel/mce.c +++ b/arch/x86_64/kernel/mce.c @@ -327,7 +327,7 @@ static void mce_init(void *dummy) } /* Add per CPU specific workarounds here */ -static void __init mce_cpu_quirks(struct cpuinfo_x86 *c) +static void __cpuinit mce_cpu_quirks(struct cpuinfo_x86 *c) { /* This should be disabled by the BIOS, but isn't always */ if (c->x86_vendor == X86_VENDOR_AMD && c->x86 == 15) { @@ -337,7 +337,7 @@ static void __init mce_cpu_quirks(struct cpuinfo_x86 *c) } } -static void __init mce_cpu_features(struct cpuinfo_x86 *c) +static void __cpuinit mce_cpu_features(struct cpuinfo_x86 *c) { switch (c->x86_vendor) { case X86_VENDOR_INTEL: @@ -352,7 +352,7 @@ static void __init mce_cpu_features(struct cpuinfo_x86 *c) * Called for each booted CPU to set up machine checks. * Must be called with preempt off. */ -void __init mcheck_init(struct cpuinfo_x86 *c) +void __cpuinit mcheck_init(struct cpuinfo_x86 *c) { static cpumask_t mce_cpus __initdata = CPU_MASK_NONE; @@ -411,7 +411,7 @@ static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize, loff memset(mcelog.entry, 0, next * sizeof(struct mce)); mcelog.next = 0; - synchronize_kernel(); + synchronize_sched(); /* Collect entries that were still getting written before the synchronize. */ @@ -542,7 +542,7 @@ ACCESSOR(bank4ctl,bank[4],mce_restart()) ACCESSOR(tolerant,tolerant,) ACCESSOR(check_interval,check_interval,mce_restart()) -static __init int mce_init_device(void) +static __cpuinit int mce_init_device(void) { int err; if (!mce_available(&boot_cpu_data)) diff --git a/arch/x86_64/kernel/mce_intel.c b/arch/x86_64/kernel/mce_intel.c index 4db9a640069..0be0a795981 100644 --- a/arch/x86_64/kernel/mce_intel.c +++ b/arch/x86_64/kernel/mce_intel.c @@ -42,7 +42,7 @@ done: irq_exit(); } -static void __init intel_init_thermal(struct cpuinfo_x86 *c) +static void __cpuinit intel_init_thermal(struct cpuinfo_x86 *c) { u32 l, h; int tm2 = 0; @@ -93,7 +93,7 @@ static void __init intel_init_thermal(struct cpuinfo_x86 *c) return; } -void __init mce_intel_feature_init(struct cpuinfo_x86 *c) +void __cpuinit mce_intel_feature_init(struct cpuinfo_x86 *c) { intel_init_thermal(c); } diff --git a/arch/x86_64/kernel/mpparse.c b/arch/x86_64/kernel/mpparse.c index 61a63be6b29..9c5aa2a790c 100644 --- a/arch/x86_64/kernel/mpparse.c +++ b/arch/x86_64/kernel/mpparse.c @@ -23,6 +23,7 @@ #include <linux/kernel_stat.h> #include <linux/mc146818rtc.h> #include <linux/acpi.h> +#include <linux/module.h> #include <asm/smp.h> #include <asm/mtrr.h> @@ -45,7 +46,8 @@ int acpi_found_madt; int apic_version [MAX_APICS]; unsigned char mp_bus_id_to_type [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 }; int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 }; -cpumask_t pci_bus_to_cpumask [256] = { [0 ... 255] = CPU_MASK_ALL }; +unsigned char pci_bus_to_node [256]; +EXPORT_SYMBOL(pci_bus_to_node); static int mp_current_pci_id = 0; /* I/O APIC entries */ @@ -904,11 +906,20 @@ void __init mp_config_acpi_legacy_irqs (void) return; } +#define MAX_GSI_NUM 4096 + int mp_register_gsi(u32 gsi, int edge_level, int active_high_low) { int ioapic = -1; int ioapic_pin = 0; int idx, bit = 0; + static int pci_irq = 16; + /* + * Mapping between Global System Interrupts, which + * represent all possible interrupts, to the IRQs + * assigned to actual devices. + */ + static int gsi_to_irq[MAX_GSI_NUM]; if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) return gsi; @@ -943,11 +954,21 @@ int mp_register_gsi(u32 gsi, int edge_level, int active_high_low) if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) { Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", mp_ioapic_routing[ioapic].apic_id, ioapic_pin); - return gsi; + return gsi_to_irq[gsi]; } mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit); + if (edge_level) { + /* + * For PCI devices assign IRQs in order, avoiding gaps + * due to unused I/O APIC pins. + */ + int irq = gsi; + gsi = pci_irq++; + gsi_to_irq[irq] = gsi; + } + io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, edge_level == ACPI_EDGE_SENSITIVE ? 0 : 1, active_high_low == ACPI_ACTIVE_HIGH ? 0 : 1); diff --git a/arch/x86_64/kernel/nmi.c b/arch/x86_64/kernel/nmi.c index 31c0f2e6ac9..4e44d6e6b7e 100644 --- a/arch/x86_64/kernel/nmi.c +++ b/arch/x86_64/kernel/nmi.c @@ -98,7 +98,7 @@ static unsigned int nmi_p4_cccr_val; (P4_CCCR_OVF_PMI0|P4_CCCR_THRESHOLD(15)|P4_CCCR_COMPLEMENT| \ P4_CCCR_COMPARE|P4_CCCR_REQUIRED|P4_CCCR_ESCR_SELECT(4)|P4_CCCR_ENABLE) -static __init inline int nmi_known_cpu(void) +static __cpuinit inline int nmi_known_cpu(void) { switch (boot_cpu_data.x86_vendor) { case X86_VENDOR_AMD: @@ -110,7 +110,7 @@ static __init inline int nmi_known_cpu(void) } /* Run after command line and cpu_init init, but before all other checks */ -void __init nmi_watchdog_default(void) +void __cpuinit nmi_watchdog_default(void) { if (nmi_watchdog != NMI_DEFAULT) return; diff --git a/arch/x86_64/kernel/process.c b/arch/x86_64/kernel/process.c index dce8bab4306..1d91271796e 100644 --- a/arch/x86_64/kernel/process.c +++ b/arch/x86_64/kernel/process.c @@ -8,7 +8,8 @@ * * X86-64 port * Andi Kleen. - * + * + * CPU hotplug support - ashok.raj@intel.com * $Id: process.c,v 1.38 2002/01/15 10:08:03 ak Exp $ */ @@ -18,6 +19,7 @@ #include <stdarg.h> +#include <linux/cpu.h> #include <linux/errno.h> #include <linux/sched.h> #include <linux/kernel.h> @@ -34,6 +36,7 @@ #include <linux/ptrace.h> #include <linux/utsname.h> #include <linux/random.h> +#include <linux/kprobes.h> #include <asm/uaccess.h> #include <asm/pgtable.h> @@ -153,6 +156,29 @@ void cpu_idle_wait(void) } EXPORT_SYMBOL_GPL(cpu_idle_wait); +#ifdef CONFIG_HOTPLUG_CPU +DECLARE_PER_CPU(int, cpu_state); + +#include <asm/nmi.h> +/* We don't actually take CPU down, just spin without interrupts. */ +static inline void play_dead(void) +{ + idle_task_exit(); + wbinvd(); + mb(); + /* Ack it */ + __get_cpu_var(cpu_state) = CPU_DEAD; + + while (1) + safe_halt(); +} +#else +static inline void play_dead(void) +{ + BUG(); +} +#endif /* CONFIG_HOTPLUG_CPU */ + /* * The idle thread. There's no useful work to be * done, so just try to conserve power and have a @@ -173,6 +199,8 @@ void cpu_idle (void) idle = pm_idle; if (!idle) idle = default_idle; + if (cpu_is_offline(smp_processor_id())) + play_dead(); idle(); } @@ -203,7 +231,7 @@ static void mwait_idle(void) } } -void __init select_idle_routine(const struct cpuinfo_x86 *c) +void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) { static int printed; if (cpu_has(c, X86_FEATURE_MWAIT)) { @@ -293,6 +321,14 @@ void exit_thread(void) { struct task_struct *me = current; struct thread_struct *t = &me->thread; + + /* + * Remove function-return probe instances associated with this task + * and put them back on the free list. Do not insert an exit probe for + * this function, it will be disabled by kprobe_flush_task if you do. + */ + kprobe_flush_task(me); + if (me->thread.io_bitmap_ptr) { struct tss_struct *tss = &per_cpu(init_tss, get_cpu()); @@ -312,6 +348,13 @@ void flush_thread(void) struct task_struct *tsk = current; struct thread_info *t = current_thread_info(); + /* + * Remove function-return probe instances associated with this task + * and put them back on the free list. Do not insert an exit probe for + * this function, it will be disabled by kprobe_flush_task if you do. + */ + kprobe_flush_task(tsk); + if (t->flags & _TIF_ABI_PENDING) t->flags ^= (_TIF_ABI_PENDING | _TIF_IA32); diff --git a/arch/x86_64/kernel/reboot.c b/arch/x86_64/kernel/reboot.c index be4b36f762c..57e71dbdfd6 100644 --- a/arch/x86_64/kernel/reboot.c +++ b/arch/x86_64/kernel/reboot.c @@ -66,41 +66,47 @@ static int __init reboot_setup(char *str) __setup("reboot=", reboot_setup); -#ifdef CONFIG_SMP -static void smp_halt(void) +static inline void kb_wait(void) { - int cpuid = safe_smp_processor_id(); - static int first_entry = 1; + int i; - if (reboot_force) - return; + for (i=0; i<0x10000; i++) + if ((inb_p(0x64) & 0x02) == 0) + break; +} - if (first_entry) { - first_entry = 0; - smp_call_function((void *)machine_restart, NULL, 1, 0); - } - - smp_stop_cpu(); +void machine_shutdown(void) +{ + /* Stop the cpus and apics */ +#ifdef CONFIG_SMP + int reboot_cpu_id; - /* AP calling this. Just halt */ - if (cpuid != boot_cpu_id) { - for (;;) - asm("hlt"); + /* The boot cpu is always logical cpu 0 */ + reboot_cpu_id = 0; + + /* Make certain the cpu I'm about to reboot on is online */ + if (!cpu_isset(reboot_cpu_id, cpu_online_map)) { + reboot_cpu_id = smp_processor_id(); } - /* Wait for all other CPUs to have run smp_stop_cpu */ - while (!cpus_empty(cpu_online_map)) - rep_nop(); -} + /* Make certain I only run on the appropriate processor */ + set_cpus_allowed(current, cpumask_of_cpu(reboot_cpu_id)); + + /* O.K Now that I'm on the appropriate processor, + * stop all of the others. + */ + smp_send_stop(); #endif -static inline void kb_wait(void) -{ - int i; + local_irq_disable(); - for (i=0; i<0x10000; i++) - if ((inb_p(0x64) & 0x02) == 0) - break; +#ifndef CONFIG_SMP + disable_local_APIC(); +#endif + + disable_IO_APIC(); + + local_irq_enable(); } void machine_restart(char * __unused) @@ -109,9 +115,7 @@ void machine_restart(char * __unused) printk("machine restart\n"); -#ifdef CONFIG_SMP - smp_halt(); -#endif + machine_shutdown(); if (!reboot_force) { local_irq_disable(); diff --git a/arch/x86_64/kernel/relocate_kernel.S b/arch/x86_64/kernel/relocate_kernel.S new file mode 100644 index 00000000000..d24fa9b72a2 --- /dev/null +++ b/arch/x86_64/kernel/relocate_kernel.S @@ -0,0 +1,143 @@ +/* + * relocate_kernel.S - put the kernel image in place to boot + * Copyright (C) 2002-2005 Eric Biederman <ebiederm@xmission.com> + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include <linux/linkage.h> + + /* + * Must be relocatable PIC code callable as a C function, that once + * it starts can not use the previous processes stack. + */ + .globl relocate_new_kernel + .code64 +relocate_new_kernel: + /* %rdi page_list + * %rsi reboot_code_buffer + * %rdx start address + * %rcx page_table + * %r8 arg5 + * %r9 arg6 + */ + + /* zero out flags, and disable interrupts */ + pushq $0 + popfq + + /* set a new stack at the bottom of our page... */ + lea 4096(%rsi), %rsp + + /* store the parameters back on the stack */ + pushq %rdx /* store the start address */ + + /* Set cr0 to a known state: + * 31 1 == Paging enabled + * 18 0 == Alignment check disabled + * 16 0 == Write protect disabled + * 3 0 == No task switch + * 2 0 == Don't do FP software emulation. + * 0 1 == Proctected mode enabled + */ + movq %cr0, %rax + andq $~((1<<18)|(1<<16)|(1<<3)|(1<<2)), %rax + orl $((1<<31)|(1<<0)), %eax + movq %rax, %cr0 + + /* Set cr4 to a known state: + * 10 0 == xmm exceptions disabled + * 9 0 == xmm registers instructions disabled + * 8 0 == performance monitoring counter disabled + * 7 0 == page global disabled + * 6 0 == machine check exceptions disabled + * 5 1 == physical address extension enabled + * 4 0 == page size extensions disabled + * 3 0 == Debug extensions disabled + * 2 0 == Time stamp disable (disabled) + * 1 0 == Protected mode virtual interrupts disabled + * 0 0 == VME disabled + */ + + movq $((1<<5)), %rax + movq %rax, %cr4 + + jmp 1f +1: + + /* Switch to the identity mapped page tables, + * and flush the TLB. + */ + movq %rcx, %cr3 + + /* Do the copies */ + movq %rdi, %rcx /* Put the page_list in %rcx */ + xorq %rdi, %rdi + xorq %rsi, %rsi + jmp 1f + +0: /* top, read another word for the indirection page */ + + movq (%rbx), %rcx + addq $8, %rbx +1: + testq $0x1, %rcx /* is it a destination page? */ + jz 2f + movq %rcx, %rdi + andq $0xfffffffffffff000, %rdi + jmp 0b +2: + testq $0x2, %rcx /* is it an indirection page? */ + jz 2f + movq %rcx, %rbx + andq $0xfffffffffffff000, %rbx + jmp 0b +2: + testq $0x4, %rcx /* is it the done indicator? */ + jz 2f + jmp 3f +2: + testq $0x8, %rcx /* is it the source indicator? */ + jz 0b /* Ignore it otherwise */ + movq %rcx, %rsi /* For ever source page do a copy */ + andq $0xfffffffffffff000, %rsi + + movq $512, %rcx + rep ; movsq + jmp 0b +3: + + /* To be certain of avoiding problems with self-modifying code + * I need to execute a serializing instruction here. + * So I flush the TLB by reloading %cr3 here, it's handy, + * and not processor dependent. + */ + movq %cr3, %rax + movq %rax, %cr3 + + /* set all of the registers to known values */ + /* leave %rsp alone */ + + xorq %rax, %rax + xorq %rbx, %rbx + xorq %rcx, %rcx + xorq %rdx, %rdx + xorq %rsi, %rsi + xorq %rdi, %rdi + xorq %rbp, %rbp + xorq %r8, %r8 + xorq %r9, %r9 + xorq %r10, %r9 + xorq %r11, %r11 + xorq %r12, %r12 + xorq %r13, %r13 + xorq %r14, %r14 + xorq %r15, %r15 + + ret +relocate_new_kernel_end: + + .globl relocate_new_kernel_size +relocate_new_kernel_size: + .quad relocate_new_kernel_end - relocate_new_kernel diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c index 99f038ede23..b02d921da4f 100644 --- a/arch/x86_64/kernel/setup.c +++ b/arch/x86_64/kernel/setup.c @@ -40,6 +40,9 @@ #include <linux/acpi.h> #include <linux/kallsyms.h> #include <linux/edd.h> +#include <linux/mmzone.h> +#include <linux/kexec.h> + #include <asm/mtrr.h> #include <asm/uaccess.h> #include <asm/system.h> @@ -348,7 +351,7 @@ static __init void parse_cmdline_early (char ** cmdline_p) if (!memcmp(from, "mem=", 4)) parse_memopt(from+4, &from); -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA if (!memcmp(from, "numa=", 5)) numa_setup(from+5); #endif @@ -365,6 +368,27 @@ static __init void parse_cmdline_early (char ** cmdline_p) if (!memcmp(from, "noexec=", 7)) nonx_setup(from + 7); +#ifdef CONFIG_KEXEC + /* crashkernel=size@addr specifies the location to reserve for + * a crash kernel. By reserving this memory we guarantee + * that linux never set's it up as a DMA target. + * Useful for holding code to do something appropriate + * after a kernel panic. + */ + else if (!memcmp(from, "crashkernel=", 12)) { + unsigned long size, base; + size = memparse(from+12, &from); + if (*from == '@') { + base = memparse(from+1, &from); + /* FIXME: Do I want a sanity check + * to validate the memory range? + */ + crashk_res.start = base; + crashk_res.end = base + size - 1; + } + } +#endif + next_char: c = *(from++); if (!c) @@ -377,17 +401,20 @@ static __init void parse_cmdline_early (char ** cmdline_p) *cmdline_p = command_line; } -#ifndef CONFIG_DISCONTIGMEM -static void __init contig_initmem_init(void) +#ifndef CONFIG_NUMA +static void __init +contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn) { - unsigned long bootmap_size, bootmap; - bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT; - bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size); - if (bootmap == -1L) - panic("Cannot find bootmem map of size %ld\n",bootmap_size); - bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn); - e820_bootmem_free(&contig_page_data, 0, end_pfn << PAGE_SHIFT); - reserve_bootmem(bootmap, bootmap_size); + unsigned long bootmap_size, bootmap; + + memory_present(0, start_pfn, end_pfn); + bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT; + bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size); + if (bootmap == -1L) + panic("Cannot find bootmem map of size %ld\n",bootmap_size); + bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn); + e820_bootmem_free(NODE_DATA(0), 0, end_pfn << PAGE_SHIFT); + reserve_bootmem(bootmap, bootmap_size); } #endif @@ -554,10 +581,10 @@ void __init setup_arch(char **cmdline_p) acpi_numa_init(); #endif -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA numa_initmem_init(0, end_pfn); #else - contig_initmem_init(); + contig_initmem_init(0, end_pfn); #endif /* Reserve direct mapping */ @@ -618,6 +645,15 @@ void __init setup_arch(char **cmdline_p) } } #endif + + sparse_init(); + +#ifdef CONFIG_KEXEC + if (crashk_res.start != crashk_res.end) { + reserve_bootmem(crashk_res.start, + crashk_res.end - crashk_res.start + 1); + } +#endif paging_init(); check_ioapic(); @@ -669,7 +705,7 @@ void __init setup_arch(char **cmdline_p) #endif } -static int __init get_model_name(struct cpuinfo_x86 *c) +static int __cpuinit get_model_name(struct cpuinfo_x86 *c) { unsigned int *v; @@ -685,7 +721,7 @@ static int __init get_model_name(struct cpuinfo_x86 *c) } -static void __init display_cacheinfo(struct cpuinfo_x86 *c) +static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c) { unsigned int n, dummy, eax, ebx, ecx, edx; @@ -796,7 +832,7 @@ static int __init init_amd(struct cpuinfo_x86 *c) return r; } -static void __init detect_ht(struct cpuinfo_x86 *c) +static void __cpuinit detect_ht(struct cpuinfo_x86 *c) { #ifdef CONFIG_SMP u32 eax, ebx, ecx, edx; @@ -857,7 +893,7 @@ static void __init detect_ht(struct cpuinfo_x86 *c) /* * find out the number of processor cores on the die */ -static int __init intel_num_cpu_cores(struct cpuinfo_x86 *c) +static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c) { unsigned int eax; @@ -875,7 +911,7 @@ static int __init intel_num_cpu_cores(struct cpuinfo_x86 *c) return 1; } -static void __init init_intel(struct cpuinfo_x86 *c) +static void __cpuinit init_intel(struct cpuinfo_x86 *c) { /* Cache sizes */ unsigned n; @@ -895,7 +931,7 @@ static void __init init_intel(struct cpuinfo_x86 *c) c->x86_num_cores = intel_num_cpu_cores(c); } -void __init get_cpu_vendor(struct cpuinfo_x86 *c) +void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) { char *v = c->x86_vendor_id; @@ -916,7 +952,7 @@ struct cpu_model_info { /* Do some early cpuid on the boot CPU to get some parameter that are needed before check_bugs. Everything advanced is in identify_cpu below. */ -void __init early_identify_cpu(struct cpuinfo_x86 *c) +void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c) { u32 tfms; @@ -970,7 +1006,7 @@ void __init early_identify_cpu(struct cpuinfo_x86 *c) /* * This does the hard work of actually picking apart the CPU stuff... */ -void __init identify_cpu(struct cpuinfo_x86 *c) +void __cpuinit identify_cpu(struct cpuinfo_x86 *c) { int i; u32 xlvl; @@ -1047,7 +1083,7 @@ void __init identify_cpu(struct cpuinfo_x86 *c) } -void __init print_cpu_info(struct cpuinfo_x86 *c) +void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) { if (c->x86_model_id[0]) printk("%s", c->x86_model_id); diff --git a/arch/x86_64/kernel/setup64.c b/arch/x86_64/kernel/setup64.c index 678b7ac33b8..f2b9c6bc999 100644 --- a/arch/x86_64/kernel/setup64.c +++ b/arch/x86_64/kernel/setup64.c @@ -29,7 +29,7 @@ char x86_boot_params[BOOT_PARAM_SIZE] __initdata = {0,}; -cpumask_t cpu_initialized __initdata = CPU_MASK_NONE; +cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE; struct x8664_pda cpu_pda[NR_CPUS] __cacheline_aligned; @@ -171,7 +171,7 @@ void syscall_init(void) wrmsrl(MSR_SYSCALL_MASK, EF_TF|EF_DF|EF_IE|0x3000); } -void __init check_efer(void) +void __cpuinit check_efer(void) { unsigned long efer; @@ -188,7 +188,7 @@ void __init check_efer(void) * 'CPU state barrier', nothing should get across. * A lot of state is already set up in PDA init. */ -void __init cpu_init (void) +void __cpuinit cpu_init (void) { #ifdef CONFIG_SMP int cpu = stack_smp_processor_id(); diff --git a/arch/x86_64/kernel/signal.c b/arch/x86_64/kernel/signal.c index 429c0269dc4..98590a989f3 100644 --- a/arch/x86_64/kernel/signal.c +++ b/arch/x86_64/kernel/signal.c @@ -34,9 +34,9 @@ #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) -void ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, +int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set, struct pt_regs * regs); -void ia32_setup_frame(int sig, struct k_sigaction *ka, +int ia32_setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, struct pt_regs * regs); asmlinkage long @@ -238,7 +238,7 @@ get_stack(struct k_sigaction *ka, struct pt_regs *regs, unsigned long size) return (void __user *)round_down(rsp - size, 16); } -static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, +static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set, struct pt_regs * regs) { struct rt_sigframe __user *frame; @@ -327,20 +327,23 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, current->comm, current->pid, frame, regs->rip, frame->pretcode); #endif - return; + return 1; give_sigsegv: force_sigsegv(sig, current); + return 0; } /* * OK, we're invoking a handler */ -static void +static int handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, sigset_t *oldset, struct pt_regs *regs) { + int ret; + #ifdef DEBUG_SIG printk("handle_signal pid:%d sig:%lu rip:%lx rsp:%lx regs=%p\n", current->pid, sig, @@ -384,20 +387,22 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, #ifdef CONFIG_IA32_EMULATION if (test_thread_flag(TIF_IA32)) { if (ka->sa.sa_flags & SA_SIGINFO) - ia32_setup_rt_frame(sig, ka, info, oldset, regs); + ret = ia32_setup_rt_frame(sig, ka, info, oldset, regs); else - ia32_setup_frame(sig, ka, oldset, regs); + ret = ia32_setup_frame(sig, ka, oldset, regs); } else #endif - setup_rt_frame(sig, ka, info, oldset, regs); + ret = setup_rt_frame(sig, ka, info, oldset, regs); - if (!(ka->sa.sa_flags & SA_NODEFER)) { + if (ret && !(ka->sa.sa_flags & SA_NODEFER)) { spin_lock_irq(¤t->sighand->siglock); sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask); sigaddset(¤t->blocked,sig); recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); } + + return ret; } /* @@ -417,10 +422,10 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset) * kernel mode. Just return without doing anything * if so. */ - if ((regs->cs & 3) != 3) + if (!user_mode(regs)) return 1; - if (try_to_freeze(0)) + if (try_to_freeze()) goto no_signal; if (!oldset) @@ -434,11 +439,10 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset) * inside the kernel. */ if (current->thread.debugreg7) - asm volatile("movq %0,%%db7" : : "r" (current->thread.debugreg7)); + set_debugreg(current->thread.debugreg7, 7); /* Whee! Actually deliver the signal. */ - handle_signal(signr, &info, &ka, oldset, regs); - return 1; + return handle_signal(signr, &info, &ka, oldset, regs); } no_signal: diff --git a/arch/x86_64/kernel/smp.c b/arch/x86_64/kernel/smp.c index 1e379ed17b1..ccae392886a 100644 --- a/arch/x86_64/kernel/smp.c +++ b/arch/x86_64/kernel/smp.c @@ -283,6 +283,16 @@ struct call_data_struct { static struct call_data_struct * call_data; +void lock_ipi_call_lock(void) +{ + spin_lock_irq(&call_lock); +} + +void unlock_ipi_call_lock(void) +{ + spin_unlock_irq(&call_lock); +} + /* * this function sends a 'generic call function' IPI to all other CPUs * in the system. diff --git a/arch/x86_64/kernel/smpboot.c b/arch/x86_64/kernel/smpboot.c index f1ec0f34594..b969ee12872 100644 --- a/arch/x86_64/kernel/smpboot.c +++ b/arch/x86_64/kernel/smpboot.c @@ -34,6 +34,7 @@ * Andi Kleen : Converted to new state machine. * Various cleanups. * Probably mostly hotplug CPU ready now. + * Ashok Raj : CPU hotplug support */ @@ -58,11 +59,6 @@ #include <asm/proto.h> #include <asm/nmi.h> -/* Change for real CPU hotplug. Note other files need to be fixed - first too. */ -#define __cpuinit __init -#define __cpuinitdata __initdata - /* Number of siblings per CPU package */ int smp_num_siblings = 1; /* Package ID of each logical CPU */ @@ -103,6 +99,37 @@ EXPORT_SYMBOL(cpu_core_map); extern unsigned char trampoline_data[]; extern unsigned char trampoline_end[]; +/* State of each CPU */ +DEFINE_PER_CPU(int, cpu_state) = { 0 }; + +/* + * Store all idle threads, this can be reused instead of creating + * a new thread. Also avoids complicated thread destroy functionality + * for idle threads. + */ +struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ; + +#define get_idle_for_cpu(x) (idle_thread_array[(x)]) +#define set_idle_for_cpu(x,p) (idle_thread_array[(x)] = (p)) + +/* + * cpu_possible_map should be static, it cannot change as cpu's + * are onlined, or offlined. The reason is per-cpu data-structures + * are allocated by some modules at init time, and dont expect to + * do this dynamically on cpu arrival/departure. + * cpu_present_map on the other hand can change dynamically. + * In case when cpu_hotplug is not compiled, then we resort to current + * behaviour, which is cpu_possible == cpu_present. + * If cpu-hotplug is supported, then we need to preallocate for all + * those NR_CPUS, hence cpu_possible_map represents entire NR_CPUS range. + * - Ashok Raj + */ +#ifdef CONFIG_HOTPLUG_CPU +#define fixup_cpu_possible_map(x) cpu_set((x), cpu_possible_map) +#else +#define fixup_cpu_possible_map(x) +#endif + /* * Currently trivial. Write the real->protected mode * bootstrap into the page concerned. The caller @@ -418,6 +445,33 @@ void __cpuinit smp_callin(void) cpu_set(cpuid, cpu_callin_map); } +static inline void set_cpu_sibling_map(int cpu) +{ + int i; + + if (smp_num_siblings > 1) { + for_each_cpu(i) { + if (cpu_core_id[cpu] == cpu_core_id[i]) { + cpu_set(i, cpu_sibling_map[cpu]); + cpu_set(cpu, cpu_sibling_map[i]); + } + } + } else { + cpu_set(cpu, cpu_sibling_map[cpu]); + } + + if (current_cpu_data.x86_num_cores > 1) { + for_each_cpu(i) { + if (phys_proc_id[cpu] == phys_proc_id[i]) { + cpu_set(i, cpu_core_map[cpu]); + cpu_set(cpu, cpu_core_map[i]); + } + } + } else { + cpu_core_map[cpu] = cpu_sibling_map[cpu]; + } +} + /* * Setup code on secondary processor (after comming out of the trampoline) */ @@ -448,9 +502,28 @@ void __cpuinit start_secondary(void) enable_APIC_timer(); /* + * The sibling maps must be set before turing the online map on for + * this cpu + */ + set_cpu_sibling_map(smp_processor_id()); + + /* + * We need to hold call_lock, so there is no inconsistency + * between the time smp_call_function() determines number of + * IPI receipients, and the time when the determination is made + * for which cpus receive the IPI in genapic_flat.c. Holding this + * lock helps us to not include this cpu in a currently in progress + * smp_call_function(). + */ + lock_ipi_call_lock(); + + /* * Allow the master to continue. */ cpu_set(smp_processor_id(), cpu_online_map); + per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; + unlock_ipi_call_lock(); + mb(); /* Wait for TSC sync to not schedule things before. @@ -628,33 +701,77 @@ static int __cpuinit wakeup_secondary_via_INIT(int phys_apicid, unsigned int sta return (send_status | accept_status); } +struct create_idle { + struct task_struct *idle; + struct completion done; + int cpu; +}; + +void do_fork_idle(void *_c_idle) +{ + struct create_idle *c_idle = _c_idle; + + c_idle->idle = fork_idle(c_idle->cpu); + complete(&c_idle->done); +} + /* * Boot one CPU. */ static int __cpuinit do_boot_cpu(int cpu, int apicid) { - struct task_struct *idle; unsigned long boot_error; int timeout; unsigned long start_rip; + struct create_idle c_idle = { + .cpu = cpu, + .done = COMPLETION_INITIALIZER(c_idle.done), + }; + DECLARE_WORK(work, do_fork_idle, &c_idle); + + c_idle.idle = get_idle_for_cpu(cpu); + + if (c_idle.idle) { + c_idle.idle->thread.rsp = (unsigned long) (((struct pt_regs *) + (THREAD_SIZE + (unsigned long) c_idle.idle->thread_info)) - 1); + init_idle(c_idle.idle, cpu); + goto do_rest; + } + /* - * We can't use kernel_thread since we must avoid to - * reschedule the child. + * During cold boot process, keventd thread is not spun up yet. + * When we do cpu hot-add, we create idle threads on the fly, we should + * not acquire any attributes from the calling context. Hence the clean + * way to create kernel_threads() is to do that from keventd(). + * We do the current_is_keventd() due to the fact that ACPI notifier + * was also queuing to keventd() and when the caller is already running + * in context of keventd(), we would end up with locking up the keventd + * thread. */ - idle = fork_idle(cpu); - if (IS_ERR(idle)) { + if (!keventd_up() || current_is_keventd()) + work.func(work.data); + else { + schedule_work(&work); + wait_for_completion(&c_idle.done); + } + + if (IS_ERR(c_idle.idle)) { printk("failed fork for CPU %d\n", cpu); - return PTR_ERR(idle); + return PTR_ERR(c_idle.idle); } - cpu_pda[cpu].pcurrent = idle; + set_idle_for_cpu(cpu, c_idle.idle); + +do_rest: + + cpu_pda[cpu].pcurrent = c_idle.idle; start_rip = setup_trampoline(); - init_rsp = idle->thread.rsp; + init_rsp = c_idle.idle->thread.rsp; per_cpu(init_tss,cpu).rsp0 = init_rsp; initial_code = start_secondary; - clear_ti_thread_flag(idle->thread_info, TIF_FORK); + clear_ti_thread_flag(c_idle.idle->thread_info, TIF_FORK); printk(KERN_INFO "Booting processor %d/%d rip %lx rsp %lx\n", cpu, apicid, start_rip, init_rsp); @@ -746,51 +863,6 @@ cycles_t cacheflush_time; unsigned long cache_decay_ticks; /* - * Construct cpu_sibling_map[], so that we can tell the sibling CPU - * on SMT systems efficiently. - */ -static __cpuinit void detect_siblings(void) -{ - int cpu; - - for (cpu = 0; cpu < NR_CPUS; cpu++) { - cpus_clear(cpu_sibling_map[cpu]); - cpus_clear(cpu_core_map[cpu]); - } - - for_each_online_cpu (cpu) { - struct cpuinfo_x86 *c = cpu_data + cpu; - int siblings = 0; - int i; - if (smp_num_siblings > 1) { - for_each_online_cpu (i) { - if (cpu_core_id[cpu] == cpu_core_id[i]) { - siblings++; - cpu_set(i, cpu_sibling_map[cpu]); - } - } - } else { - siblings++; - cpu_set(cpu, cpu_sibling_map[cpu]); - } - - if (siblings != smp_num_siblings) { - printk(KERN_WARNING - "WARNING: %d siblings found for CPU%d, should be %d\n", - siblings, cpu, smp_num_siblings); - smp_num_siblings = siblings; - } - if (c->x86_num_cores > 1) { - for_each_online_cpu(i) { - if (phys_proc_id[cpu] == phys_proc_id[i]) - cpu_set(i, cpu_core_map[cpu]); - } - } else - cpu_core_map[cpu] = cpu_sibling_map[cpu]; - } -} - -/* * Cleanup possible dangling ends... */ static __cpuinit void smp_cleanup_boot(void) @@ -823,7 +895,7 @@ static __cpuinit void smp_cleanup_boot(void) * * RED-PEN audit/test this more. I bet there is more state messed up here. */ -static __cpuinit void disable_smp(void) +static __init void disable_smp(void) { cpu_present_map = cpumask_of_cpu(0); cpu_possible_map = cpumask_of_cpu(0); @@ -838,7 +910,7 @@ static __cpuinit void disable_smp(void) /* * Handle user cpus=... parameter. */ -static __cpuinit void enforce_max_cpus(unsigned max_cpus) +static __init void enforce_max_cpus(unsigned max_cpus) { int i, k; k = 0; @@ -855,7 +927,7 @@ static __cpuinit void enforce_max_cpus(unsigned max_cpus) /* * Various sanity checks. */ -static int __cpuinit smp_sanity_check(unsigned max_cpus) +static int __init smp_sanity_check(unsigned max_cpus) { if (!physid_isset(hard_smp_processor_id(), phys_cpu_present_map)) { printk("weird, boot CPU (#%d) not listed by the BIOS.\n", @@ -913,7 +985,7 @@ static int __cpuinit smp_sanity_check(unsigned max_cpus) * Prepare for SMP bootup. The MP table or ACPI has been read * earlier. Just do some sanity checking here and enable APIC mode. */ -void __cpuinit smp_prepare_cpus(unsigned int max_cpus) +void __init smp_prepare_cpus(unsigned int max_cpus) { int i; @@ -930,10 +1002,9 @@ void __cpuinit smp_prepare_cpus(unsigned int max_cpus) int apicid = cpu_present_to_apicid(i); if (physid_isset(apicid, phys_cpu_present_map)) { cpu_set(i, cpu_present_map); - /* possible map would be different if we supported real - CPU hotplug. */ cpu_set(i, cpu_possible_map); } + fixup_cpu_possible_map(i); } if (smp_sanity_check(max_cpus) < 0) { @@ -978,13 +1049,13 @@ void __init smp_prepare_boot_cpu(void) int me = smp_processor_id(); cpu_set(me, cpu_online_map); cpu_set(me, cpu_callout_map); + cpu_set(0, cpu_sibling_map[0]); + cpu_set(0, cpu_core_map[0]); + per_cpu(cpu_state, me) = CPU_ONLINE; } /* * Entry point to boot a CPU. - * - * This is all __cpuinit, not __devinit for now because we don't support - * CPU hotplug (yet). */ int __cpuinit __cpu_up(unsigned int cpu) { @@ -1001,6 +1072,15 @@ int __cpuinit __cpu_up(unsigned int cpu) return -EINVAL; } + /* + * Already booted CPU? + */ + if (cpu_isset(cpu, cpu_callin_map)) { + Dprintk("do_boot_cpu %d Already started\n", cpu); + return -ENOSYS; + } + + per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; /* Boot it! */ err = do_boot_cpu(cpu, apicid); if (err < 0) { @@ -1013,23 +1093,118 @@ int __cpuinit __cpu_up(unsigned int cpu) while (!cpu_isset(cpu, cpu_online_map)) cpu_relax(); - return 0; + err = 0; + + return err; } /* * Finish the SMP boot. */ -void __cpuinit smp_cpus_done(unsigned int max_cpus) +void __init smp_cpus_done(unsigned int max_cpus) { +#ifndef CONFIG_HOTPLUG_CPU zap_low_mappings(); +#endif smp_cleanup_boot(); #ifdef CONFIG_X86_IO_APIC setup_ioapic_dest(); #endif - detect_siblings(); time_init_gtod(); check_nmi_watchdog(); } + +#ifdef CONFIG_HOTPLUG_CPU + +static void remove_siblinginfo(int cpu) +{ + int sibling; + + for_each_cpu_mask(sibling, cpu_sibling_map[cpu]) + cpu_clear(cpu, cpu_sibling_map[sibling]); + for_each_cpu_mask(sibling, cpu_core_map[cpu]) + cpu_clear(cpu, cpu_core_map[sibling]); + cpus_clear(cpu_sibling_map[cpu]); + cpus_clear(cpu_core_map[cpu]); + phys_proc_id[cpu] = BAD_APICID; + cpu_core_id[cpu] = BAD_APICID; +} + +void remove_cpu_from_maps(void) +{ + int cpu = smp_processor_id(); + + cpu_clear(cpu, cpu_callout_map); + cpu_clear(cpu, cpu_callin_map); + clear_bit(cpu, &cpu_initialized); /* was set by cpu_init() */ +} + +int __cpu_disable(void) +{ + int cpu = smp_processor_id(); + + /* + * Perhaps use cpufreq to drop frequency, but that could go + * into generic code. + * + * We won't take down the boot processor on i386 due to some + * interrupts only being able to be serviced by the BSP. + * Especially so if we're not using an IOAPIC -zwane + */ + if (cpu == 0) + return -EBUSY; + + disable_APIC_timer(); + + /* + * HACK: + * Allow any queued timer interrupts to get serviced + * This is only a temporary solution until we cleanup + * fixup_irqs as we do for IA64. + */ + local_irq_enable(); + mdelay(1); + + local_irq_disable(); + remove_siblinginfo(cpu); + + /* It's now safe to remove this processor from the online map */ + cpu_clear(cpu, cpu_online_map); + remove_cpu_from_maps(); + fixup_irqs(cpu_online_map); + return 0; +} + +void __cpu_die(unsigned int cpu) +{ + /* We don't do anything here: idle task is faking death itself. */ + unsigned int i; + + for (i = 0; i < 10; i++) { + /* They ack this in play_dead by setting CPU_DEAD */ + if (per_cpu(cpu_state, cpu) == CPU_DEAD) { + printk ("CPU %d is now offline\n", cpu); + return; + } + current->state = TASK_UNINTERRUPTIBLE; + schedule_timeout(HZ/10); + } + printk(KERN_ERR "CPU %u didn't die...\n", cpu); +} + +#else /* ... !CONFIG_HOTPLUG_CPU */ + +int __cpu_disable(void) +{ + return -ENOSYS; +} + +void __cpu_die(unsigned int cpu) +{ + /* We said "no" in __cpu_disable */ + BUG(); +} +#endif /* CONFIG_HOTPLUG_CPU */ diff --git a/arch/x86_64/kernel/suspend.c b/arch/x86_64/kernel/suspend.c index ebaa1e37d65..6c0f402e3a8 100644 --- a/arch/x86_64/kernel/suspend.c +++ b/arch/x86_64/kernel/suspend.c @@ -44,7 +44,6 @@ void __save_processor_state(struct saved_context *ctxt) */ asm volatile ("sgdt %0" : "=m" (ctxt->gdt_limit)); asm volatile ("sidt %0" : "=m" (ctxt->idt_limit)); - asm volatile ("sldt %0" : "=m" (ctxt->ldt)); asm volatile ("str %0" : "=m" (ctxt->tr)); /* XMM0..XMM15 should be handled by kernel_fpu_begin(). */ @@ -69,6 +68,7 @@ void __save_processor_state(struct saved_context *ctxt) asm volatile ("movq %%cr2, %0" : "=r" (ctxt->cr2)); asm volatile ("movq %%cr3, %0" : "=r" (ctxt->cr3)); asm volatile ("movq %%cr4, %0" : "=r" (ctxt->cr4)); + asm volatile ("movq %%cr8, %0" : "=r" (ctxt->cr8)); } void save_processor_state(void) @@ -90,12 +90,20 @@ void __restore_processor_state(struct saved_context *ctxt) /* * control registers */ + asm volatile ("movq %0, %%cr8" :: "r" (ctxt->cr8)); asm volatile ("movq %0, %%cr4" :: "r" (ctxt->cr4)); asm volatile ("movq %0, %%cr3" :: "r" (ctxt->cr3)); asm volatile ("movq %0, %%cr2" :: "r" (ctxt->cr2)); asm volatile ("movq %0, %%cr0" :: "r" (ctxt->cr0)); /* + * now restore the descriptor tables to their proper values + * ltr is done i fix_processor_context(). + */ + asm volatile ("lgdt %0" :: "m" (ctxt->gdt_limit)); + asm volatile ("lidt %0" :: "m" (ctxt->idt_limit)); + + /* * segment registers */ asm volatile ("movw %0, %%ds" :: "r" (ctxt->ds)); @@ -108,14 +116,6 @@ void __restore_processor_state(struct saved_context *ctxt) wrmsrl(MSR_GS_BASE, ctxt->gs_base); wrmsrl(MSR_KERNEL_GS_BASE, ctxt->gs_kernel_base); - /* - * now restore the descriptor tables to their proper values - * ltr is done i fix_processor_context(). - */ - asm volatile ("lgdt %0" :: "m" (ctxt->gdt_limit)); - asm volatile ("lidt %0" :: "m" (ctxt->idt_limit)); - asm volatile ("lldt %0" :: "m" (ctxt->ldt)); - fix_processor_context(); do_fpu_end(); diff --git a/arch/x86_64/kernel/time.c b/arch/x86_64/kernel/time.c index fb8c809b4cd..66bf6ddeb0c 100644 --- a/arch/x86_64/kernel/time.c +++ b/arch/x86_64/kernel/time.c @@ -64,6 +64,7 @@ static int notsc __initdata = 0; unsigned int cpu_khz; /* TSC clocks / usec, not used here */ static unsigned long hpet_period; /* fsecs / HPET clock */ unsigned long hpet_tick; /* HPET clocks / interrupt */ +static int hpet_use_timer; unsigned long vxtime_hz = PIT_TICK_RATE; int report_lost_ticks; /* command line option */ unsigned long long monotonic_base; @@ -105,7 +106,9 @@ static inline unsigned int do_gettimeoffset_tsc(void) static inline unsigned int do_gettimeoffset_hpet(void) { - return ((hpet_readl(HPET_COUNTER) - vxtime.last) * vxtime.quot) >> 32; + /* cap counter read to one tick to avoid inconsistencies */ + unsigned long counter = hpet_readl(HPET_COUNTER) - vxtime.last; + return (min(counter,hpet_tick) * vxtime.quot) >> 32; } unsigned int (*do_gettimeoffset)(void) = do_gettimeoffset_tsc; @@ -301,7 +304,7 @@ unsigned long long monotonic_clock(void) last_offset = vxtime.last; base = monotonic_base; - this_offset = hpet_readl(HPET_T0_CMP) - hpet_tick; + this_offset = hpet_readl(HPET_COUNTER); } while (read_seqretry(&xtime_lock, seq)); offset = (this_offset - last_offset); @@ -377,7 +380,14 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs) write_seqlock(&xtime_lock); - if (vxtime.hpet_address) { + if (vxtime.hpet_address) + offset = hpet_readl(HPET_COUNTER); + + if (hpet_use_timer) { + /* if we're using the hpet timer functionality, + * we can more accurately know the counter value + * when the timer interrupt occured. + */ offset = hpet_readl(HPET_T0_CMP) - hpet_tick; delay = hpet_readl(HPET_COUNTER) - offset; } else { @@ -803,17 +813,18 @@ static int hpet_timer_stop_set_go(unsigned long tick) * Set up timer 0, as periodic with first interrupt to happen at hpet_tick, * and period also hpet_tick. */ - - hpet_writel(HPET_TN_ENABLE | HPET_TN_PERIODIC | HPET_TN_SETVAL | + if (hpet_use_timer) { + hpet_writel(HPET_TN_ENABLE | HPET_TN_PERIODIC | HPET_TN_SETVAL | HPET_TN_32BIT, HPET_T0_CFG); - hpet_writel(hpet_tick, HPET_T0_CMP); - hpet_writel(hpet_tick, HPET_T0_CMP); /* AK: why twice? */ - + hpet_writel(hpet_tick, HPET_T0_CMP); + hpet_writel(hpet_tick, HPET_T0_CMP); /* AK: why twice? */ + cfg |= HPET_CFG_LEGACY; + } /* * Go! */ - cfg |= HPET_CFG_ENABLE | HPET_CFG_LEGACY; + cfg |= HPET_CFG_ENABLE; hpet_writel(cfg, HPET_CFG); return 0; @@ -834,8 +845,7 @@ static int hpet_init(void) id = hpet_readl(HPET_ID); - if (!(id & HPET_ID_VENDOR) || !(id & HPET_ID_NUMBER) || - !(id & HPET_ID_LEGSUP)) + if (!(id & HPET_ID_VENDOR) || !(id & HPET_ID_NUMBER)) return -1; hpet_period = hpet_readl(HPET_PERIOD); @@ -845,6 +855,8 @@ static int hpet_init(void) hpet_tick = (1000000000L * (USEC_PER_SEC / HZ) + hpet_period / 2) / hpet_period; + hpet_use_timer = (id & HPET_ID_LEGSUP); + return hpet_timer_stop_set_go(hpet_tick); } @@ -901,9 +913,11 @@ void __init time_init(void) set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); - if (!hpet_init()) { + if (!hpet_init()) vxtime_hz = (1000000000000000L + hpet_period / 2) / hpet_period; + + if (hpet_use_timer) { cpu_khz = hpet_calibrate_tsc(); timename = "HPET"; #ifdef CONFIG_X86_PM_TIMER @@ -968,7 +982,7 @@ void __init time_init_gtod(void) if (unsynchronized_tsc()) notsc = 1; if (vxtime.hpet_address && notsc) { - timetype = "HPET"; + timetype = hpet_use_timer ? "HPET" : "PIT/HPET"; vxtime.last = hpet_readl(HPET_T0_CMP) - hpet_tick; vxtime.mode = VXTIME_HPET; do_gettimeoffset = do_gettimeoffset_hpet; @@ -983,7 +997,7 @@ void __init time_init_gtod(void) printk(KERN_INFO "Disabling vsyscall due to use of PM timer\n"); #endif } else { - timetype = vxtime.hpet_address ? "HPET/TSC" : "PIT/TSC"; + timetype = hpet_use_timer ? "HPET/TSC" : "PIT/TSC"; vxtime.mode = VXTIME_TSC; } diff --git a/arch/x86_64/kernel/traps.c b/arch/x86_64/kernel/traps.c index 3dfec8fdabc..10273663000 100644 --- a/arch/x86_64/kernel/traps.c +++ b/arch/x86_64/kernel/traps.c @@ -274,7 +274,7 @@ EXPORT_SYMBOL(dump_stack); void show_registers(struct pt_regs *regs) { int i; - int in_kernel = (regs->cs & 3) == 0; + int in_kernel = !user_mode(regs); unsigned long rsp; const int cpu = safe_smp_processor_id(); struct task_struct *cur = cpu_pda[cpu].pcurrent; @@ -318,7 +318,7 @@ void handle_BUG(struct pt_regs *regs) struct bug_frame f; char tmp; - if (regs->cs & 3) + if (user_mode(regs)) return; if (__copy_from_user(&f, (struct bug_frame *) regs->rip, sizeof(struct bug_frame))) @@ -437,7 +437,7 @@ static void do_trap(int trapnr, int signr, char *str, } #endif - if ((regs->cs & 3) != 0) { + if (user_mode(regs)) { struct task_struct *tsk = current; if (exception_trace && unhandled_signal(tsk, signr)) @@ -522,7 +522,7 @@ asmlinkage void do_general_protection(struct pt_regs * regs, long error_code) } #endif - if ((regs->cs & 3)!=0) { + if (user_mode(regs)) { struct task_struct *tsk = current; if (exception_trace && unhandled_signal(tsk, SIGSEGV)) @@ -586,11 +586,17 @@ static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs) asmlinkage void default_do_nmi(struct pt_regs *regs) { unsigned char reason = 0; + int cpu; + + cpu = smp_processor_id(); /* Only the BSP gets external NMIs from the system. */ - if (!smp_processor_id()) + if (!cpu) reason = get_nmi_reason(); + if (!cpu_online(cpu)) + return; + if (!(reason & 0xc0)) { if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 0, SIGINT) == NOTIFY_STOP) @@ -638,7 +644,7 @@ asmlinkage struct pt_regs *sync_regs(struct pt_regs *eregs) if (eregs == (struct pt_regs *)eregs->rsp) ; /* Exception from user space */ - else if (eregs->cs & 3) + else if (user_mode(eregs)) regs = ((struct pt_regs *)current->thread.rsp0) - 1; /* Exception from kernel and interrupts are enabled. Move to kernel process stack. */ @@ -669,7 +675,7 @@ asmlinkage void do_debug(struct pt_regs * regs, unsigned long error_code) } #endif - asm("movq %%db6,%0" : "=r" (condition)); + get_debugreg(condition, 6); if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code, SIGTRAP) == NOTIFY_STOP) @@ -697,7 +703,7 @@ asmlinkage void do_debug(struct pt_regs * regs, unsigned long error_code) * allowing programs to debug themselves without the ptrace() * interface. */ - if ((regs->cs & 3) == 0) + if (!user_mode(regs)) goto clear_TF_reenable; /* * Was the TF flag set by a debugger? If so, clear it now, @@ -715,13 +721,13 @@ asmlinkage void do_debug(struct pt_regs * regs, unsigned long error_code) info.si_signo = SIGTRAP; info.si_errno = 0; info.si_code = TRAP_BRKPT; - if ((regs->cs & 3) == 0) + if (!user_mode(regs)) goto clear_dr7; info.si_addr = (void __user *)regs->rip; force_sig_info(SIGTRAP, &info, tsk); clear_dr7: - asm volatile("movq %0,%%db7"::"r"(0UL)); + set_debugreg(0UL, 7); return; clear_TF_reenable: @@ -756,7 +762,7 @@ asmlinkage void do_coprocessor_error(struct pt_regs *regs) unsigned short cwd, swd; conditional_sti(regs); - if ((regs->cs & 3) == 0 && + if (!user_mode(regs) && kernel_math_error(regs, "kernel x87 math error")) return; @@ -822,7 +828,7 @@ asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs) unsigned short mxcsr; conditional_sti(regs); - if ((regs->cs & 3) == 0 && + if (!user_mode(regs) && kernel_math_error(regs, "kernel simd math error")) return; diff --git a/arch/x86_64/kernel/vmlinux.lds.S b/arch/x86_64/kernel/vmlinux.lds.S index 59ebd5beda8..73389f51c4e 100644 --- a/arch/x86_64/kernel/vmlinux.lds.S +++ b/arch/x86_64/kernel/vmlinux.lds.S @@ -2,7 +2,10 @@ * Written by Martin Mares <mj@atrey.karlin.mff.cuni.cz>; */ +#define LOAD_OFFSET __START_KERNEL_map + #include <asm-generic/vmlinux.lds.h> +#include <asm/page.h> #include <linux/config.h> OUTPUT_FORMAT("elf64-x86-64", "elf64-x86-64", "elf64-x86-64") @@ -11,28 +14,30 @@ ENTRY(phys_startup_64) jiffies_64 = jiffies; SECTIONS { - . = 0xffffffff80100000; + . = __START_KERNEL; phys_startup_64 = startup_64 - LOAD_OFFSET; _text = .; /* Text and read-only data */ - .text : { + .text : AT(ADDR(.text) - LOAD_OFFSET) { *(.text) SCHED_TEXT LOCK_TEXT *(.fixup) *(.gnu.warning) } = 0x9090 - .text.lock : { *(.text.lock) } /* out-of-line lock text */ + /* out-of-line lock text */ + .text.lock : AT(ADDR(.text.lock) - LOAD_OFFSET) { *(.text.lock) } _etext = .; /* End of text section */ . = ALIGN(16); /* Exception table */ __start___ex_table = .; - __ex_table : { *(__ex_table) } + __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { *(__ex_table) } __stop___ex_table = .; RODATA - .data : { /* Data */ + /* Data */ + .data : AT(ADDR(.data) - LOAD_OFFSET) { *(.data) CONSTRUCTORS } @@ -40,62 +45,95 @@ SECTIONS _edata = .; /* End of data section */ __bss_start = .; /* BSS */ - .bss : { + .bss : AT(ADDR(.bss) - LOAD_OFFSET) { *(.bss.page_aligned) *(.bss) } __bss_end = .; + . = ALIGN(PAGE_SIZE); . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - .data.cacheline_aligned : { *(.data.cacheline_aligned) } + .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { + *(.data.cacheline_aligned) + } + +#define VSYSCALL_ADDR (-10*1024*1024) +#define VSYSCALL_PHYS_ADDR ((LOADADDR(.data.cacheline_aligned) + SIZEOF(.data.cacheline_aligned) + 4095) & ~(4095)) +#define VSYSCALL_VIRT_ADDR ((ADDR(.data.cacheline_aligned) + SIZEOF(.data.cacheline_aligned) + 4095) & ~(4095)) + +#define VLOAD_OFFSET (VSYSCALL_ADDR - VSYSCALL_PHYS_ADDR) +#define VLOAD(x) (ADDR(x) - VLOAD_OFFSET) + +#define VVIRT_OFFSET (VSYSCALL_ADDR - VSYSCALL_VIRT_ADDR) +#define VVIRT(x) (ADDR(x) - VVIRT_OFFSET) -#define AFTER(x) BINALIGN(LOADADDR(x) + SIZEOF(x), 16) -#define BINALIGN(x,y) (((x) + (y) - 1) & ~((y) - 1)) -#define CACHE_ALIGN(x) BINALIGN(x, CONFIG_X86_L1_CACHE_BYTES) + . = VSYSCALL_ADDR; + .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) } + __vsyscall_0 = VSYSCALL_VIRT_ADDR; - .vsyscall_0 -10*1024*1024: AT ((LOADADDR(.data.cacheline_aligned) + SIZEOF(.data.cacheline_aligned) + 4095) & ~(4095)) { *(.vsyscall_0) } - __vsyscall_0 = LOADADDR(.vsyscall_0); . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - .xtime_lock : AT CACHE_ALIGN(AFTER(.vsyscall_0)) { *(.xtime_lock) } - xtime_lock = LOADADDR(.xtime_lock); - .vxtime : AT AFTER(.xtime_lock) { *(.vxtime) } - vxtime = LOADADDR(.vxtime); - .wall_jiffies : AT AFTER(.vxtime) { *(.wall_jiffies) } - wall_jiffies = LOADADDR(.wall_jiffies); - .sys_tz : AT AFTER(.wall_jiffies) { *(.sys_tz) } - sys_tz = LOADADDR(.sys_tz); - .sysctl_vsyscall : AT AFTER(.sys_tz) { *(.sysctl_vsyscall) } - sysctl_vsyscall = LOADADDR(.sysctl_vsyscall); - .xtime : AT AFTER(.sysctl_vsyscall) { *(.xtime) } - xtime = LOADADDR(.xtime); + .xtime_lock : AT(VLOAD(.xtime_lock)) { *(.xtime_lock) } + xtime_lock = VVIRT(.xtime_lock); + + .vxtime : AT(VLOAD(.vxtime)) { *(.vxtime) } + vxtime = VVIRT(.vxtime); + + .wall_jiffies : AT(VLOAD(.wall_jiffies)) { *(.wall_jiffies) } + wall_jiffies = VVIRT(.wall_jiffies); + + .sys_tz : AT(VLOAD(.sys_tz)) { *(.sys_tz) } + sys_tz = VVIRT(.sys_tz); + + .sysctl_vsyscall : AT(VLOAD(.sysctl_vsyscall)) { *(.sysctl_vsyscall) } + sysctl_vsyscall = VVIRT(.sysctl_vsyscall); + + .xtime : AT(VLOAD(.xtime)) { *(.xtime) } + xtime = VVIRT(.xtime); + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - .jiffies : AT CACHE_ALIGN(AFTER(.xtime)) { *(.jiffies) } - jiffies = LOADADDR(.jiffies); - .vsyscall_1 ADDR(.vsyscall_0) + 1024: AT (LOADADDR(.vsyscall_0) + 1024) { *(.vsyscall_1) } - . = LOADADDR(.vsyscall_0) + 4096; + .jiffies : AT(VLOAD(.jiffies)) { *(.jiffies) } + jiffies = VVIRT(.jiffies); + + .vsyscall_1 ADDR(.vsyscall_0) + 1024: AT(VLOAD(.vsyscall_1)) { *(.vsyscall_1) } + .vsyscall_2 ADDR(.vsyscall_0) + 2048: AT(VLOAD(.vsyscall_2)) { *(.vsyscall_2) } + .vsyscall_3 ADDR(.vsyscall_0) + 3072: AT(VLOAD(.vsyscall_3)) { *(.vsyscall_3) } + + . = VSYSCALL_VIRT_ADDR + 4096; + +#undef VSYSCALL_ADDR +#undef VSYSCALL_PHYS_ADDR +#undef VSYSCALL_VIRT_ADDR +#undef VLOAD_OFFSET +#undef VLOAD +#undef VVIRT_OFFSET +#undef VVIRT . = ALIGN(8192); /* init_task */ - .data.init_task : { *(.data.init_task) } + .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { + *(.data.init_task) + } . = ALIGN(4096); - .data.page_aligned : { *(.data.page_aligned) } + .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { + *(.data.page_aligned) + } . = ALIGN(4096); /* Init code and data */ __init_begin = .; - .init.text : { + .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { _sinittext = .; *(.init.text) _einittext = .; } __initdata_begin = .; - .init.data : { *(.init.data) } + .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { *(.init.data) } __initdata_end = .; . = ALIGN(16); __setup_start = .; - .init.setup : { *(.init.setup) } + .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { *(.init.setup) } __setup_end = .; __initcall_start = .; - .initcall.init : { + .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) { *(.initcall1.init) *(.initcall2.init) *(.initcall3.init) @@ -106,32 +144,38 @@ SECTIONS } __initcall_end = .; __con_initcall_start = .; - .con_initcall.init : { *(.con_initcall.init) } + .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { + *(.con_initcall.init) + } __con_initcall_end = .; SECURITY_INIT . = ALIGN(8); __alt_instructions = .; - .altinstructions : { *(.altinstructions) } + .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { + *(.altinstructions) + } __alt_instructions_end = .; - .altinstr_replacement : { *(.altinstr_replacement) } + .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) { + *(.altinstr_replacement) + } /* .exit.text is discard at runtime, not link time, to deal with references from .altinstructions and .eh_frame */ - .exit.text : { *(.exit.text) } - .exit.data : { *(.exit.data) } + .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { *(.exit.text) } + .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { *(.exit.data) } . = ALIGN(4096); __initramfs_start = .; - .init.ramfs : { *(.init.ramfs) } + .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { *(.init.ramfs) } __initramfs_end = .; . = ALIGN(32); __per_cpu_start = .; - .data.percpu : { *(.data.percpu) } + .data.percpu : AT(ADDR(.data.percpu) - LOAD_OFFSET) { *(.data.percpu) } __per_cpu_end = .; . = ALIGN(4096); __init_end = .; . = ALIGN(4096); __nosave_begin = .; - .data_nosave : { *(.data.nosave) } + .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { *(.data.nosave) } . = ALIGN(4096); __nosave_end = .; diff --git a/arch/x86_64/lib/delay.c b/arch/x86_64/lib/delay.c index aed61a668a1..33a873a3c22 100644 --- a/arch/x86_64/lib/delay.c +++ b/arch/x86_64/lib/delay.c @@ -12,6 +12,7 @@ #include <linux/sched.h> #include <linux/delay.h> #include <asm/delay.h> +#include <asm/msr.h> #ifdef CONFIG_SMP #include <asm/smp.h> @@ -19,6 +20,12 @@ int x86_udelay_tsc = 0; /* Delay via TSC */ +int read_current_timer(unsigned long *timer_value) +{ + rdtscll(*timer_value); + return 0; +} + void __delay(unsigned long loops) { unsigned bclock, now; diff --git a/arch/x86_64/mm/Makefile b/arch/x86_64/mm/Makefile index 66c354ad80c..1d232a87f11 100644 --- a/arch/x86_64/mm/Makefile +++ b/arch/x86_64/mm/Makefile @@ -4,7 +4,7 @@ obj-y := init.o fault.o ioremap.o extable.o pageattr.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o -obj-$(CONFIG_DISCONTIGMEM) += numa.o +obj-$(CONFIG_NUMA) += numa.o obj-$(CONFIG_K8_NUMA) += k8topology.o obj-$(CONFIG_ACPI_NUMA) += srat.o diff --git a/arch/x86_64/mm/fault.c b/arch/x86_64/mm/fault.c index 57d3ab15a5c..2f187986f94 100644 --- a/arch/x86_64/mm/fault.c +++ b/arch/x86_64/mm/fault.c @@ -74,7 +74,7 @@ static noinline int is_prefetch(struct pt_regs *regs, unsigned long addr, instr = (unsigned char *)convert_rip_to_linear(current, regs); max_instr = instr + 15; - if ((regs->cs & 3) != 0 && instr >= (unsigned char *)TASK_SIZE) + if (user_mode(regs) && instr >= (unsigned char *)TASK_SIZE) return 0; while (scan_more && instr < max_instr) { @@ -106,7 +106,7 @@ static noinline int is_prefetch(struct pt_regs *regs, unsigned long addr, /* Could check the LDT for lm, but for now it's good enough to assume that long mode only uses well known segments or kernel. */ - scan_more = ((regs->cs & 3) == 0) || (regs->cs == __USER_CS); + scan_more = (!user_mode(regs)) || (regs->cs == __USER_CS); break; case 0x60: diff --git a/arch/x86_64/mm/init.c b/arch/x86_64/mm/init.c index dbe53b4c7e6..72e4b364ed7 100644 --- a/arch/x86_64/mm/init.c +++ b/arch/x86_64/mm/init.c @@ -318,7 +318,7 @@ void zap_low_mappings(void) flush_tlb_all(); } -#ifndef CONFIG_DISCONTIGMEM +#ifndef CONFIG_NUMA void __init paging_init(void) { { @@ -427,13 +427,16 @@ void __init mem_init(void) reservedpages = 0; /* this will put all low memory onto the freelists */ -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA totalram_pages += numa_free_all_bootmem(); tmp = 0; /* should count reserved pages here for all nodes */ #else + +#ifdef CONFIG_FLATMEM max_mapnr = end_pfn; if (!mem_map) BUG(); +#endif totalram_pages += free_all_bootmem(); @@ -515,7 +518,7 @@ void free_initrd_mem(unsigned long start, unsigned long end) void __init reserve_bootmem_generic(unsigned long phys, unsigned len) { /* Should check here against the e820 map to avoid double free */ -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA int nid = phys_to_nid(phys); reserve_bootmem_node(NODE_DATA(nid), phys, len); #else diff --git a/arch/x86_64/mm/ioremap.c b/arch/x86_64/mm/ioremap.c index 58aac23760e..6972df480d2 100644 --- a/arch/x86_64/mm/ioremap.c +++ b/arch/x86_64/mm/ioremap.c @@ -178,7 +178,7 @@ void __iomem * __ioremap(unsigned long phys_addr, unsigned long size, unsigned l if (phys_addr >= ISA_START_ADDRESS && last_addr < ISA_END_ADDRESS) return (__force void __iomem *)phys_to_virt(phys_addr); -#ifndef CONFIG_DISCONTIGMEM +#ifdef CONFIG_FLATMEM /* * Don't allow anybody to remap normal RAM that we're using.. */ diff --git a/arch/x86_64/mm/numa.c b/arch/x86_64/mm/numa.c index fd9f25d7a6c..ac61c186eb0 100644 --- a/arch/x86_64/mm/numa.c +++ b/arch/x86_64/mm/numa.c @@ -66,6 +66,13 @@ int __init compute_hash_shift(struct node *nodes, int numnodes) return -1; } +#ifdef CONFIG_SPARSEMEM +int early_pfn_to_nid(unsigned long pfn) +{ + return phys_to_nid(pfn << PAGE_SHIFT); +} +#endif + /* Initialize bootmem allocator for a node */ void __init setup_node_bootmem(int nodeid, unsigned long start, unsigned long end) { @@ -80,6 +87,7 @@ void __init setup_node_bootmem(int nodeid, unsigned long start, unsigned long en start_pfn = start >> PAGE_SHIFT; end_pfn = end >> PAGE_SHIFT; + memory_present(nodeid, start_pfn, end_pfn); nodedata_phys = find_e820_area(start, end, pgdat_size); if (nodedata_phys == -1L) panic("Cannot find memory pgdat in node %d\n", nodeid); @@ -243,7 +251,7 @@ void __init numa_initmem_init(unsigned long start_pfn, unsigned long end_pfn) setup_node_bootmem(0, start_pfn << PAGE_SHIFT, end_pfn << PAGE_SHIFT); } -__init void numa_add_cpu(int cpu) +__cpuinit void numa_add_cpu(int cpu) { /* BP is initialized elsewhere */ if (cpu) diff --git a/arch/x86_64/pci/k8-bus.c b/arch/x86_64/pci/k8-bus.c index 62349c78db5..7e7d0c2a002 100644 --- a/arch/x86_64/pci/k8-bus.c +++ b/arch/x86_64/pci/k8-bus.c @@ -53,25 +53,11 @@ fill_mp_bus_to_cpumask(void) for (j = SECONDARY_LDT_BUS_NUMBER(ldtbus); j <= SUBORDINATE_LDT_BUS_NUMBER(ldtbus); j++) - pci_bus_to_cpumask[j] = - node_to_cpumask(NODE_ID(nid)); + pci_bus_to_node[j] = NODE_ID(nid); } } } - /* quick sanity check */ - printed = 0; - for (i = 0; i < 256; i++) { - if (cpus_empty(pci_bus_to_cpumask[i])) { - pci_bus_to_cpumask[i] = CPU_MASK_ALL; - if (printed) - continue; - printk(KERN_ERR - "k8-bus.c: some busses have empty cpu mask\n"); - printed = 1; - } - } - return 0; } diff --git a/arch/xtensa/Kconfig b/arch/xtensa/Kconfig new file mode 100644 index 00000000000..3e89767cea7 --- /dev/null +++ b/arch/xtensa/Kconfig @@ -0,0 +1,258 @@ +# For a description of the syntax of this configuration file, +# see Documentation/kbuild/config-language.txt. + +mainmenu "Linux/Xtensa Kernel Configuration" + +config FRAME_POINTER + bool + default n + +config XTENSA + bool + default y + help + Xtensa processors are 32-bit RISC machines designed by Tensilica + primarily for embedded systems. These processors are both + configurable and extensible. The Linux port to the Xtensa + architecture supports all processor configurations and extensions, + with reasonable minimum requirements. The Xtensa Linux project has + a home page at <http://xtensa.sourceforge.net/>. + +config UID16 + bool + default n + +config RWSEM_XCHGADD_ALGORITHM + bool + default y + +config HAVE_DEC_LOCK + bool + default y + +config GENERIC_HARDIRQS + bool + default y + +source "init/Kconfig" + +menu "Processor type and features" + +choice + prompt "Xtensa Processor Configuration" + default XTENSA_CPU_LINUX_BE + +config XTENSA_CPU_LINUX_BE + bool "linux_be" + ---help--- + The linux_be processor configuration is the baseline Xtensa + configurations included in this kernel and also used by + binutils, gcc, and gdb. It contains no TIE, no coprocessors, + and the following configuration options: + + Code Density Option 2 Misc Special Registers + NSA/NSAU Instructions 128-bit Data Bus Width + Processor ID 8K, 2-way I and D Caches + Zero-Overhead Loops 2 Inst Address Break Registers + Big Endian 2 Data Address Break Registers + 64 General-Purpose Registers JTAG Interface and Trace Port + 17 Interrupts MMU w/ TLBs and Autorefill + 3 Interrupt Levels 8 Autorefill Ways (I/D TLBs) + 3 Timers Unaligned Exceptions +endchoice + +config MMU + bool + default y + +config XTENSA_UNALIGNED_USER + bool "Unaligned memory access in use space" + ---help--- + The Xtensa architecture currently does not handle unaligned + memory accesses in hardware but through an exception handler. + Per default, unaligned memory accesses are disabled in user space. + + Say Y here to enable unaligned memory access in user space. + +config PREEMPT + bool "Preemptible Kernel" + ---help--- + This option reduces the latency of the kernel when reacting to + real-time or interactive events by allowing a low priority process to + be preempted even if it is in kernel mode executing a system call. + Unfortunately the kernel code has some race conditions if both + CONFIG_SMP and CONFIG_PREEMPT are enabled, so this option is + currently disabled if you are building an SMP kernel. + + Say Y here if you are building a kernel for a desktop, embedded + or real-time system. Say N if you are unsure. + +config MATH_EMULATION + bool "Math emulation" + help + Can we use information of configuration file? + +config HIGHMEM + bool "High memory support" + +endmenu + +menu "Platform options" + +choice + prompt "Xtensa System Type" + default XTENSA_PLATFORM_ISS + +config XTENSA_PLATFORM_ISS + bool "ISS" + help + ISS is an acronym for Tensilica's Instruction Set Simulator. + +config XTENSA_PLATFORM_XT2000 + bool "XT2000" + help + XT2000 is the name of Tensilica's feature-rich emulation platform. + This hardware is capable of running a full Linux distribution. + +endchoice + + +config XTENSA_CALIBRATE_CCOUNT + bool "Auto calibration of the CPU clock rate" + ---help--- + On some platforms (XT2000, for example), the CPU clock rate can + vary. The frequency can be determined, however, by measuring + against a well known, fixed frequency, such as an UART oscillator. + +config XTENSA_CPU_CLOCK + int "CPU clock rate [MHz]" + depends on !XTENSA_CALIBRATE_CCOUNT + default "16" + +config GENERIC_CALIBRATE_DELAY + bool "Auto calibration of the BogoMIPS value" + ---help--- + The BogoMIPS value can easily derived from the CPU frequency. + +config CMDLINE_BOOL + bool "Default bootloader kernel arguments" + +config CMDLINE + string "Initial kernel command string" + depends on CMDLINE_BOOL + default "console=ttyS0,38400 root=/dev/ram" + help + On some architectures (EBSA110 and CATS), there is currently no way + for the boot loader to pass arguments to the kernel. For these + architectures, you should supply some command-line options at build + time by entering them here. As a minimum, you should specify the + memory size and the root device (e.g., mem=64M root=/dev/nfs). + +config SERIAL_CONSOLE + bool + depends on XTENSA_PLATFORM_ISS + default y + +config XTENSA_ISS_NETWORK + bool + depends on XTENSA_PLATFORM_ISS + default y + +endmenu + +menu "Bus options" + +config PCI + bool "PCI support" if !XTENSA_PLATFORM_ISS + depends on !XTENSA_PLATFORM_ISS + default y + help + Find out whether you have a PCI motherboard. PCI is the name of a + bus system, i.e. the way the CPU talks to the other stuff inside + your box. Other bus systems are ISA, EISA, MicroChannel (MCA) or + VESA. If you have PCI, say Y, otherwise N. + + The PCI-HOWTO, available from + <http://www.linuxdoc.org/docs.html#howto>, contains valuable + information about which PCI hardware does work under Linux and which + doesn't + +source "drivers/pci/Kconfig" + +config HOTPLUG + + bool "Support for hot-pluggable devices" + ---help--- + Say Y here if you want to plug devices into your computer while + the system is running, and be able to use them quickly. In many + cases, the devices can likewise be unplugged at any time too. + + One well known example of this is PCMCIA- or PC-cards, credit-card + size devices such as network cards, modems or hard drives which are + plugged into slots found on all modern laptop computers. Another + example, used on modern desktops as well as laptops, is USB. + + Enable HOTPLUG and KMOD, and build a modular kernel. Get agent + software (at <http://linux-hotplug.sourceforge.net/>) and install it. + Then your kernel will automatically call out to a user mode "policy + agent" (/sbin/hotplug) to load modules and set up software needed + to use devices as you hotplug them. + +source "drivers/pcmcia/Kconfig" + +source "drivers/pci/hotplug/Kconfig" + +endmenu + +menu "Exectuable file formats" + +# only elf supported +config KCORE_ELF + bool + depends on PROC_FS + default y + help + If you enabled support for /proc file system then the file + /proc/kcore will contain the kernel core image in ELF format. This + can be used in gdb: + + $ cd /usr/src/linux ; gdb vmlinux /proc/kcore + + This is especially useful if you have compiled the kernel with the + "-g" option to preserve debugging information. It is mainly used + for examining kernel data structures on the live kernel. + +source "fs/Kconfig.binfmt" + +endmenu + +source "drivers/Kconfig" + +source "fs/Kconfig" + +menu "Xtensa initrd options" + depends on BLK_DEV_INITRD + + config EMBEDDED_RAMDISK + bool "Embed root filesystem ramdisk into the kernel" + +config EMBEDDED_RAMDISK_IMAGE + string "Filename of gziped ramdisk image" + depends on EMBEDDED_RAMDISK + default "ramdisk.gz" + help + This is the filename of the ramdisk image to be built into the + kernel. Relative pathnames are relative to arch/xtensa/boot/ramdisk/. + The ramdisk image is not part of the kernel distribution; you must + provide one yourself. +endmenu + +source "arch/xtensa/Kconfig.debug" + +source "security/Kconfig" + +source "crypto/Kconfig" + +source "lib/Kconfig" + + diff --git a/arch/xtensa/Kconfig.debug b/arch/xtensa/Kconfig.debug new file mode 100644 index 00000000000..11c585295dd --- /dev/null +++ b/arch/xtensa/Kconfig.debug @@ -0,0 +1,7 @@ +menu "Kernel hacking" + +source "lib/Kconfig.debug" + +endmenu + + diff --git a/arch/xtensa/Makefile b/arch/xtensa/Makefile new file mode 100644 index 00000000000..4fa27453b1f --- /dev/null +++ b/arch/xtensa/Makefile @@ -0,0 +1,102 @@ +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# +# Copyright (C) 2001 - 2005 Tensilica Inc. +# +# This file is included by the global makefile so that you can add your own +# architecture-specific flags and dependencies. Remember to do have actions +# for "archclean" and "archdep" for cleaning up and making dependencies for +# this architecture + +# Core configuration. +# (Use CPU=<xtensa_config> to use another default compiler.) + +cpu-$(CONFIG_XTENSA_CPU_LINUX_BE) := linux_be +cpu-$(CONFIG_XTENSA_CPU_LINUX_CUSTOM) := linux_custom + +CPU = $(cpu-y) +export CPU + +# Platform configuration + +platform-y := common +platform-$(CONFIG_XTENSA_PLATFORM_XT2000) := xt2000 +platform-$(CONFIG_XTENSA_PLATFORM_ISS) := iss + +PLATFORM = $(platform-y) +export PLATFORM + +#LDFLAGS_vmlinux := -T$(word 1,$(LINKSCRIPT)) +AFLAGS_vmlinux.lds.o := -Uxtensa +CPPFLAGS += -Iarch/xtensa -Iinclude/asm -mlongcalls -g +AFLAGS += -Iarch/xtensa -Iinclude/asm +CPP = $(CC) -E $(CFLAGS) + +cflags-y += -Iarch/xtensa -pipe -mlongcalls + + +KBUILD_DEFCONFIG := common_defconfig + +# ramdisk/initrd support +# You need a compressed ramdisk image, named ramdisk.gz in +# arch/xtensa/boot/ramdisk + +core-$(CONFIG_EMBEDDED_RAMDISK) += arch/xtensa/boot/ramdisk/ + +# Test for cross compiling + +ifneq ($(CPU),) + COMPILE_ARCH = $(shell uname -m) + + ifneq ($(COMPILE_ARCH), xtensa) + ifndef CROSS_COMPILE + CROSS_COMPILE = xtensa_$(CPU)- + endif + endif +endif + +# + +LIBGCC := $(shell $(CC) $(CFLAGS) -print-libgcc-file-name) + +head-y := arch/xtensa/kernel/head.o +core-y += arch/xtensa/kernel/ \ + arch/xtensa/mm/ arch/xtensa/platform-$(PLATFORM)/ +libs-y += arch/xtensa/lib/ $(LIBGCC) + +boot := arch/xtensa/boot + +arch/xtensa/kernel/asm-offsets.s: \ + arch/xtensa/kernel/asm-offsets.c \ + include/asm-xtensa/.platform + +include/asm-xtensa/offsets.h: arch/xtensa/kernel/asm-offsets.s + $(call filechk,gen-asm-offsets) + +prepare: include/asm-xtensa/.platform include/asm-xtensa/offsets.h + +# Update machine cpu and platform symlinks if something which affects +# them changed. + +include/asm-xtensa/.platform: $(wildcard include/config/arch/*.h) + @echo ' Setting up cpu ($(CPU)) and platform ($(PLATFORM)) symlinks' + $(Q)rm -f include/asm-xtensa/platform + $(Q)rm -f include/asm-xtensa/xtensa/config + $(Q)(cd include/asm-xtensa/; ln -sf platform-$(PLATFORM) platform) + $(Q)(cd include/asm-xtensa/xtensa; ln -sf config-$(CPU) config) + +all: zImage + +bzImage : zImage + +zImage zImage.initrd: vmlinux + $(Q)$(MAKE) $(build)=$(boot) $@ + +CLEAN_FILES += arch/xtensa/vmlinux.lds include/asm-xtensa/offset.h + +define archhelp + @echo '* zImage - Compressed kernel image (arch/xtensa/boot/images/zImage.*)' +endef + diff --git a/arch/xtensa/boot/Makefile b/arch/xtensa/boot/Makefile new file mode 100644 index 00000000000..260f456ccf0 --- /dev/null +++ b/arch/xtensa/boot/Makefile @@ -0,0 +1,37 @@ +# +# arch/xtensa/boot/Makefile +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# +# + + +CFLAGS += -fno-builtin -Iarch/$(ARCH)/boot/include +HOSTFLAGS += -Iarch/$(ARCH)/boot/include + +BIG_ENDIAN := $(shell echo -e "\#ifdef __XTENSA_EL__\nint little;\n\#else\nint big;\n\#endif" | $(CC) -E -|grep -c big) + + +export CFLAGS +export AFLAGS +export BIG_ENDIAN + +# Subdirs for the boot loader(s) + +bootdir-$(CONFIG_XTENSA_PLATFORM_ISS) += boot-elf +bootdir-$(CONFIG_XTENSA_PLATFORM_XT2000) += boot-redboot boot-elf + +subdir-y := lib/ + +subdir-y += boot-elf/ boot-redboot/ + +zImage zImage.initrd Image Image.initrd: $(bootdir-y) + +$(bootdir-y): $(addprefix $(obj)/,$(subdir-y)) \ + $(addprefix $(obj)/,$(host-progs)) + $(Q)$(MAKE) $(build)=$(obj)/$@ $(MAKECMDGOALS) + + + diff --git a/arch/xtensa/boot/boot-elf/Makefile b/arch/xtensa/boot/boot-elf/Makefile new file mode 100644 index 00000000000..f6ef6a36966 --- /dev/null +++ b/arch/xtensa/boot/boot-elf/Makefile @@ -0,0 +1,52 @@ +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# + +GZIP = gzip +GZIP_FLAGS = -v9fc + +ifeq ($(BIG_ENDIAN),1) +OBJCOPY_ARGS := -O elf32-xtensa-be +else +OBJCOPY_ARGS := -O elf32-xtensa-le +endif + +export OBJCOPY_ARGS + +boot-y := bootstrap.o + +OBJS := $(addprefix $(obj)/,$(boot-y)) + +Image: vmlinux $(OBJS) + $(OBJCOPY) --strip-all -R .comment -R .xt.insn -O binary \ + vmlinux vmlinux.tmp + $(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \ + --add-section image=vmlinux.tmp \ + --set-section-flags image=contents,alloc,load,load,data \ + $(OBJS) $@.tmp + $(LD) $(LDFLAGS) $(LDFLAGS_vmlinux) \ + -T arch/$(ARCH)/boot/boot-elf/boot.ld \ + -o arch/$(ARCH)/boot/$@.elf $@.tmp + rm -f $@.tmp vmlinux.tmp + +Image.initrd: vmlinux $(OBJS) + $(OBJCOPY) --strip-all -R .comment -R .xt.insn -O binary \ + --add-section .initrd=arch/$(ARCH)/boot/ramdisk \ + --set-section-flags .initrd=contents,alloc,load,load,data \ + vmlinux vmlinux.tmp + $(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \ + --add-section image=vmlinux.tmp \ + --set-section-flags image=contents,alloc,load,load,data \ + $(OBJS) $@.tmp + $(LD) $(LDFLAGS) $(LDFLAGS_vmlinux) \ + -T arch/$(ARCH)/boot/boot-elf/boot.ld \ + -o arch/$(ARCH)/boot/$@.elf $@.tmp + rm -f $@.tmp vmlinux.tmp + + +zImage: Image + +zImage.initrd: Image.initrd + diff --git a/arch/xtensa/boot/boot-elf/boot.ld b/arch/xtensa/boot/boot-elf/boot.ld new file mode 100644 index 00000000000..4ab06a0a7a6 --- /dev/null +++ b/arch/xtensa/boot/boot-elf/boot.ld @@ -0,0 +1,71 @@ +OUTPUT_ARCH(xtensa) + +SECTIONS +{ + .start 0xD0000000 : { *(.start) } + + .text 0xD0000000: + { + __reloc_start = . ; + _text_start = . ; + *(.literal .text.literal .text) + _text_end = . ; + } + + .rodata ALIGN(0x04): + { + *(.rodata) + *(.rodata1) + } + + .data ALIGN(0x04): + { + *(.data) + *(.data1) + *(.sdata) + *(.sdata2) + *(.got.plt) + *(.got) + *(.dynamic) + } + + __reloc_end = . ; + + .initrd ALIGN(0x10) : + { + boot_initrd_start = . ; + *(.initrd) + boot_initrd_end = .; + } + + . = ALIGN(0x10); + __image_load = . ; + .image 0xd0001000: + { + _image_start = .; + *(image) + . = (. + 3) & ~ 3; + _image_end = . ; + } + + + .bss ((LOADADDR(.image) + SIZEOF(.image) + 3) & ~ 3): + { + __bss_start = .; + *(.sbss) + *(.scommon) + *(.dynbss) + *(.bss) + __bss_end = .; + } + _end = .; + _param_start = .; + + .ResetVector.text 0xfe000020 : + { + *(.ResetVector.text) + } + + + PROVIDE (end = .); +} diff --git a/arch/xtensa/boot/boot-elf/bootstrap.S b/arch/xtensa/boot/boot-elf/bootstrap.S new file mode 100644 index 00000000000..7cba94abdab --- /dev/null +++ b/arch/xtensa/boot/boot-elf/bootstrap.S @@ -0,0 +1,37 @@ + +#include <xtensa/config/specreg.h> +#include <xtensa/config/core.h> + +#include <linux/config.h> +#include <asm/bootparam.h> + + +/* ResetVector + */ + .section .ResetVector.text, "ax" + .global _ResetVector +_ResetVector: + _j reset + .align 4 +RomInitAddr: + .word 0xd0001000 +RomBootParam: + .word _bootparam +reset: + l32r a0, RomInitAddr + l32r a2, RomBootParam + movi a3, 0 + movi a4, 0 + jx a0 + + .align 4 + .section .bootstrap.data, "aw" + + .globl _bootparam +_bootparam: + .short BP_TAG_FIRST + .short 4 + .long BP_VERSION + .short BP_TAG_LAST + .short 0 + .long 0 diff --git a/arch/xtensa/boot/boot-redboot/Makefile b/arch/xtensa/boot/boot-redboot/Makefile new file mode 100644 index 00000000000..ca8a68bc847 --- /dev/null +++ b/arch/xtensa/boot/boot-redboot/Makefile @@ -0,0 +1,35 @@ +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# + +GZIP = gzip +GZIP_FLAGS = -v9fc +ifeq ($(BIG_ENDIAN),1) +OBJCOPY_ARGS := -O elf32-xtensa-be +else +OBJCOPY_ARGS := -O elf32-xtensa-le +endif + +LD_ARGS = -T $(obj)/boot.ld + +boot-y := bootstrap.o + +OBJS := $(addprefix $(obj)/,$(boot-y)) +LIBS := arch/$(ARCH)/boot/lib/lib.a arch/$(ARCH)/lib/lib.a + +LIBGCC := $(shell $(CC) $(CFLAGS) -print-libgcc-file-name) + +zImage: vmlinux $(OBJS) $(LIBS) + $(OBJCOPY) --strip-all -R .comment -R .xt.insn -O binary \ + $(TOPDIR)/vmlinux vmlinux.tmp + gzip -vf9 vmlinux.tmp + $(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \ + --add-section image=vmlinux.tmp.gz \ + --set-section-flags image=contents,alloc,load,load,data \ + $(OBJS) $@.tmp + $(LD) $(LD_ARGS) -o $@.elf $@.tmp $(LIBS) -L/xtensa-elf/lib $(LIBGCC) + $(OBJCOPY) -S -O binary $@.elf arch/$(ARCH)/boot/images/$@.redboot +# rm -f $@.tmp $@.elf vmlinux.tmp.gz + diff --git a/arch/xtensa/boot/boot-redboot/boot.ld b/arch/xtensa/boot/boot-redboot/boot.ld new file mode 100644 index 00000000000..65b726410e8 --- /dev/null +++ b/arch/xtensa/boot/boot-redboot/boot.ld @@ -0,0 +1,66 @@ +OUTPUT_ARCH(xtensa) + +SECTIONS +{ + .start 0xD0200000 : { *(.start) } + + .text : + { + __reloc_start = . ; + _text_start = . ; + *(.literal .text.literal .text) + _text_end = . ; + } + + .rodata ALIGN(0x04): + { + *(.rodata) + *(.rodata1) + } + + .data ALIGN(0x04): + { + *(.data) + *(.data1) + *(.sdata) + *(.sdata2) + *(.got.plt) + *(.got) + *(.dynamic) + } + + __reloc_end = . ; + + .initrd ALIGN(0x10) : + { + boot_initrd_start = . ; + *(.initrd) + boot_initrd_end = .; + } + + . = ALIGN(0x10); + __image_load = . ; + .image 0xd0001000: AT(__image_load) + { + _image_start = .; + *(image) + . = (. + 3) & ~ 3; + _image_end = . ; + } + + + .bss ((LOADADDR(.image) + SIZEOF(.image) + 3) & ~ 3): + { + __bss_start = .; + *(.sbss) + *(.scommon) + *(.dynbss) + *(.bss) + __bss_end = .; + } + _end = .; + _param_start = .; + + + PROVIDE (end = .); +} diff --git a/arch/xtensa/boot/boot-redboot/bootstrap.S b/arch/xtensa/boot/boot-redboot/bootstrap.S new file mode 100644 index 00000000000..ee636b0da81 --- /dev/null +++ b/arch/xtensa/boot/boot-redboot/bootstrap.S @@ -0,0 +1,246 @@ + +#define _ASMLANGUAGE +#include <xtensa/config/specreg.h> +#include <xtensa/config/core.h> +#include <xtensa/cacheasm.h> + + /* + * RB-Data: RedBoot data/bss + * P: Boot-Parameters + * L: Kernel-Loader + * + * The Linux-Kernel image including the loader must be loaded + * to a position so that the kernel and the boot parameters + * can fit in the space before the load address. + * ______________________________________________________ + * |_RB-Data_|_P_|__________|_L_|___Linux-Kernel___|______| + * ^ + * ^ Load address + * ______________________________________________________ + * |___Linux-Kernel___|_P_|_L_|___________________________| + * + * The loader copies the parameter to the position that will + * be the end of the kernel and itself to the end of the + * parameter list. + */ + +/* Make sure we have enough space for the 'uncompressor' */ + +#define STACK_SIZE 32768 +#define HEAP_SIZE (131072*4) + + # a2: Parameter list + # a3: Size of parameter list + + .section .start, "ax" + + .globl __start + /* this must be the first byte of the loader! */ +__start: + entry sp, 32 # we do not intend to return + _call0 _start +__start_a0: + .align 4 + + .section .text, "ax" + .begin literal_prefix .text + + /* put literals in here! */ + + .globl _start +_start: + + /* 'reset' window registers */ + + movi a4, 1 + wsr a4, PS + rsync + + rsr a5, WINDOWBASE + ssl a5 + sll a4, a4 + wsr a4, WINDOWSTART + rsync + + movi a4, 0x00040000 + wsr a4, PS + rsync + + /* copy the loader to its address + * Note: The loader itself is a very small piece, so we assume we + * don't partially overlap. We also assume (even more important) + * that the kernel image is out of the way. Usually, when the + * load address of this image is not at an arbitrary address, + * but aligned to some 10K's we shouldn't overlap. + */ + + /* Note: The assembler cannot relax "addi a0, a0, ..." to an + l32r, so we load to a4 first. */ + + addi a4, a0, __start - __start_a0 + mov a0, a4 + movi a4, __start + movi a5, __reloc_end + + # a0: address where this code has been loaded + # a4: compiled address of __start + # a5: compiled end address + + mov.n a7, a0 + mov.n a8, a4 + +1: + l32i a10, a7, 0 + l32i a11, a7, 4 + s32i a10, a8, 0 + s32i a11, a8, 4 + l32i a10, a7, 8 + l32i a11, a7, 12 + s32i a10, a8, 8 + s32i a11, a8, 12 + addi a8, a8, 16 + addi a7, a7, 16 + blt a8, a5, 1b + + + /* We have to flush and invalidate the caches here before we jump. */ + +#if XCHAL_DCACHE_IS_WRITEBACK + dcache_writeback_all a5, a6 +#endif + icache_invalidate_all a5, a6 + + movi a11, _reloc + jx a11 + + .globl _reloc +_reloc: + + /* RedBoot is now at the end of the memory, so we don't have + * to copy the parameter list. Keep the code around; in case + * we need it again. */ +#if 0 + # a0: load address + # a2: start address of parameter list + # a3: length of parameter list + # a4: __start + + /* copy the parameter list out of the way */ + + movi a6, _param_start + add a3, a2, a3 +2: + l32i a8, a2, 0 + s32i a8, a6, 0 + addi a2, a2, 4 + addi a6, a6, 4 + blt a2, a3, 2b +#endif + + /* clear BSS section */ + movi a6, __bss_start + movi a7, __bss_end + movi.n a5, 0 +3: + s32i a5, a6, 0 + addi a6, a6, 4 + blt a6, a7, 3b + + movi a5, -16 + movi a1, _stack + STACK_SIZE + and a1, a1, a5 + + /* Uncompress the kernel */ + + # a0: load address + # a2: boot parameter + # a4: __start + + movi a3, __image_load + sub a4, a3, a4 + add a8, a0, a4 + + # a1 Stack + # a8(a4) Load address of the image + + movi a6, _image_start + movi a10, _image_end + movi a7, 0x1000000 + sub a11, a10, a6 + movi a9, complen + s32i a11, a9, 0 + + movi a0, 0 + + # a6 destination + # a7 maximum size of destination + # a8 source + # a9 ptr to length + + .extern gunzip + movi a4, gunzip + beqz a4, 1f + + callx4 a4 + + j 2f + + + # a6 destination start + # a7 maximum size of destination + # a8 source start + # a9 ptr to length + # a10 destination end + +1: + l32i a9, a8, 0 + l32i a11, a8, 4 + s32i a9, a6, 0 + s32i a11, a6, 4 + l32i a9, a8, 8 + l32i a11, a8, 12 + s32i a9, a6, 8 + s32i a11, a6, 12 + addi a6, a6, 16 + addi a8, a8, 16 + blt a6, a10, 1b + + + /* jump to the kernel */ +2: +#if XCHAL_DCACHE_IS_WRITEBACK + dcache_writeback_all a5, a6 +#endif + icache_invalidate_all a5, a6 + + movi a5, __start + movi a3, boot_initrd_start + movi a4, boot_initrd_end + sub a3, a3, a5 + sub a4, a4, a5 + add a3, a0, a3 + add a4, a0, a4 + + # a2 Boot parameter list + # a3 initrd_start (virtual load address) + # a4 initrd_end (virtual load address) + + movi a0, _image_start + jx a0 + + .align 16 + .data + .globl avail_ram +avail_ram: + .long _heap + .globl end_avail +end_avail: + .long _heap + HEAP_SIZE + + .comm _stack, STACK_SIZE + .comm _heap, HEAP_SIZE + + .globl end_avail + .comm complen, 4 + + .end literal_prefix diff --git a/arch/xtensa/boot/include/zlib.h b/arch/xtensa/boot/include/zlib.h new file mode 100644 index 00000000000..ea29b623785 --- /dev/null +++ b/arch/xtensa/boot/include/zlib.h @@ -0,0 +1,433 @@ +/* + * BK Id: SCCS/s.zlib.h 1.8 05/18/01 15:17:23 cort + */ +/* + * This file is derived from zlib.h and zconf.h from the zlib-0.95 + * distribution by Jean-loup Gailly and Mark Adler, with some additions + * by Paul Mackerras to aid in implementing Deflate compression and + * decompression for PPP packets. + */ + +/* + * ==FILEVERSION 960122== + * + * This marker is used by the Linux installation script to determine + * whether an up-to-date version of this file is already installed. + */ + +/* zlib.h -- interface of the 'zlib' general purpose compression library + version 0.95, Aug 16th, 1995. + + Copyright (C) 1995 Jean-loup Gailly and Mark Adler + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + Jean-loup Gailly Mark Adler + gzip@prep.ai.mit.edu madler@alumni.caltech.edu + */ + +#ifndef _ZLIB_H +#define _ZLIB_H + +/* #include "zconf.h" */ /* included directly here */ + +/* zconf.h -- configuration of the zlib compression library + * Copyright (C) 1995 Jean-loup Gailly. + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* From: zconf.h,v 1.12 1995/05/03 17:27:12 jloup Exp */ + +/* + The library does not install any signal handler. It is recommended to + add at least a handler for SIGSEGV when decompressing; the library checks + the consistency of the input data whenever possible but may go nuts + for some forms of corrupted input. + */ + +/* + * Compile with -DMAXSEG_64K if the alloc function cannot allocate more + * than 64k bytes at a time (needed on systems with 16-bit int). + * Compile with -DUNALIGNED_OK if it is OK to access shorts or ints + * at addresses which are not a multiple of their size. + * Under DOS, -DFAR=far or -DFAR=__far may be needed. + */ + +#ifndef STDC +# if defined(MSDOS) || defined(__STDC__) || defined(__cplusplus) +# define STDC +# endif +#endif + +#ifdef __MWERKS__ /* Metrowerks CodeWarrior declares fileno() in unix.h */ +# include <unix.h> +#endif + +/* Maximum value for memLevel in deflateInit2 */ +#ifndef MAX_MEM_LEVEL +# ifdef MAXSEG_64K +# define MAX_MEM_LEVEL 8 +# else +# define MAX_MEM_LEVEL 9 +# endif +#endif + +#ifndef FAR +# define FAR +#endif + +/* Maximum value for windowBits in deflateInit2 and inflateInit2 */ +#ifndef MAX_WBITS +# define MAX_WBITS 15 /* 32K LZ77 window */ +#endif + +/* The memory requirements for deflate are (in bytes): + 1 << (windowBits+2) + 1 << (memLevel+9) + that is: 128K for windowBits=15 + 128K for memLevel = 8 (default values) + plus a few kilobytes for small objects. For example, if you want to reduce + the default memory requirements from 256K to 128K, compile with + make CFLAGS="-O -DMAX_WBITS=14 -DMAX_MEM_LEVEL=7" + Of course this will generally degrade compression (there's no free lunch). + + The memory requirements for inflate are (in bytes) 1 << windowBits + that is, 32K for windowBits=15 (default value) plus a few kilobytes + for small objects. +*/ + + /* Type declarations */ + +#ifndef OF /* function prototypes */ +# ifdef STDC +# define OF(args) args +# else +# define OF(args) () +# endif +#endif + +typedef unsigned char Byte; /* 8 bits */ +typedef unsigned int uInt; /* 16 bits or more */ +typedef unsigned long uLong; /* 32 bits or more */ + +typedef Byte FAR Bytef; +typedef char FAR charf; +typedef int FAR intf; +typedef uInt FAR uIntf; +typedef uLong FAR uLongf; + +#ifdef STDC + typedef void FAR *voidpf; + typedef void *voidp; +#else + typedef Byte FAR *voidpf; + typedef Byte *voidp; +#endif + +/* end of original zconf.h */ + +#define ZLIB_VERSION "0.95P" + +/* + The 'zlib' compression library provides in-memory compression and + decompression functions, including integrity checks of the uncompressed + data. This version of the library supports only one compression method + (deflation) but other algorithms may be added later and will have the same + stream interface. + + For compression the application must provide the output buffer and + may optionally provide the input buffer for optimization. For decompression, + the application must provide the input buffer and may optionally provide + the output buffer for optimization. + + Compression can be done in a single step if the buffers are large + enough (for example if an input file is mmap'ed), or can be done by + repeated calls of the compression function. In the latter case, the + application must provide more input and/or consume the output + (providing more output space) before each call. +*/ + +typedef voidpf (*alloc_func) OF((voidpf opaque, uInt items, uInt size)); +typedef void (*free_func) OF((voidpf opaque, voidpf address, uInt nbytes)); + +struct internal_state; + +typedef struct z_stream_s { + Bytef *next_in; /* next input byte */ + uInt avail_in; /* number of bytes available at next_in */ + uLong total_in; /* total nb of input bytes read so far */ + + Bytef *next_out; /* next output byte should be put there */ + uInt avail_out; /* remaining free space at next_out */ + uLong total_out; /* total nb of bytes output so far */ + + char *msg; /* last error message, NULL if no error */ + struct internal_state FAR *state; /* not visible by applications */ + + alloc_func zalloc; /* used to allocate the internal state */ + free_func zfree; /* used to free the internal state */ + voidp opaque; /* private data object passed to zalloc and zfree */ + + Byte data_type; /* best guess about the data type: ascii or binary */ + +} z_stream; + +/* + The application must update next_in and avail_in when avail_in has + dropped to zero. It must update next_out and avail_out when avail_out + has dropped to zero. The application must initialize zalloc, zfree and + opaque before calling the init function. All other fields are set by the + compression library and must not be updated by the application. + + The opaque value provided by the application will be passed as the first + parameter for calls of zalloc and zfree. This can be useful for custom + memory management. The compression library attaches no meaning to the + opaque value. + + zalloc must return Z_NULL if there is not enough memory for the object. + On 16-bit systems, the functions zalloc and zfree must be able to allocate + exactly 65536 bytes, but will not be required to allocate more than this + if the symbol MAXSEG_64K is defined (see zconf.h). WARNING: On MSDOS, + pointers returned by zalloc for objects of exactly 65536 bytes *must* + have their offset normalized to zero. The default allocation function + provided by this library ensures this (see zutil.c). To reduce memory + requirements and avoid any allocation of 64K objects, at the expense of + compression ratio, compile the library with -DMAX_WBITS=14 (see zconf.h). + + The fields total_in and total_out can be used for statistics or + progress reports. After compression, total_in holds the total size of + the uncompressed data and may be saved for use in the decompressor + (particularly if the decompressor wants to decompress everything in + a single step). +*/ + + /* constants */ + +#define Z_NO_FLUSH 0 +#define Z_PARTIAL_FLUSH 1 +#define Z_FULL_FLUSH 2 +#define Z_SYNC_FLUSH 3 /* experimental: partial_flush + byte align */ +#define Z_FINISH 4 +#define Z_PACKET_FLUSH 5 +/* See deflate() below for the usage of these constants */ + +#define Z_OK 0 +#define Z_STREAM_END 1 +#define Z_ERRNO (-1) +#define Z_STREAM_ERROR (-2) +#define Z_DATA_ERROR (-3) +#define Z_MEM_ERROR (-4) +#define Z_BUF_ERROR (-5) +/* error codes for the compression/decompression functions */ + +#define Z_BEST_SPEED 1 +#define Z_BEST_COMPRESSION 9 +#define Z_DEFAULT_COMPRESSION (-1) +/* compression levels */ + +#define Z_FILTERED 1 +#define Z_HUFFMAN_ONLY 2 +#define Z_DEFAULT_STRATEGY 0 + +#define Z_BINARY 0 +#define Z_ASCII 1 +#define Z_UNKNOWN 2 +/* Used to set the data_type field */ + +#define Z_NULL 0 /* for initializing zalloc, zfree, opaque */ + +extern char *zlib_version; +/* The application can compare zlib_version and ZLIB_VERSION for consistency. + If the first character differs, the library code actually used is + not compatible with the zlib.h header file used by the application. + */ + + /* basic functions */ + +extern int inflateInit OF((z_stream *strm)); +/* + Initializes the internal stream state for decompression. The fields + zalloc and zfree must be initialized before by the caller. If zalloc and + zfree are set to Z_NULL, inflateInit updates them to use default allocation + functions. + + inflateInit returns Z_OK if success, Z_MEM_ERROR if there was not + enough memory. msg is set to null if there is no error message. + inflateInit does not perform any decompression: this will be done by + inflate(). +*/ + + +extern int inflate OF((z_stream *strm, int flush)); +/* + Performs one or both of the following actions: + + - Decompress more input starting at next_in and update next_in and avail_in + accordingly. If not all input can be processed (because there is not + enough room in the output buffer), next_in is updated and processing + will resume at this point for the next call of inflate(). + + - Provide more output starting at next_out and update next_out and avail_out + accordingly. inflate() always provides as much output as possible + (until there is no more input data or no more space in the output buffer). + + Before the call of inflate(), the application should ensure that at least + one of the actions is possible, by providing more input and/or consuming + more output, and updating the next_* and avail_* values accordingly. + The application can consume the uncompressed output when it wants, for + example when the output buffer is full (avail_out == 0), or after each + call of inflate(). + + If the parameter flush is set to Z_PARTIAL_FLUSH or Z_PACKET_FLUSH, + inflate flushes as much output as possible to the output buffer. The + flushing behavior of inflate is not specified for values of the flush + parameter other than Z_PARTIAL_FLUSH, Z_PACKET_FLUSH or Z_FINISH, but the + current implementation actually flushes as much output as possible + anyway. For Z_PACKET_FLUSH, inflate checks that once all the input data + has been consumed, it is expecting to see the length field of a stored + block; if not, it returns Z_DATA_ERROR. + + inflate() should normally be called until it returns Z_STREAM_END or an + error. However if all decompression is to be performed in a single step + (a single call of inflate), the parameter flush should be set to + Z_FINISH. In this case all pending input is processed and all pending + output is flushed; avail_out must be large enough to hold all the + uncompressed data. (The size of the uncompressed data may have been saved + by the compressor for this purpose.) The next operation on this stream must + be inflateEnd to deallocate the decompression state. The use of Z_FINISH + is never required, but can be used to inform inflate that a faster routine + may be used for the single inflate() call. + + inflate() returns Z_OK if some progress has been made (more input + processed or more output produced), Z_STREAM_END if the end of the + compressed data has been reached and all uncompressed output has been + produced, Z_DATA_ERROR if the input data was corrupted, Z_STREAM_ERROR if + the stream structure was inconsistent (for example if next_in or next_out + was NULL), Z_MEM_ERROR if there was not enough memory, Z_BUF_ERROR if no + progress is possible or if there was not enough room in the output buffer + when Z_FINISH is used. In the Z_DATA_ERROR case, the application may then + call inflateSync to look for a good compression block. */ + + +extern int inflateEnd OF((z_stream *strm)); +/* + All dynamically allocated data structures for this stream are freed. + This function discards any unprocessed input and does not flush any + pending output. + + inflateEnd returns Z_OK if success, Z_STREAM_ERROR if the stream state + was inconsistent. In the error case, msg may be set but then points to a + static string (which must not be deallocated). +*/ + + /* advanced functions */ + +extern int inflateInit2 OF((z_stream *strm, + int windowBits)); +/* + This is another version of inflateInit with more compression options. The + fields next_out, zalloc and zfree must be initialized before by the caller. + + The windowBits parameter is the base two logarithm of the maximum window + size (the size of the history buffer). It should be in the range 8..15 for + this version of the library (the value 16 will be allowed soon). The + default value is 15 if inflateInit is used instead. If a compressed stream + with a larger window size is given as input, inflate() will return with + the error code Z_DATA_ERROR instead of trying to allocate a larger window. + + If next_out is not null, the library will use this buffer for the history + buffer; the buffer must either be large enough to hold the entire output + data, or have at least 1<<windowBits bytes. If next_out is null, the + library will allocate its own buffer (and leave next_out null). next_in + need not be provided here but must be provided by the application for the + next call of inflate(). + + If the history buffer is provided by the application, next_out must + never be changed by the application since the decompressor maintains + history information inside this buffer from call to call; the application + can only reset next_out to the beginning of the history buffer when + avail_out is zero and all output has been consumed. + + inflateInit2 returns Z_OK if success, Z_MEM_ERROR if there was + not enough memory, Z_STREAM_ERROR if a parameter is invalid (such as + windowBits < 8). msg is set to null if there is no error message. + inflateInit2 does not perform any decompression: this will be done by + inflate(). +*/ + +extern int inflateSync OF((z_stream *strm)); +/* + Skips invalid compressed data until the special marker (see deflate() + above) can be found, or until all available input is skipped. No output + is provided. + + inflateSync returns Z_OK if the special marker has been found, Z_BUF_ERROR + if no more input was provided, Z_DATA_ERROR if no marker has been found, + or Z_STREAM_ERROR if the stream structure was inconsistent. In the success + case, the application may save the current current value of total_in which + indicates where valid compressed data was found. In the error case, the + application may repeatedly call inflateSync, providing more input each time, + until success or end of the input data. +*/ + +extern int inflateReset OF((z_stream *strm)); +/* + This function is equivalent to inflateEnd followed by inflateInit, + but does not free and reallocate all the internal decompression state. + The stream will keep attributes that may have been set by inflateInit2. + + inflateReset returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent (such as zalloc or state being NULL). +*/ + +extern int inflateIncomp OF((z_stream *strm)); +/* + This function adds the data at next_in (avail_in bytes) to the output + history without performing any output. There must be no pending output, + and the decompressor must be expecting to see the start of a block. + Calling this function is equivalent to decompressing a stored block + containing the data at next_in (except that the data is not output). +*/ + + /* checksum functions */ + +/* + This function is not related to compression but is exported + anyway because it might be useful in applications using the + compression library. +*/ + +extern uLong adler32 OF((uLong adler, Bytef *buf, uInt len)); + +/* + Update a running Adler-32 checksum with the bytes buf[0..len-1] and + return the updated checksum. If buf is NULL, this function returns + the required initial value for the checksum. + An Adler-32 checksum is almost as reliable as a CRC32 but can be computed + much faster. Usage example: + + uLong adler = adler32(0L, Z_NULL, 0); + + while (read_buffer(buffer, length) != EOF) { + adler = adler32(adler, buffer, length); + } + if (adler != original_adler) error(); +*/ + +#ifndef _Z_UTIL_H + struct internal_state {int dummy;}; /* hack for buggy compilers */ +#endif + +#endif /* _ZLIB_H */ diff --git a/arch/xtensa/boot/lib/Makefile b/arch/xtensa/boot/lib/Makefile new file mode 100644 index 00000000000..c0a74dc3a0d --- /dev/null +++ b/arch/xtensa/boot/lib/Makefile @@ -0,0 +1,6 @@ +# +# Makefile for some libs needed by zImage. +# + + +lib-y := zlib.o zmem.o diff --git a/arch/xtensa/boot/lib/memcpy.S b/arch/xtensa/boot/lib/memcpy.S new file mode 100644 index 00000000000..a029f5df2d5 --- /dev/null +++ b/arch/xtensa/boot/lib/memcpy.S @@ -0,0 +1,36 @@ +/* + * arch/xtensa/lib/memcpy.S + * + * ANSI C standard library function memcpy + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Copyright (C) 2002 Tensilica Inc. + */ + +#define _ASMLANGUAGE +#include <xtensa/config/core.h> + +.text +.align 4 +.global bcopy +.type bcopy,@function +bcopy: + movi a14, xthal_bcopy // a14 safe to use regardless of whether caller + // used call4 or call8 (can't have used call12) + jx a14 // let the Core HAL do the work + +.text +.align 4 +.global memcpy +.type memcpy,@function +memcpy: +.global memmove +.type memmove,@function +memmove: + movi a14, xthal_memcpy // a14 safe to use regardless of whether caller + // used call4 or call8 (can't have used call12) + jx a14 // let the Core HAL do the work + diff --git a/arch/xtensa/boot/lib/zlib.c b/arch/xtensa/boot/lib/zlib.c new file mode 100644 index 00000000000..e3859f63107 --- /dev/null +++ b/arch/xtensa/boot/lib/zlib.c @@ -0,0 +1,2150 @@ +/* + * BK Id: SCCS/s.zlib.c 1.8 05/18/01 15:17:24 cort + */ +/* + * This file is derived from various .h and .c files from the zlib-0.95 + * distribution by Jean-loup Gailly and Mark Adler, with some additions + * by Paul Mackerras to aid in implementing Deflate compression and + * decompression for PPP packets. See zlib.h for conditions of + * distribution and use. + * + * Changes that have been made include: + * - changed functions not used outside this file to "local" + * - added minCompression parameter to deflateInit2 + * - added Z_PACKET_FLUSH (see zlib.h for details) + * - added inflateIncomp + * + */ + +/*+++++*/ +/* zutil.h -- internal interface and configuration of the compression library + * Copyright (C) 1995 Jean-loup Gailly. + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +/* From: zutil.h,v 1.9 1995/05/03 17:27:12 jloup Exp */ + +#define _Z_UTIL_H + +#include "zlib.h" + +#ifndef local +# define local static +#endif +/* compile with -Dlocal if your debugger can't find static symbols */ + +#define FAR + +typedef unsigned char uch; +typedef uch FAR uchf; +typedef unsigned short ush; +typedef ush FAR ushf; +typedef unsigned long ulg; + +extern char *z_errmsg[]; /* indexed by 1-zlib_error */ + +#define ERR_RETURN(strm,err) return (strm->msg=z_errmsg[1-err], err) +/* To be used only when the state is known to be valid */ + +#ifndef NULL +#define NULL ((void *) 0) +#endif + + /* common constants */ + +#define DEFLATED 8 + +#ifndef DEF_WBITS +# define DEF_WBITS MAX_WBITS +#endif +/* default windowBits for decompression. MAX_WBITS is for compression only */ + +#if MAX_MEM_LEVEL >= 8 +# define DEF_MEM_LEVEL 8 +#else +# define DEF_MEM_LEVEL MAX_MEM_LEVEL +#endif +/* default memLevel */ + +#define STORED_BLOCK 0 +#define STATIC_TREES 1 +#define DYN_TREES 2 +/* The three kinds of block type */ + +#define MIN_MATCH 3 +#define MAX_MATCH 258 +/* The minimum and maximum match lengths */ + + /* functions */ + +#include <linux/string.h> +#define zmemcpy memcpy +#define zmemzero(dest, len) memset(dest, 0, len) + +/* Diagnostic functions */ +#ifdef DEBUG_ZLIB +# include <stdio.h> +# ifndef verbose +# define verbose 0 +# endif +# define Assert(cond,msg) {if(!(cond)) z_error(msg);} +# define Trace(x) fprintf x +# define Tracev(x) {if (verbose) fprintf x ;} +# define Tracevv(x) {if (verbose>1) fprintf x ;} +# define Tracec(c,x) {if (verbose && (c)) fprintf x ;} +# define Tracecv(c,x) {if (verbose>1 && (c)) fprintf x ;} +#else +# define Assert(cond,msg) +# define Trace(x) +# define Tracev(x) +# define Tracevv(x) +# define Tracec(c,x) +# define Tracecv(c,x) +#endif + + +typedef uLong (*check_func) OF((uLong check, Bytef *buf, uInt len)); + +/* voidpf zcalloc OF((voidpf opaque, unsigned items, unsigned size)); */ +/* void zcfree OF((voidpf opaque, voidpf ptr)); */ + +#define ZALLOC(strm, items, size) \ + (*((strm)->zalloc))((strm)->opaque, (items), (size)) +#define ZFREE(strm, addr, size) \ + (*((strm)->zfree))((strm)->opaque, (voidpf)(addr), (size)) +#define TRY_FREE(s, p, n) {if (p) ZFREE(s, p, n);} + +/* deflate.h -- internal compression state + * Copyright (C) 1995 Jean-loup Gailly + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +/*+++++*/ +/* infblock.h -- header to use infblock.c + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +struct inflate_blocks_state; +typedef struct inflate_blocks_state FAR inflate_blocks_statef; + +local inflate_blocks_statef * inflate_blocks_new OF(( + z_stream *z, + check_func c, /* check function */ + uInt w)); /* window size */ + +local int inflate_blocks OF(( + inflate_blocks_statef *, + z_stream *, + int)); /* initial return code */ + +local void inflate_blocks_reset OF(( + inflate_blocks_statef *, + z_stream *, + uLongf *)); /* check value on output */ + +local int inflate_blocks_free OF(( + inflate_blocks_statef *, + z_stream *, + uLongf *)); /* check value on output */ + +local int inflate_addhistory OF(( + inflate_blocks_statef *, + z_stream *)); + +local int inflate_packet_flush OF(( + inflate_blocks_statef *)); + +/*+++++*/ +/* inftrees.h -- header to use inftrees.c + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +/* Huffman code lookup table entry--this entry is four bytes for machines + that have 16-bit pointers (e.g. PC's in the small or medium model). */ + +typedef struct inflate_huft_s FAR inflate_huft; + +struct inflate_huft_s { + union { + struct { + Byte Exop; /* number of extra bits or operation */ + Byte Bits; /* number of bits in this code or subcode */ + } what; + uInt Nalloc; /* number of these allocated here */ + Bytef *pad; /* pad structure to a power of 2 (4 bytes for */ + } word; /* 16-bit, 8 bytes for 32-bit machines) */ + union { + uInt Base; /* literal, length base, or distance base */ + inflate_huft *Next; /* pointer to next level of table */ + } more; +}; + +#ifdef DEBUG_ZLIB + local uInt inflate_hufts; +#endif + +local int inflate_trees_bits OF(( + uIntf *, /* 19 code lengths */ + uIntf *, /* bits tree desired/actual depth */ + inflate_huft * FAR *, /* bits tree result */ + z_stream *)); /* for zalloc, zfree functions */ + +local int inflate_trees_dynamic OF(( + uInt, /* number of literal/length codes */ + uInt, /* number of distance codes */ + uIntf *, /* that many (total) code lengths */ + uIntf *, /* literal desired/actual bit depth */ + uIntf *, /* distance desired/actual bit depth */ + inflate_huft * FAR *, /* literal/length tree result */ + inflate_huft * FAR *, /* distance tree result */ + z_stream *)); /* for zalloc, zfree functions */ + +local int inflate_trees_fixed OF(( + uIntf *, /* literal desired/actual bit depth */ + uIntf *, /* distance desired/actual bit depth */ + inflate_huft * FAR *, /* literal/length tree result */ + inflate_huft * FAR *)); /* distance tree result */ + +local int inflate_trees_free OF(( + inflate_huft *, /* tables to free */ + z_stream *)); /* for zfree function */ + + +/*+++++*/ +/* infcodes.h -- header to use infcodes.c + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +struct inflate_codes_state; +typedef struct inflate_codes_state FAR inflate_codes_statef; + +local inflate_codes_statef *inflate_codes_new OF(( + uInt, uInt, + inflate_huft *, inflate_huft *, + z_stream *)); + +local int inflate_codes OF(( + inflate_blocks_statef *, + z_stream *, + int)); + +local void inflate_codes_free OF(( + inflate_codes_statef *, + z_stream *)); + + +/*+++++*/ +/* inflate.c -- zlib interface to inflate modules + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* inflate private state */ +struct internal_state { + + /* mode */ + enum { + METHOD, /* waiting for method byte */ + FLAG, /* waiting for flag byte */ + BLOCKS, /* decompressing blocks */ + CHECK4, /* four check bytes to go */ + CHECK3, /* three check bytes to go */ + CHECK2, /* two check bytes to go */ + CHECK1, /* one check byte to go */ + DONE, /* finished check, done */ + BAD} /* got an error--stay here */ + mode; /* current inflate mode */ + + /* mode dependent information */ + union { + uInt method; /* if FLAGS, method byte */ + struct { + uLong was; /* computed check value */ + uLong need; /* stream check value */ + } check; /* if CHECK, check values to compare */ + uInt marker; /* if BAD, inflateSync's marker bytes count */ + } sub; /* submode */ + + /* mode independent information */ + int nowrap; /* flag for no wrapper */ + uInt wbits; /* log2(window size) (8..15, defaults to 15) */ + inflate_blocks_statef + *blocks; /* current inflate_blocks state */ + +}; + + +int inflateReset(z) +z_stream *z; +{ + uLong c; + + if (z == Z_NULL || z->state == Z_NULL) + return Z_STREAM_ERROR; + z->total_in = z->total_out = 0; + z->msg = Z_NULL; + z->state->mode = z->state->nowrap ? BLOCKS : METHOD; + inflate_blocks_reset(z->state->blocks, z, &c); + Trace((stderr, "inflate: reset\n")); + return Z_OK; +} + + +int inflateEnd(z) +z_stream *z; +{ + uLong c; + + if (z == Z_NULL || z->state == Z_NULL || z->zfree == Z_NULL) + return Z_STREAM_ERROR; + if (z->state->blocks != Z_NULL) + inflate_blocks_free(z->state->blocks, z, &c); + ZFREE(z, z->state, sizeof(struct internal_state)); + z->state = Z_NULL; + Trace((stderr, "inflate: end\n")); + return Z_OK; +} + + +int inflateInit2(z, w) +z_stream *z; +int w; +{ + /* initialize state */ + if (z == Z_NULL) + return Z_STREAM_ERROR; +/* if (z->zalloc == Z_NULL) z->zalloc = zcalloc; */ +/* if (z->zfree == Z_NULL) z->zfree = zcfree; */ + if ((z->state = (struct internal_state FAR *) + ZALLOC(z,1,sizeof(struct internal_state))) == Z_NULL) + return Z_MEM_ERROR; + z->state->blocks = Z_NULL; + + /* handle undocumented nowrap option (no zlib header or check) */ + z->state->nowrap = 0; + if (w < 0) + { + w = - w; + z->state->nowrap = 1; + } + + /* set window size */ + if (w < 8 || w > 15) + { + inflateEnd(z); + return Z_STREAM_ERROR; + } + z->state->wbits = (uInt)w; + + /* create inflate_blocks state */ + if ((z->state->blocks = + inflate_blocks_new(z, z->state->nowrap ? Z_NULL : adler32, 1 << w)) + == Z_NULL) + { + inflateEnd(z); + return Z_MEM_ERROR; + } + Trace((stderr, "inflate: allocated\n")); + + /* reset state */ + inflateReset(z); + return Z_OK; +} + + +int inflateInit(z) +z_stream *z; +{ + return inflateInit2(z, DEF_WBITS); +} + + +#define NEEDBYTE {if(z->avail_in==0)goto empty;r=Z_OK;} +#define NEXTBYTE (z->avail_in--,z->total_in++,*z->next_in++) + +int inflate(z, f) +z_stream *z; +int f; +{ + int r; + uInt b; + + if (z == Z_NULL || z->next_in == Z_NULL) + return Z_STREAM_ERROR; + r = Z_BUF_ERROR; + while (1) switch (z->state->mode) + { + case METHOD: + NEEDBYTE + if (((z->state->sub.method = NEXTBYTE) & 0xf) != DEFLATED) + { + z->state->mode = BAD; + z->msg = "unknown compression method"; + z->state->sub.marker = 5; /* can't try inflateSync */ + break; + } + if ((z->state->sub.method >> 4) + 8 > z->state->wbits) + { + z->state->mode = BAD; + z->msg = "invalid window size"; + z->state->sub.marker = 5; /* can't try inflateSync */ + break; + } + z->state->mode = FLAG; + case FLAG: + NEEDBYTE + if ((b = NEXTBYTE) & 0x20) + { + z->state->mode = BAD; + z->msg = "invalid reserved bit"; + z->state->sub.marker = 5; /* can't try inflateSync */ + break; + } + if (((z->state->sub.method << 8) + b) % 31) + { + z->state->mode = BAD; + z->msg = "incorrect header check"; + z->state->sub.marker = 5; /* can't try inflateSync */ + break; + } + Trace((stderr, "inflate: zlib header ok\n")); + z->state->mode = BLOCKS; + case BLOCKS: + r = inflate_blocks(z->state->blocks, z, r); + if (f == Z_PACKET_FLUSH && z->avail_in == 0 && z->avail_out != 0) + r = inflate_packet_flush(z->state->blocks); + if (r == Z_DATA_ERROR) + { + z->state->mode = BAD; + z->state->sub.marker = 0; /* can try inflateSync */ + break; + } + if (r != Z_STREAM_END) + return r; + r = Z_OK; + inflate_blocks_reset(z->state->blocks, z, &z->state->sub.check.was); + if (z->state->nowrap) + { + z->state->mode = DONE; + break; + } + z->state->mode = CHECK4; + case CHECK4: + NEEDBYTE + z->state->sub.check.need = (uLong)NEXTBYTE << 24; + z->state->mode = CHECK3; + case CHECK3: + NEEDBYTE + z->state->sub.check.need += (uLong)NEXTBYTE << 16; + z->state->mode = CHECK2; + case CHECK2: + NEEDBYTE + z->state->sub.check.need += (uLong)NEXTBYTE << 8; + z->state->mode = CHECK1; + case CHECK1: + NEEDBYTE + z->state->sub.check.need += (uLong)NEXTBYTE; + + if (z->state->sub.check.was != z->state->sub.check.need) + { + z->state->mode = BAD; + z->msg = "incorrect data check"; + z->state->sub.marker = 5; /* can't try inflateSync */ + break; + } + Trace((stderr, "inflate: zlib check ok\n")); + z->state->mode = DONE; + case DONE: + return Z_STREAM_END; + case BAD: + return Z_DATA_ERROR; + default: + return Z_STREAM_ERROR; + } + + empty: + if (f != Z_PACKET_FLUSH) + return r; + z->state->mode = BAD; + z->state->sub.marker = 0; /* can try inflateSync */ + return Z_DATA_ERROR; +} + +/* + * This subroutine adds the data at next_in/avail_in to the output history + * without performing any output. The output buffer must be "caught up"; + * i.e. no pending output (hence s->read equals s->write), and the state must + * be BLOCKS (i.e. we should be willing to see the start of a series of + * BLOCKS). On exit, the output will also be caught up, and the checksum + * will have been updated if need be. + */ + +int inflateIncomp(z) +z_stream *z; +{ + if (z->state->mode != BLOCKS) + return Z_DATA_ERROR; + return inflate_addhistory(z->state->blocks, z); +} + + +int inflateSync(z) +z_stream *z; +{ + uInt n; /* number of bytes to look at */ + Bytef *p; /* pointer to bytes */ + uInt m; /* number of marker bytes found in a row */ + uLong r, w; /* temporaries to save total_in and total_out */ + + /* set up */ + if (z == Z_NULL || z->state == Z_NULL) + return Z_STREAM_ERROR; + if (z->state->mode != BAD) + { + z->state->mode = BAD; + z->state->sub.marker = 0; + } + if ((n = z->avail_in) == 0) + return Z_BUF_ERROR; + p = z->next_in; + m = z->state->sub.marker; + + /* search */ + while (n && m < 4) + { + if (*p == (Byte)(m < 2 ? 0 : 0xff)) + m++; + else if (*p) + m = 0; + else + m = 4 - m; + p++, n--; + } + + /* restore */ + z->total_in += p - z->next_in; + z->next_in = p; + z->avail_in = n; + z->state->sub.marker = m; + + /* return no joy or set up to restart on a new block */ + if (m != 4) + return Z_DATA_ERROR; + r = z->total_in; w = z->total_out; + inflateReset(z); + z->total_in = r; z->total_out = w; + z->state->mode = BLOCKS; + return Z_OK; +} + +#undef NEEDBYTE +#undef NEXTBYTE + +/*+++++*/ +/* infutil.h -- types and macros common to blocks and codes + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +/* inflate blocks semi-private state */ +struct inflate_blocks_state { + + /* mode */ + enum { + TYPE, /* get type bits (3, including end bit) */ + LENS, /* get lengths for stored */ + STORED, /* processing stored block */ + TABLE, /* get table lengths */ + BTREE, /* get bit lengths tree for a dynamic block */ + DTREE, /* get length, distance trees for a dynamic block */ + CODES, /* processing fixed or dynamic block */ + DRY, /* output remaining window bytes */ + DONEB, /* finished last block, done */ + BADB} /* got a data error--stuck here */ + mode; /* current inflate_block mode */ + + /* mode dependent information */ + union { + uInt left; /* if STORED, bytes left to copy */ + struct { + uInt table; /* table lengths (14 bits) */ + uInt index; /* index into blens (or border) */ + uIntf *blens; /* bit lengths of codes */ + uInt bb; /* bit length tree depth */ + inflate_huft *tb; /* bit length decoding tree */ + int nblens; /* # elements allocated at blens */ + } trees; /* if DTREE, decoding info for trees */ + struct { + inflate_huft *tl, *td; /* trees to free */ + inflate_codes_statef + *codes; + } decode; /* if CODES, current state */ + } sub; /* submode */ + uInt last; /* true if this block is the last block */ + + /* mode independent information */ + uInt bitk; /* bits in bit buffer */ + uLong bitb; /* bit buffer */ + Bytef *window; /* sliding window */ + Bytef *end; /* one byte after sliding window */ + Bytef *read; /* window read pointer */ + Bytef *write; /* window write pointer */ + check_func checkfn; /* check function */ + uLong check; /* check on output */ + +}; + + +/* defines for inflate input/output */ +/* update pointers and return */ +#define UPDBITS {s->bitb=b;s->bitk=k;} +#define UPDIN {z->avail_in=n;z->total_in+=p-z->next_in;z->next_in=p;} +#define UPDOUT {s->write=q;} +#define UPDATE {UPDBITS UPDIN UPDOUT} +#define LEAVE {UPDATE return inflate_flush(s,z,r);} +/* get bytes and bits */ +#define LOADIN {p=z->next_in;n=z->avail_in;b=s->bitb;k=s->bitk;} +#define NEEDBYTE {if(n)r=Z_OK;else LEAVE} +#define NEXTBYTE (n--,*p++) +#define NEEDBITS(j) {while(k<(j)){NEEDBYTE;b|=((uLong)NEXTBYTE)<<k;k+=8;}} +#define DUMPBITS(j) {b>>=(j);k-=(j);} +/* output bytes */ +#define WAVAIL (q<s->read?s->read-q-1:s->end-q) +#define LOADOUT {q=s->write;m=WAVAIL;} +#define WRAP {if(q==s->end&&s->read!=s->window){q=s->window;m=WAVAIL;}} +#define FLUSH {UPDOUT r=inflate_flush(s,z,r); LOADOUT} +#define NEEDOUT {if(m==0){WRAP if(m==0){FLUSH WRAP if(m==0) LEAVE}}r=Z_OK;} +#define OUTBYTE(a) {*q++=(Byte)(a);m--;} +/* load local pointers */ +#define LOAD {LOADIN LOADOUT} + +/* + * The IBM 150 firmware munges the data right after _etext[]. This + * protects it. -- Cort + */ +local uInt protect_mask[] = {0, 0, 0, 0, 0, 0, 0, 0, 0 ,0 ,0 ,0}; +/* And'ing with mask[n] masks the lower n bits */ +local uInt inflate_mask[] = { + 0x0000, + 0x0001, 0x0003, 0x0007, 0x000f, 0x001f, 0x003f, 0x007f, 0x00ff, + 0x01ff, 0x03ff, 0x07ff, 0x0fff, 0x1fff, 0x3fff, 0x7fff, 0xffff +}; + +/* copy as much as possible from the sliding window to the output area */ +local int inflate_flush OF(( + inflate_blocks_statef *, + z_stream *, + int)); + +/*+++++*/ +/* inffast.h -- header to use inffast.c + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* WARNING: this file should *not* be used by applications. It is + part of the implementation of the compression library and is + subject to change. Applications should only use zlib.h. + */ + +local int inflate_fast OF(( + uInt, + uInt, + inflate_huft *, + inflate_huft *, + inflate_blocks_statef *, + z_stream *)); + + +/*+++++*/ +/* infblock.c -- interpret and process block types to last block + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* Table for deflate from PKZIP's appnote.txt. */ +local uInt border[] = { /* Order of the bit length code lengths */ + 16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15}; + +/* + Notes beyond the 1.93a appnote.txt: + + 1. Distance pointers never point before the beginning of the output + stream. + 2. Distance pointers can point back across blocks, up to 32k away. + 3. There is an implied maximum of 7 bits for the bit length table and + 15 bits for the actual data. + 4. If only one code exists, then it is encoded using one bit. (Zero + would be more efficient, but perhaps a little confusing.) If two + codes exist, they are coded using one bit each (0 and 1). + 5. There is no way of sending zero distance codes--a dummy must be + sent if there are none. (History: a pre 2.0 version of PKZIP would + store blocks with no distance codes, but this was discovered to be + too harsh a criterion.) Valid only for 1.93a. 2.04c does allow + zero distance codes, which is sent as one code of zero bits in + length. + 6. There are up to 286 literal/length codes. Code 256 represents the + end-of-block. Note however that the static length tree defines + 288 codes just to fill out the Huffman codes. Codes 286 and 287 + cannot be used though, since there is no length base or extra bits + defined for them. Similarily, there are up to 30 distance codes. + However, static trees define 32 codes (all 5 bits) to fill out the + Huffman codes, but the last two had better not show up in the data. + 7. Unzip can check dynamic Huffman blocks for complete code sets. + The exception is that a single code would not be complete (see #4). + 8. The five bits following the block type is really the number of + literal codes sent minus 257. + 9. Length codes 8,16,16 are interpreted as 13 length codes of 8 bits + (1+6+6). Therefore, to output three times the length, you output + three codes (1+1+1), whereas to output four times the same length, + you only need two codes (1+3). Hmm. + 10. In the tree reconstruction algorithm, Code = Code + Increment + only if BitLength(i) is not zero. (Pretty obvious.) + 11. Correction: 4 Bits: # of Bit Length codes - 4 (4 - 19) + 12. Note: length code 284 can represent 227-258, but length code 285 + really is 258. The last length deserves its own, short code + since it gets used a lot in very redundant files. The length + 258 is special since 258 - 3 (the min match length) is 255. + 13. The literal/length and distance code bit lengths are read as a + single stream of lengths. It is possible (and advantageous) for + a repeat code (16, 17, or 18) to go across the boundary between + the two sets of lengths. + */ + + +local void inflate_blocks_reset(s, z, c) +inflate_blocks_statef *s; +z_stream *z; +uLongf *c; +{ + if (s->checkfn != Z_NULL) + *c = s->check; + if (s->mode == BTREE || s->mode == DTREE) + ZFREE(z, s->sub.trees.blens, s->sub.trees.nblens * sizeof(uInt)); + if (s->mode == CODES) + { + inflate_codes_free(s->sub.decode.codes, z); + inflate_trees_free(s->sub.decode.td, z); + inflate_trees_free(s->sub.decode.tl, z); + } + s->mode = TYPE; + s->bitk = 0; + s->bitb = 0; + s->read = s->write = s->window; + if (s->checkfn != Z_NULL) + s->check = (*s->checkfn)(0L, Z_NULL, 0); + Trace((stderr, "inflate: blocks reset\n")); +} + + +local inflate_blocks_statef *inflate_blocks_new(z, c, w) +z_stream *z; +check_func c; +uInt w; +{ + inflate_blocks_statef *s; + + if ((s = (inflate_blocks_statef *)ZALLOC + (z,1,sizeof(struct inflate_blocks_state))) == Z_NULL) + return s; + if ((s->window = (Bytef *)ZALLOC(z, 1, w)) == Z_NULL) + { + ZFREE(z, s, sizeof(struct inflate_blocks_state)); + return Z_NULL; + } + s->end = s->window + w; + s->checkfn = c; + s->mode = TYPE; + Trace((stderr, "inflate: blocks allocated\n")); + inflate_blocks_reset(s, z, &s->check); + return s; +} + + +local int inflate_blocks(s, z, r) +inflate_blocks_statef *s; +z_stream *z; +int r; +{ + uInt t; /* temporary storage */ + uLong b; /* bit buffer */ + uInt k; /* bits in bit buffer */ + Bytef *p; /* input data pointer */ + uInt n; /* bytes available there */ + Bytef *q; /* output window write pointer */ + uInt m; /* bytes to end of window or read pointer */ + + /* copy input/output information to locals (UPDATE macro restores) */ + LOAD + + /* process input based on current state */ + while (1) switch (s->mode) + { + case TYPE: + NEEDBITS(3) + t = (uInt)b & 7; + s->last = t & 1; + switch (t >> 1) + { + case 0: /* stored */ + Trace((stderr, "inflate: stored block%s\n", + s->last ? " (last)" : "")); + DUMPBITS(3) + t = k & 7; /* go to byte boundary */ + DUMPBITS(t) + s->mode = LENS; /* get length of stored block */ + break; + case 1: /* fixed */ + Trace((stderr, "inflate: fixed codes block%s\n", + s->last ? " (last)" : "")); + { + uInt bl, bd; + inflate_huft *tl, *td; + + inflate_trees_fixed(&bl, &bd, &tl, &td); + s->sub.decode.codes = inflate_codes_new(bl, bd, tl, td, z); + if (s->sub.decode.codes == Z_NULL) + { + r = Z_MEM_ERROR; + LEAVE + } + s->sub.decode.tl = Z_NULL; /* don't try to free these */ + s->sub.decode.td = Z_NULL; + } + DUMPBITS(3) + s->mode = CODES; + break; + case 2: /* dynamic */ + Trace((stderr, "inflate: dynamic codes block%s\n", + s->last ? " (last)" : "")); + DUMPBITS(3) + s->mode = TABLE; + break; + case 3: /* illegal */ + DUMPBITS(3) + s->mode = BADB; + z->msg = "invalid block type"; + r = Z_DATA_ERROR; + LEAVE + } + break; + case LENS: + NEEDBITS(32) + if (((~b) >> 16) != (b & 0xffff)) + { + s->mode = BADB; + z->msg = "invalid stored block lengths"; + r = Z_DATA_ERROR; + LEAVE + } + s->sub.left = (uInt)b & 0xffff; + b = k = 0; /* dump bits */ + Tracev((stderr, "inflate: stored length %u\n", s->sub.left)); + s->mode = s->sub.left ? STORED : TYPE; + break; + case STORED: + if (n == 0) + LEAVE + NEEDOUT + t = s->sub.left; + if (t > n) t = n; + if (t > m) t = m; + zmemcpy(q, p, t); + p += t; n -= t; + q += t; m -= t; + if ((s->sub.left -= t) != 0) + break; + Tracev((stderr, "inflate: stored end, %lu total out\n", + z->total_out + (q >= s->read ? q - s->read : + (s->end - s->read) + (q - s->window)))); + s->mode = s->last ? DRY : TYPE; + break; + case TABLE: + NEEDBITS(14) + s->sub.trees.table = t = (uInt)b & 0x3fff; +#ifndef PKZIP_BUG_WORKAROUND + if ((t & 0x1f) > 29 || ((t >> 5) & 0x1f) > 29) + { + s->mode = BADB; + z->msg = "too many length or distance symbols"; + r = Z_DATA_ERROR; + LEAVE + } +#endif + t = 258 + (t & 0x1f) + ((t >> 5) & 0x1f); + if (t < 19) + t = 19; + if ((s->sub.trees.blens = (uIntf*)ZALLOC(z, t, sizeof(uInt))) == Z_NULL) + { + r = Z_MEM_ERROR; + LEAVE + } + s->sub.trees.nblens = t; + DUMPBITS(14) + s->sub.trees.index = 0; + Tracev((stderr, "inflate: table sizes ok\n")); + s->mode = BTREE; + case BTREE: + while (s->sub.trees.index < 4 + (s->sub.trees.table >> 10)) + { + NEEDBITS(3) + s->sub.trees.blens[border[s->sub.trees.index++]] = (uInt)b & 7; + DUMPBITS(3) + } + while (s->sub.trees.index < 19) + s->sub.trees.blens[border[s->sub.trees.index++]] = 0; + s->sub.trees.bb = 7; + t = inflate_trees_bits(s->sub.trees.blens, &s->sub.trees.bb, + &s->sub.trees.tb, z); + if (t != Z_OK) + { + r = t; + if (r == Z_DATA_ERROR) + s->mode = BADB; + LEAVE + } + s->sub.trees.index = 0; + Tracev((stderr, "inflate: bits tree ok\n")); + s->mode = DTREE; + case DTREE: + while (t = s->sub.trees.table, + s->sub.trees.index < 258 + (t & 0x1f) + ((t >> 5) & 0x1f)) + { + inflate_huft *h; + uInt i, j, c; + + t = s->sub.trees.bb; + NEEDBITS(t) + h = s->sub.trees.tb + ((uInt)b & inflate_mask[t]); + t = h->word.what.Bits; + c = h->more.Base; + if (c < 16) + { + DUMPBITS(t) + s->sub.trees.blens[s->sub.trees.index++] = c; + } + else /* c == 16..18 */ + { + i = c == 18 ? 7 : c - 14; + j = c == 18 ? 11 : 3; + NEEDBITS(t + i) + DUMPBITS(t) + j += (uInt)b & inflate_mask[i]; + DUMPBITS(i) + i = s->sub.trees.index; + t = s->sub.trees.table; + if (i + j > 258 + (t & 0x1f) + ((t >> 5) & 0x1f) || + (c == 16 && i < 1)) + { + s->mode = BADB; + z->msg = "invalid bit length repeat"; + r = Z_DATA_ERROR; + LEAVE + } + c = c == 16 ? s->sub.trees.blens[i - 1] : 0; + do { + s->sub.trees.blens[i++] = c; + } while (--j); + s->sub.trees.index = i; + } + } + inflate_trees_free(s->sub.trees.tb, z); + s->sub.trees.tb = Z_NULL; + { + uInt bl, bd; + inflate_huft *tl, *td; + inflate_codes_statef *c; + + bl = 9; /* must be <= 9 for lookahead assumptions */ + bd = 6; /* must be <= 9 for lookahead assumptions */ + t = s->sub.trees.table; + t = inflate_trees_dynamic(257 + (t & 0x1f), 1 + ((t >> 5) & 0x1f), + s->sub.trees.blens, &bl, &bd, &tl, &td, z); + if (t != Z_OK) + { + if (t == (uInt)Z_DATA_ERROR) + s->mode = BADB; + r = t; + LEAVE + } + Tracev((stderr, "inflate: trees ok\n")); + if ((c = inflate_codes_new(bl, bd, tl, td, z)) == Z_NULL) + { + inflate_trees_free(td, z); + inflate_trees_free(tl, z); + r = Z_MEM_ERROR; + LEAVE + } + ZFREE(z, s->sub.trees.blens, s->sub.trees.nblens * sizeof(uInt)); + s->sub.decode.codes = c; + s->sub.decode.tl = tl; + s->sub.decode.td = td; + } + s->mode = CODES; + case CODES: + UPDATE + if ((r = inflate_codes(s, z, r)) != Z_STREAM_END) + return inflate_flush(s, z, r); + r = Z_OK; + inflate_codes_free(s->sub.decode.codes, z); + inflate_trees_free(s->sub.decode.td, z); + inflate_trees_free(s->sub.decode.tl, z); + LOAD + Tracev((stderr, "inflate: codes end, %lu total out\n", + z->total_out + (q >= s->read ? q - s->read : + (s->end - s->read) + (q - s->window)))); + if (!s->last) + { + s->mode = TYPE; + break; + } + if (k > 7) /* return unused byte, if any */ + { + Assert(k < 16, "inflate_codes grabbed too many bytes") + k -= 8; + n++; + p--; /* can always return one */ + } + s->mode = DRY; + case DRY: + FLUSH + if (s->read != s->write) + LEAVE + s->mode = DONEB; + case DONEB: + r = Z_STREAM_END; + LEAVE + case BADB: + r = Z_DATA_ERROR; + LEAVE + default: + r = Z_STREAM_ERROR; + LEAVE + } +} + + +local int inflate_blocks_free(s, z, c) +inflate_blocks_statef *s; +z_stream *z; +uLongf *c; +{ + inflate_blocks_reset(s, z, c); + ZFREE(z, s->window, s->end - s->window); + ZFREE(z, s, sizeof(struct inflate_blocks_state)); + Trace((stderr, "inflate: blocks freed\n")); + return Z_OK; +} + +/* + * This subroutine adds the data at next_in/avail_in to the output history + * without performing any output. The output buffer must be "caught up"; + * i.e. no pending output (hence s->read equals s->write), and the state must + * be BLOCKS (i.e. we should be willing to see the start of a series of + * BLOCKS). On exit, the output will also be caught up, and the checksum + * will have been updated if need be. + */ +local int inflate_addhistory(s, z) +inflate_blocks_statef *s; +z_stream *z; +{ + uLong b; /* bit buffer */ /* NOT USED HERE */ + uInt k; /* bits in bit buffer */ /* NOT USED HERE */ + uInt t; /* temporary storage */ + Bytef *p; /* input data pointer */ + uInt n; /* bytes available there */ + Bytef *q; /* output window write pointer */ + uInt m; /* bytes to end of window or read pointer */ + + if (s->read != s->write) + return Z_STREAM_ERROR; + if (s->mode != TYPE) + return Z_DATA_ERROR; + + /* we're ready to rock */ + LOAD + /* while there is input ready, copy to output buffer, moving + * pointers as needed. + */ + while (n) { + t = n; /* how many to do */ + /* is there room until end of buffer? */ + if (t > m) t = m; + /* update check information */ + if (s->checkfn != Z_NULL) + s->check = (*s->checkfn)(s->check, q, t); + zmemcpy(q, p, t); + q += t; + p += t; + n -= t; + z->total_out += t; + s->read = q; /* drag read pointer forward */ +/* WRAP */ /* expand WRAP macro by hand to handle s->read */ + if (q == s->end) { + s->read = q = s->window; + m = WAVAIL; + } + } + UPDATE + return Z_OK; +} + + +/* + * At the end of a Deflate-compressed PPP packet, we expect to have seen + * a `stored' block type value but not the (zero) length bytes. + */ +local int inflate_packet_flush(s) + inflate_blocks_statef *s; +{ + if (s->mode != LENS) + return Z_DATA_ERROR; + s->mode = TYPE; + return Z_OK; +} + + +/*+++++*/ +/* inftrees.c -- generate Huffman trees for efficient decoding + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* simplify the use of the inflate_huft type with some defines */ +#define base more.Base +#define next more.Next +#define exop word.what.Exop +#define bits word.what.Bits + + +local int huft_build OF(( + uIntf *, /* code lengths in bits */ + uInt, /* number of codes */ + uInt, /* number of "simple" codes */ + uIntf *, /* list of base values for non-simple codes */ + uIntf *, /* list of extra bits for non-simple codes */ + inflate_huft * FAR*,/* result: starting table */ + uIntf *, /* maximum lookup bits (returns actual) */ + z_stream *)); /* for zalloc function */ + +local voidpf falloc OF(( + voidpf, /* opaque pointer (not used) */ + uInt, /* number of items */ + uInt)); /* size of item */ + +local void ffree OF(( + voidpf q, /* opaque pointer (not used) */ + voidpf p, /* what to free (not used) */ + uInt n)); /* number of bytes (not used) */ + +/* Tables for deflate from PKZIP's appnote.txt. */ +local uInt cplens[] = { /* Copy lengths for literal codes 257..285 */ + 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31, + 35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0}; + /* actually lengths - 2; also see note #13 above about 258 */ +local uInt cplext[] = { /* Extra bits for literal codes 257..285 */ + 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, + 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 0, 192, 192}; /* 192==invalid */ +local uInt cpdist[] = { /* Copy offsets for distance codes 0..29 */ + 1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193, + 257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145, + 8193, 12289, 16385, 24577}; +local uInt cpdext[] = { /* Extra bits for distance codes */ + 0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, + 7, 7, 8, 8, 9, 9, 10, 10, 11, 11, + 12, 12, 13, 13}; + +/* + Huffman code decoding is performed using a multi-level table lookup. + The fastest way to decode is to simply build a lookup table whose + size is determined by the longest code. However, the time it takes + to build this table can also be a factor if the data being decoded + is not very long. The most common codes are necessarily the + shortest codes, so those codes dominate the decoding time, and hence + the speed. The idea is you can have a shorter table that decodes the + shorter, more probable codes, and then point to subsidiary tables for + the longer codes. The time it costs to decode the longer codes is + then traded against the time it takes to make longer tables. + + This results of this trade are in the variables lbits and dbits + below. lbits is the number of bits the first level table for literal/ + length codes can decode in one step, and dbits is the same thing for + the distance codes. Subsequent tables are also less than or equal to + those sizes. These values may be adjusted either when all of the + codes are shorter than that, in which case the longest code length in + bits is used, or when the shortest code is *longer* than the requested + table size, in which case the length of the shortest code in bits is + used. + + There are two different values for the two tables, since they code a + different number of possibilities each. The literal/length table + codes 286 possible values, or in a flat code, a little over eight + bits. The distance table codes 30 possible values, or a little less + than five bits, flat. The optimum values for speed end up being + about one bit more than those, so lbits is 8+1 and dbits is 5+1. + The optimum values may differ though from machine to machine, and + possibly even between compilers. Your mileage may vary. + */ + + +/* If BMAX needs to be larger than 16, then h and x[] should be uLong. */ +#define BMAX 15 /* maximum bit length of any code */ +#define N_MAX 288 /* maximum number of codes in any set */ + +#ifdef DEBUG_ZLIB + uInt inflate_hufts; +#endif + +local int huft_build(b, n, s, d, e, t, m, zs) +uIntf *b; /* code lengths in bits (all assumed <= BMAX) */ +uInt n; /* number of codes (assumed <= N_MAX) */ +uInt s; /* number of simple-valued codes (0..s-1) */ +uIntf *d; /* list of base values for non-simple codes */ +uIntf *e; /* list of extra bits for non-simple codes */ +inflate_huft * FAR *t; /* result: starting table */ +uIntf *m; /* maximum lookup bits, returns actual */ +z_stream *zs; /* for zalloc function */ +/* Given a list of code lengths and a maximum table size, make a set of + tables to decode that set of codes. Return Z_OK on success, Z_BUF_ERROR + if the given code set is incomplete (the tables are still built in this + case), Z_DATA_ERROR if the input is invalid (all zero length codes or an + over-subscribed set of lengths), or Z_MEM_ERROR if not enough memory. */ +{ + + uInt a; /* counter for codes of length k */ + uInt c[BMAX+1]; /* bit length count table */ + uInt f; /* i repeats in table every f entries */ + int g; /* maximum code length */ + int h; /* table level */ + register uInt i; /* counter, current code */ + register uInt j; /* counter */ + register int k; /* number of bits in current code */ + int l; /* bits per table (returned in m) */ + register uIntf *p; /* pointer into c[], b[], or v[] */ + inflate_huft *q; /* points to current table */ + struct inflate_huft_s r; /* table entry for structure assignment */ + inflate_huft *u[BMAX]; /* table stack */ + uInt v[N_MAX]; /* values in order of bit length */ + register int w; /* bits before this table == (l * h) */ + uInt x[BMAX+1]; /* bit offsets, then code stack */ + uIntf *xp; /* pointer into x */ + int y; /* number of dummy codes added */ + uInt z; /* number of entries in current table */ + + + /* Generate counts for each bit length */ + p = c; +#define C0 *p++ = 0; +#define C2 C0 C0 C0 C0 +#define C4 C2 C2 C2 C2 + C4 /* clear c[]--assume BMAX+1 is 16 */ + p = b; i = n; + do { + c[*p++]++; /* assume all entries <= BMAX */ + } while (--i); + if (c[0] == n) /* null input--all zero length codes */ + { + *t = (inflate_huft *)Z_NULL; + *m = 0; + return Z_OK; + } + + + /* Find minimum and maximum length, bound *m by those */ + l = *m; + for (j = 1; j <= BMAX; j++) + if (c[j]) + break; + k = j; /* minimum code length */ + if ((uInt)l < j) + l = j; + for (i = BMAX; i; i--) + if (c[i]) + break; + g = i; /* maximum code length */ + if ((uInt)l > i) + l = i; + *m = l; + + + /* Adjust last length count to fill out codes, if needed */ + for (y = 1 << j; j < i; j++, y <<= 1) + if ((y -= c[j]) < 0) + return Z_DATA_ERROR; + if ((y -= c[i]) < 0) + return Z_DATA_ERROR; + c[i] += y; + + + /* Generate starting offsets into the value table for each length */ + x[1] = j = 0; + p = c + 1; xp = x + 2; + while (--i) { /* note that i == g from above */ + *xp++ = (j += *p++); + } + + + /* Make a table of values in order of bit lengths */ + p = b; i = 0; + do { + if ((j = *p++) != 0) + v[x[j]++] = i; + } while (++i < n); + + + /* Generate the Huffman codes and for each, make the table entries */ + x[0] = i = 0; /* first Huffman code is zero */ + p = v; /* grab values in bit order */ + h = -1; /* no tables yet--level -1 */ + w = -l; /* bits decoded == (l * h) */ + u[0] = (inflate_huft *)Z_NULL; /* just to keep compilers happy */ + q = (inflate_huft *)Z_NULL; /* ditto */ + z = 0; /* ditto */ + + /* go through the bit lengths (k already is bits in shortest code) */ + for (; k <= g; k++) + { + a = c[k]; + while (a--) + { + /* here i is the Huffman code of length k bits for value *p */ + /* make tables up to required level */ + while (k > w + l) + { + h++; + w += l; /* previous table always l bits */ + + /* compute minimum size table less than or equal to l bits */ + z = (z = g - w) > (uInt)l ? l : z; /* table size upper limit */ + if ((f = 1 << (j = k - w)) > a + 1) /* try a k-w bit table */ + { /* too few codes for k-w bit table */ + f -= a + 1; /* deduct codes from patterns left */ + xp = c + k; + if (j < z) + while (++j < z) /* try smaller tables up to z bits */ + { + if ((f <<= 1) <= *++xp) + break; /* enough codes to use up j bits */ + f -= *xp; /* else deduct codes from patterns */ + } + } + z = 1 << j; /* table entries for j-bit table */ + + /* allocate and link in new table */ + if ((q = (inflate_huft *)ZALLOC + (zs,z + 1,sizeof(inflate_huft))) == Z_NULL) + { + if (h) + inflate_trees_free(u[0], zs); + return Z_MEM_ERROR; /* not enough memory */ + } + q->word.Nalloc = z + 1; +#ifdef DEBUG_ZLIB + inflate_hufts += z + 1; +#endif + *t = q + 1; /* link to list for huft_free() */ + *(t = &(q->next)) = Z_NULL; + u[h] = ++q; /* table starts after link */ + + /* connect to last table, if there is one */ + if (h) + { + x[h] = i; /* save pattern for backing up */ + r.bits = (Byte)l; /* bits to dump before this table */ + r.exop = (Byte)j; /* bits in this table */ + r.next = q; /* pointer to this table */ + j = i >> (w - l); /* (get around Turbo C bug) */ + u[h-1][j] = r; /* connect to last table */ + } + } + + /* set up table entry in r */ + r.bits = (Byte)(k - w); + if (p >= v + n) + r.exop = 128 + 64; /* out of values--invalid code */ + else if (*p < s) + { + r.exop = (Byte)(*p < 256 ? 0 : 32 + 64); /* 256 is end-of-block */ + r.base = *p++; /* simple code is just the value */ + } + else + { + r.exop = (Byte)e[*p - s] + 16 + 64; /* non-simple--look up in lists */ + r.base = d[*p++ - s]; + } + + /* fill code-like entries with r */ + f = 1 << (k - w); + for (j = i >> w; j < z; j += f) + q[j] = r; + + /* backwards increment the k-bit code i */ + for (j = 1 << (k - 1); i & j; j >>= 1) + i ^= j; + i ^= j; + + /* backup over finished tables */ + while ((i & ((1 << w) - 1)) != x[h]) + { + h--; /* don't need to update q */ + w -= l; + } + } + } + + + /* Return Z_BUF_ERROR if we were given an incomplete table */ + return y != 0 && g != 1 ? Z_BUF_ERROR : Z_OK; +} + + +local int inflate_trees_bits(c, bb, tb, z) +uIntf *c; /* 19 code lengths */ +uIntf *bb; /* bits tree desired/actual depth */ +inflate_huft * FAR *tb; /* bits tree result */ +z_stream *z; /* for zfree function */ +{ + int r; + + r = huft_build(c, 19, 19, (uIntf*)Z_NULL, (uIntf*)Z_NULL, tb, bb, z); + if (r == Z_DATA_ERROR) + z->msg = "oversubscribed dynamic bit lengths tree"; + else if (r == Z_BUF_ERROR) + { + inflate_trees_free(*tb, z); + z->msg = "incomplete dynamic bit lengths tree"; + r = Z_DATA_ERROR; + } + return r; +} + + +local int inflate_trees_dynamic(nl, nd, c, bl, bd, tl, td, z) +uInt nl; /* number of literal/length codes */ +uInt nd; /* number of distance codes */ +uIntf *c; /* that many (total) code lengths */ +uIntf *bl; /* literal desired/actual bit depth */ +uIntf *bd; /* distance desired/actual bit depth */ +inflate_huft * FAR *tl; /* literal/length tree result */ +inflate_huft * FAR *td; /* distance tree result */ +z_stream *z; /* for zfree function */ +{ + int r; + + /* build literal/length tree */ + if ((r = huft_build(c, nl, 257, cplens, cplext, tl, bl, z)) != Z_OK) + { + if (r == Z_DATA_ERROR) + z->msg = "oversubscribed literal/length tree"; + else if (r == Z_BUF_ERROR) + { + inflate_trees_free(*tl, z); + z->msg = "incomplete literal/length tree"; + r = Z_DATA_ERROR; + } + return r; + } + + /* build distance tree */ + if ((r = huft_build(c + nl, nd, 0, cpdist, cpdext, td, bd, z)) != Z_OK) + { + if (r == Z_DATA_ERROR) + z->msg = "oversubscribed literal/length tree"; + else if (r == Z_BUF_ERROR) { +#ifdef PKZIP_BUG_WORKAROUND + r = Z_OK; + } +#else + inflate_trees_free(*td, z); + z->msg = "incomplete literal/length tree"; + r = Z_DATA_ERROR; + } + inflate_trees_free(*tl, z); + return r; +#endif + } + + /* done */ + return Z_OK; +} + + +/* build fixed tables only once--keep them here */ +local int fixed_lock = 0; +local int fixed_built = 0; +#define FIXEDH 530 /* number of hufts used by fixed tables */ +local uInt fixed_left = FIXEDH; +local inflate_huft fixed_mem[FIXEDH]; +local uInt fixed_bl; +local uInt fixed_bd; +local inflate_huft *fixed_tl; +local inflate_huft *fixed_td; + + +local voidpf falloc(q, n, s) +voidpf q; /* opaque pointer (not used) */ +uInt n; /* number of items */ +uInt s; /* size of item */ +{ + Assert(s == sizeof(inflate_huft) && n <= fixed_left, + "inflate_trees falloc overflow"); + if (q) s++; /* to make some compilers happy */ + fixed_left -= n; + return (voidpf)(fixed_mem + fixed_left); +} + + +local void ffree(q, p, n) +voidpf q; +voidpf p; +uInt n; +{ + Assert(0, "inflate_trees ffree called!"); + if (q) q = p; /* to make some compilers happy */ +} + + +local int inflate_trees_fixed(bl, bd, tl, td) +uIntf *bl; /* literal desired/actual bit depth */ +uIntf *bd; /* distance desired/actual bit depth */ +inflate_huft * FAR *tl; /* literal/length tree result */ +inflate_huft * FAR *td; /* distance tree result */ +{ + /* build fixed tables if not built already--lock out other instances */ + while (++fixed_lock > 1) + fixed_lock--; + if (!fixed_built) + { + int k; /* temporary variable */ + unsigned c[288]; /* length list for huft_build */ + z_stream z; /* for falloc function */ + + /* set up fake z_stream for memory routines */ + z.zalloc = falloc; + z.zfree = ffree; + z.opaque = Z_NULL; + + /* literal table */ + for (k = 0; k < 144; k++) + c[k] = 8; + for (; k < 256; k++) + c[k] = 9; + for (; k < 280; k++) + c[k] = 7; + for (; k < 288; k++) + c[k] = 8; + fixed_bl = 7; + huft_build(c, 288, 257, cplens, cplext, &fixed_tl, &fixed_bl, &z); + + /* distance table */ + for (k = 0; k < 30; k++) + c[k] = 5; + fixed_bd = 5; + huft_build(c, 30, 0, cpdist, cpdext, &fixed_td, &fixed_bd, &z); + + /* done */ + fixed_built = 1; + } + fixed_lock--; + *bl = fixed_bl; + *bd = fixed_bd; + *tl = fixed_tl; + *td = fixed_td; + return Z_OK; +} + + +local int inflate_trees_free(t, z) +inflate_huft *t; /* table to free */ +z_stream *z; /* for zfree function */ +/* Free the malloc'ed tables built by huft_build(), which makes a linked + list of the tables it made, with the links in a dummy first entry of + each table. */ +{ + register inflate_huft *p, *q; + + /* Go through linked list, freeing from the malloced (t[-1]) address. */ + p = t; + while (p != Z_NULL) + { + q = (--p)->next; + ZFREE(z, p, p->word.Nalloc * sizeof(inflate_huft)); + p = q; + } + return Z_OK; +} + +/*+++++*/ +/* infcodes.c -- process literals and length/distance pairs + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* simplify the use of the inflate_huft type with some defines */ +#define base more.Base +#define next more.Next +#define exop word.what.Exop +#define bits word.what.Bits + +/* inflate codes private state */ +struct inflate_codes_state { + + /* mode */ + enum { /* waiting for "i:"=input, "o:"=output, "x:"=nothing */ + START, /* x: set up for LEN */ + LEN, /* i: get length/literal/eob next */ + LENEXT, /* i: getting length extra (have base) */ + DIST, /* i: get distance next */ + DISTEXT, /* i: getting distance extra */ + COPY, /* o: copying bytes in window, waiting for space */ + LIT, /* o: got literal, waiting for output space */ + WASH, /* o: got eob, possibly still output waiting */ + END, /* x: got eob and all data flushed */ + BADCODE} /* x: got error */ + mode; /* current inflate_codes mode */ + + /* mode dependent information */ + uInt len; + union { + struct { + inflate_huft *tree; /* pointer into tree */ + uInt need; /* bits needed */ + } code; /* if LEN or DIST, where in tree */ + uInt lit; /* if LIT, literal */ + struct { + uInt get; /* bits to get for extra */ + uInt dist; /* distance back to copy from */ + } copy; /* if EXT or COPY, where and how much */ + } sub; /* submode */ + + /* mode independent information */ + Byte lbits; /* ltree bits decoded per branch */ + Byte dbits; /* dtree bits decoder per branch */ + inflate_huft *ltree; /* literal/length/eob tree */ + inflate_huft *dtree; /* distance tree */ + +}; + + +local inflate_codes_statef *inflate_codes_new(bl, bd, tl, td, z) +uInt bl, bd; +inflate_huft *tl, *td; +z_stream *z; +{ + inflate_codes_statef *c; + + if ((c = (inflate_codes_statef *) + ZALLOC(z,1,sizeof(struct inflate_codes_state))) != Z_NULL) + { + c->mode = START; + c->lbits = (Byte)bl; + c->dbits = (Byte)bd; + c->ltree = tl; + c->dtree = td; + Tracev((stderr, "inflate: codes new\n")); + } + return c; +} + + +local int inflate_codes(s, z, r) +inflate_blocks_statef *s; +z_stream *z; +int r; +{ + uInt j; /* temporary storage */ + inflate_huft *t; /* temporary pointer */ + uInt e; /* extra bits or operation */ + uLong b; /* bit buffer */ + uInt k; /* bits in bit buffer */ + Bytef *p; /* input data pointer */ + uInt n; /* bytes available there */ + Bytef *q; /* output window write pointer */ + uInt m; /* bytes to end of window or read pointer */ + Bytef *f; /* pointer to copy strings from */ + inflate_codes_statef *c = s->sub.decode.codes; /* codes state */ + + /* copy input/output information to locals (UPDATE macro restores) */ + LOAD + + /* process input and output based on current state */ + while (1) switch (c->mode) + { /* waiting for "i:"=input, "o:"=output, "x:"=nothing */ + case START: /* x: set up for LEN */ +#ifndef SLOW + if (m >= 258 && n >= 10) + { + UPDATE + r = inflate_fast(c->lbits, c->dbits, c->ltree, c->dtree, s, z); + LOAD + if (r != Z_OK) + { + c->mode = r == Z_STREAM_END ? WASH : BADCODE; + break; + } + } +#endif /* !SLOW */ + c->sub.code.need = c->lbits; + c->sub.code.tree = c->ltree; + c->mode = LEN; + case LEN: /* i: get length/literal/eob next */ + j = c->sub.code.need; + NEEDBITS(j) + t = c->sub.code.tree + ((uInt)b & inflate_mask[j]); + DUMPBITS(t->bits) + e = (uInt)(t->exop); + if (e == 0) /* literal */ + { + c->sub.lit = t->base; + Tracevv((stderr, t->base >= 0x20 && t->base < 0x7f ? + "inflate: literal '%c'\n" : + "inflate: literal 0x%02x\n", t->base)); + c->mode = LIT; + break; + } + if (e & 16) /* length */ + { + c->sub.copy.get = e & 15; + c->len = t->base; + c->mode = LENEXT; + break; + } + if ((e & 64) == 0) /* next table */ + { + c->sub.code.need = e; + c->sub.code.tree = t->next; + break; + } + if (e & 32) /* end of block */ + { + Tracevv((stderr, "inflate: end of block\n")); + c->mode = WASH; + break; + } + c->mode = BADCODE; /* invalid code */ + z->msg = "invalid literal/length code"; + r = Z_DATA_ERROR; + LEAVE + case LENEXT: /* i: getting length extra (have base) */ + j = c->sub.copy.get; + NEEDBITS(j) + c->len += (uInt)b & inflate_mask[j]; + DUMPBITS(j) + c->sub.code.need = c->dbits; + c->sub.code.tree = c->dtree; + Tracevv((stderr, "inflate: length %u\n", c->len)); + c->mode = DIST; + case DIST: /* i: get distance next */ + j = c->sub.code.need; + NEEDBITS(j) + t = c->sub.code.tree + ((uInt)b & inflate_mask[j]); + DUMPBITS(t->bits) + e = (uInt)(t->exop); + if (e & 16) /* distance */ + { + c->sub.copy.get = e & 15; + c->sub.copy.dist = t->base; + c->mode = DISTEXT; + break; + } + if ((e & 64) == 0) /* next table */ + { + c->sub.code.need = e; + c->sub.code.tree = t->next; + break; + } + c->mode = BADCODE; /* invalid code */ + z->msg = "invalid distance code"; + r = Z_DATA_ERROR; + LEAVE + case DISTEXT: /* i: getting distance extra */ + j = c->sub.copy.get; + NEEDBITS(j) + c->sub.copy.dist += (uInt)b & inflate_mask[j]; + DUMPBITS(j) + Tracevv((stderr, "inflate: distance %u\n", c->sub.copy.dist)); + c->mode = COPY; + case COPY: /* o: copying bytes in window, waiting for space */ +#ifndef __TURBOC__ /* Turbo C bug for following expression */ + f = (uInt)(q - s->window) < c->sub.copy.dist ? + s->end - (c->sub.copy.dist - (q - s->window)) : + q - c->sub.copy.dist; +#else + f = q - c->sub.copy.dist; + if ((uInt)(q - s->window) < c->sub.copy.dist) + f = s->end - (c->sub.copy.dist - (q - s->window)); +#endif + while (c->len) + { + NEEDOUT + OUTBYTE(*f++) + if (f == s->end) + f = s->window; + c->len--; + } + c->mode = START; + break; + case LIT: /* o: got literal, waiting for output space */ + NEEDOUT + OUTBYTE(c->sub.lit) + c->mode = START; + break; + case WASH: /* o: got eob, possibly more output */ + FLUSH + if (s->read != s->write) + LEAVE + c->mode = END; + case END: + r = Z_STREAM_END; + LEAVE + case BADCODE: /* x: got error */ + r = Z_DATA_ERROR; + LEAVE + default: + r = Z_STREAM_ERROR; + LEAVE + } +} + + +local void inflate_codes_free(c, z) +inflate_codes_statef *c; +z_stream *z; +{ + ZFREE(z, c, sizeof(struct inflate_codes_state)); + Tracev((stderr, "inflate: codes free\n")); +} + +/*+++++*/ +/* inflate_util.c -- data and routines common to blocks and codes + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* copy as much as possible from the sliding window to the output area */ +local int inflate_flush(s, z, r) +inflate_blocks_statef *s; +z_stream *z; +int r; +{ + uInt n; + Bytef *p, *q; + + /* local copies of source and destination pointers */ + p = z->next_out; + q = s->read; + + /* compute number of bytes to copy as far as end of window */ + n = (uInt)((q <= s->write ? s->write : s->end) - q); + if (n > z->avail_out) n = z->avail_out; + if (n && r == Z_BUF_ERROR) r = Z_OK; + + /* update counters */ + z->avail_out -= n; + z->total_out += n; + + /* update check information */ + if (s->checkfn != Z_NULL) + s->check = (*s->checkfn)(s->check, q, n); + + /* copy as far as end of window */ + zmemcpy(p, q, n); + p += n; + q += n; + + /* see if more to copy at beginning of window */ + if (q == s->end) + { + /* wrap pointers */ + q = s->window; + if (s->write == s->end) + s->write = s->window; + + /* compute bytes to copy */ + n = (uInt)(s->write - q); + if (n > z->avail_out) n = z->avail_out; + if (n && r == Z_BUF_ERROR) r = Z_OK; + + /* update counters */ + z->avail_out -= n; + z->total_out += n; + + /* update check information */ + if (s->checkfn != Z_NULL) + s->check = (*s->checkfn)(s->check, q, n); + + /* copy */ + zmemcpy(p, q, n); + p += n; + q += n; + } + + /* update pointers */ + z->next_out = p; + s->read = q; + + /* done */ + return r; +} + + +/*+++++*/ +/* inffast.c -- process literals and length/distance pairs fast + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* simplify the use of the inflate_huft type with some defines */ +#define base more.Base +#define next more.Next +#define exop word.what.Exop +#define bits word.what.Bits + +/* macros for bit input with no checking and for returning unused bytes */ +#define GRABBITS(j) {while(k<(j)){b|=((uLong)NEXTBYTE)<<k;k+=8;}} +#define UNGRAB {n+=(c=k>>3);p-=c;k&=7;} + +/* Called with number of bytes left to write in window at least 258 + (the maximum string length) and number of input bytes available + at least ten. The ten bytes are six bytes for the longest length/ + distance pair plus four bytes for overloading the bit buffer. */ + +local int inflate_fast(bl, bd, tl, td, s, z) +uInt bl, bd; +inflate_huft *tl, *td; +inflate_blocks_statef *s; +z_stream *z; +{ + inflate_huft *t; /* temporary pointer */ + uInt e; /* extra bits or operation */ + uLong b; /* bit buffer */ + uInt k; /* bits in bit buffer */ + Bytef *p; /* input data pointer */ + uInt n; /* bytes available there */ + Bytef *q; /* output window write pointer */ + uInt m; /* bytes to end of window or read pointer */ + uInt ml; /* mask for literal/length tree */ + uInt md; /* mask for distance tree */ + uInt c; /* bytes to copy */ + uInt d; /* distance back to copy from */ + Bytef *r; /* copy source pointer */ + + /* load input, output, bit values */ + LOAD + + /* initialize masks */ + ml = inflate_mask[bl]; + md = inflate_mask[bd]; + + /* do until not enough input or output space for fast loop */ + do { /* assume called with m >= 258 && n >= 10 */ + /* get literal/length code */ + GRABBITS(20) /* max bits for literal/length code */ + if ((e = (t = tl + ((uInt)b & ml))->exop) == 0) + { + DUMPBITS(t->bits) + Tracevv((stderr, t->base >= 0x20 && t->base < 0x7f ? + "inflate: * literal '%c'\n" : + "inflate: * literal 0x%02x\n", t->base)); + *q++ = (Byte)t->base; + m--; + continue; + } + do { + DUMPBITS(t->bits) + if (e & 16) + { + /* get extra bits for length */ + e &= 15; + c = t->base + ((uInt)b & inflate_mask[e]); + DUMPBITS(e) + Tracevv((stderr, "inflate: * length %u\n", c)); + + /* decode distance base of block to copy */ + GRABBITS(15); /* max bits for distance code */ + e = (t = td + ((uInt)b & md))->exop; + do { + DUMPBITS(t->bits) + if (e & 16) + { + /* get extra bits to add to distance base */ + e &= 15; + GRABBITS(e) /* get extra bits (up to 13) */ + d = t->base + ((uInt)b & inflate_mask[e]); + DUMPBITS(e) + Tracevv((stderr, "inflate: * distance %u\n", d)); + + /* do the copy */ + m -= c; + if ((uInt)(q - s->window) >= d) /* offset before dest */ + { /* just copy */ + r = q - d; + *q++ = *r++; c--; /* minimum count is three, */ + *q++ = *r++; c--; /* so unroll loop a little */ + } + else /* else offset after destination */ + { + e = d - (q - s->window); /* bytes from offset to end */ + r = s->end - e; /* pointer to offset */ + if (c > e) /* if source crosses, */ + { + c -= e; /* copy to end of window */ + do { + *q++ = *r++; + } while (--e); + r = s->window; /* copy rest from start of window */ + } + } + do { /* copy all or what's left */ + *q++ = *r++; + } while (--c); + break; + } + else if ((e & 64) == 0) + e = (t = t->next + ((uInt)b & inflate_mask[e]))->exop; + else + { + z->msg = "invalid distance code"; + UNGRAB + UPDATE + return Z_DATA_ERROR; + } + } while (1); + break; + } + if ((e & 64) == 0) + { + if ((e = (t = t->next + ((uInt)b & inflate_mask[e]))->exop) == 0) + { + DUMPBITS(t->bits) + Tracevv((stderr, t->base >= 0x20 && t->base < 0x7f ? + "inflate: * literal '%c'\n" : + "inflate: * literal 0x%02x\n", t->base)); + *q++ = (Byte)t->base; + m--; + break; + } + } + else if (e & 32) + { + Tracevv((stderr, "inflate: * end of block\n")); + UNGRAB + UPDATE + return Z_STREAM_END; + } + else + { + z->msg = "invalid literal/length code"; + UNGRAB + UPDATE + return Z_DATA_ERROR; + } + } while (1); + } while (m >= 258 && n >= 10); + + /* not enough input or output--restore pointers and return */ + UNGRAB + UPDATE + return Z_OK; +} + + +/*+++++*/ +/* zutil.c -- target dependent utility functions for the compression library + * Copyright (C) 1995 Jean-loup Gailly. + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* From: zutil.c,v 1.8 1995/05/03 17:27:12 jloup Exp */ + +char *zlib_version = ZLIB_VERSION; + +char *z_errmsg[] = { +"stream end", /* Z_STREAM_END 1 */ +"", /* Z_OK 0 */ +"file error", /* Z_ERRNO (-1) */ +"stream error", /* Z_STREAM_ERROR (-2) */ +"data error", /* Z_DATA_ERROR (-3) */ +"insufficient memory", /* Z_MEM_ERROR (-4) */ +"buffer error", /* Z_BUF_ERROR (-5) */ +""}; + + +/*+++++*/ +/* adler32.c -- compute the Adler-32 checksum of a data stream + * Copyright (C) 1995 Mark Adler + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* From: adler32.c,v 1.6 1995/05/03 17:27:08 jloup Exp */ + +#define BASE 65521L /* largest prime smaller than 65536 */ +#define NMAX 5552 +/* NMAX is the largest n such that 255n(n+1)/2 + (n+1)(BASE-1) <= 2^32-1 */ + +#define DO1(buf) {s1 += *buf++; s2 += s1;} +#define DO2(buf) DO1(buf); DO1(buf); +#define DO4(buf) DO2(buf); DO2(buf); +#define DO8(buf) DO4(buf); DO4(buf); +#define DO16(buf) DO8(buf); DO8(buf); + +/* ========================================================================= */ +uLong adler32(adler, buf, len) + uLong adler; + Bytef *buf; + uInt len; +{ + unsigned long s1 = adler & 0xffff; + unsigned long s2 = (adler >> 16) & 0xffff; + int k; + + if (buf == Z_NULL) return 1L; + + while (len > 0) { + k = len < NMAX ? len : NMAX; + len -= k; + while (k >= 16) { + DO16(buf); + k -= 16; + } + if (k != 0) do { + DO1(buf); + } while (--k); + s1 %= BASE; + s2 %= BASE; + } + return (s2 << 16) | s1; +} diff --git a/arch/xtensa/boot/lib/zmem.c b/arch/xtensa/boot/lib/zmem.c new file mode 100644 index 00000000000..7848f126d67 --- /dev/null +++ b/arch/xtensa/boot/lib/zmem.c @@ -0,0 +1,87 @@ +#include "zlib.h" + +/* bits taken from ppc */ + +extern void *avail_ram, *end_avail; + +void exit (void) +{ + for (;;); +} + +void *zalloc(void *x, unsigned items, unsigned size) +{ + void *p = avail_ram; + + size *= items; + size = (size + 7) & -8; + avail_ram += size; + if (avail_ram > end_avail) { + //puts("oops... out of memory\n"); + //pause(); + exit (); + } + return p; +} + +void zfree(void *x, void *addr, unsigned nb) +{ +} + + +#define HEAD_CRC 2 +#define EXTRA_FIELD 4 +#define ORIG_NAME 8 +#define COMMENT 0x10 +#define RESERVED 0xe0 + +#define DEFLATED 8 + +void gunzip (void *dst, int dstlen, unsigned char *src, int *lenp) +{ + z_stream s; + int r, i, flags; + + /* skip header */ + + i = 10; + flags = src[3]; + if (src[2] != DEFLATED || (flags & RESERVED) != 0) { + //puts("bad gzipped data\n"); + exit(); + } + if ((flags & EXTRA_FIELD) != 0) + i = 12 + src[10] + (src[11] << 8); + if ((flags & ORIG_NAME) != 0) + while (src[i++] != 0) + ; + if ((flags & COMMENT) != 0) + while (src[i++] != 0) + ; + if ((flags & HEAD_CRC) != 0) + i += 2; + if (i >= *lenp) { + //puts("gunzip: ran out of data in header\n"); + exit(); + } + + s.zalloc = zalloc; + s.zfree = zfree; + r = inflateInit2(&s, -MAX_WBITS); + if (r != Z_OK) { + //puts("inflateInit2 returned "); puthex(r); puts("\n"); + exit(); + } + s.next_in = src + i; + s.avail_in = *lenp - i; + s.next_out = dst; + s.avail_out = dstlen; + r = inflate(&s, Z_FINISH); + if (r != Z_OK && r != Z_STREAM_END) { + //puts("inflate returned "); puthex(r); puts("\n"); + exit(); + } + *lenp = s.next_out - (unsigned char *) dst; + inflateEnd(&s); +} + diff --git a/arch/xtensa/boot/ramdisk/Makefile b/arch/xtensa/boot/ramdisk/Makefile new file mode 100644 index 00000000000..b12f7635243 --- /dev/null +++ b/arch/xtensa/boot/ramdisk/Makefile @@ -0,0 +1,23 @@ +# +# Makefile for a ramdisk image +# + +BIG_ENDIAN := $(shell echo -e "\#ifdef __XTENSA_EL__\nint little;\n\#else\nint big;\n\#endif" | $(CC) -E -|grep -c big) + +ifeq ($(BIG_ENDIAN),1) +OBJCOPY_ARGS := -O elf32-xtensa-be +else +OBJCOPY_ARGS := -O elf32-xtensa-le +endif + +obj-y = ramdisk.o + +RAMDISK_IMAGE = arch/$(ARCH)/boot/ramdisk/$(CONFIG_EMBEDDED_RAMDISK_IMAGE) + +arch/$(ARCH)/boot/ramdisk/ramdisk.o: + $(Q)echo -e "dummy:" | $(AS) -o $@; + $(Q)$(OBJCOPY) $(OBJCOPY_ARGS) \ + --add-section .initrd=$(RAMDISK_IMAGE) \ + --set-section-flags .initrd=contents,alloc,load,load,data \ + arch/$(ARCH)/boot/ramdisk/ramdisk.o $@ + diff --git a/arch/xtensa/configs/common_defconfig b/arch/xtensa/configs/common_defconfig new file mode 100644 index 00000000000..1d230ee081b --- /dev/null +++ b/arch/xtensa/configs/common_defconfig @@ -0,0 +1,662 @@ +# +# Automatically generated make config: don't edit +# Linux kernel version: 2.6.11-rc2 +# Tue Mar 1 16:36:53 2005 +# +# CONFIG_FRAME_POINTER is not set +CONFIG_XTENSA=y +# CONFIG_UID16 is not set +CONFIG_RWSEM_XCHGADD_ALGORITHM=y +CONFIG_HAVE_DEC_LOCK=y +CONFIG_GENERIC_HARDIRQS=y + +# +# Code maturity level options +# +CONFIG_EXPERIMENTAL=y +CONFIG_CLEAN_COMPILE=y +CONFIG_BROKEN_ON_SMP=y + +# +# General setup +# +CONFIG_LOCALVERSION="" +CONFIG_SWAP=y +CONFIG_SYSVIPC=y +# CONFIG_POSIX_MQUEUE is not set +CONFIG_BSD_PROCESS_ACCT=y +# CONFIG_BSD_PROCESS_ACCT_V3 is not set +CONFIG_SYSCTL=y +# CONFIG_AUDIT is not set +CONFIG_LOG_BUF_SHIFT=14 +# CONFIG_HOTPLUG is not set +CONFIG_KOBJECT_UEVENT=y +# CONFIG_IKCONFIG is not set +# CONFIG_EMBEDDED is not set +CONFIG_KALLSYMS=y +# CONFIG_KALLSYMS_ALL is not set +# CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_FUTEX=y +CONFIG_EPOLL=y +# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set +CONFIG_SHMEM=y +CONFIG_CC_ALIGN_FUNCTIONS=0 +CONFIG_CC_ALIGN_LABELS=0 +CONFIG_CC_ALIGN_LOOPS=0 +CONFIG_CC_ALIGN_JUMPS=0 +# CONFIG_TINY_SHMEM is not set + +# +# Loadable module support +# +CONFIG_MODULES=y +# CONFIG_MODULE_UNLOAD is not set +CONFIG_OBSOLETE_MODPARM=y +CONFIG_MODVERSIONS=y +# CONFIG_MODULE_SRCVERSION_ALL is not set +CONFIG_KMOD=y + +# +# Processor type and features +# +CONFIG_XTENSA_ARCH_LINUX_BE=y +# CONFIG_XTENSA_ARCH_LINUX_LE is not set +# CONFIG_XTENSA_ARCH_LINUX_TEST is not set +# CONFIG_XTENSA_ARCH_S5 is not set +# CONFIG_XTENSA_CUSTOM is not set +CONFIG_MMU=y +# CONFIG_XTENSA_UNALIGNED_USER is not set +# CONFIG_PREEMPT is not set +# CONFIG_MATH_EMULATION is not set +# CONFIG_HIGHMEM is not set + +# +# Platform options +# +# CONFIG_XTENSA_PLATFORM_ISS is not set +CONFIG_XTENSA_PLATFORM_XT2000=y +CONFIG_XTENSA_CALIBRATE_CCOUNT=y +CONFIG_GENERIC_CALIBRATE_DELAY=y +CONFIG_CMDLINE_BOOL=y +CONFIG_CMDLINE="console=ttyS0,38400 ip=bootp root=nfs nfsroot=/opt/montavista/pro/devkit/xtensa/linux_be/target" + +# +# Bus options +# +CONFIG_PCI=y +# CONFIG_PCI_LEGACY_PROC is not set +# CONFIG_PCI_NAMES is not set + +# +# PCCARD (PCMCIA/CardBus) support +# +# CONFIG_PCCARD is not set + +# +# PC-card bridges +# + +# +# PCI Hotplug Support +# +# CONFIG_HOTPLUG_PCI is not set + +# +# Exectuable file formats +# +CONFIG_KCORE_ELF=y +CONFIG_BINFMT_ELF=y +CONFIG_BINFMT_MISC=y + +# +# Device Drivers +# + +# +# Generic Driver Options +# +CONFIG_STANDALONE=y +CONFIG_PREVENT_FIRMWARE_BUILD=y +# CONFIG_FW_LOADER is not set +# CONFIG_DEBUG_DRIVER is not set + +# +# Memory Technology Devices (MTD) +# +# CONFIG_MTD is not set + +# +# Parallel port support +# +# CONFIG_PARPORT is not set + +# +# Plug and Play support +# + +# +# Block devices +# +# CONFIG_BLK_DEV_FD is not set +# CONFIG_BLK_CPQ_DA is not set +# CONFIG_BLK_CPQ_CISS_DA is not set +# CONFIG_BLK_DEV_DAC960 is not set +# CONFIG_BLK_DEV_UMEM is not set +# CONFIG_BLK_DEV_COW_COMMON is not set +# CONFIG_BLK_DEV_LOOP is not set +# CONFIG_BLK_DEV_NBD is not set +# CONFIG_BLK_DEV_SX8 is not set +# CONFIG_BLK_DEV_RAM is not set +CONFIG_BLK_DEV_RAM_COUNT=16 +CONFIG_INITRAMFS_SOURCE="" +# CONFIG_CDROM_PKTCDVD is not set + +# +# IO Schedulers +# +CONFIG_IOSCHED_NOOP=y +CONFIG_IOSCHED_AS=y +CONFIG_IOSCHED_DEADLINE=y +CONFIG_IOSCHED_CFQ=y +# CONFIG_ATA_OVER_ETH is not set + +# +# ATA/ATAPI/MFM/RLL support +# +# CONFIG_IDE is not set + +# +# SCSI device support +# +# CONFIG_SCSI is not set + +# +# Multi-device support (RAID and LVM) +# +# CONFIG_MD is not set + +# +# Fusion MPT device support +# + +# +# IEEE 1394 (FireWire) support +# +# CONFIG_IEEE1394 is not set + +# +# I2O device support +# +# CONFIG_I2O is not set + +# +# Networking support +# +CONFIG_NET=y + +# +# Networking options +# +# CONFIG_PACKET is not set +# CONFIG_NETLINK_DEV is not set +CONFIG_UNIX=y +# CONFIG_NET_KEY is not set +CONFIG_INET=y +CONFIG_IP_MULTICAST=y +CONFIG_IP_ADVANCED_ROUTER=y +CONFIG_IP_MULTIPLE_TABLES=y +CONFIG_IP_ROUTE_MULTIPATH=y +CONFIG_IP_ROUTE_VERBOSE=y +CONFIG_IP_PNP=y +CONFIG_IP_PNP_DHCP=y +CONFIG_IP_PNP_BOOTP=y +CONFIG_IP_PNP_RARP=y +# CONFIG_NET_IPIP is not set +# CONFIG_NET_IPGRE is not set +# CONFIG_IP_MROUTE is not set +# CONFIG_ARPD is not set +# CONFIG_SYN_COOKIES is not set +# CONFIG_INET_AH is not set +# CONFIG_INET_ESP is not set +# CONFIG_INET_IPCOMP is not set +# CONFIG_INET_TUNNEL is not set +# CONFIG_IP_TCPDIAG is not set +# CONFIG_IP_TCPDIAG_IPV6 is not set +# CONFIG_IPV6 is not set +# CONFIG_NETFILTER is not set + +# +# SCTP Configuration (EXPERIMENTAL) +# +# CONFIG_IP_SCTP is not set +# CONFIG_ATM is not set +# CONFIG_BRIDGE is not set +# CONFIG_VLAN_8021Q is not set +# CONFIG_DECNET is not set +# CONFIG_LLC2 is not set +# CONFIG_IPX is not set +# CONFIG_ATALK is not set +# CONFIG_X25 is not set +# CONFIG_LAPB is not set +# CONFIG_NET_DIVERT is not set +# CONFIG_ECONET is not set +# CONFIG_WAN_ROUTER is not set + +# +# QoS and/or fair queueing +# +CONFIG_NET_SCHED=y +CONFIG_NET_SCH_CLK_JIFFIES=y +# CONFIG_NET_SCH_CLK_GETTIMEOFDAY is not set +# CONFIG_NET_SCH_CLK_CPU is not set +CONFIG_NET_SCH_CBQ=m +CONFIG_NET_SCH_HTB=m +# CONFIG_NET_SCH_HFSC is not set +CONFIG_NET_SCH_PRIO=m +CONFIG_NET_SCH_RED=m +CONFIG_NET_SCH_SFQ=m +CONFIG_NET_SCH_TEQL=m +CONFIG_NET_SCH_TBF=m +CONFIG_NET_SCH_GRED=m +CONFIG_NET_SCH_DSMARK=m +# CONFIG_NET_SCH_NETEM is not set +CONFIG_NET_SCH_INGRESS=m +CONFIG_NET_QOS=y +CONFIG_NET_ESTIMATOR=y +CONFIG_NET_CLS=y +CONFIG_NET_CLS_TCINDEX=m +CONFIG_NET_CLS_ROUTE4=m +CONFIG_NET_CLS_ROUTE=y +CONFIG_NET_CLS_FW=m +CONFIG_NET_CLS_U32=m +# CONFIG_CLS_U32_PERF is not set +# CONFIG_NET_CLS_IND is not set +CONFIG_NET_CLS_RSVP=m +CONFIG_NET_CLS_RSVP6=m +# CONFIG_NET_CLS_ACT is not set +CONFIG_NET_CLS_POLICE=y + +# +# Network testing +# +# CONFIG_NET_PKTGEN is not set +# CONFIG_NETPOLL is not set +# CONFIG_NET_POLL_CONTROLLER is not set +# CONFIG_HAMRADIO is not set +# CONFIG_IRDA is not set +# CONFIG_BT is not set +CONFIG_NETDEVICES=y +CONFIG_DUMMY=y +# CONFIG_BONDING is not set +# CONFIG_EQUALIZER is not set +# CONFIG_TUN is not set + +# +# ARCnet devices +# +# CONFIG_ARCNET is not set + +# +# Ethernet (10 or 100Mbit) +# +CONFIG_NET_ETHERNET=y +# CONFIG_MII is not set +CONFIG_XT2000_SONIC=y +# CONFIG_HAPPYMEAL is not set +# CONFIG_SUNGEM is not set +# CONFIG_NET_VENDOR_3COM is not set + +# +# Tulip family network device support +# +# CONFIG_NET_TULIP is not set +# CONFIG_HP100 is not set +# CONFIG_NET_PCI is not set + +# +# Ethernet (1000 Mbit) +# +# CONFIG_ACENIC is not set +# CONFIG_DL2K is not set +# CONFIG_E1000 is not set +# CONFIG_NS83820 is not set +# CONFIG_HAMACHI is not set +# CONFIG_YELLOWFIN is not set +# CONFIG_R8169 is not set +# CONFIG_SK98LIN is not set +# CONFIG_TIGON3 is not set + +# +# Ethernet (10000 Mbit) +# +# CONFIG_IXGB is not set +# CONFIG_S2IO is not set + +# +# Token Ring devices +# +# CONFIG_TR is not set + +# +# Wireless LAN (non-hamradio) +# +CONFIG_NET_RADIO=y + +# +# Obsolete Wireless cards support (pre-802.11) +# +CONFIG_STRIP=m + +# +# Wireless 802.11b ISA/PCI cards support +# +CONFIG_HERMES=m +# CONFIG_PLX_HERMES is not set +# CONFIG_TMD_HERMES is not set +# CONFIG_PCI_HERMES is not set +# CONFIG_ATMEL is not set + +# +# Prism GT/Duette 802.11(a/b/g) PCI/Cardbus support +# +# CONFIG_PRISM54 is not set +CONFIG_NET_WIRELESS=y + +# +# Wan interfaces +# +# CONFIG_WAN is not set +# CONFIG_FDDI is not set +# CONFIG_HIPPI is not set +# CONFIG_PPP is not set +# CONFIG_SLIP is not set +# CONFIG_SHAPER is not set +# CONFIG_NETCONSOLE is not set + +# +# ISDN subsystem +# +# CONFIG_ISDN is not set + +# +# Telephony Support +# +# CONFIG_PHONE is not set + +# +# Input device support +# +CONFIG_INPUT=y + +# +# Userland interfaces +# +CONFIG_INPUT_MOUSEDEV=y +# CONFIG_INPUT_MOUSEDEV_PSAUX is not set +CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024 +CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768 +# CONFIG_INPUT_JOYDEV is not set +# CONFIG_INPUT_TSDEV is not set +# CONFIG_INPUT_EVDEV is not set +# CONFIG_INPUT_EVBUG is not set + +# +# Input I/O drivers +# +# CONFIG_GAMEPORT is not set +CONFIG_SOUND_GAMEPORT=y +CONFIG_SERIO=y +# CONFIG_SERIO_I8042 is not set +# CONFIG_SERIO_SERPORT is not set +# CONFIG_SERIO_CT82C710 is not set +# CONFIG_SERIO_PCIPS2 is not set +# CONFIG_SERIO_RAW is not set + +# +# Input Device Drivers +# +# CONFIG_INPUT_KEYBOARD is not set +# CONFIG_INPUT_MOUSE is not set +# CONFIG_INPUT_JOYSTICK is not set +# CONFIG_INPUT_TOUCHSCREEN is not set +# CONFIG_INPUT_MISC is not set + +# +# Character devices +# +CONFIG_VT=y +CONFIG_VT_CONSOLE=y +CONFIG_HW_CONSOLE=y +# CONFIG_SERIAL_NONSTANDARD is not set + +# +# Serial drivers +# +CONFIG_SERIAL_8250=y +CONFIG_SERIAL_8250_CONSOLE=y +CONFIG_SERIAL_8250_NR_UARTS=4 +# CONFIG_SERIAL_8250_EXTENDED is not set + +# +# Non-8250 serial port support +# +CONFIG_SERIAL_CORE=y +CONFIG_SERIAL_CORE_CONSOLE=y +CONFIG_UNIX98_PTYS=y +CONFIG_LEGACY_PTYS=y +CONFIG_LEGACY_PTY_COUNT=256 + +# +# IPMI +# +# CONFIG_IPMI_HANDLER is not set + +# +# Watchdog Cards +# +# CONFIG_WATCHDOG is not set +# CONFIG_RTC is not set +# CONFIG_GEN_RTC is not set +# CONFIG_DTLK is not set +# CONFIG_R3964 is not set +# CONFIG_APPLICOM is not set + +# +# Ftape, the floppy tape device driver +# +# CONFIG_DRM is not set +# CONFIG_RAW_DRIVER is not set + +# +# I2C support +# +# CONFIG_I2C is not set + +# +# Dallas's 1-wire bus +# +# CONFIG_W1 is not set + +# +# Misc devices +# + +# +# Multimedia devices +# +# CONFIG_VIDEO_DEV is not set + +# +# Digital Video Broadcasting Devices +# +# CONFIG_DVB is not set + +# +# Graphics support +# +# CONFIG_FB is not set + +# +# Console display driver support +# +# CONFIG_VGA_CONSOLE is not set +CONFIG_DUMMY_CONSOLE=y +# CONFIG_BACKLIGHT_LCD_SUPPORT is not set + +# +# Sound +# +# CONFIG_SOUND is not set + +# +# USB support +# +# CONFIG_USB is not set +CONFIG_USB_ARCH_HAS_HCD=y +CONFIG_USB_ARCH_HAS_OHCI=y + +# +# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support' may also be needed; see USB_STORAGE Help for more information +# + +# +# USB Gadget Support +# +# CONFIG_USB_GADGET is not set + +# +# MMC/SD Card support +# +# CONFIG_MMC is not set + +# +# InfiniBand support +# +# CONFIG_INFINIBAND is not set + +# +# File systems +# +# CONFIG_EXT2_FS is not set +# CONFIG_EXT3_FS is not set +# CONFIG_JBD is not set +# CONFIG_REISERFS_FS is not set +# CONFIG_JFS_FS is not set +# CONFIG_XFS_FS is not set +# CONFIG_MINIX_FS is not set +# CONFIG_ROMFS_FS is not set +# CONFIG_QUOTA is not set +CONFIG_DNOTIFY=y +# CONFIG_AUTOFS_FS is not set +# CONFIG_AUTOFS4_FS is not set + +# +# CD-ROM/DVD Filesystems +# +# CONFIG_ISO9660_FS is not set +# CONFIG_UDF_FS is not set + +# +# DOS/FAT/NT Filesystems +# +# CONFIG_MSDOS_FS is not set +# CONFIG_VFAT_FS is not set +# CONFIG_NTFS_FS is not set + +# +# Pseudo filesystems +# +CONFIG_PROC_FS=y +# CONFIG_PROC_KCORE is not set +CONFIG_SYSFS=y +CONFIG_DEVFS_FS=y +# CONFIG_DEVFS_MOUNT is not set +# CONFIG_DEVFS_DEBUG is not set +# CONFIG_DEVPTS_FS_XATTR is not set +# CONFIG_TMPFS is not set +# CONFIG_HUGETLB_PAGE is not set +CONFIG_RAMFS=y + +# +# Miscellaneous filesystems +# +# CONFIG_ADFS_FS is not set +# CONFIG_AFFS_FS is not set +# CONFIG_HFS_FS is not set +# CONFIG_HFSPLUS_FS is not set +# CONFIG_BEFS_FS is not set +# CONFIG_BFS_FS is not set +# CONFIG_EFS_FS is not set +# CONFIG_CRAMFS is not set +# CONFIG_VXFS_FS is not set +# CONFIG_HPFS_FS is not set +# CONFIG_QNX4FS_FS is not set +# CONFIG_SYSV_FS is not set +# CONFIG_UFS_FS is not set + +# +# Network File Systems +# +CONFIG_NFS_FS=y +CONFIG_NFS_V3=y +# CONFIG_NFS_V4 is not set +# CONFIG_NFS_DIRECTIO is not set +# CONFIG_NFSD is not set +CONFIG_ROOT_NFS=y +CONFIG_LOCKD=y +CONFIG_LOCKD_V4=y +# CONFIG_EXPORTFS is not set +CONFIG_SUNRPC=y +# CONFIG_RPCSEC_GSS_KRB5 is not set +# CONFIG_RPCSEC_GSS_SPKM3 is not set +# CONFIG_SMB_FS is not set +# CONFIG_CIFS is not set +# CONFIG_NCP_FS is not set +# CONFIG_CODA_FS is not set +# CONFIG_AFS_FS is not set + +# +# Partition Types +# +# CONFIG_PARTITION_ADVANCED is not set +CONFIG_MSDOS_PARTITION=y + +# +# Native Language Support +# +# CONFIG_NLS is not set + +# +# Kernel hacking +# +CONFIG_DEBUG_KERNEL=y +# CONFIG_DEBUG_STACKOVERFLOW is not set +# CONFIG_DEBUG_SLAB is not set +CONFIG_MAGIC_SYSRQ=y +# CONFIG_DEBUG_SPINLOCK is not set +# CONFIG_DEBUG_PAGEALLOC is not set +# CONFIG_DEBUG_INFO is not set +# CONFIG_DEBUG_SPINLOCK_SLEEP is not set +# CONFIG_KGDB is not set + +# +# Security options +# +# CONFIG_KEYS is not set +# CONFIG_SECURITY is not set + +# +# Cryptographic options +# +# CONFIG_CRYPTO is not set + +# +# Hardware crypto devices +# + +# +# Library routines +# +# CONFIG_CRC_CCITT is not set +# CONFIG_CRC32 is not set +# CONFIG_LIBCRC32C is not set diff --git a/arch/xtensa/configs/iss_defconfig b/arch/xtensa/configs/iss_defconfig new file mode 100644 index 00000000000..802621dd486 --- /dev/null +++ b/arch/xtensa/configs/iss_defconfig @@ -0,0 +1,531 @@ +# +# Automatically generated make config: don't edit +# Linux kernel version: 2.6.11-rc2 +# Fri Feb 25 19:21:24 2005 +# +CONFIG_FRAME_POINTER=y +CONFIG_XTENSA=y +# CONFIG_UID16 is not set +CONFIG_RWSEM_XCHGADD_ALGORITHM=y +CONFIG_HAVE_DEC_LOCK=y +CONFIG_GENERIC_HARDIRQS=y + +# +# Code maturity level options +# +CONFIG_EXPERIMENTAL=y +CONFIG_CLEAN_COMPILE=y +CONFIG_BROKEN_ON_SMP=y + +# +# General setup +# +CONFIG_LOCALVERSION="" +CONFIG_SWAP=y +CONFIG_SYSVIPC=y +# CONFIG_POSIX_MQUEUE is not set +# CONFIG_BSD_PROCESS_ACCT is not set +CONFIG_SYSCTL=y +# CONFIG_AUDIT is not set +CONFIG_LOG_BUF_SHIFT=14 +# CONFIG_HOTPLUG is not set +# CONFIG_KOBJECT_UEVENT is not set +# CONFIG_IKCONFIG is not set +CONFIG_EMBEDDED=y +CONFIG_KALLSYMS=y +# CONFIG_KALLSYMS_ALL is not set +# CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_FUTEX=y +CONFIG_EPOLL=y +# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set +CONFIG_SHMEM=y +CONFIG_CC_ALIGN_FUNCTIONS=0 +CONFIG_CC_ALIGN_LABELS=0 +CONFIG_CC_ALIGN_LOOPS=0 +CONFIG_CC_ALIGN_JUMPS=0 +# CONFIG_TINY_SHMEM is not set + +# +# Loadable module support +# +# CONFIG_MODULES is not set + +# +# Processor type and features +# +CONFIG_XTENSA_ARCH_LINUX_BE=y +# CONFIG_XTENSA_ARCH_LINUX_LE is not set +# CONFIG_XTENSA_ARCH_LINUX_TEST is not set +# CONFIG_XTENSA_ARCH_S5 is not set +# CONFIG_XTENSA_CUSTOM is not set +CONFIG_MMU=y +# CONFIG_XTENSA_UNALIGNED_USER is not set +# CONFIG_PREEMPT is not set +# CONFIG_MATH_EMULATION is not set +# CONFIG_HIGHMEM is not set + +# +# Platform options +# +CONFIG_XTENSA_PLATFORM_ISS=y +# CONFIG_XTENSA_PLATFORM_XT2000 is not set +# CONFIG_XTENSA_PLATFORM_ARUBA is not set +# CONFIG_XTENSA_CALIBRATE_CCOUNT is not set +CONFIG_XTENSA_CPU_CLOCK=10 +# CONFIG_GENERIC_CALIBRATE_DELAY is not set +CONFIG_CMDLINE_BOOL=y +CONFIG_CMDLINE="console=ttyS0,38400 eth0=tuntap,,tap0 ip=192.168.168.5:192.168.168.1 root=nfs nfsroot=192.168.168.1:/opt/montavista/pro/devkit/xtensa/linux_be/target" +CONFIG_SERIAL_CONSOLE=y +CONFIG_XTENSA_ISS_NETWORK=y + +# +# Bus options +# + +# +# PCCARD (PCMCIA/CardBus) support +# +# CONFIG_PCCARD is not set + +# +# PC-card bridges +# + +# +# PCI Hotplug Support +# + +# +# Exectuable file formats +# +CONFIG_KCORE_ELF=y +CONFIG_BINFMT_ELF=y +# CONFIG_BINFMT_MISC is not set + +# +# Device Drivers +# + +# +# Generic Driver Options +# +# CONFIG_STANDALONE is not set +CONFIG_PREVENT_FIRMWARE_BUILD=y +# CONFIG_FW_LOADER is not set +# CONFIG_DEBUG_DRIVER is not set + +# +# Memory Technology Devices (MTD) +# +# CONFIG_MTD is not set + +# +# Parallel port support +# +# CONFIG_PARPORT is not set + +# +# Plug and Play support +# + +# +# Block devices +# +# CONFIG_BLK_DEV_FD is not set +# CONFIG_BLK_DEV_COW_COMMON is not set +# CONFIG_BLK_DEV_LOOP is not set +# CONFIG_BLK_DEV_NBD is not set +# CONFIG_BLK_DEV_RAM is not set +CONFIG_BLK_DEV_RAM_COUNT=16 +CONFIG_INITRAMFS_SOURCE="" +# CONFIG_CDROM_PKTCDVD is not set + +# +# IO Schedulers +# +CONFIG_IOSCHED_NOOP=y +# CONFIG_IOSCHED_AS is not set +# CONFIG_IOSCHED_DEADLINE is not set +# CONFIG_IOSCHED_CFQ is not set +# CONFIG_ATA_OVER_ETH is not set + +# +# ATA/ATAPI/MFM/RLL support +# +# CONFIG_IDE is not set + +# +# SCSI device support +# +# CONFIG_SCSI is not set + +# +# Multi-device support (RAID and LVM) +# +# CONFIG_MD is not set + +# +# Fusion MPT device support +# + +# +# IEEE 1394 (FireWire) support +# + +# +# I2O device support +# + +# +# Networking support +# +CONFIG_NET=y + +# +# Networking options +# +CONFIG_PACKET=y +# CONFIG_PACKET_MMAP is not set +# CONFIG_NETLINK_DEV is not set +CONFIG_UNIX=y +# CONFIG_NET_KEY is not set +CONFIG_INET=y +# CONFIG_IP_MULTICAST is not set +# CONFIG_IP_ADVANCED_ROUTER is not set +CONFIG_IP_PNP=y +CONFIG_IP_PNP_DHCP=y +CONFIG_IP_PNP_BOOTP=y +CONFIG_IP_PNP_RARP=y +# CONFIG_NET_IPIP is not set +# CONFIG_NET_IPGRE is not set +# CONFIG_ARPD is not set +# CONFIG_SYN_COOKIES is not set +# CONFIG_INET_AH is not set +# CONFIG_INET_ESP is not set +# CONFIG_INET_IPCOMP is not set +# CONFIG_INET_TUNNEL is not set +# CONFIG_IP_TCPDIAG is not set +# CONFIG_IP_TCPDIAG_IPV6 is not set +# CONFIG_IPV6 is not set +# CONFIG_NETFILTER is not set + +# +# SCTP Configuration (EXPERIMENTAL) +# +# CONFIG_IP_SCTP is not set +# CONFIG_SCTP_HMAC_NONE is not set +# CONFIG_SCTP_HMAC_SHA1 is not set +# CONFIG_SCTP_HMAC_MD5 is not set +# CONFIG_ATM is not set +# CONFIG_BRIDGE is not set +# CONFIG_VLAN_8021Q is not set +# CONFIG_DECNET is not set +# CONFIG_LLC2 is not set +# CONFIG_IPX is not set +# CONFIG_ATALK is not set +# CONFIG_X25 is not set +# CONFIG_LAPB is not set +# CONFIG_NET_DIVERT is not set +# CONFIG_ECONET is not set +# CONFIG_WAN_ROUTER is not set + +# +# QoS and/or fair queueing +# +# CONFIG_NET_SCHED is not set +# CONFIG_NET_SCH_CLK_JIFFIES is not set +# CONFIG_NET_SCH_CLK_GETTIMEOFDAY is not set +# CONFIG_NET_SCH_CLK_CPU is not set +# CONFIG_NET_CLS_ROUTE is not set + +# +# Network testing +# +# CONFIG_NET_PKTGEN is not set +# CONFIG_NETPOLL is not set +# CONFIG_NET_POLL_CONTROLLER is not set +# CONFIG_HAMRADIO is not set +# CONFIG_IRDA is not set +# CONFIG_BT is not set +# CONFIG_NETDEVICES is not set + +# +# ISDN subsystem +# +# CONFIG_ISDN is not set + +# +# Telephony Support +# +# CONFIG_PHONE is not set + +# +# Input device support +# +CONFIG_INPUT=y + +# +# Userland interfaces +# +# CONFIG_INPUT_MOUSEDEV is not set +# CONFIG_INPUT_JOYDEV is not set +# CONFIG_INPUT_TSDEV is not set +# CONFIG_INPUT_EVDEV is not set +# CONFIG_INPUT_EVBUG is not set + +# +# Input I/O drivers +# +# CONFIG_GAMEPORT is not set +CONFIG_SOUND_GAMEPORT=y +# CONFIG_SERIO is not set +# CONFIG_SERIO_I8042 is not set + +# +# Input Device Drivers +# +# CONFIG_INPUT_KEYBOARD is not set +# CONFIG_INPUT_MOUSE is not set +# CONFIG_INPUT_JOYSTICK is not set +# CONFIG_INPUT_TOUCHSCREEN is not set +# CONFIG_INPUT_MISC is not set + +# +# Character devices +# +CONFIG_VT=y +CONFIG_VT_CONSOLE=y +CONFIG_HW_CONSOLE=y +# CONFIG_SERIAL_NONSTANDARD is not set + +# +# Serial drivers +# +# CONFIG_SERIAL_8250 is not set + +# +# Non-8250 serial port support +# +CONFIG_UNIX98_PTYS=y +CONFIG_LEGACY_PTYS=y +CONFIG_LEGACY_PTY_COUNT=256 + +# +# IPMI +# +# CONFIG_IPMI_HANDLER is not set + +# +# Watchdog Cards +# +CONFIG_WATCHDOG=y +CONFIG_WATCHDOG_NOWAYOUT=y + +# +# Watchdog Device Drivers +# +CONFIG_SOFT_WATCHDOG=y +# CONFIG_RTC is not set +# CONFIG_GEN_RTC is not set +# CONFIG_DTLK is not set +# CONFIG_R3964 is not set + +# +# Ftape, the floppy tape device driver +# +# CONFIG_DRM is not set +# CONFIG_RAW_DRIVER is not set + +# +# I2C support +# +# CONFIG_I2C is not set + +# +# Dallas's 1-wire bus +# +# CONFIG_W1 is not set + +# +# Misc devices +# + +# +# Multimedia devices +# +# CONFIG_VIDEO_DEV is not set + +# +# Digital Video Broadcasting Devices +# +# CONFIG_DVB is not set + +# +# Graphics support +# +# CONFIG_FB is not set + +# +# Console display driver support +# +# CONFIG_VGA_CONSOLE is not set +CONFIG_DUMMY_CONSOLE=y +# CONFIG_BACKLIGHT_LCD_SUPPORT is not set + +# +# Sound +# +# CONFIG_SOUND is not set + +# +# USB support +# +# CONFIG_USB_ARCH_HAS_HCD is not set +# CONFIG_USB_ARCH_HAS_OHCI is not set + +# +# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support' may also be needed; see USB_STORAGE Help for more information +# + +# +# USB Gadget Support +# +# CONFIG_USB_GADGET is not set + +# +# MMC/SD Card support +# +# CONFIG_MMC is not set + +# +# InfiniBand support +# +# CONFIG_INFINIBAND is not set + +# +# File systems +# +# CONFIG_EXT2_FS is not set +# CONFIG_EXT3_FS is not set +# CONFIG_JBD is not set +# CONFIG_REISERFS_FS is not set +# CONFIG_JFS_FS is not set +# CONFIG_XFS_FS is not set +# CONFIG_MINIX_FS is not set +# CONFIG_ROMFS_FS is not set +# CONFIG_QUOTA is not set +# CONFIG_DNOTIFY is not set +# CONFIG_AUTOFS_FS is not set +# CONFIG_AUTOFS4_FS is not set + +# +# CD-ROM/DVD Filesystems +# +# CONFIG_ISO9660_FS is not set +# CONFIG_UDF_FS is not set + +# +# DOS/FAT/NT Filesystems +# +# CONFIG_MSDOS_FS is not set +# CONFIG_VFAT_FS is not set +# CONFIG_NTFS_FS is not set + +# +# Pseudo filesystems +# +CONFIG_PROC_FS=y +CONFIG_PROC_KCORE=y +CONFIG_SYSFS=y +CONFIG_DEVFS_FS=y +CONFIG_DEVFS_MOUNT=y +# CONFIG_DEVFS_DEBUG is not set +# CONFIG_DEVPTS_FS_XATTR is not set +CONFIG_TMPFS=y +# CONFIG_TMPFS_XATTR is not set +# CONFIG_HUGETLB_PAGE is not set +CONFIG_RAMFS=y + +# +# Miscellaneous filesystems +# +# CONFIG_ADFS_FS is not set +# CONFIG_AFFS_FS is not set +# CONFIG_HFS_FS is not set +# CONFIG_HFSPLUS_FS is not set +# CONFIG_BEFS_FS is not set +# CONFIG_BFS_FS is not set +# CONFIG_EFS_FS is not set +# CONFIG_CRAMFS is not set +# CONFIG_VXFS_FS is not set +# CONFIG_HPFS_FS is not set +# CONFIG_QNX4FS_FS is not set +# CONFIG_SYSV_FS is not set +# CONFIG_UFS_FS is not set + +# +# Network File Systems +# +CONFIG_NFS_FS=y +CONFIG_NFS_V3=y +# CONFIG_NFS_V4 is not set +CONFIG_NFS_DIRECTIO=y +# CONFIG_NFSD is not set +CONFIG_ROOT_NFS=y +CONFIG_LOCKD=y +CONFIG_LOCKD_V4=y +# CONFIG_EXPORTFS is not set +CONFIG_SUNRPC=y +# CONFIG_RPCSEC_GSS_KRB5 is not set +# CONFIG_RPCSEC_GSS_SPKM3 is not set +# CONFIG_SMB_FS is not set +# CONFIG_CIFS is not set +# CONFIG_NCP_FS is not set +# CONFIG_CODA_FS is not set +# CONFIG_AFS_FS is not set + +# +# Partition Types +# +# CONFIG_PARTITION_ADVANCED is not set +CONFIG_MSDOS_PARTITION=y + +# +# Native Language Support +# +# CONFIG_NLS is not set + +# +# Kernel hacking +# +CONFIG_DEBUG_KERNEL=y +# CONFIG_DEBUG_STACKOVERFLOW is not set +# CONFIG_DEBUG_SLAB is not set +# CONFIG_MAGIC_SYSRQ is not set +# CONFIG_DEBUG_SPINLOCK is not set +# CONFIG_DEBUG_PAGEALLOC is not set +# CONFIG_DEBUG_INFO is not set +# CONFIG_DEBUG_SPINLOCK_SLEEP is not set +# CONFIG_KGDB is not set + +# +# Security options +# +# CONFIG_KEYS is not set +# CONFIG_SECURITY is not set + +# +# Cryptographic options +# +# CONFIG_CRYPTO is not set + +# +# Hardware crypto devices +# + +# +# Library routines +# +# CONFIG_CRC_CCITT is not set +# CONFIG_CRC32 is not set +# CONFIG_LIBCRC32C is not set diff --git a/arch/xtensa/kernel/Makefile b/arch/xtensa/kernel/Makefile new file mode 100644 index 00000000000..d573017a5dd --- /dev/null +++ b/arch/xtensa/kernel/Makefile @@ -0,0 +1,18 @@ +# +# Makefile for the Linux/Xtensa kernel. +# + +extra-y := head.o vmlinux.lds + + +obj-y := align.o entry.o irq.o coprocessor.o process.o ptrace.o semaphore.o \ + setup.o signal.o syscalls.o time.o traps.o vectors.o platform.o \ + pci-dma.o + +## windowspill.o + +obj-$(CONFIG_KGDB) += xtensa-stub.o +obj-$(CONFIG_PCI) += pci.o +obj-$(CONFIG_MODULES) += xtensa_ksyms.o module.o + + diff --git a/arch/xtensa/kernel/align.S b/arch/xtensa/kernel/align.S new file mode 100644 index 00000000000..74b1e90ef08 --- /dev/null +++ b/arch/xtensa/kernel/align.S @@ -0,0 +1,459 @@ +/* + * arch/xtensa/kernel/align.S + * + * Handle unalignment exceptions in kernel space. + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Copyright (C) 2001 - 2005 Tensilica, Inc. + * + * Rewritten by Chris Zankel <chris@zankel.net> + * + * Based on work from Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * and Marc Gauthier <marc@tensilica.com, marc@alimni.uwaterloo.ca> + */ + +#include <linux/linkage.h> +#include <asm/ptrace.h> +#include <asm/ptrace.h> +#include <asm/current.h> +#include <asm/offsets.h> +#include <asm/pgtable.h> +#include <asm/processor.h> +#include <asm/page.h> +#include <asm/thread_info.h> + +#if XCHAL_UNALIGNED_LOAD_EXCEPTION || XCHAL_UNALIGNED_STORE_EXCEPTION + +/* First-level exception handler for unaligned exceptions. + * + * Note: This handler works only for kernel exceptions. Unaligned user + * access should get a seg fault. + */ + +/* Big and little endian 16-bit values are located in + * different halves of a register. HWORD_START helps to + * abstract the notion of extracting a 16-bit value from a + * register. + * We also have to define new shifting instructions because + * lsb and msb are on 'opposite' ends in a register for + * different endian machines. + * + * Assume a memory region in ascending address: + * 0 1 2 3|4 5 6 7 + * + * When loading one word into a register, the content of that register is: + * LE 3 2 1 0, 7 6 5 4 + * BE 0 1 2 3, 4 5 6 7 + * + * Masking the bits of the higher/lower address means: + * LE X X 0 0, 0 0 X X + * BE 0 0 X X, X X 0 0 + * + * Shifting to higher/lower addresses, means: + * LE shift left / shift right + * BE shift right / shift left + * + * Extracting 16 bits from a 32 bit reg. value to higher/lower address means: + * LE mask 0 0 X X / shift left + * BE shift left / mask 0 0 X X + */ + +#define UNALIGNED_USER_EXCEPTION + +#if XCHAL_HAVE_BE + +#define HWORD_START 16 +#define INSN_OP0 28 +#define INSN_T 24 +#define INSN_OP1 16 + +.macro __src_b r, w0, w1; src \r, \w0, \w1; .endm +.macro __ssa8 r; ssa8b \r; .endm +.macro __ssa8r r; ssa8l \r; .endm +.macro __sh r, s; srl \r, \s; .endm +.macro __sl r, s; sll \r, \s; .endm +.macro __exth r, s; extui \r, \s, 0, 16; .endm +.macro __extl r, s; slli \r, \s, 16; .endm + +#else + +#define HWORD_START 0 +#define INSN_OP0 0 +#define INSN_T 4 +#define INSN_OP1 12 + +.macro __src_b r, w0, w1; src \r, \w1, \w0; .endm +.macro __ssa8 r; ssa8l \r; .endm +.macro __ssa8r r; ssa8b \r; .endm +.macro __sh r, s; sll \r, \s; .endm +.macro __sl r, s; srl \r, \s; .endm +.macro __exth r, s; slli \r, \s, 16; .endm +.macro __extl r, s; extui \r, \s, 0, 16; .endm + +#endif + +/* + * xxxx xxxx = imm8 field + * yyyy = imm4 field + * ssss = s field + * tttt = t field + * + * 16 0 + * ------------------- + * L32I.N yyyy ssss tttt 1000 + * S32I.N yyyy ssss tttt 1001 + * + * 23 0 + * ----------------------------- + * res 0000 0010 + * L16UI xxxx xxxx 0001 ssss tttt 0010 + * L32I xxxx xxxx 0010 ssss tttt 0010 + * XXX 0011 ssss tttt 0010 + * XXX 0100 ssss tttt 0010 + * S16I xxxx xxxx 0101 ssss tttt 0010 + * S32I xxxx xxxx 0110 ssss tttt 0010 + * XXX 0111 ssss tttt 0010 + * XXX 1000 ssss tttt 0010 + * L16SI xxxx xxxx 1001 ssss tttt 0010 + * XXX 1010 0010 + * **L32AI xxxx xxxx 1011 ssss tttt 0010 unsupported + * XXX 1100 0010 + * XXX 1101 0010 + * XXX 1110 0010 + * **S32RI xxxx xxxx 1111 ssss tttt 0010 unsupported + * ----------------------------- + * ^ ^ ^ + * sub-opcode (NIBBLE_R) -+ | | + * t field (NIBBLE_T) -----------+ | + * major opcode (NIBBLE_OP0) --------------+ + */ + +#define OP0_L32I_N 0x8 /* load immediate narrow */ +#define OP0_S32I_N 0x9 /* store immediate narrow */ +#define OP1_SI_MASK 0x4 /* OP1 bit set for stores */ +#define OP1_SI_BIT 2 /* OP1 bit number for stores */ + +#define OP1_L32I 0x2 +#define OP1_L16UI 0x1 +#define OP1_L16SI 0x9 +#define OP1_L32AI 0xb + +#define OP1_S32I 0x6 +#define OP1_S16I 0x5 +#define OP1_S32RI 0xf + +/* + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + */ + + +ENTRY(fast_unaligned) + + /* Note: We don't expect the address to be aligned on a word + * boundary. After all, the processor generated that exception + * and it would be a hardware fault. + */ + + /* Save some working register */ + + s32i a4, a2, PT_AREG4 + s32i a5, a2, PT_AREG5 + s32i a6, a2, PT_AREG6 + s32i a7, a2, PT_AREG7 + s32i a8, a2, PT_AREG8 + + rsr a0, DEPC + xsr a3, EXCSAVE_1 + s32i a0, a2, PT_AREG2 + s32i a3, a2, PT_AREG3 + + /* Keep value of SAR in a0 */ + + rsr a0, SAR + rsr a8, EXCVADDR # load unaligned memory address + + /* Now, identify one of the following load/store instructions. + * + * The only possible danger of a double exception on the + * following l32i instructions is kernel code in vmalloc + * memory. The processor was just executing at the EPC_1 + * address, and indeed, already fetched the instruction. That + * guarantees a TLB mapping, which hasn't been replaced by + * this unaligned exception handler that uses only static TLB + * mappings. However, high-level interrupt handlers might + * modify TLB entries, so for the generic case, we register a + * TABLE_FIXUP handler here, too. + */ + + /* a3...a6 saved on stack, a2 = SP */ + + /* Extract the instruction that caused the unaligned access. */ + + rsr a7, EPC_1 # load exception address + movi a3, ~3 + and a3, a3, a7 # mask lower bits + + l32i a4, a3, 0 # load 2 words + l32i a5, a3, 4 + + __ssa8 a7 + __src_b a4, a4, a5 # a4 has the instruction + + /* Analyze the instruction (load or store?). */ + + extui a5, a4, INSN_OP0, 4 # get insn.op0 nibble + +#if XCHAL_HAVE_NARROW + _beqi a5, OP0_L32I_N, .Lload # L32I.N, jump + addi a6, a5, -OP0_S32I_N + _beqz a6, .Lstore # S32I.N, do a store +#endif + /* 'store indicator bit' not set, jump */ + _bbci.l a4, OP1_SI_BIT + INSN_OP1, .Lload + + /* Store: Jump to table entry to get the value in the source register.*/ + +.Lstore:movi a5, .Lstore_table # table + extui a6, a4, INSN_T, 4 # get source register + addx8 a5, a6, a5 + jx a5 # jump into table + + /* Invalid instruction, CRITICAL! */ +.Linvalid_instruction_load: + j .Linvalid_instruction + + /* Load: Load memory address. */ + +.Lload: movi a3, ~3 + and a3, a3, a8 # align memory address + + __ssa8 a8 +#ifdef UNALIGNED_USER_EXCEPTION + addi a3, a3, 8 + l32e a5, a3, -8 + l32e a6, a3, -4 +#else + l32i a5, a3, 0 + l32i a6, a3, 4 +#endif + __src_b a3, a5, a6 # a3 has the data word + +#if XCHAL_HAVE_NARROW + addi a7, a7, 2 # increment PC (assume 16-bit insn) + + extui a5, a4, INSN_OP0, 4 + _beqi a5, OP0_L32I_N, 1f # l32i.n: jump + + addi a7, a7, 1 +#else + addi a7, a7, 3 +#endif + + extui a5, a4, INSN_OP1, 4 + _beqi a5, OP1_L32I, 1f # l32i: jump + + extui a3, a3, 0, 16 # extract lower 16 bits + _beqi a5, OP1_L16UI, 1f + addi a5, a5, -OP1_L16SI + _bnez a5, .Linvalid_instruction_load + + /* sign extend value */ + + slli a3, a3, 16 + srai a3, a3, 16 + + /* Set target register. */ + +1: + +#if XCHAL_HAVE_LOOP + rsr a3, LEND # check if we reached LEND + bne a7, a3, 1f + rsr a3, LCOUNT # and LCOUNT != 0 + beqz a3, 1f + addi a3, a3, -1 # decrement LCOUNT and set + rsr a7, LBEG # set PC to LBEGIN + wsr a3, LCOUNT +#endif + +1: wsr a7, EPC_1 # skip load instruction + extui a4, a4, INSN_T, 4 # extract target register + movi a5, .Lload_table + addx8 a4, a4, a5 + jx a4 # jump to entry for target register + + .align 8 +.Lload_table: + s32i a3, a2, PT_AREG0; _j .Lexit; .align 8 + mov a1, a3; _j .Lexit; .align 8 # fishy?? + s32i a3, a2, PT_AREG2; _j .Lexit; .align 8 + s32i a3, a2, PT_AREG3; _j .Lexit; .align 8 + s32i a3, a2, PT_AREG4; _j .Lexit; .align 8 + s32i a3, a2, PT_AREG5; _j .Lexit; .align 8 + s32i a3, a2, PT_AREG6; _j .Lexit; .align 8 + s32i a3, a2, PT_AREG7; _j .Lexit; .align 8 + s32i a3, a2, PT_AREG8; _j .Lexit; .align 8 + mov a9, a3 ; _j .Lexit; .align 8 + mov a10, a3 ; _j .Lexit; .align 8 + mov a11, a3 ; _j .Lexit; .align 8 + mov a12, a3 ; _j .Lexit; .align 8 + mov a13, a3 ; _j .Lexit; .align 8 + mov a14, a3 ; _j .Lexit; .align 8 + mov a15, a3 ; _j .Lexit; .align 8 + +.Lstore_table: + l32i a3, a2, PT_AREG0; _j 1f; .align 8 + mov a3, a1; _j 1f; .align 8 # fishy?? + l32i a3, a2, PT_AREG2; _j 1f; .align 8 + l32i a3, a2, PT_AREG3; _j 1f; .align 8 + l32i a3, a2, PT_AREG4; _j 1f; .align 8 + l32i a3, a2, PT_AREG5; _j 1f; .align 8 + l32i a3, a2, PT_AREG6; _j 1f; .align 8 + l32i a3, a2, PT_AREG7; _j 1f; .align 8 + l32i a3, a2, PT_AREG8; _j 1f; .align 8 + mov a3, a9 ; _j 1f; .align 8 + mov a3, a10 ; _j 1f; .align 8 + mov a3, a11 ; _j 1f; .align 8 + mov a3, a12 ; _j 1f; .align 8 + mov a3, a13 ; _j 1f; .align 8 + mov a3, a14 ; _j 1f; .align 8 + mov a3, a15 ; _j 1f; .align 8 + +1: # a7: instruction pointer, a4: instruction, a3: value + + movi a6, 0 # mask: ffffffff:00000000 + +#if XCHAL_HAVE_NARROW + addi a7, a7, 2 # incr. PC,assume 16-bit instruction + + extui a5, a4, INSN_OP0, 4 # extract OP0 + addi a5, a5, -OP0_S32I_N + _beqz a5, 1f # s32i.n: jump + + addi a7, a7, 1 # increment PC, 32-bit instruction +#else + addi a7, a7, 3 # increment PC, 32-bit instruction +#endif + + extui a5, a4, INSN_OP1, 4 # extract OP1 + _beqi a5, OP1_S32I, 1f # jump if 32 bit store + _bnei a5, OP1_S16I, .Linvalid_instruction_store + + movi a5, -1 + __extl a3, a3 # get 16-bit value + __exth a6, a5 # get 16-bit mask ffffffff:ffff0000 + + /* Get memory address */ + +1: +#if XCHAL_HAVE_LOOP + rsr a3, LEND # check if we reached LEND + bne a7, a3, 1f + rsr a3, LCOUNT # and LCOUNT != 0 + beqz a3, 1f + addi a3, a3, -1 # decrement LCOUNT and set + rsr a7, LBEG # set PC to LBEGIN + wsr a3, LCOUNT +#endif + +1: wsr a7, EPC_1 # skip store instruction + movi a4, ~3 + and a4, a4, a8 # align memory address + + /* Insert value into memory */ + + movi a5, -1 # mask: ffffffff:XXXX0000 +#ifdef UNALIGNED_USER_EXCEPTION + addi a4, a4, 8 +#endif + + __ssa8r a8 + __src_b a7, a5, a6 # lo-mask F..F0..0 (BE) 0..0F..F (LE) + __src_b a6, a6, a5 # hi-mask 0..0F..F (BE) F..F0..0 (LE) +#ifdef UNALIGNED_USER_EXCEPTION + l32e a5, a4, -8 +#else + l32i a5, a4, 0 # load lower address word +#endif + and a5, a5, a7 # mask + __sh a7, a3 # shift value + or a5, a5, a7 # or with original value +#ifdef UNALIGNED_USER_EXCEPTION + s32e a5, a4, -8 + l32e a7, a4, -4 +#else + s32i a5, a4, 0 # store + l32i a7, a4, 4 # same for upper address word +#endif + __sl a5, a3 + and a6, a7, a6 + or a6, a6, a5 +#ifdef UNALIGNED_USER_EXCEPTION + s32e a6, a4, -4 +#else + s32i a6, a4, 4 +#endif + + /* Done. restore stack and return */ + +.Lexit: + movi a4, 0 + rsr a3, EXCSAVE_1 + s32i a4, a3, EXC_TABLE_FIXUP + + /* Restore working register */ + + l32i a7, a2, PT_AREG7 + l32i a6, a2, PT_AREG6 + l32i a5, a2, PT_AREG5 + l32i a4, a2, PT_AREG4 + l32i a3, a2, PT_AREG3 + + /* restore SAR and return */ + + wsr a0, SAR + l32i a0, a2, PT_AREG0 + l32i a2, a2, PT_AREG2 + rfe + + /* We cannot handle this exception. */ + + .extern _kernel_exception +.Linvalid_instruction_store: +.Linvalid_instruction: + + /* Restore a4...a8 and SAR, set SP, and jump to default exception. */ + + l32i a8, a2, PT_AREG8 + l32i a7, a2, PT_AREG7 + l32i a6, a2, PT_AREG6 + l32i a5, a2, PT_AREG5 + l32i a4, a2, PT_AREG4 + wsr a0, SAR + mov a1, a2 + + rsr a0, PS + bbsi.l a2, PS_UM_SHIFT, 1f # jump if user mode + + movi a0, _kernel_exception + jx a0 + +1: movi a0, _user_exception + jx a0 + + +#endif /* XCHAL_UNALIGNED_LOAD_EXCEPTION || XCHAL_UNALIGNED_STORE_EXCEPTION */ + diff --git a/arch/xtensa/kernel/asm-offsets.c b/arch/xtensa/kernel/asm-offsets.c new file mode 100644 index 00000000000..840cd9a1d3d --- /dev/null +++ b/arch/xtensa/kernel/asm-offsets.c @@ -0,0 +1,94 @@ +/* + * arch/xtensa/kernel/asm-offsets.c + * + * Generates definitions from c-type structures used by assembly sources. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + */ + +#include <asm/processor.h> + +#include <linux/types.h> +#include <linux/sched.h> +#include <linux/stddef.h> +#include <linux/thread_info.h> +#include <linux/ptrace.h> +#include <asm/ptrace.h> +#include <asm/processor.h> +#include <asm/uaccess.h> + +#define DEFINE(sym, val) asm volatile("\n->" #sym " %0 " #val : : "i" (val)) +#define BLANK() asm volatile("\n->" : : ) + +int main(void) +{ + /* struct pt_regs */ + DEFINE(PT_PC, offsetof (struct pt_regs, pc)); + DEFINE(PT_PS, offsetof (struct pt_regs, ps)); + DEFINE(PT_DEPC, offsetof (struct pt_regs, depc)); + DEFINE(PT_EXCCAUSE, offsetof (struct pt_regs, exccause)); + DEFINE(PT_EXCVADDR, offsetof (struct pt_regs, excvaddr)); + DEFINE(PT_DEBUGCAUSE, offsetof (struct pt_regs, debugcause)); + DEFINE(PT_WMASK, offsetof (struct pt_regs, wmask)); + DEFINE(PT_LBEG, offsetof (struct pt_regs, lbeg)); + DEFINE(PT_LEND, offsetof (struct pt_regs, lend)); + DEFINE(PT_LCOUNT, offsetof (struct pt_regs, lcount)); + DEFINE(PT_SAR, offsetof (struct pt_regs, sar)); + DEFINE(PT_SYSCALL, offsetof (struct pt_regs, syscall)); + DEFINE(PT_AREG, offsetof (struct pt_regs, areg[0])); + DEFINE(PT_AREG0, offsetof (struct pt_regs, areg[0])); + DEFINE(PT_AREG1, offsetof (struct pt_regs, areg[1])); + DEFINE(PT_AREG2, offsetof (struct pt_regs, areg[2])); + DEFINE(PT_AREG3, offsetof (struct pt_regs, areg[3])); + DEFINE(PT_AREG4, offsetof (struct pt_regs, areg[4])); + DEFINE(PT_AREG5, offsetof (struct pt_regs, areg[5])); + DEFINE(PT_AREG6, offsetof (struct pt_regs, areg[6])); + DEFINE(PT_AREG7, offsetof (struct pt_regs, areg[7])); + DEFINE(PT_AREG8, offsetof (struct pt_regs, areg[8])); + DEFINE(PT_AREG9, offsetof (struct pt_regs, areg[9])); + DEFINE(PT_AREG10, offsetof (struct pt_regs, areg[10])); + DEFINE(PT_AREG11, offsetof (struct pt_regs, areg[11])); + DEFINE(PT_AREG12, offsetof (struct pt_regs, areg[12])); + DEFINE(PT_AREG13, offsetof (struct pt_regs, areg[13])); + DEFINE(PT_AREG14, offsetof (struct pt_regs, areg[14])); + DEFINE(PT_AREG15, offsetof (struct pt_regs, areg[15])); + DEFINE(PT_WINDOWBASE, offsetof (struct pt_regs, windowbase)); + DEFINE(PT_WINDOWSTART, offsetof(struct pt_regs, windowstart)); + DEFINE(PT_SIZE, sizeof(struct pt_regs)); + DEFINE(PT_AREG_END, offsetof (struct pt_regs, areg[XCHAL_NUM_AREGS])); + DEFINE(PT_USER_SIZE, offsetof(struct pt_regs, areg[XCHAL_NUM_AREGS])); + BLANK(); + + /* struct task_struct */ + DEFINE(TASK_PTRACE, offsetof (struct task_struct, ptrace)); + DEFINE(TASK_MM, offsetof (struct task_struct, mm)); + DEFINE(TASK_ACTIVE_MM, offsetof (struct task_struct, active_mm)); + DEFINE(TASK_PID, offsetof (struct task_struct, pid)); + DEFINE(TASK_THREAD, offsetof (struct task_struct, thread)); + DEFINE(TASK_THREAD_INFO, offsetof (struct task_struct, thread_info)); + DEFINE(TASK_STRUCT_SIZE, sizeof (struct task_struct)); + BLANK(); + + /* struct thread_info (offset from start_struct) */ + DEFINE(THREAD_RA, offsetof (struct task_struct, thread.ra)); + DEFINE(THREAD_SP, offsetof (struct task_struct, thread.sp)); + DEFINE(THREAD_CP_SAVE, offsetof (struct task_struct, thread.cp_save)); + DEFINE(THREAD_CURRENT_DS, offsetof (struct task_struct, thread.current_ds)); + BLANK(); + + /* struct mm_struct */ + DEFINE(MM_USERS, offsetof(struct mm_struct, mm_users)); + DEFINE(MM_PGD, offsetof (struct mm_struct, pgd)); + DEFINE(MM_CONTEXT, offsetof (struct mm_struct, context)); + BLANK(); + DEFINE(PT_SINGLESTEP_BIT, PT_SINGLESTEP_BIT); + return 0; +} + + diff --git a/arch/xtensa/kernel/coprocessor.S b/arch/xtensa/kernel/coprocessor.S new file mode 100644 index 00000000000..356192a4d39 --- /dev/null +++ b/arch/xtensa/kernel/coprocessor.S @@ -0,0 +1,201 @@ +/* + * arch/xtensa/kernel/coprocessor.S + * + * Xtensa processor configuration-specific table of coprocessor and + * other custom register layout information. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2003 - 2005 Tensilica Inc. + * + * Marc Gauthier <marc@tensilica.com> <marc@alumni.uwaterloo.ca> + */ + +/* + * This module contains a table that describes the layout of the various + * custom registers and states associated with each coprocessor, as well + * as those not associated with any coprocessor ("extra state"). + * This table is included with core dumps and is available via the ptrace + * interface, allowing the layout of such register/state information to + * be modified in the kernel without affecting the debugger. Each + * register or state is identified using a 32-bit "libdb target number" + * assigned when the Xtensa processor is generated. + */ + +#include <linux/config.h> +#include <linux/linkage.h> +#include <asm/processor.h> + +#if XCHAL_HAVE_CP + +#define CP_LAST ((XCHAL_CP_MAX - 1) * COPROCESSOR_INFO_SIZE) + +ENTRY(release_coprocessors) + + entry a1, 16 + # a2: task + movi a3, 1 << XCHAL_CP_MAX # a3: coprocessor-bit + movi a4, coprocessor_info+CP_LAST # a4: owner-table + # a5: tmp + movi a6, 0 # a6: 0 + rsil a7, LOCKLEVEL # a7: PS + +1: /* Check if task is coprocessor owner of coprocessor[i]. */ + + l32i a5, a4, COPROCESSOR_INFO_OWNER + srli a3, a3, 1 + beqz a3, 1f + addi a4, a4, -8 + beq a2, a5, 1b + + /* Found an entry: Clear entry CPENABLE bit to disable CP. */ + + rsr a5, CPENABLE + s32i a6, a4, COPROCESSOR_INFO_OWNER + xor a5, a3, a5 + wsr a5, CPENABLE + + bnez a3, 1b + +1: wsr a7, PS + rsync + retw + + +ENTRY(disable_coprocessor) + entry sp, 16 + rsil a7, LOCKLEVEL + rsr a3, CPENABLE + movi a4, 1 + ssl a2 + sll a4, a4 + and a4, a3, a4 + xor a3, a3, a4 + wsr a3, CPENABLE + wsr a7, PS + rsync + retw + +ENTRY(enable_coprocessor) + entry sp, 16 + rsil a7, LOCKLEVEL + rsr a3, CPENABLE + movi a4, 1 + ssl a2 + sll a4, a4 + or a3, a3, a4 + wsr a3, CPENABLE + wsr a7, PS + rsync + retw + +#endif + +ENTRY(save_coprocessor_extra) + entry sp, 16 + xchal_extra_store_funcbody + retw + +ENTRY(restore_coprocessor_extra) + entry sp, 16 + xchal_extra_load_funcbody + retw + +ENTRY(save_coprocessor_registers) + entry sp, 16 + xchal_cpi_store_funcbody + retw + +ENTRY(restore_coprocessor_registers) + entry sp, 16 + xchal_cpi_load_funcbody + retw + + +/* + * The Xtensa compile-time HAL (core.h) XCHAL_*_SA_CONTENTS_LIBDB macros + * describe the contents of coprocessor & extra save areas in terms of + * undefined CONTENTS_LIBDB_{SREG,UREG,REGF} macros. We define these + * latter macros here; they expand into a table of the format we want. + * The general format is: + * + * CONTENTS_LIBDB_SREG(libdbnum, offset, size, align, rsv1, name, sregnum, + * bitmask, rsv2, rsv3) + * CONTENTS_LIBDB_UREG(libdbnum, offset, size, align, rsv1, name, uregnum, + * bitmask, rsv2, rsv3) + * CONTENTS_LIBDB_REGF(libdbnum, offset, size, align, rsv1, name, index, + * numentries, contentsize, regname_base, + * regfile_name, rsv2, rsv3) + * + * For this table, we only care about the <libdbnum>, <offset> and <size> + * fields. + */ + +/* Map all XCHAL CONTENTS macros to the reg_entry asm macro defined below: */ + +#define CONTENTS_LIBDB_SREG(libdbnum,offset,size,align,rsv1,name,sregnum, \ + bitmask, rsv2, rsv3) \ + reg_entry libdbnum, offset, size ; +#define CONTENTS_LIBDB_UREG(libdbnum,offset,size,align,rsv1,name,uregnum, \ + bitmask, rsv2, rsv3) \ + reg_entry libdbnum, offset, size ; +#define CONTENTS_LIBDB_REGF(libdbnum, offset, size, align, rsv1, name, index, \ + numentries, contentsize, regname_base, \ + regfile_name, rsv2, rsv3) \ + reg_entry libdbnum, offset, size ; + +/* A single table entry: */ + .macro reg_entry libdbnum, offset, size + .ifne (__last_offset-(__last_group_offset+\offset)) + /* padding entry */ + .word (0xFC000000+__last_offset-(__last_group_offset+\offset)) + .endif + .word \libdbnum /* actual entry */ + .set __last_offset, __last_group_offset+\offset+\size + .endm /* reg_entry */ + + +/* Table entry that marks the beginning of a group (coprocessor or "extra"): */ + .macro reg_group cpnum, num_entries, align + .set __last_group_offset, (__last_offset + \align- 1) & -\align + .ifne \num_entries + .word 0xFD000000+(\cpnum<<16)+\num_entries + .endif + .endm /* reg_group */ + +/* + * Register info tables. + */ + + .section .rodata, "a" + .globl _xtensa_reginfo_tables + .globl _xtensa_reginfo_table_size + .align 4 +_xtensa_reginfo_table_size: + .word _xtensa_reginfo_table_end - _xtensa_reginfo_tables + +_xtensa_reginfo_tables: + .set __last_offset, 0 + reg_group 0xFF, XCHAL_EXTRA_SA_CONTENTS_LIBDB_NUM, XCHAL_EXTRA_SA_ALIGN + XCHAL_EXTRA_SA_CONTENTS_LIBDB + reg_group 0, XCHAL_CP0_SA_CONTENTS_LIBDB_NUM, XCHAL_CP0_SA_ALIGN + XCHAL_CP0_SA_CONTENTS_LIBDB + reg_group 1, XCHAL_CP1_SA_CONTENTS_LIBDB_NUM, XCHAL_CP1_SA_ALIGN + XCHAL_CP1_SA_CONTENTS_LIBDB + reg_group 2, XCHAL_CP2_SA_CONTENTS_LIBDB_NUM, XCHAL_CP2_SA_ALIGN + XCHAL_CP2_SA_CONTENTS_LIBDB + reg_group 3, XCHAL_CP3_SA_CONTENTS_LIBDB_NUM, XCHAL_CP3_SA_ALIGN + XCHAL_CP3_SA_CONTENTS_LIBDB + reg_group 4, XCHAL_CP4_SA_CONTENTS_LIBDB_NUM, XCHAL_CP4_SA_ALIGN + XCHAL_CP4_SA_CONTENTS_LIBDB + reg_group 5, XCHAL_CP5_SA_CONTENTS_LIBDB_NUM, XCHAL_CP5_SA_ALIGN + XCHAL_CP5_SA_CONTENTS_LIBDB + reg_group 6, XCHAL_CP6_SA_CONTENTS_LIBDB_NUM, XCHAL_CP6_SA_ALIGN + XCHAL_CP6_SA_CONTENTS_LIBDB + reg_group 7, XCHAL_CP7_SA_CONTENTS_LIBDB_NUM, XCHAL_CP7_SA_ALIGN + XCHAL_CP7_SA_CONTENTS_LIBDB + .word 0xFC000000 /* invalid register number,marks end of table*/ +_xtensa_reginfo_table_end: + diff --git a/arch/xtensa/kernel/entry.S b/arch/xtensa/kernel/entry.S new file mode 100644 index 00000000000..c64a01f71de --- /dev/null +++ b/arch/xtensa/kernel/entry.S @@ -0,0 +1,1996 @@ +/* + * arch/xtensa/kernel/entry.S + * + * Low-level exception handling + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2004-2005 by Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * + */ + +#include <linux/linkage.h> +#include <asm/offsets.h> +#include <asm/processor.h> +#include <asm/thread_info.h> +#include <asm/uaccess.h> +#include <asm/unistd.h> +#include <asm/ptrace.h> +#include <asm/current.h> +#include <asm/pgtable.h> +#include <asm/page.h> +#include <asm/signal.h> +#include <xtensa/coreasm.h> + +/* Unimplemented features. */ + +#undef SIGNAL_HANDLING_IN_DOUBLE_EXCEPTION +#undef KERNEL_STACK_OVERFLOW_CHECK +#undef PREEMPTIBLE_KERNEL +#undef ALLOCA_EXCEPTION_IN_IRAM + +/* Not well tested. + * + * - fast_coprocessor + */ + +/* + * Macro to find first bit set in WINDOWBASE from the left + 1 + * + * 100....0 -> 1 + * 010....0 -> 2 + * 000....1 -> WSBITS + */ + + .macro ffs_ws bit mask + +#if XCHAL_HAVE_NSA + nsau \bit, \mask # 32-WSBITS ... 31 (32 iff 0) + addi \bit, \bit, WSBITS - 32 + 1 # uppest bit set -> return 1 +#else + movi \bit, WSBITS +#if WSBITS > 16 + _bltui \mask, 0x10000, 99f + addi \bit, \bit, -16 + extui \mask, \mask, 16, 16 +#endif +#if WSBITS > 8 +99: _bltui \mask, 0x100, 99f + addi \bit, \bit, -8 + srli \mask, \mask, 8 +#endif +99: _bltui \mask, 0x10, 99f + addi \bit, \bit, -4 + srli \mask, \mask, 4 +99: _bltui \mask, 0x4, 99f + addi \bit, \bit, -2 + srli \mask, \mask, 2 +99: _bltui \mask, 0x2, 99f + addi \bit, \bit, -1 +99: + +#endif + .endm + +/* ----------------- DEFAULT FIRST LEVEL EXCEPTION HANDLERS ----------------- */ + +/* + * First-level exception handler for user exceptions. + * Save some special registers, extra states and all registers in the AR + * register file that were in use in the user task, and jump to the common + * exception code. + * We save SAR (used to calculate WMASK), and WB and WS (we don't have to + * save them for kernel exceptions). + * + * Entry condition for user_exception: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original value in depc + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + * + * Entry condition for _user_exception: + * + * a0-a3 and depc have been saved to PT_AREG0...PT_AREG3 and PT_DEPC + * excsave has been restored, and + * stack pointer (a1) has been set. + * + * Note: _user_exception might be at an odd adress. Don't use call0..call12 + */ + +ENTRY(user_exception) + + /* Save a2, a3, and depc, restore excsave_1 and set SP. */ + + xsr a3, EXCSAVE_1 + rsr a0, DEPC + s32i a1, a2, PT_AREG1 + s32i a0, a2, PT_AREG2 + s32i a3, a2, PT_AREG3 + mov a1, a2 + + .globl _user_exception +_user_exception: + + /* Save SAR and turn off single stepping */ + + movi a2, 0 + rsr a3, SAR + wsr a2, ICOUNTLEVEL + s32i a3, a1, PT_SAR + + /* Rotate ws so that the current windowbase is at bit0. */ + /* Assume ws = xxwww1yyyy. Rotate ws right, so that a2 = yyyyxxwww1 */ + + rsr a2, WINDOWBASE + rsr a3, WINDOWSTART + ssr a2 + s32i a2, a1, PT_WINDOWBASE + s32i a3, a1, PT_WINDOWSTART + slli a2, a3, 32-WSBITS + src a2, a3, a2 + srli a2, a2, 32-WSBITS + s32i a2, a1, PT_WMASK # needed for restoring registers + + /* Save only live registers. */ + + _bbsi.l a2, 1, 1f + s32i a4, a1, PT_AREG4 + s32i a5, a1, PT_AREG5 + s32i a6, a1, PT_AREG6 + s32i a7, a1, PT_AREG7 + _bbsi.l a2, 2, 1f + s32i a8, a1, PT_AREG8 + s32i a9, a1, PT_AREG9 + s32i a10, a1, PT_AREG10 + s32i a11, a1, PT_AREG11 + _bbsi.l a2, 3, 1f + s32i a12, a1, PT_AREG12 + s32i a13, a1, PT_AREG13 + s32i a14, a1, PT_AREG14 + s32i a15, a1, PT_AREG15 + _bnei a2, 1, 1f # only one valid frame? + + /* Only one valid frame, skip saving regs. */ + + j 2f + + /* Save the remaining registers. + * We have to save all registers up to the first '1' from + * the right, except the current frame (bit 0). + * Assume a2 is: 001001000110001 + * All regiser frames starting from the top fiel to the marked '1' + * must be saved. + */ + +1: addi a3, a2, -1 # eliminate '1' in bit 0: yyyyxxww0 + neg a3, a3 # yyyyxxww0 -> YYYYXXWW1+1 + and a3, a3, a2 # max. only one bit is set + + /* Find number of frames to save */ + + ffs_ws a0, a3 # number of frames to the '1' from left + + /* Store information into WMASK: + * bits 0..3: xxx1 masked lower 4 bits of the rotated windowstart, + * bits 4...: number of valid 4-register frames + */ + + slli a3, a0, 4 # number of frames to save in bits 8..4 + extui a2, a2, 0, 4 # mask for the first 16 registers + or a2, a3, a2 + s32i a2, a1, PT_WMASK # needed when we restore the reg-file + + /* Save 4 registers at a time */ + +1: rotw -1 + s32i a0, a5, PT_AREG_END - 16 + s32i a1, a5, PT_AREG_END - 12 + s32i a2, a5, PT_AREG_END - 8 + s32i a3, a5, PT_AREG_END - 4 + addi a0, a4, -1 + addi a1, a5, -16 + _bnez a0, 1b + + /* WINDOWBASE still in SAR! */ + + rsr a2, SAR # original WINDOWBASE + movi a3, 1 + ssl a2 + sll a3, a3 + wsr a3, WINDOWSTART # set corresponding WINDOWSTART bit + wsr a2, WINDOWBASE # and WINDOWSTART + rsync + + /* We are back to the original stack pointer (a1) */ + +2: +#if XCHAL_EXTRA_SA_SIZE + + /* For user exceptions, save the extra state into the user's TCB. + * Note: We must assume that xchal_extra_store_funcbody destroys a2..a15 + */ + + GET_CURRENT(a2,a1) + addi a2, a2, THREAD_CP_SAVE + xchal_extra_store_funcbody +#endif + + /* Now, jump to the common exception handler. */ + + j common_exception + + +/* + * First-level exit handler for kernel exceptions + * Save special registers and the live window frame. + * Note: Even though we changes the stack pointer, we don't have to do a + * MOVSP here, as we do that when we return from the exception. + * (See comment in the kernel exception exit code) + * + * Entry condition for kernel_exception: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + * + * Entry condition for _kernel_exception: + * + * a0-a3 and depc have been saved to PT_AREG0...PT_AREG3 and PT_DEPC + * excsave has been restored, and + * stack pointer (a1) has been set. + * + * Note: _kernel_exception might be at an odd adress. Don't use call0..call12 + */ + +ENTRY(kernel_exception) + + /* Save a0, a2, a3, DEPC and set SP. */ + + xsr a3, EXCSAVE_1 # restore a3, excsave_1 + rsr a0, DEPC # get a2 + s32i a1, a2, PT_AREG1 + s32i a0, a2, PT_AREG2 + s32i a3, a2, PT_AREG3 + mov a1, a2 + + .globl _kernel_exception +_kernel_exception: + + /* Save SAR and turn off single stepping */ + + movi a2, 0 + rsr a3, SAR + wsr a2, ICOUNTLEVEL + s32i a3, a1, PT_SAR + + /* Rotate ws so that the current windowbase is at bit0. */ + /* Assume ws = xxwww1yyyy. Rotate ws right, so that a2 = yyyyxxwww1 */ + + rsr a2, WINDOWBASE # don't need to save these, we only + rsr a3, WINDOWSTART # need shifted windowstart: windowmask + ssr a2 + slli a2, a3, 32-WSBITS + src a2, a3, a2 + srli a2, a2, 32-WSBITS + s32i a2, a1, PT_WMASK # needed for kernel_exception_exit + + /* Save only the live window-frame */ + + _bbsi.l a2, 1, 1f + s32i a4, a1, PT_AREG4 + s32i a5, a1, PT_AREG5 + s32i a6, a1, PT_AREG6 + s32i a7, a1, PT_AREG7 + _bbsi.l a2, 2, 1f + s32i a8, a1, PT_AREG8 + s32i a9, a1, PT_AREG9 + s32i a10, a1, PT_AREG10 + s32i a11, a1, PT_AREG11 + _bbsi.l a2, 3, 1f + s32i a12, a1, PT_AREG12 + s32i a13, a1, PT_AREG13 + s32i a14, a1, PT_AREG14 + s32i a15, a1, PT_AREG15 + +1: + +#ifdef KERNEL_STACK_OVERFLOW_CHECK + + /* Stack overflow check, for debugging */ + extui a2, a1, TASK_SIZE_BITS,XX + movi a3, SIZE?? + _bge a2, a3, out_of_stack_panic + +#endif + +/* + * This is the common exception handler. + * We get here from the user exception handler or simply by falling through + * from the kernel exception handler. + * Save the remaining special registers, switch to kernel mode, and jump + * to the second-level exception handler. + * + */ + +common_exception: + + /* Save EXCVADDR, DEBUGCAUSE, and PC, and clear LCOUNT */ + + rsr a2, DEBUGCAUSE + rsr a3, EPC_1 + s32i a2, a1, PT_DEBUGCAUSE + s32i a3, a1, PT_PC + + rsr a3, EXCVADDR + movi a2, 0 + s32i a3, a1, PT_EXCVADDR + xsr a2, LCOUNT + s32i a2, a1, PT_LCOUNT + + /* It is now save to restore the EXC_TABLE_FIXUP variable. */ + + rsr a0, EXCCAUSE + movi a3, 0 + rsr a2, EXCSAVE_1 + s32i a0, a1, PT_EXCCAUSE + s32i a3, a2, EXC_TABLE_FIXUP + + /* All unrecoverable states are saved on stack, now, and a1 is valid, + * so we can allow exceptions and interrupts (*) again. + * Set PS(EXCM = 0, UM = 0, RING = 0, OWB = 0, WOE = 1, INTLEVEL = X) + * + * (*) We only allow interrupts if PS.INTLEVEL was not set to 1 before + * (interrupts disabled) and if this exception is not an interrupt. + */ + + rsr a3, PS + addi a0, a0, -4 + movi a2, 1 + extui a3, a3, 0, 1 # a3 = PS.INTLEVEL[0] + moveqz a3, a2, a0 # a3 = 1 iff interrupt exception + movi a2, PS_WOE_MASK + or a3, a3, a2 + rsr a0, EXCCAUSE + xsr a3, PS + + s32i a3, a1, PT_PS # save ps + + /* Save LBEG, LEND */ + + rsr a2, LBEG + rsr a3, LEND + s32i a2, a1, PT_LBEG + s32i a3, a1, PT_LEND + + /* Go to second-level dispatcher. Set up parameters to pass to the + * exception handler and call the exception handler. + */ + + movi a4, exc_table + mov a6, a1 # pass stack frame + mov a7, a0 # pass EXCCAUSE + addx4 a4, a0, a4 + l32i a4, a4, EXC_TABLE_DEFAULT # load handler + + /* Call the second-level handler */ + + callx4 a4 + + /* Jump here for exception exit */ + +common_exception_return: + + /* Jump if we are returning from kernel exceptions. */ + +1: l32i a3, a1, PT_PS + _bbsi.l a3, PS_UM_SHIFT, 2f + j kernel_exception_exit + + /* Specific to a user exception exit: + * We need to check some flags for signal handling and rescheduling, + * and have to restore WB and WS, extra states, and all registers + * in the register file that were in use in the user task. + */ + +2: wsr a3, PS /* disable interrupts */ + + /* Check for signals (keep interrupts disabled while we read TI_FLAGS) + * Note: PS.INTLEVEL = 0, PS.EXCM = 1 + */ + + GET_THREAD_INFO(a2,a1) + l32i a4, a2, TI_FLAGS + + /* Enable interrupts again. + * Note: When we get here, we certainly have handled any interrupts. + * (Hint: There is only one user exception frame on stack) + */ + + movi a3, PS_WOE_MASK + + _bbsi.l a4, TIF_NEED_RESCHED, 3f + _bbci.l a4, TIF_SIGPENDING, 4f + +#ifndef SIGNAL_HANDLING_IN_DOUBLE_EXCEPTION + l32i a4, a1, PT_DEPC + bgeui a4, VALID_DOUBLE_EXCEPTION_ADDRESS, 4f +#endif + + /* Reenable interrupts and call do_signal() */ + + wsr a3, PS + movi a4, do_signal # int do_signal(struct pt_regs*, sigset_t*) + mov a6, a1 + movi a7, 0 + callx4 a4 + j 1b + +3: /* Reenable interrupts and reschedule */ + + wsr a3, PS + movi a4, schedule # void schedule (void) + callx4 a4 + j 1b + + /* Restore the state of the task and return from the exception. */ + + + /* If we are returning from a user exception, and the process + * to run next has PT_SINGLESTEP set, we want to setup + * ICOUNT and ICOUNTLEVEL to step one instruction. + * PT_SINGLESTEP is set by sys_ptrace (ptrace.c) + */ + +4: /* a2 holds GET_CURRENT(a2,a1) */ + + l32i a3, a2, TI_TASK + l32i a3, a3, TASK_PTRACE + bbci.l a3, PT_SINGLESTEP_BIT, 1f # jump if single-step flag is not set + + movi a3, -2 # PT_SINGLESTEP flag is set, + movi a4, 1 # icountlevel of 1 means it won't + wsr a3, ICOUNT # start counting until after rfe + wsr a4, ICOUNTLEVEL # so setup icount & icountlevel. + isync + +1: + +#if XCHAL_EXTRA_SA_SIZE + + /* For user exceptions, restore the extra state from the user's TCB. */ + + /* Note: a2 still contains GET_CURRENT(a2,a1) */ + addi a2, a2, THREAD_CP_SAVE + xchal_extra_load_funcbody + + /* We must assume that xchal_extra_store_funcbody destroys + * registers a2..a15. FIXME, this list can eventually be + * reduced once real register requirements of the macro are + * finalized. */ + +#endif /* XCHAL_EXTRA_SA_SIZE */ + + + /* Switch to the user thread WINDOWBASE. Save SP temporarily in DEPC */ + + l32i a2, a1, PT_WINDOWBASE + l32i a3, a1, PT_WINDOWSTART + wsr a1, DEPC # use DEPC as temp storage + wsr a3, WINDOWSTART # restore WINDOWSTART + ssr a2 # preserve user's WB in the SAR + wsr a2, WINDOWBASE # switch to user's saved WB + rsync + rsr a1, DEPC # restore stack pointer + l32i a2, a1, PT_WMASK # register frames saved (in bits 4...9) + rotw -1 # we restore a4..a7 + _bltui a6, 16, 1f # only have to restore current window? + + /* The working registers are a0 and a3. We are restoring to + * a4..a7. Be careful not to destroy what we have just restored. + * Note: wmask has the format YYYYM: + * Y: number of registers saved in groups of 4 + * M: 4 bit mask of first 16 registers + */ + + mov a2, a6 + mov a3, a5 + +2: rotw -1 # a0..a3 become a4..a7 + addi a3, a7, -4*4 # next iteration + addi a2, a6, -16 # decrementing Y in WMASK + l32i a4, a3, PT_AREG_END + 0 + l32i a5, a3, PT_AREG_END + 4 + l32i a6, a3, PT_AREG_END + 8 + l32i a7, a3, PT_AREG_END + 12 + _bgeui a2, 16, 2b + + /* Clear unrestored registers (don't leak anything to user-land */ + +1: rsr a0, WINDOWBASE + rsr a3, SAR + sub a3, a0, a3 + beqz a3, 2f + extui a3, a3, 0, WBBITS + +1: rotw -1 + addi a3, a7, -1 + movi a4, 0 + movi a5, 0 + movi a6, 0 + movi a7, 0 + bgei a3, 1, 1b + + /* We are back were we were when we started. + * Note: a2 still contains WMASK (if we've returned to the original + * frame where we had loaded a2), or at least the lower 4 bits + * (if we have restored WSBITS-1 frames). + */ + +2: j common_exception_exit + + /* This is the kernel exception exit. + * We avoided to do a MOVSP when we entered the exception, but we + * have to do it here. + */ + +kernel_exception_exit: + + /* Disable interrupts (a3 holds PT_PS) */ + + wsr a3, PS + +#ifdef PREEMPTIBLE_KERNEL + +#ifdef CONFIG_PREEMPT + + /* + * Note: We've just returned from a call4, so we have + * at least 4 addt'l regs. + */ + + /* Check current_thread_info->preempt_count */ + + GET_THREAD_INFO(a2) + l32i a3, a2, TI_PREEMPT + bnez a3, 1f + + l32i a2, a2, TI_FLAGS + +1: + +#endif + +#endif + + /* Check if we have to do a movsp. + * + * We only have to do a movsp if the previous window-frame has + * been spilled to the *temporary* exception stack instead of the + * task's stack. This is the case if the corresponding bit in + * WINDOWSTART for the previous window-frame was set before + * (not spilled) but is zero now (spilled). + * If this bit is zero, all other bits except the one for the + * current window frame are also zero. So, we can use a simple test: + * 'and' WINDOWSTART and WINDOWSTART-1: + * + * (XXXXXX1[0]* - 1) AND XXXXXX1[0]* = XXXXXX0[0]* + * + * The result is zero only if one bit was set. + * + * (Note: We might have gone through several task switches before + * we come back to the current task, so WINDOWBASE might be + * different from the time the exception occurred.) + */ + + /* Test WINDOWSTART before and after the exception. + * We actually have WMASK, so we only have to test if it is 1 or not. + */ + + l32i a2, a1, PT_WMASK + _beqi a2, 1, common_exception_exit # Spilled before exception,jump + + /* Test WINDOWSTART now. If spilled, do the movsp */ + + rsr a3, WINDOWSTART + addi a0, a3, -1 + and a3, a3, a0 + _bnez a3, common_exception_exit + + /* Do a movsp (we returned from a call4, so we have at least a0..a7) */ + + addi a0, a1, -16 + l32i a3, a0, 0 + l32i a4, a0, 4 + s32i a3, a1, PT_SIZE+0 + s32i a4, a1, PT_SIZE+4 + l32i a3, a0, 8 + l32i a4, a0, 12 + s32i a3, a1, PT_SIZE+8 + s32i a4, a1, PT_SIZE+12 + + /* Common exception exit. + * We restore the special register and the current window frame, and + * return from the exception. + * + * Note: We expect a2 to hold PT_WMASK + */ + +common_exception_exit: + + _bbsi.l a2, 1, 1f + l32i a4, a1, PT_AREG4 + l32i a5, a1, PT_AREG5 + l32i a6, a1, PT_AREG6 + l32i a7, a1, PT_AREG7 + _bbsi.l a2, 2, 1f + l32i a8, a1, PT_AREG8 + l32i a9, a1, PT_AREG9 + l32i a10, a1, PT_AREG10 + l32i a11, a1, PT_AREG11 + _bbsi.l a2, 3, 1f + l32i a12, a1, PT_AREG12 + l32i a13, a1, PT_AREG13 + l32i a14, a1, PT_AREG14 + l32i a15, a1, PT_AREG15 + + /* Restore PC, SAR */ + +1: l32i a2, a1, PT_PC + l32i a3, a1, PT_SAR + wsr a2, EPC_1 + wsr a3, SAR + + /* Restore LBEG, LEND, LCOUNT */ + + l32i a2, a1, PT_LBEG + l32i a3, a1, PT_LEND + wsr a2, LBEG + l32i a2, a1, PT_LCOUNT + wsr a3, LEND + wsr a2, LCOUNT + + /* Check if it was double exception. */ + + l32i a0, a1, PT_DEPC + l32i a3, a1, PT_AREG3 + l32i a2, a1, PT_AREG2 + _bgeui a0, VALID_DOUBLE_EXCEPTION_ADDRESS, 1f + + /* Restore a0...a3 and return */ + + l32i a0, a1, PT_AREG0 + l32i a1, a1, PT_AREG1 + rfe + +1: wsr a0, DEPC + l32i a0, a1, PT_AREG0 + l32i a1, a1, PT_AREG1 + rfde + +/* + * Debug exception handler. + * + * Currently, we don't support KGDB, so only user application can be debugged. + * + * When we get here, a0 is trashed and saved to excsave[debuglevel] + */ + +ENTRY(debug_exception) + + rsr a0, EPS + XCHAL_DEBUGLEVEL + bbsi.l a0, PS_EXCM_SHIFT, 1f # exception mode + + /* Set EPC_1 and EXCCAUSE */ + + wsr a2, DEPC # save a2 temporarily + rsr a2, EPC + XCHAL_DEBUGLEVEL + wsr a2, EPC_1 + + movi a2, EXCCAUSE_MAPPED_DEBUG + wsr a2, EXCCAUSE + + /* Restore PS to the value before the debug exc but with PS.EXCM set.*/ + + movi a2, 1 << PS_EXCM_SHIFT + or a2, a0, a2 + movi a0, debug_exception # restore a3, debug jump vector + wsr a2, PS + xsr a0, EXCSAVE + XCHAL_DEBUGLEVEL + + /* Switch to kernel/user stack, restore jump vector, and save a0 */ + + bbsi.l a2, PS_UM_SHIFT, 2f # jump if user mode + + addi a2, a1, -16-PT_SIZE # assume kernel stack + s32i a0, a2, PT_AREG0 + movi a0, 0 + s32i a1, a2, PT_AREG1 + s32i a0, a2, PT_DEPC # mark it as a regular exception + xsr a0, DEPC + s32i a3, a2, PT_AREG3 + s32i a0, a2, PT_AREG2 + mov a1, a2 + j _kernel_exception + +2: rsr a2, EXCSAVE_1 + l32i a2, a2, EXC_TABLE_KSTK # load kernel stack pointer + s32i a0, a2, PT_AREG0 + movi a0, 0 + s32i a1, a2, PT_AREG1 + s32i a0, a2, PT_DEPC + xsr a0, DEPC + s32i a3, a2, PT_AREG3 + s32i a0, a2, PT_AREG2 + mov a1, a2 + j _user_exception + + /* Debug exception while in exception mode. */ +1: j 1b // FIXME!! + + +/* + * We get here in case of an unrecoverable exception. + * The only thing we can do is to be nice and print a panic message. + * We only produce a single stack frame for panic, so ??? + * + * + * Entry conditions: + * + * - a0 contains the caller address; original value saved in excsave1. + * - the original a0 contains a valid return address (backtrace) or 0. + * - a2 contains a valid stackpointer + * + * Notes: + * + * - If the stack pointer could be invalid, the caller has to setup a + * dummy stack pointer (e.g. the stack of the init_task) + * + * - If the return address could be invalid, the caller has to set it + * to 0, so the backtrace would stop. + * + */ + .align 4 +unrecoverable_text: + .ascii "Unrecoverable error in exception handler\0" + +ENTRY(unrecoverable_exception) + + movi a0, 1 + movi a1, 0 + + wsr a0, WINDOWSTART + wsr a1, WINDOWBASE + rsync + + movi a1, PS_WOE_MASK | 1 + wsr a1, PS + rsync + + movi a1, init_task + movi a0, 0 + addi a1, a1, PT_REGS_OFFSET + + movi a4, panic + movi a6, unrecoverable_text + + callx4 a4 + +1: j 1b + + +/* -------------------------- FAST EXCEPTION HANDLERS ----------------------- */ + +/* + * Fast-handler for alloca exceptions + * + * The ALLOCA handler is entered when user code executes the MOVSP + * instruction and the caller's frame is not in the register file. + * In this case, the caller frame's a0..a3 are on the stack just + * below sp (a1), and this handler moves them. + * + * For "MOVSP <ar>,<as>" without destination register a1, this routine + * simply moves the value from <as> to <ar> without moving the save area. + * + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + */ + +#if XCHAL_HAVE_BE +#define _EXTUI_MOVSP_SRC(ar) extui ar, ar, 4, 4 +#define _EXTUI_MOVSP_DST(ar) extui ar, ar, 0, 4 +#else +#define _EXTUI_MOVSP_SRC(ar) extui ar, ar, 0, 4 +#define _EXTUI_MOVSP_DST(ar) extui ar, ar, 4, 4 +#endif + +ENTRY(fast_alloca) + + /* We shouldn't be in a double exception. */ + + l32i a0, a2, PT_DEPC + _bgeui a0, VALID_DOUBLE_EXCEPTION_ADDRESS, .Lunhandled_double + + rsr a0, DEPC # get a2 + s32i a4, a2, PT_AREG4 # save a4 and + s32i a0, a2, PT_AREG2 # a2 to stack + + /* Exit critical section. */ + + movi a0, 0 + s32i a0, a3, EXC_TABLE_FIXUP + + /* Restore a3, excsave_1 */ + + xsr a3, EXCSAVE_1 # make sure excsave_1 is valid for dbl. + rsr a4, EPC_1 # get exception address + s32i a3, a2, PT_AREG3 # save a3 to stack + +#ifdef ALLOCA_EXCEPTION_IN_IRAM +#error iram not supported +#else + /* Note: l8ui not allowed in IRAM/IROM!! */ + l8ui a0, a4, 1 # read as(src) from MOVSP instruction +#endif + movi a3, .Lmovsp_src + _EXTUI_MOVSP_SRC(a0) # extract source register number + addx8 a3, a0, a3 + jx a3 + +.Lunhandled_double: + wsr a0, EXCSAVE_1 + movi a0, unrecoverable_exception + callx0 a0 + + .align 8 +.Lmovsp_src: + l32i a3, a2, PT_AREG0; _j 1f; .align 8 + mov a3, a1; _j 1f; .align 8 + l32i a3, a2, PT_AREG2; _j 1f; .align 8 + l32i a3, a2, PT_AREG3; _j 1f; .align 8 + l32i a3, a2, PT_AREG4; _j 1f; .align 8 + mov a3, a5; _j 1f; .align 8 + mov a3, a6; _j 1f; .align 8 + mov a3, a7; _j 1f; .align 8 + mov a3, a8; _j 1f; .align 8 + mov a3, a9; _j 1f; .align 8 + mov a3, a10; _j 1f; .align 8 + mov a3, a11; _j 1f; .align 8 + mov a3, a12; _j 1f; .align 8 + mov a3, a13; _j 1f; .align 8 + mov a3, a14; _j 1f; .align 8 + mov a3, a15; _j 1f; .align 8 + +1: + +#ifdef ALLOCA_EXCEPTION_IN_IRAM +#error iram not supported +#else + l8ui a0, a4, 0 # read ar(dst) from MOVSP instruction +#endif + addi a4, a4, 3 # step over movsp + _EXTUI_MOVSP_DST(a0) # extract destination register + wsr a4, EPC_1 # save new epc_1 + + _bnei a0, 1, 1f # no 'movsp a1, ax': jump + + /* Move the save area. This implies the use of the L32E + * and S32E instructions, because this move must be done with + * the user's PS.RING privilege levels, not with ring 0 + * (kernel's) privileges currently active with PS.EXCM + * set. Note that we have stil registered a fixup routine with the + * double exception vector in case a double exception occurs. + */ + + /* a0,a4:avail a1:old user stack a2:exc. stack a3:new user stack. */ + + l32e a0, a1, -16 + l32e a4, a1, -12 + s32e a0, a3, -16 + s32e a4, a3, -12 + l32e a0, a1, -8 + l32e a4, a1, -4 + s32e a0, a3, -8 + s32e a4, a3, -4 + + /* Restore stack-pointer and all the other saved registers. */ + + mov a1, a3 + + l32i a4, a2, PT_AREG4 + l32i a3, a2, PT_AREG3 + l32i a0, a2, PT_AREG0 + l32i a2, a2, PT_AREG2 + rfe + + /* MOVSP <at>,<as> was invoked with <at> != a1. + * Because the stack pointer is not being modified, + * we should be able to just modify the pointer + * without moving any save area. + * The processor only traps these occurrences if the + * caller window isn't live, so unfortunately we can't + * use this as an alternate trap mechanism. + * So we just do the move. This requires that we + * resolve the destination register, not just the source, + * so there's some extra work. + * (PERHAPS NOT REALLY NEEDED, BUT CLEANER...) + */ + + /* a0 dst-reg, a1 user-stack, a2 stack, a3 value of src reg. */ + +1: movi a4, .Lmovsp_dst + addx8 a4, a0, a4 + jx a4 + + .align 8 +.Lmovsp_dst: + s32i a3, a2, PT_AREG0; _j 1f; .align 8 + mov a1, a3; _j 1f; .align 8 + s32i a3, a2, PT_AREG2; _j 1f; .align 8 + s32i a3, a2, PT_AREG3; _j 1f; .align 8 + s32i a3, a2, PT_AREG4; _j 1f; .align 8 + mov a5, a3; _j 1f; .align 8 + mov a6, a3; _j 1f; .align 8 + mov a7, a3; _j 1f; .align 8 + mov a8, a3; _j 1f; .align 8 + mov a9, a3; _j 1f; .align 8 + mov a10, a3; _j 1f; .align 8 + mov a11, a3; _j 1f; .align 8 + mov a12, a3; _j 1f; .align 8 + mov a13, a3; _j 1f; .align 8 + mov a14, a3; _j 1f; .align 8 + mov a15, a3; _j 1f; .align 8 + +1: l32i a4, a2, PT_AREG4 + l32i a3, a2, PT_AREG3 + l32i a0, a2, PT_AREG0 + l32i a2, a2, PT_AREG2 + rfe + + +/* + * fast system calls. + * + * WARNING: The kernel doesn't save the entire user context before + * handling a fast system call. These functions are small and short, + * usually offering some functionality not available to user tasks. + * + * BE CAREFUL TO PRESERVE THE USER'S CONTEXT. + * + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + */ + +ENTRY(fast_syscall_kernel) + + /* Skip syscall. */ + + rsr a0, EPC_1 + addi a0, a0, 3 + wsr a0, EPC_1 + + l32i a0, a2, PT_DEPC + bgeui a0, VALID_DOUBLE_EXCEPTION_ADDRESS, fast_syscall_unrecoverable + + rsr a0, DEPC # get syscall-nr + _beqz a0, fast_syscall_spill_registers + + addi a0, a0, -__NR_sysxtensa + _beqz a0, fast_syscall_sysxtensa + + j kernel_exception + + +ENTRY(fast_syscall_user) + + /* Skip syscall. */ + + rsr a0, EPC_1 + addi a0, a0, 3 + wsr a0, EPC_1 + + l32i a0, a2, PT_DEPC + bgeui a0, VALID_DOUBLE_EXCEPTION_ADDRESS, fast_syscall_unrecoverable + + rsr a0, DEPC # get syscall-nr + _beqz a0, fast_syscall_spill_registers + + addi a0, a0, -__NR_sysxtensa + _beqz a0, fast_syscall_sysxtensa + + j user_exception + +ENTRY(fast_syscall_unrecoverable) + + /* Restore all states. */ + + l32i a0, a2, PT_AREG0 # restore a0 + xsr a2, DEPC # restore a2, depc + rsr a3, EXCSAVE_1 + + wsr a0, EXCSAVE_1 + movi a0, unrecoverable_exception + callx0 a0 + + + +/* + * sysxtensa syscall handler + * + * int sysxtensa (XTENSA_ATOMIC_SET, ptr, val, unused); + * int sysxtensa (XTENSA_ATOMIC_ADD, ptr, val, unused); + * int sysxtensa (XTENSA_ATOMIC_EXG_ADD, ptr, val, unused); + * int sysxtensa (XTENSA_ATOMIC_CMP_SWP, ptr, oldval, newval); + * a2 a6 a3 a4 a5 + * + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + * + * Note: we don't have to save a2; a2 holds the return value + * + * We use the two macros TRY and CATCH: + * + * TRY adds an entry to the __ex_table fixup table for the immediately + * following instruction. + * + * CATCH catches any exception that occurred at one of the preceeding TRY + * statements and continues from there + * + * Usage TRY l32i a0, a1, 0 + * <other code> + * done: rfe + * CATCH <set return code> + * j done + */ + +#define TRY \ + .section __ex_table, "a"; \ + .word 66f, 67f; \ + .text; \ +66: + +#define CATCH \ +67: + +ENTRY(fast_syscall_sysxtensa) + + _beqz a6, 1f + _blti a6, SYSXTENSA_COUNT, 2f + +1: j user_exception + +2: xsr a3, EXCSAVE_1 # restore a3, excsave1 + s32i a7, a2, PT_AREG7 + + movi a7, 4 # sizeof(unsigned int) + verify_area a3, a7, a0, a2, .Leac + + _beqi a6, SYSXTENSA_ATOMIC_SET, .Lset + _beqi a6, SYSXTENSA_ATOMIC_EXG_ADD, .Lexg + _beqi a6, SYSXTENSA_ATOMIC_ADD, .Ladd + + /* Fall through for SYSXTENSA_ATOMIC_CMP_SWP */ + +.Lswp: /* Atomic compare and swap */ + +TRY l32i a7, a3, 0 # read old value + bne a7, a4, 1f # same as old value? jump + s32i a5, a3, 0 # different, modify value + movi a7, 1 # and return 1 + j .Lret + +1: movi a7, 0 # same values: return 0 + j .Lret + +.Ladd: /* Atomic add */ +.Lexg: /* Atomic (exchange) add */ + +TRY l32i a7, a3, 0 + add a4, a4, a7 + s32i a4, a3, 0 + j .Lret + +.Lset: /* Atomic set */ + +TRY l32i a7, a3, 0 # read old value as return value + s32i a4, a3, 0 # write new value + +.Lret: mov a0, a2 + mov a2, a7 + l32i a7, a0, PT_AREG7 + l32i a3, a0, PT_AREG3 + l32i a0, a0, PT_AREG0 + rfe + +CATCH +.Leac: movi a7, -EFAULT + j .Lret + + + +/* fast_syscall_spill_registers. + * + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * Note: We assume the stack pointer is EXC_TABLE_KSTK in the fixup handler. + * Note: We don't need to save a2 in depc (return value) + */ + +ENTRY(fast_syscall_spill_registers) + + /* Register a FIXUP handler (pass current wb as a parameter) */ + + movi a0, fast_syscall_spill_registers_fixup + s32i a0, a3, EXC_TABLE_FIXUP + rsr a0, WINDOWBASE + s32i a0, a3, EXC_TABLE_PARAM + + /* Save a3 and SAR on stack. */ + + rsr a0, SAR + xsr a3, EXCSAVE_1 # restore a3 and excsave_1 + s32i a0, a2, PT_AREG4 # store SAR to PT_AREG4 + s32i a3, a2, PT_AREG3 + + /* The spill routine might clobber a7, a11, and a15. */ + + s32i a7, a2, PT_AREG5 + s32i a11, a2, PT_AREG6 + s32i a15, a2, PT_AREG7 + + call0 _spill_registers # destroys a3, DEPC, and SAR + + /* Advance PC, restore registers and SAR, and return from exception. */ + + l32i a3, a2, PT_AREG4 + l32i a0, a2, PT_AREG0 + wsr a3, SAR + l32i a3, a2, PT_AREG3 + + /* Restore clobbered registers. */ + + l32i a7, a2, PT_AREG5 + l32i a11, a2, PT_AREG6 + l32i a15, a2, PT_AREG7 + + movi a2, 0 + rfe + +/* Fixup handler. + * + * We get here if the spill routine causes an exception, e.g. tlb miss. + * We basically restore WINDOWBASE and WINDOWSTART to the condition when + * we entered the spill routine and jump to the user exception handler. + * + * a0: value of depc, original value in depc + * a2: trashed, original value in EXC_TABLE_DOUBLE_SAVE + * a3: exctable, original value in excsave1 + */ + +fast_syscall_spill_registers_fixup: + + rsr a2, WINDOWBASE # get current windowbase (a2 is saved) + xsr a0, DEPC # restore depc and a0 + ssl a2 # set shift (32 - WB) + + /* We need to make sure the current registers (a0-a3) are preserved. + * To do this, we simply set the bit for the current window frame + * in WS, so that the exception handlers save them to the task stack. + */ + + rsr a3, EXCSAVE_1 # get spill-mask + slli a2, a3, 1 # shift left by one + + slli a3, a2, 32-WSBITS + src a2, a2, a3 # a1 = xxwww1yyxxxwww1yy...... + wsr a2, WINDOWSTART # set corrected windowstart + + movi a3, exc_table + l32i a2, a3, EXC_TABLE_DOUBLE_SAVE # restore a2 + l32i a3, a3, EXC_TABLE_PARAM # original WB (in user task) + + /* Return to the original (user task) WINDOWBASE. + * We leave the following frame behind: + * a0, a1, a2 same + * a3: trashed (saved in excsave_1) + * depc: depc (we have to return to that address) + * excsave_1: a3 + */ + + wsr a3, WINDOWBASE + rsync + + /* We are now in the original frame when we entered _spill_registers: + * a0: return address + * a1: used, stack pointer + * a2: kernel stack pointer + * a3: available, saved in EXCSAVE_1 + * depc: exception address + * excsave: a3 + * Note: This frame might be the same as above. + */ + +#ifdef SIGNAL_HANDLING_IN_DOUBLE_EXCEPTION + /* Restore registers we precautiously saved. + * We have the value of the 'right' a3 + */ + + l32i a7, a2, PT_AREG5 + l32i a11, a2, PT_AREG6 + l32i a15, a2, PT_AREG7 +#endif + + /* Setup stack pointer. */ + + addi a2, a2, -PT_USER_SIZE + s32i a0, a2, PT_AREG0 + + /* Make sure we return to this fixup handler. */ + + movi a3, fast_syscall_spill_registers_fixup_return + s32i a3, a2, PT_DEPC # setup depc + + /* Jump to the exception handler. */ + + movi a3, exc_table + rsr a0, EXCCAUSE + addx4 a0, a0, a3 # find entry in table + l32i a0, a0, EXC_TABLE_FAST_USER # load handler + jx a0 + +fast_syscall_spill_registers_fixup_return: + + /* When we return here, all registers have been restored (a2: DEPC) */ + + wsr a2, DEPC # exception address + + /* Restore fixup handler. */ + + xsr a3, EXCSAVE_1 + movi a2, fast_syscall_spill_registers_fixup + s32i a2, a3, EXC_TABLE_FIXUP + rsr a2, WINDOWBASE + s32i a2, a3, EXC_TABLE_PARAM + l32i a2, a3, EXC_TABLE_KSTK + +#ifdef SIGNAL_HANDLING_IN_DOUBLE_EXCEPTION + /* Save registers again that might be clobbered. */ + + s32i a7, a2, PT_AREG5 + s32i a11, a2, PT_AREG6 + s32i a15, a2, PT_AREG7 +#endif + + /* Load WB at the time the exception occurred. */ + + rsr a3, SAR # WB is still in SAR + neg a3, a3 + wsr a3, WINDOWBASE + rsync + + /* Restore a3 and return. */ + + movi a3, exc_table + xsr a3, EXCSAVE_1 + + rfde + + +/* + * spill all registers. + * + * This is not a real function. The following conditions must be met: + * + * - must be called with call0. + * - uses DEPC, a3 and SAR. + * - the last 'valid' register of each frame are clobbered. + * - the caller must have registered a fixup handler + * (or be inside a critical section) + * - PS_EXCM must be set (PS_WOE cleared?) + */ + +ENTRY(_spill_registers) + + /* + * Rotate ws so that the current windowbase is at bit 0. + * Assume ws = xxxwww1yy (www1 current window frame). + * Rotate ws right so that a2 = yyxxxwww1. + */ + + wsr a2, DEPC # preserve a2 + rsr a2, WINDOWBASE + rsr a3, WINDOWSTART + ssr a2 # holds WB + slli a2, a3, WSBITS + or a3, a3, a2 # a2 = xxxwww1yyxxxwww1yy + srl a3, a3 + + /* We are done if there are no more than the current register frame. */ + + extui a3, a3, 1, WSBITS-2 # a3 = 0yyxxxwww + movi a2, (1 << (WSBITS-1)) + _beqz a3, .Lnospill # only one active frame? jump + + /* We want 1 at the top, so that we return to the current windowbase */ + + or a3, a3, a2 # 1yyxxxwww + + /* Skip empty frames - get 'oldest' WINDOWSTART-bit. */ + + wsr a3, WINDOWSTART # save shifted windowstart + neg a2, a3 + and a3, a2, a3 # first bit set from right: 000010000 + + ffs_ws a2, a3 # a2: shifts to skip empty frames + movi a3, WSBITS + sub a2, a3, a2 # WSBITS-a2:number of 0-bits from right + ssr a2 # save in SAR for later. + + rsr a3, WINDOWBASE + add a3, a3, a2 + rsr a2, DEPC # restore a2 + wsr a3, WINDOWBASE + rsync + + rsr a3, WINDOWSTART + srl a3, a3 # shift windowstart + + /* WB is now just one frame below the oldest frame in the register + window. WS is shifted so the oldest frame is in bit 0, thus, WB + and WS differ by one 4-register frame. */ + + /* Save frames. Depending what call was used (call4, call8, call12), + * we have to save 4,8. or 12 registers. + */ + + _bbsi.l a3, 1, .Lc4 + _bbsi.l a3, 2, .Lc8 + + /* Special case: we have a call12-frame starting at a4. */ + + _bbci.l a3, 3, .Lc12 # bit 3 shouldn't be zero! (Jump to Lc12 first) + + s32e a4, a1, -16 # a1 is valid with an empty spill area + l32e a4, a5, -12 + s32e a8, a4, -48 + mov a8, a4 + l32e a4, a1, -16 + j .Lc12c + +.Lloop: _bbsi.l a3, 1, .Lc4 + _bbci.l a3, 2, .Lc12 + +.Lc8: s32e a4, a13, -16 + l32e a4, a5, -12 + s32e a8, a4, -32 + s32e a5, a13, -12 + s32e a6, a13, -8 + s32e a7, a13, -4 + s32e a9, a4, -28 + s32e a10, a4, -24 + s32e a11, a4, -20 + + srli a11, a3, 2 # shift windowbase by 2 + rotw 2 + _bnei a3, 1, .Lloop + +.Lexit: /* Done. Do the final rotation, set WS, and return. */ + + rotw 1 + rsr a3, WINDOWBASE + ssl a3 + movi a3, 1 + sll a3, a3 + wsr a3, WINDOWSTART + +.Lnospill: + jx a0 + +.Lc4: s32e a4, a9, -16 + s32e a5, a9, -12 + s32e a6, a9, -8 + s32e a7, a9, -4 + + srli a7, a3, 1 + rotw 1 + _bnei a3, 1, .Lloop + j .Lexit + +.Lc12: _bbci.l a3, 3, .Linvalid_mask # bit 2 shouldn't be zero! + + /* 12-register frame (call12) */ + + l32e a2, a5, -12 + s32e a8, a2, -48 + mov a8, a2 + +.Lc12c: s32e a9, a8, -44 + s32e a10, a8, -40 + s32e a11, a8, -36 + s32e a12, a8, -32 + s32e a13, a8, -28 + s32e a14, a8, -24 + s32e a15, a8, -20 + srli a15, a3, 3 + + /* The stack pointer for a4..a7 is out of reach, so we rotate the + * window, grab the stackpointer, and rotate back. + * Alternatively, we could also use the following approach, but that + * makes the fixup routine much more complicated: + * rotw 1 + * s32e a0, a13, -16 + * ... + * rotw 2 + */ + + rotw 1 + mov a5, a13 + rotw -1 + + s32e a4, a9, -16 + s32e a5, a9, -12 + s32e a6, a9, -8 + s32e a7, a9, -4 + + rotw 3 + + _beqi a3, 1, .Lexit + j .Lloop + +.Linvalid_mask: + + /* We get here because of an unrecoverable error in the window + * registers. If we are in user space, we kill the application, + * however, this condition is unrecoverable in kernel space. + */ + + rsr a0, PS + _bbci.l a0, PS_UM_SHIFT, 1f + + /* User space: Setup a dummy frame and kill application. + * Note: We assume EXC_TABLE_KSTK contains a valid stack pointer. + */ + + movi a0, 1 + movi a1, 0 + + wsr a0, WINDOWSTART + wsr a1, WINDOWBASE + rsync + + movi a0, 0 + + movi a3, exc_table + l32i a1, a3, EXC_TABLE_KSTK + wsr a3, EXCSAVE_1 + + movi a4, PS_WOE_MASK | 1 + wsr a4, PS + rsync + + movi a6, SIGSEGV + movi a4, do_exit + callx4 a4 + +1: /* Kernel space: PANIC! */ + + wsr a0, EXCSAVE_1 + movi a0, unrecoverable_exception + callx0 a0 # should not return +1: j 1b + +/* + * We should never get here. Bail out! + */ + +ENTRY(fast_second_level_miss_double_kernel) + +1: movi a0, unrecoverable_exception + callx0 a0 # should not return +1: j 1b + +/* First-level entry handler for user, kernel, and double 2nd-level + * TLB miss exceptions. Note that for now, user and kernel miss + * exceptions share the same entry point and are handled identically. + * + * An old, less-efficient C version of this function used to exist. + * We include it below, interleaved as comments, for reference. + * + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + */ + +ENTRY(fast_second_level_miss) + + /* Save a1. Note: we don't expect a double exception. */ + + s32i a1, a2, PT_AREG1 + + /* We need to map the page of PTEs for the user task. Find + * the pointer to that page. Also, it's possible for tsk->mm + * to be NULL while tsk->active_mm is nonzero if we faulted on + * a vmalloc address. In that rare case, we must use + * active_mm instead to avoid a fault in this handler. See + * + * http://mail.nl.linux.org/linux-mm/2002-08/msg00258.html + * (or search Internet on "mm vs. active_mm") + * + * if (!mm) + * mm = tsk->active_mm; + * pgd = pgd_offset (mm, regs->excvaddr); + * pmd = pmd_offset (pgd, regs->excvaddr); + * pmdval = *pmd; + */ + + GET_CURRENT(a1,a2) + l32i a0, a1, TASK_MM # tsk->mm + beqz a0, 9f + +8: rsr a1, EXCVADDR # fault address + _PGD_OFFSET(a0, a1, a1) + l32i a0, a0, 0 # read pmdval + //beqi a0, _PAGE_USER, 2f + beqz a0, 2f + + /* Read ptevaddr and convert to top of page-table page. + * + * vpnval = read_ptevaddr_register() & PAGE_MASK; + * vpnval += DTLB_WAY_PGTABLE; + * pteval = mk_pte (virt_to_page(pmd_val(pmdval)), PAGE_KERNEL); + * write_dtlb_entry (pteval, vpnval); + * + * The messy computation for 'pteval' above really simplifies + * into the following: + * + * pteval = ((pmdval - PAGE_OFFSET) & PAGE_MASK) | PAGE_KERNEL + */ + + movi a1, -PAGE_OFFSET + add a0, a0, a1 # pmdval - PAGE_OFFSET + extui a1, a0, 0, PAGE_SHIFT # ... & PAGE_MASK + xor a0, a0, a1 + + + movi a1, PAGE_DIRECTORY + or a0, a0, a1 # ... | PAGE_DIRECTORY + + rsr a1, PTEVADDR + srli a1, a1, PAGE_SHIFT + slli a1, a1, PAGE_SHIFT # ptevaddr & PAGE_MASK + addi a1, a1, DTLB_WAY_PGTABLE # ... + way_number + + wdtlb a0, a1 + dsync + + /* Exit critical section. */ + + movi a0, 0 + s32i a0, a3, EXC_TABLE_FIXUP + + /* Restore the working registers, and return. */ + + l32i a0, a2, PT_AREG0 + l32i a1, a2, PT_AREG1 + l32i a2, a2, PT_DEPC + xsr a3, EXCSAVE_1 + + bgeui a2, VALID_DOUBLE_EXCEPTION_ADDRESS, 1f + + /* Restore excsave1 and return. */ + + rsr a2, DEPC + rfe + + /* Return from double exception. */ + +1: xsr a2, DEPC + esync + rfde + +9: l32i a0, a1, TASK_ACTIVE_MM # unlikely case mm == 0 + j 8b + +2: /* Invalid PGD, default exception handling */ + + rsr a1, DEPC + xsr a3, EXCSAVE_1 + s32i a1, a2, PT_AREG2 + s32i a3, a2, PT_AREG3 + mov a1, a2 + + rsr a2, PS + bbsi.l a2, PS_UM_SHIFT, 1f + j _kernel_exception +1: j _user_exception + + +/* + * StoreProhibitedException + * + * Update the pte and invalidate the itlb mapping for this pte. + * + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + */ + +ENTRY(fast_store_prohibited) + + /* Save a1 and a4. */ + + s32i a1, a2, PT_AREG1 + s32i a4, a2, PT_AREG4 + + GET_CURRENT(a1,a2) + l32i a0, a1, TASK_MM # tsk->mm + beqz a0, 9f + +8: rsr a1, EXCVADDR # fault address + _PGD_OFFSET(a0, a1, a4) + l32i a0, a0, 0 + //beqi a0, _PAGE_USER, 2f # FIXME use _PAGE_INVALID + beqz a0, 2f + + _PTE_OFFSET(a0, a1, a4) + l32i a4, a0, 0 # read pteval + movi a1, _PAGE_VALID | _PAGE_RW + bnall a4, a1, 2f + + movi a1, _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_WRENABLE + or a4, a4, a1 + rsr a1, EXCVADDR + s32i a4, a0, 0 + + /* We need to flush the cache if we have page coloring. */ +#if (DCACHE_WAY_SIZE > PAGE_SIZE) && XCHAL_DCACHE_IS_WRITEBACK + dhwb a0, 0 +#endif + pdtlb a0, a1 + beqz a0, 1f + idtlb a0 // FIXME do we need this? + wdtlb a4, a0 +1: + + /* Exit critical section. */ + + movi a0, 0 + s32i a0, a3, EXC_TABLE_FIXUP + + /* Restore the working registers, and return. */ + + l32i a4, a2, PT_AREG4 + l32i a1, a2, PT_AREG1 + l32i a0, a2, PT_AREG0 + l32i a2, a2, PT_DEPC + + /* Restore excsave1 and a3. */ + + xsr a3, EXCSAVE_1 + bgeui a2, VALID_DOUBLE_EXCEPTION_ADDRESS, 1f + + rsr a2, DEPC + rfe + + /* Double exception. Restore FIXUP handler and return. */ + +1: xsr a2, DEPC + esync + rfde + +9: l32i a0, a1, TASK_ACTIVE_MM # unlikely case mm == 0 + j 8b + +2: /* If there was a problem, handle fault in C */ + + rsr a4, DEPC # still holds a2 + xsr a3, EXCSAVE_1 + s32i a4, a2, PT_AREG2 + s32i a3, a2, PT_AREG3 + l32i a4, a2, PT_AREG4 + mov a1, a2 + + rsr a2, PS + bbsi.l a2, PS_UM_SHIFT, 1f + j _kernel_exception +1: j _user_exception + + +#if XCHAL_EXTRA_SA_SIZE + +#warning fast_coprocessor untested + +/* + * Entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original in DEPC + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + */ + +ENTRY(fast_coprocessor_double) + wsr a0, EXCSAVE_1 + movi a0, unrecoverable_exception + callx0 a0 + +ENTRY(fast_coprocessor) + + /* Fatal if we are in a double exception. */ + + l32i a0, a2, PT_DEPC + _bgeui a0, VALID_DOUBLE_EXCEPTION_ADDRESS, fast_coprocessor_double + + /* Save some registers a1, a3, a4, SAR */ + + xsr a3, EXCSAVE_1 + s32i a3, a2, PT_AREG3 + rsr a3, SAR + s32i a4, a2, PT_AREG4 + s32i a1, a2, PT_AREG1 + s32i a5, a1, PT_AREG5 + s32i a3, a2, PT_SAR + mov a1, a2 + + /* Currently, the HAL macros only guarantee saving a0 and a1. + * These can and will be refined in the future, but for now, + * just save the remaining registers of a2...a15. + */ + s32i a6, a1, PT_AREG6 + s32i a7, a1, PT_AREG7 + s32i a8, a1, PT_AREG8 + s32i a9, a1, PT_AREG9 + s32i a10, a1, PT_AREG10 + s32i a11, a1, PT_AREG11 + s32i a12, a1, PT_AREG12 + s32i a13, a1, PT_AREG13 + s32i a14, a1, PT_AREG14 + s32i a15, a1, PT_AREG15 + + /* Find coprocessor number. Subtract first CP EXCCAUSE from EXCCAUSE */ + + rsr a0, EXCCAUSE + addi a3, a0, -XCHAL_EXCCAUSE_COPROCESSOR0_DISABLED + + /* Set corresponding CPENABLE bit */ + + movi a4, 1 + ssl a3 # SAR: 32 - coprocessor_number + rsr a5, CPENABLE + sll a4, a4 + or a4, a5, a4 + wsr a4, CPENABLE + rsync + movi a5, coprocessor_info # list of owner and offset into cp_save + addx8 a0, a4, a5 # entry for CP + + bne a4, a5, .Lload # bit wasn't set before, cp not in use + + /* Now compare the current task with the owner of the coprocessor. + * If they are the same, there is no reason to save or restore any + * coprocessor state. Having already enabled the coprocessor, + * branch ahead to return. + */ + GET_CURRENT(a5,a1) + l32i a4, a0, COPROCESSOR_INFO_OWNER # a4: current owner for this CP + beq a4, a5, .Ldone + + /* Find location to dump current coprocessor state: + * task_struct->task_cp_save_offset + coprocessor_offset[coprocessor] + * + * Note: a0 pointer to the entry in the coprocessor owner table, + * a3 coprocessor number, + * a4 current owner of coprocessor. + */ + l32i a5, a0, COPROCESSOR_INFO_OFFSET + addi a2, a4, THREAD_CP_SAVE + add a2, a2, a5 + + /* Store current coprocessor states. (a5 still has CP number) */ + + xchal_cpi_store_funcbody + + /* The macro might have destroyed a3 (coprocessor number), but + * SAR still has 32 - coprocessor_number! + */ + movi a3, 32 + rsr a4, SAR + sub a3, a3, a4 + +.Lload: /* A new task now owns the corpocessors. Save its TCB pointer into + * the coprocessor owner table. + * + * Note: a0 pointer to the entry in the coprocessor owner table, + * a3 coprocessor number. + */ + GET_CURRENT(a4,a1) + s32i a4, a0, 0 + + /* Find location from where to restore the current coprocessor state.*/ + + l32i a5, a0, COPROCESSOR_INFO_OFFSET + addi a2, a4, THREAD_CP_SAVE + add a2, a2, a4 + + xchal_cpi_load_funcbody + + /* We must assume that the xchal_cpi_store_funcbody macro destroyed + * registers a2..a15. + */ + +.Ldone: l32i a15, a1, PT_AREG15 + l32i a14, a1, PT_AREG14 + l32i a13, a1, PT_AREG13 + l32i a12, a1, PT_AREG12 + l32i a11, a1, PT_AREG11 + l32i a10, a1, PT_AREG10 + l32i a9, a1, PT_AREG9 + l32i a8, a1, PT_AREG8 + l32i a7, a1, PT_AREG7 + l32i a6, a1, PT_AREG6 + l32i a5, a1, PT_AREG5 + l32i a4, a1, PT_AREG4 + l32i a3, a1, PT_AREG3 + l32i a2, a1, PT_AREG2 + l32i a0, a1, PT_AREG0 + l32i a1, a1, PT_AREG1 + + rfe + +#endif /* XCHAL_EXTRA_SA_SIZE */ + +/* + * Task switch. + * + * struct task* _switch_to (struct task* prev, struct task* next) + * a2 a2 a3 + */ + +ENTRY(_switch_to) + + entry a1, 16 + + mov a4, a3 # preserve a3 + + s32i a0, a2, THREAD_RA # save return address + s32i a1, a2, THREAD_SP # save stack pointer + + /* Disable ints while we manipulate the stack pointer; spill regs. */ + + movi a5, PS_EXCM_MASK | LOCKLEVEL + xsr a5, PS + rsr a3, EXCSAVE_1 + rsync + s32i a3, a3, EXC_TABLE_FIXUP /* enter critical section */ + + call0 _spill_registers + + /* Set kernel stack (and leave critical section) + * Note: It's save to set it here. The stack will not be overwritten + * because the kernel stack will only be loaded again after + * we return from kernel space. + */ + + l32i a0, a4, TASK_THREAD_INFO + rsr a3, EXCSAVE_1 # exc_table + movi a1, 0 + addi a0, a0, PT_REGS_OFFSET + s32i a1, a3, EXC_TABLE_FIXUP + s32i a0, a3, EXC_TABLE_KSTK + + /* restore context of the task that 'next' addresses */ + + l32i a0, a4, THREAD_RA /* restore return address */ + l32i a1, a4, THREAD_SP /* restore stack pointer */ + + wsr a5, PS + rsync + + retw + + +ENTRY(ret_from_fork) + + /* void schedule_tail (struct task_struct *prev) + * Note: prev is still in a6 (return value from fake call4 frame) + */ + movi a4, schedule_tail + callx4 a4 + + movi a4, do_syscall_trace + callx4 a4 + + j common_exception_return + + + +/* + * Table of syscalls + */ + +.data +.align 4 +.global sys_call_table +sys_call_table: + +#define SYSCALL(call, narg) .word call +#include "syscalls.h" + +/* + * Number of arguments of each syscall + */ + +.global sys_narg_table +sys_narg_table: + +#undef SYSCALL +#define SYSCALL(call, narg) .byte narg +#include "syscalls.h" + diff --git a/arch/xtensa/kernel/head.S b/arch/xtensa/kernel/head.S new file mode 100644 index 00000000000..6e9b5225b8f --- /dev/null +++ b/arch/xtensa/kernel/head.S @@ -0,0 +1,237 @@ +/* + * arch/xtensa/kernel/head.S + * + * Xtensa Processor startup code. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca> + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Kevin Chea + */ + +#include <xtensa/cacheasm.h> +#include <linux/config.h> +#include <asm/processor.h> +#include <asm/page.h> + +/* + * This module contains the entry code for kernel images. It performs the + * minimal setup needed to call the generic C routines. + * + * Prerequisites: + * + * - The kernel image has been loaded to the actual address where it was + * compiled to. + * - a2 contains either 0 or a pointer to a list of boot parameters. + * (see setup.c for more details) + * + */ + + .macro iterate from, to , cmd + .ifeq ((\to - \from) & ~0xfff) + \cmd \from + iterate "(\from+1)", \to, \cmd + .endif + .endm + +/* + * _start + * + * The bootloader passes a pointer to a list of boot parameters in a2. + */ + + /* The first bytes of the kernel image must be an instruction, so we + * manually allocate and define the literal constant we need for a jx + * instruction. + */ + + .section .head.text, "ax" + .globl _start +_start: _j 2f + .align 4 +1: .word _startup +2: l32r a0, 1b + jx a0 + + .text + .align 4 +_startup: + + /* Disable interrupts and exceptions. */ + + movi a0, XCHAL_PS_EXCM_MASK + wsr a0, PS + + /* Preserve the pointer to the boot parameter list in EXCSAVE_1 */ + + wsr a2, EXCSAVE_1 + + /* Start with a fresh windowbase and windowstart. */ + + movi a1, 1 + movi a0, 0 + wsr a1, WINDOWSTART + wsr a0, WINDOWBASE + rsync + + /* Set a0 to 0 for the remaining initialization. */ + + movi a0, 0 + + /* Clear debugging registers. */ + +#if XCHAL_HAVE_DEBUG + wsr a0, IBREAKENABLE + wsr a0, ICOUNT + movi a1, 15 + wsr a0, ICOUNTLEVEL + + .macro reset_dbreak num + wsr a0, DBREAKC + \num + .endm + + iterate 0, XCHAL_NUM_IBREAK-1, reset_dbreak +#endif + + /* Clear CCOUNT (not really necessary, but nice) */ + + wsr a0, CCOUNT # not really necessary, but nice + + /* Disable zero-loops. */ + +#if XCHAL_HAVE_LOOPS + wsr a0, LCOUNT +#endif + + /* Disable all timers. */ + + .macro reset_timer num + wsr a0, CCOMPARE_0 + \num + .endm + iterate 0, XCHAL_NUM_TIMERS-1, reset_timer + + /* Interrupt initialization. */ + + movi a2, XCHAL_INTTYPE_MASK_SOFTWARE | XCHAL_INTTYPE_MASK_EXTERN_EDGE + wsr a0, INTENABLE + wsr a2, INTCLEAR + + /* Disable coprocessors. */ + +#if XCHAL_CP_NUM > 0 + wsr a0, CPENABLE +#endif + + /* Set PS.INTLEVEL=1, PS.WOE=0, kernel stack, PS.EXCM=0 + * + * Note: PS.EXCM must be cleared before using any loop + * instructions; otherwise, they are silently disabled, and + * at most one iteration of the loop is executed. + */ + + movi a1, 1 + wsr a1, PS + rsync + + /* Initialize the caches. + * Does not include flushing writeback d-cache. + * a6, a7 are just working registers (clobbered). + */ + + icache_reset a2, a3 + dcache_reset a2, a3 + + /* Unpack data sections + * + * The linker script used to build the Linux kernel image + * creates a table located at __boot_reloc_table_start + * that contans the information what data needs to be unpacked. + * + * Uses a2-a7. + */ + + movi a2, __boot_reloc_table_start + movi a3, __boot_reloc_table_end + +1: beq a2, a3, 3f # no more entries? + l32i a4, a2, 0 # start destination (in RAM) + l32i a5, a2, 4 # end desination (in RAM) + l32i a6, a2, 8 # start source (in ROM) + addi a2, a2, 12 # next entry + beq a4, a5, 1b # skip, empty entry + beq a4, a6, 1b # skip, source and dest. are the same + +2: l32i a7, a6, 0 # load word + addi a6, a6, 4 + s32i a7, a4, 0 # store word + addi a4, a4, 4 + bltu a4, a5, 2b + j 1b + +3: + /* All code and initialized data segments have been copied. + * Now clear the BSS segment. + */ + + movi a2, _bss_start # start of BSS + movi a3, _bss_end # end of BSS + +1: addi a2, a2, 4 + s32i a0, a2, 0 + blt a2, a3, 1b + +#if XCHAL_DCACHE_IS_WRITEBACK + + /* After unpacking, flush the writeback cache to memory so the + * instructions/data are available. + */ + + dcache_writeback_all a2, a3 +#endif + + /* Setup stack and enable window exceptions (keep irqs disabled) */ + + movi a1, init_thread_union + addi a1, a1, KERNEL_STACK_SIZE + + movi a2, 0x00040001 # WOE=1, INTLEVEL=1, UM=0 + wsr a2, PS # (enable reg-windows; progmode stack) + rsync + + /* Set up EXCSAVE[DEBUGLEVEL] to point to the Debug Exception Handler.*/ + + movi a2, debug_exception + wsr a2, EXCSAVE + XCHAL_DEBUGLEVEL + + /* Set up EXCSAVE[1] to point to the exc_table. */ + + movi a6, exc_table + xsr a6, EXCSAVE_1 + + /* init_arch kick-starts the linux kernel */ + + movi a4, init_arch + callx4 a4 + + movi a4, start_kernel + callx4 a4 + +should_never_return: + j should_never_return + + /* Define some common data structures here. We define them + * here in this assembly file due to their unusual alignment + * requirements. + */ + + .comm swapper_pg_dir,PAGE_SIZE,PAGE_SIZE + .comm empty_bad_page_table,PAGE_SIZE,PAGE_SIZE + .comm empty_bad_page,PAGE_SIZE,PAGE_SIZE + .comm empty_zero_page,PAGE_SIZE,PAGE_SIZE + diff --git a/arch/xtensa/kernel/irq.c b/arch/xtensa/kernel/irq.c new file mode 100644 index 00000000000..4cbf6d91571 --- /dev/null +++ b/arch/xtensa/kernel/irq.c @@ -0,0 +1,192 @@ +/* + * linux/arch/xtensa/kernel/irq.c + * + * Xtensa built-in interrupt controller and some generic functions copied + * from i386. + * + * Copyright (C) 2002 - 2005 Tensilica, Inc. + * Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar + * + * + * Chris Zankel <chris@zankel.net> + * Kevin Chea + * + */ + +#include <linux/module.h> +#include <linux/seq_file.h> +#include <linux/interrupt.h> +#include <linux/irq.h> +#include <linux/kernel_stat.h> + +#include <asm/uaccess.h> +#include <asm/platform.h> + +static void enable_xtensa_irq(unsigned int irq); +static void disable_xtensa_irq(unsigned int irq); +static void mask_and_ack_xtensa(unsigned int irq); +static void end_xtensa_irq(unsigned int irq); + +static unsigned int cached_irq_mask; + +atomic_t irq_err_count; + +/* + * 'what should we do if we get a hw irq event on an illegal vector'. + * each architecture has to answer this themselves. + */ +void ack_bad_irq(unsigned int irq) +{ + printk("unexpected IRQ trap at vector %02x\n", irq); +} + +/* + * do_IRQ handles all normal device IRQ's (the special + * SMP cross-CPU interrupts have their own specific + * handlers). + */ + +unsigned int do_IRQ(int irq, struct pt_regs *regs) +{ + irq_enter(); + +#ifdef CONFIG_DEBUG_STACKOVERFLOW + /* Debugging check for stack overflow: is there less than 1KB free? */ + { + unsigned long sp; + + __asm__ __volatile__ ("mov %0, a1\n" : "=a" (sp)); + sp &= THREAD_SIZE - 1; + + if (unlikely(sp < (sizeof(thread_info) + 1024))) + printk("Stack overflow in do_IRQ: %ld\n", + sp - sizeof(struct thread_info)); + } +#endif + + __do_IRQ(irq, regs); + + irq_exit(); + + return 1; +} + +/* + * Generic, controller-independent functions: + */ + +int show_interrupts(struct seq_file *p, void *v) +{ + int i = *(loff_t *) v, j; + struct irqaction * action; + unsigned long flags; + + if (i == 0) { + seq_printf(p, " "); + for (j=0; j<NR_CPUS; j++) + if (cpu_online(j)) + seq_printf(p, "CPU%d ",j); + seq_putc(p, '\n'); + } + + if (i < NR_IRQS) { + spin_lock_irqsave(&irq_desc[i].lock, flags); + action = irq_desc[i].action; + if (!action) + goto skip; + seq_printf(p, "%3d: ",i); +#ifndef CONFIG_SMP + seq_printf(p, "%10u ", kstat_irqs(i)); +#else + for (j = 0; j < NR_CPUS; j++) + if (cpu_online(j)) + seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]); +#endif + seq_printf(p, " %14s", irq_desc[i].handler->typename); + seq_printf(p, " %s", action->name); + + for (action=action->next; action; action = action->next) + seq_printf(p, ", %s", action->name); + + seq_putc(p, '\n'); +skip: + spin_unlock_irqrestore(&irq_desc[i].lock, flags); + } else if (i == NR_IRQS) { + seq_printf(p, "NMI: "); + for (j = 0; j < NR_CPUS; j++) + if (cpu_online(j)) + seq_printf(p, "%10u ", nmi_count(j)); + seq_putc(p, '\n'); + seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count)); + } + return 0; +} +/* shutdown is same as "disable" */ +#define shutdown_xtensa_irq disable_xtensa_irq + +static unsigned int startup_xtensa_irq(unsigned int irq) +{ + enable_xtensa_irq(irq); + return 0; /* never anything pending */ +} + +static struct hw_interrupt_type xtensa_irq_type = { + "Xtensa-IRQ", + startup_xtensa_irq, + shutdown_xtensa_irq, + enable_xtensa_irq, + disable_xtensa_irq, + mask_and_ack_xtensa, + end_xtensa_irq +}; + +static inline void mask_irq(unsigned int irq) +{ + cached_irq_mask &= ~(1 << irq); + set_sr (cached_irq_mask, INTENABLE); +} + +static inline void unmask_irq(unsigned int irq) +{ + cached_irq_mask |= 1 << irq; + set_sr (cached_irq_mask, INTENABLE); +} + +static void disable_xtensa_irq(unsigned int irq) +{ + unsigned long flags; + local_save_flags(flags); + mask_irq(irq); + local_irq_restore(flags); +} + +static void enable_xtensa_irq(unsigned int irq) +{ + unsigned long flags; + local_save_flags(flags); + unmask_irq(irq); + local_irq_restore(flags); +} + +static void mask_and_ack_xtensa(unsigned int irq) +{ + disable_xtensa_irq(irq); +} + +static void end_xtensa_irq(unsigned int irq) +{ + enable_xtensa_irq(irq); +} + + +void __init init_IRQ(void) +{ + int i; + + for (i=0; i < XTENSA_NR_IRQS; i++) + irq_desc[i].handler = &xtensa_irq_type; + + cached_irq_mask = 0; + + platform_init_irq(); +} diff --git a/arch/xtensa/kernel/module.c b/arch/xtensa/kernel/module.c new file mode 100644 index 00000000000..d1683cfa19a --- /dev/null +++ b/arch/xtensa/kernel/module.c @@ -0,0 +1,78 @@ +/* + * arch/xtensa/kernel/platform.c + * + * Module support. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * + */ + +#include <linux/module.h> +#include <linux/moduleloader.h> +#include <linux/elf.h> +#include <linux/vmalloc.h> +#include <linux/fs.h> +#include <linux/string.h> +#include <linux/kernel.h> +#include <linux/cache.h> + +LIST_HEAD(module_buf_list); + +void *module_alloc(unsigned long size) +{ + panic("module_alloc not implemented"); +} + +void module_free(struct module *mod, void *module_region) +{ + panic("module_free not implemented"); +} + +int module_frob_arch_sections(Elf32_Ehdr *hdr, + Elf32_Shdr *sechdrs, + char *secstrings, + struct module *me) +{ + panic("module_frob_arch_sections not implemented"); +} + +int apply_relocate(Elf32_Shdr *sechdrs, + const char *strtab, + unsigned int symindex, + unsigned int relsec, + struct module *module) +{ + panic ("apply_relocate not implemented"); +} + +int apply_relocate_add(Elf32_Shdr *sechdrs, + const char *strtab, + unsigned int symindex, + unsigned int relsec, + struct module *module) +{ + panic("apply_relocate_add not implemented"); +} + +int module_finalize(const Elf_Ehdr *hdr, + const Elf_Shdr *sechdrs, + struct module *me) +{ + panic ("module_finalize not implemented"); +} + +void module_arch_cleanup(struct module *mod) +{ + panic("module_arch_cleanup not implemented"); +} + +struct bug_entry *module_find_bug(unsigned long bugaddr) +{ + panic("module_find_bug not implemented"); +} diff --git a/arch/xtensa/kernel/pci-dma.c b/arch/xtensa/kernel/pci-dma.c new file mode 100644 index 00000000000..84fde258cf8 --- /dev/null +++ b/arch/xtensa/kernel/pci-dma.c @@ -0,0 +1,73 @@ +/* + * arch/xtensa/pci-dma.c + * + * DMA coherent memory allocation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * Copyright (C) 2002 - 2005 Tensilica Inc. + * + * Based on version for i386. + * + * Chris Zankel <chris@zankel.net> + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + */ + +#include <linux/types.h> +#include <linux/mm.h> +#include <linux/string.h> +#include <linux/pci.h> +#include <asm/io.h> +#include <asm/cacheflush.h> + +/* + * Note: We assume that the full memory space is always mapped to 'kseg' + * Otherwise we have to use page attributes (not implemented). + */ + +void * +dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *handle, int gfp) +{ + void *ret; + + /* ignore region speicifiers */ + gfp &= ~(__GFP_DMA | __GFP_HIGHMEM); + + if (dev == NULL || (*dev->dma_mask < 0xffffffff)) + gfp |= GFP_DMA; + ret = (void *)__get_free_pages(gfp, get_order(size)); + + if (ret != NULL) { + memset(ret, 0, size); + *handle = virt_to_bus(ret); + } + return (void*) BYPASS_ADDR((unsigned long)ret); +} + +void dma_free_coherent(struct device *hwdev, size_t size, + void *vaddr, dma_addr_t dma_handle) +{ + free_pages(CACHED_ADDR((unsigned long)vaddr), get_order(size)); +} + + +void consistent_sync(void *vaddr, size_t size, int direction) +{ + switch (direction) { + case PCI_DMA_NONE: + BUG(); + case PCI_DMA_FROMDEVICE: /* invalidate only */ + __invalidate_dcache_range((unsigned long)vaddr, + (unsigned long)size); + break; + + case PCI_DMA_TODEVICE: /* writeback only */ + case PCI_DMA_BIDIRECTIONAL: /* writeback and invalidate */ + __flush_invalidate_dcache_range((unsigned long)vaddr, + (unsigned long)size); + break; + } +} diff --git a/arch/xtensa/kernel/pci.c b/arch/xtensa/kernel/pci.c new file mode 100644 index 00000000000..d29a8164863 --- /dev/null +++ b/arch/xtensa/kernel/pci.c @@ -0,0 +1,563 @@ +/* + * arch/xtensa/pcibios.c + * + * PCI bios-type initialisation for PCI machines + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * Copyright (C) 2001-2005 Tensilica Inc. + * + * Based largely on work from Cort (ppc/kernel/pci.c) + * IO functions copied from sparc. + * + * Chris Zankel <chris@zankel.net> + * + */ + +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/pci.h> +#include <linux/delay.h> +#include <linux/string.h> +#include <linux/init.h> +#include <linux/sched.h> +#include <linux/errno.h> +#include <linux/bootmem.h> + +#include <asm/pci-bridge.h> +#include <asm/platform.h> + +#undef DEBUG + +#ifdef DEBUG +#define DBG(x...) printk(x) +#else +#define DBG(x...) +#endif + +/* PCI Controller */ + + +/* + * pcibios_alloc_controller + * pcibios_enable_device + * pcibios_fixups + * pcibios_align_resource + * pcibios_fixup_bus + * pcibios_setup + * pci_bus_add_device + * pci_mmap_page_range + */ + +struct pci_controller* pci_ctrl_head; +struct pci_controller** pci_ctrl_tail = &pci_ctrl_head; + +static int pci_bus_count; + +static void pcibios_fixup_resources(struct pci_dev* dev); + +#if 0 // FIXME +struct pci_fixup pcibios_fixups[] = { + { DECLARE_PCI_FIXUP_HEADER, PCI_ANY_ID, PCI_ANY_ID, pcibios_fixup_resources }, + { 0 } +}; +#endif + +void +pcibios_update_resource(struct pci_dev *dev, struct resource *root, + struct resource *res, int resource) +{ + u32 new, check, mask; + int reg; + struct pci_controller* pci_ctrl = dev->sysdata; + + new = res->start; + if (pci_ctrl && res->flags & IORESOURCE_IO) { + new -= pci_ctrl->io_space.base; + } + new |= (res->flags & PCI_REGION_FLAG_MASK); + if (resource < 6) { + reg = PCI_BASE_ADDRESS_0 + 4*resource; + } else if (resource == PCI_ROM_RESOURCE) { + res->flags |= PCI_ROM_ADDRESS_ENABLE; + reg = dev->rom_base_reg; + } else { + /* Somebody might have asked allocation of a non-standard resource */ + return; + } + + pci_write_config_dword(dev, reg, new); + pci_read_config_dword(dev, reg, &check); + mask = (new & PCI_BASE_ADDRESS_SPACE_IO) ? + PCI_BASE_ADDRESS_IO_MASK : PCI_BASE_ADDRESS_MEM_MASK; + + if ((new ^ check) & mask) { + printk(KERN_ERR "PCI: Error while updating region " + "%s/%d (%08x != %08x)\n", dev->slot_name, resource, + new, check); + } +} + +/* + * We need to avoid collisions with `mirrored' VGA ports + * and other strange ISA hardware, so we always want the + * addresses to be allocated in the 0x000-0x0ff region + * modulo 0x400. + * + * Why? Because some silly external IO cards only decode + * the low 10 bits of the IO address. The 0x00-0xff region + * is reserved for motherboard devices that decode all 16 + * bits, so it's ok to allocate at, say, 0x2800-0x28ff, + * but we want to try to avoid allocating at 0x2900-0x2bff + * which might have be mirrored at 0x0100-0x03ff.. + */ +void +pcibios_align_resource(void *data, struct resource *res, unsigned long size, + unsigned long align) +{ + struct pci_dev *dev = data; + + if (res->flags & IORESOURCE_IO) { + unsigned long start = res->start; + + if (size > 0x100) { + printk(KERN_ERR "PCI: I/O Region %s/%d too large" + " (%ld bytes)\n", dev->slot_name, + dev->resource - res, size); + } + + if (start & 0x300) { + start = (start + 0x3ff) & ~0x3ff; + res->start = start; + } + } +} + +int +pcibios_enable_resources(struct pci_dev *dev, int mask) +{ + u16 cmd, old_cmd; + int idx; + struct resource *r; + + pci_read_config_word(dev, PCI_COMMAND, &cmd); + old_cmd = cmd; + for(idx=0; idx<6; idx++) { + r = &dev->resource[idx]; + if (!r->start && r->end) { + printk (KERN_ERR "PCI: Device %s not available because " + "of resource collisions\n", dev->slot_name); + return -EINVAL; + } + if (r->flags & IORESOURCE_IO) + cmd |= PCI_COMMAND_IO; + if (r->flags & IORESOURCE_MEM) + cmd |= PCI_COMMAND_MEMORY; + } + if (dev->resource[PCI_ROM_RESOURCE].start) + cmd |= PCI_COMMAND_MEMORY; + if (cmd != old_cmd) { + printk("PCI: Enabling device %s (%04x -> %04x)\n", + dev->slot_name, old_cmd, cmd); + pci_write_config_word(dev, PCI_COMMAND, cmd); + } + return 0; +} + +struct pci_controller * __init pcibios_alloc_controller(void) +{ + struct pci_controller *pci_ctrl; + + pci_ctrl = (struct pci_controller *)alloc_bootmem(sizeof(*pci_ctrl)); + memset(pci_ctrl, 0, sizeof(struct pci_controller)); + + *pci_ctrl_tail = pci_ctrl; + pci_ctrl_tail = &pci_ctrl->next; + + return pci_ctrl; +} + +static int __init pcibios_init(void) +{ + struct pci_controller *pci_ctrl; + struct pci_bus *bus; + int next_busno = 0, i; + + printk("PCI: Probing PCI hardware\n"); + + /* Scan all of the recorded PCI controllers. */ + for (pci_ctrl = pci_ctrl_head; pci_ctrl; pci_ctrl = pci_ctrl->next) { + pci_ctrl->last_busno = 0xff; + bus = pci_scan_bus(pci_ctrl->first_busno, pci_ctrl->ops, + pci_ctrl); + if (pci_ctrl->io_resource.flags) { + unsigned long offs; + + offs = (unsigned long)pci_ctrl->io_space.base; + pci_ctrl->io_resource.start += offs; + pci_ctrl->io_resource.end += offs; + bus->resource[0] = &pci_ctrl->io_resource; + } + for (i = 0; i < 3; ++i) + if (pci_ctrl->mem_resources[i].flags) + bus->resource[i+1] =&pci_ctrl->mem_resources[i]; + pci_ctrl->bus = bus; + pci_ctrl->last_busno = bus->subordinate; + if (next_busno <= pci_ctrl->last_busno) + next_busno = pci_ctrl->last_busno+1; + } + pci_bus_count = next_busno; + + return platform_pcibios_fixup(); +} + +subsys_initcall(pcibios_init); + +void __init pcibios_fixup_bus(struct pci_bus *bus) +{ + struct pci_controller *pci_ctrl = bus->sysdata; + struct resource *res; + unsigned long io_offset; + int i; + + io_offset = (unsigned long)pci_ctrl->io_space.base; + if (bus->parent == NULL) { + /* this is a host bridge - fill in its resources */ + pci_ctrl->bus = bus; + + bus->resource[0] = res = &pci_ctrl->io_resource; + if (!res->flags) { + if (io_offset) + printk (KERN_ERR "I/O resource not set for host" + " bridge %d\n", pci_ctrl->index); + res->start = 0; + res->end = IO_SPACE_LIMIT; + res->flags = IORESOURCE_IO; + } + res->start += io_offset; + res->end += io_offset; + + for (i = 0; i < 3; i++) { + res = &pci_ctrl->mem_resources[i]; + if (!res->flags) { + if (i > 0) + continue; + printk(KERN_ERR "Memory resource not set for " + "host bridge %d\n", pci_ctrl->index); + res->start = 0; + res->end = ~0U; + res->flags = IORESOURCE_MEM; + } + bus->resource[i+1] = res; + } + } else { + /* This is a subordinate bridge */ + pci_read_bridge_bases(bus); + + for (i = 0; i < 4; i++) { + if ((res = bus->resource[i]) == NULL || !res->flags) + continue; + if (io_offset && (res->flags & IORESOURCE_IO)) { + res->start += io_offset; + res->end += io_offset; + } + } + } +} + +char __init *pcibios_setup(char *str) +{ + return str; +} + +/* the next one is stolen from the alpha port... */ + +void __init +pcibios_update_irq(struct pci_dev *dev, int irq) +{ + pci_write_config_byte(dev, PCI_INTERRUPT_LINE, irq); +} + +int pcibios_enable_device(struct pci_dev *dev, int mask) +{ + u16 cmd, old_cmd; + int idx; + struct resource *r; + + pci_read_config_word(dev, PCI_COMMAND, &cmd); + old_cmd = cmd; + for (idx=0; idx<6; idx++) { + r = &dev->resource[idx]; + if (!r->start && r->end) { + printk(KERN_ERR "PCI: Device %s not available because " + "of resource collisions\n", dev->slot_name); + return -EINVAL; + } + if (r->flags & IORESOURCE_IO) + cmd |= PCI_COMMAND_IO; + if (r->flags & IORESOURCE_MEM) + cmd |= PCI_COMMAND_MEMORY; + } + if (cmd != old_cmd) { + printk("PCI: Enabling device %s (%04x -> %04x)\n", + dev->slot_name, old_cmd, cmd); + pci_write_config_word(dev, PCI_COMMAND, cmd); + } + + return 0; +} + +#ifdef CONFIG_PROC_FS + +/* + * Return the index of the PCI controller for device pdev. + */ + +int +pci_controller_num(struct pci_dev *dev) +{ + struct pci_controller *pci_ctrl = (struct pci_controller*) dev->sysdata; + return pci_ctrl->index; +} + +#endif /* CONFIG_PROC_FS */ + + +static void +pcibios_fixup_resources(struct pci_dev *dev) +{ + struct pci_controller* pci_ctrl = (struct pci_controller *)dev->sysdata; + int i; + unsigned long offset; + + if (!pci_ctrl) { + printk(KERN_ERR "No pci_ctrl for PCI dev %s!\n",dev->slot_name); + return; + } + for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) { + struct resource *res = dev->resource + i; + if (!res->start || !res->flags) + continue; + if (res->end == 0xffffffff) { + DBG("PCI:%s Resource %d [%08lx-%08lx] is unassigned\n", + dev->slot_name, i, res->start, res->end); + res->end -= res->start; + res->start = 0; + continue; + } + offset = 0; + if (res->flags & IORESOURCE_IO) + offset = (unsigned long) pci_ctrl->io_space.base; + else if (res->flags & IORESOURCE_MEM) + offset = (unsigned long) pci_ctrl->mem_space.base; + + if (offset != 0) { + res->start += offset; + res->end += offset; +#ifdef DEBUG + printk("Fixup res %d (%lx) of dev %s: %lx -> %lx\n", + i, res->flags, dev->slot_name, + res->start - offset, res->start); +#endif + } + } +} + +/* + * Platform support for /proc/bus/pci/X/Y mmap()s, + * modelled on the sparc64 implementation by Dave Miller. + * -- paulus. + */ + +/* + * Adjust vm_pgoff of VMA such that it is the physical page offset + * corresponding to the 32-bit pci bus offset for DEV requested by the user. + * + * Basically, the user finds the base address for his device which he wishes + * to mmap. They read the 32-bit value from the config space base register, + * add whatever PAGE_SIZE multiple offset they wish, and feed this into the + * offset parameter of mmap on /proc/bus/pci/XXX for that device. + * + * Returns negative error code on failure, zero on success. + */ +static __inline__ int +__pci_mmap_make_offset(struct pci_dev *dev, struct vm_area_struct *vma, + enum pci_mmap_state mmap_state) +{ + struct pci_controller *pci_ctrl = (struct pci_controller*) dev->sysdata; + unsigned long offset = vma->vm_pgoff << PAGE_SHIFT; + unsigned long io_offset = 0; + int i, res_bit; + + if (pci_ctrl == 0) + return -EINVAL; /* should never happen */ + + /* If memory, add on the PCI bridge address offset */ + if (mmap_state == pci_mmap_mem) { + res_bit = IORESOURCE_MEM; + } else { + io_offset = (unsigned long)pci_ctrl->io_space.base; + offset += io_offset; + res_bit = IORESOURCE_IO; + } + + /* + * Check that the offset requested corresponds to one of the + * resources of the device. + */ + for (i = 0; i <= PCI_ROM_RESOURCE; i++) { + struct resource *rp = &dev->resource[i]; + int flags = rp->flags; + + /* treat ROM as memory (should be already) */ + if (i == PCI_ROM_RESOURCE) + flags |= IORESOURCE_MEM; + + /* Active and same type? */ + if ((flags & res_bit) == 0) + continue; + + /* In the range of this resource? */ + if (offset < (rp->start & PAGE_MASK) || offset > rp->end) + continue; + + /* found it! construct the final physical address */ + if (mmap_state == pci_mmap_io) + offset += pci_ctrl->io_space.start - io_offset; + vma->vm_pgoff = offset >> PAGE_SHIFT; + return 0; + } + + return -EINVAL; +} + +/* + * Set vm_flags of VMA, as appropriate for this architecture, for a pci device + * mapping. + */ +static __inline__ void +__pci_mmap_set_flags(struct pci_dev *dev, struct vm_area_struct *vma, + enum pci_mmap_state mmap_state) +{ + vma->vm_flags |= VM_SHM | VM_LOCKED | VM_IO; +} + +/* + * Set vm_page_prot of VMA, as appropriate for this architecture, for a pci + * device mapping. + */ +static __inline__ void +__pci_mmap_set_pgprot(struct pci_dev *dev, struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine) +{ + int prot = pgprot_val(vma->vm_page_prot); + + /* Set to write-through */ + prot &= ~_PAGE_NO_CACHE; +#if 0 + if (!write_combine) + prot |= _PAGE_WRITETHRU; +#endif + vma->vm_page_prot = __pgprot(prot); +} + +/* + * Perform the actual remap of the pages for a PCI device mapping, as + * appropriate for this architecture. The region in the process to map + * is described by vm_start and vm_end members of VMA, the base physical + * address is found in vm_pgoff. + * The pci device structure is provided so that architectures may make mapping + * decisions on a per-device or per-bus basis. + * + * Returns a negative error code on failure, zero on success. + */ +int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, + int write_combine) +{ + int ret; + + ret = __pci_mmap_make_offset(dev, vma, mmap_state); + if (ret < 0) + return ret; + + __pci_mmap_set_flags(dev, vma, mmap_state); + __pci_mmap_set_pgprot(dev, vma, mmap_state, write_combine); + + ret = io_remap_page_range(vma, vma->vm_start, vma->vm_pgoff<<PAGE_SHIFT, + vma->vm_end - vma->vm_start, vma->vm_page_prot); + + return ret; +} + +/* + * This probably belongs here rather than ioport.c because + * we do not want this crud linked into SBus kernels. + * Also, think for a moment about likes of floppy.c that + * include architecture specific parts. They may want to redefine ins/outs. + * + * We do not use horroble macroses here because we want to + * advance pointer by sizeof(size). + */ +void outsb(unsigned long addr, const void *src, unsigned long count) { + while (count) { + count -= 1; + writeb(*(const char *)src, addr); + src += 1; + addr += 1; + } +} + +void outsw(unsigned long addr, const void *src, unsigned long count) { + while (count) { + count -= 2; + writew(*(const short *)src, addr); + src += 2; + addr += 2; + } +} + +void outsl(unsigned long addr, const void *src, unsigned long count) { + while (count) { + count -= 4; + writel(*(const long *)src, addr); + src += 4; + addr += 4; + } +} + +void insb(unsigned long addr, void *dst, unsigned long count) { + while (count) { + count -= 1; + *(unsigned char *)dst = readb(addr); + dst += 1; + addr += 1; + } +} + +void insw(unsigned long addr, void *dst, unsigned long count) { + while (count) { + count -= 2; + *(unsigned short *)dst = readw(addr); + dst += 2; + addr += 2; + } +} + +void insl(unsigned long addr, void *dst, unsigned long count) { + while (count) { + count -= 4; + /* + * XXX I am sure we are in for an unaligned trap here. + */ + *(unsigned long *)dst = readl(addr); + dst += 4; + addr += 4; + } +} + + + diff --git a/arch/xtensa/kernel/platform.c b/arch/xtensa/kernel/platform.c new file mode 100644 index 00000000000..cf136278444 --- /dev/null +++ b/arch/xtensa/kernel/platform.c @@ -0,0 +1,49 @@ +/* + * arch/xtensa/kernel/platform.c + * + * Default platform functions. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + */ + +#include <linux/config.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/time.h> +#include <asm/platform.h> +#include <asm/timex.h> + +#define _F(r,f,a,b) \ + r __platform_##f a b; \ + r platform_##f a __attribute__((weak, alias("__platform_"#f))) + +/* + * Default functions that are used if no platform specific function is defined. + * (Please, refer to include/asm-xtensa/platform.h for more information) + */ + +_F(void, setup, (char** cmd), { }); +_F(void, init_irq, (void), { }); +_F(void, restart, (void), { while(1); }); +_F(void, halt, (void), { while(1); }); +_F(void, power_off, (void), { while(1); }); +_F(void, idle, (void), { __asm__ __volatile__ ("waiti 0" ::: "memory"); }); +_F(void, heartbeat, (void), { }); +_F(int, pcibios_fixup, (void), { return 0; }); +_F(int, get_rtc_time, (time_t* t), { return 0; }); +_F(int, set_rtc_time, (time_t t), { return 0; }); + +#if CONFIG_XTENSA_CALIBRATE_CCOUNT +_F(void, calibrate_ccount, (void), +{ + printk ("ERROR: Cannot calibrate cpu frequency! Assuming 100MHz.\n"); + ccount_per_jiffy = 100 * (1000000UL/HZ); +}); +#endif + diff --git a/arch/xtensa/kernel/process.c b/arch/xtensa/kernel/process.c new file mode 100644 index 00000000000..4099703b14b --- /dev/null +++ b/arch/xtensa/kernel/process.c @@ -0,0 +1,482 @@ +// TODO verify coprocessor handling +/* + * arch/xtensa/kernel/process.c + * + * Xtensa Processor version. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Chris Zankel <chris@zankel.net> + * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca> + * Kevin Chea + */ + +#include <linux/config.h> +#include <linux/errno.h> +#include <linux/sched.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/smp.h> +#include <linux/smp_lock.h> +#include <linux/stddef.h> +#include <linux/unistd.h> +#include <linux/ptrace.h> +#include <linux/slab.h> +#include <linux/elf.h> +#include <linux/init.h> +#include <linux/prctl.h> +#include <linux/init_task.h> +#include <linux/module.h> +#include <linux/mqueue.h> + +#include <asm/pgtable.h> +#include <asm/uaccess.h> +#include <asm/system.h> +#include <asm/io.h> +#include <asm/processor.h> +#include <asm/platform.h> +#include <asm/mmu.h> +#include <asm/irq.h> +#include <asm/atomic.h> +#include <asm/offsets.h> +#include <asm/coprocessor.h> + +extern void ret_from_fork(void); + +static struct fs_struct init_fs = INIT_FS; +static struct files_struct init_files = INIT_FILES; +static struct signal_struct init_signals = INIT_SIGNALS(init_signals); +static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand); +struct mm_struct init_mm = INIT_MM(init_mm); +EXPORT_SYMBOL(init_mm); + +union thread_union init_thread_union + __attribute__((__section__(".data.init_task"))) = +{ INIT_THREAD_INFO(init_task) }; + +struct task_struct init_task = INIT_TASK(init_task); +EXPORT_SYMBOL(init_task); + +struct task_struct *current_set[NR_CPUS] = {&init_task, }; + + +#if XCHAL_CP_NUM > 0 + +/* + * Coprocessor ownership. + */ + +coprocessor_info_t coprocessor_info[] = { + { 0, XTENSA_CPE_CP0_OFFSET }, + { 0, XTENSA_CPE_CP1_OFFSET }, + { 0, XTENSA_CPE_CP2_OFFSET }, + { 0, XTENSA_CPE_CP3_OFFSET }, + { 0, XTENSA_CPE_CP4_OFFSET }, + { 0, XTENSA_CPE_CP5_OFFSET }, + { 0, XTENSA_CPE_CP6_OFFSET }, + { 0, XTENSA_CPE_CP7_OFFSET }, +}; + +#endif + +/* + * Powermanagement idle function, if any is provided by the platform. + */ + +void cpu_idle(void) +{ + local_irq_enable(); + + /* endless idle loop with no priority at all */ + while (1) { + while (!need_resched()) + platform_idle(); + preempt_enable(); + schedule(); + } +} + +/* + * Free current thread data structures etc.. + */ + +void exit_thread(void) +{ + release_coprocessors(current); /* Empty macro if no CPs are defined */ +} + +void flush_thread(void) +{ + release_coprocessors(current); /* Empty macro if no CPs are defined */ +} + +/* + * Copy thread. + * + * The stack layout for the new thread looks like this: + * + * +------------------------+ <- sp in childregs (= tos) + * | childregs | + * +------------------------+ <- thread.sp = sp in dummy-frame + * | dummy-frame | (saved in dummy-frame spill-area) + * +------------------------+ + * + * We create a dummy frame to return to ret_from_fork: + * a0 points to ret_from_fork (simulating a call4) + * sp points to itself (thread.sp) + * a2, a3 are unused. + * + * Note: This is a pristine frame, so we don't need any spill region on top of + * childregs. + */ + +int copy_thread(int nr, unsigned long clone_flags, unsigned long usp, + unsigned long unused, + struct task_struct * p, struct pt_regs * regs) +{ + struct pt_regs *childregs; + unsigned long tos; + int user_mode = user_mode(regs); + + /* Set up new TSS. */ + tos = (unsigned long)p->thread_info + THREAD_SIZE; + if (user_mode) + childregs = (struct pt_regs*)(tos - PT_USER_SIZE); + else + childregs = (struct pt_regs*)tos - 1; + + *childregs = *regs; + + /* Create a call4 dummy-frame: a0 = 0, a1 = childregs. */ + *((int*)childregs - 3) = (unsigned long)childregs; + *((int*)childregs - 4) = 0; + + childregs->areg[1] = tos; + childregs->areg[2] = 0; + p->set_child_tid = p->clear_child_tid = NULL; + p->thread.ra = MAKE_RA_FOR_CALL((unsigned long)ret_from_fork, 0x1); + p->thread.sp = (unsigned long)childregs; + if (user_mode(regs)) { + + int len = childregs->wmask & ~0xf; + childregs->areg[1] = usp; + memcpy(&childregs->areg[XCHAL_NUM_AREGS - len/4], + ®s->areg[XCHAL_NUM_AREGS - len/4], len); + + if (clone_flags & CLONE_SETTLS) + childregs->areg[2] = childregs->areg[6]; + + } else { + /* In kernel space, we start a new thread with a new stack. */ + childregs->wmask = 1; + } + return 0; +} + + +/* + * Create a kernel thread + */ + +int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags) +{ + long retval; + __asm__ __volatile__ + ("mov a5, %4\n\t" /* preserve fn in a5 */ + "mov a6, %3\n\t" /* preserve and setup arg in a6 */ + "movi a2, %1\n\t" /* load __NR_clone for syscall*/ + "mov a3, sp\n\t" /* sp check and sys_clone */ + "mov a4, %5\n\t" /* load flags for syscall */ + "syscall\n\t" + "beq a3, sp, 1f\n\t" /* branch if parent */ + "callx4 a5\n\t" /* call fn */ + "movi a2, %2\n\t" /* load __NR_exit for syscall */ + "mov a3, a6\n\t" /* load fn return value */ + "syscall\n" + "1:\n\t" + "mov %0, a2\n\t" /* parent returns zero */ + :"=r" (retval) + :"i" (__NR_clone), "i" (__NR_exit), + "r" (arg), "r" (fn), + "r" (flags | CLONE_VM) + : "a2", "a3", "a4", "a5", "a6" ); + return retval; +} + + +/* + * These bracket the sleeping functions.. + */ + +unsigned long get_wchan(struct task_struct *p) +{ + unsigned long sp, pc; + unsigned long stack_page = (unsigned long) p->thread_info; + int count = 0; + + if (!p || p == current || p->state == TASK_RUNNING) + return 0; + + sp = p->thread.sp; + pc = MAKE_PC_FROM_RA(p->thread.ra, p->thread.sp); + + do { + if (sp < stack_page + sizeof(struct task_struct) || + sp >= (stack_page + THREAD_SIZE) || + pc == 0) + return 0; + if (!in_sched_functions(pc)) + return pc; + + /* Stack layout: sp-4: ra, sp-3: sp' */ + + pc = MAKE_PC_FROM_RA(*(unsigned long*)sp - 4, sp); + sp = *(unsigned long *)sp - 3; + } while (count++ < 16); + return 0; +} + +/* + * do_copy_regs() gathers information from 'struct pt_regs' and + * 'current->thread.areg[]' to fill in the xtensa_gregset_t + * structure. + * + * xtensa_gregset_t and 'struct pt_regs' are vastly different formats + * of processor registers. Besides different ordering, + * xtensa_gregset_t contains non-live register information that + * 'struct pt_regs' does not. Exception handling (primarily) uses + * 'struct pt_regs'. Core files and ptrace use xtensa_gregset_t. + * + */ + +void do_copy_regs (xtensa_gregset_t *elfregs, struct pt_regs *regs, + struct task_struct *tsk) +{ + int i, n, wb_offset; + + elfregs->xchal_config_id0 = XCHAL_HW_CONFIGID0; + elfregs->xchal_config_id1 = XCHAL_HW_CONFIGID1; + + __asm__ __volatile__ ("rsr %0, 176\n" : "=a" (i)); + elfregs->cpux = i; + __asm__ __volatile__ ("rsr %0, 208\n" : "=a" (i)); + elfregs->cpuy = i; + + /* Note: PS.EXCM is not set while user task is running; its + * being set in regs->ps is for exception handling convenience. + */ + + elfregs->pc = regs->pc; + elfregs->ps = (regs->ps & ~XCHAL_PS_EXCM_MASK); + elfregs->exccause = regs->exccause; + elfregs->excvaddr = regs->excvaddr; + elfregs->windowbase = regs->windowbase; + elfregs->windowstart = regs->windowstart; + elfregs->lbeg = regs->lbeg; + elfregs->lend = regs->lend; + elfregs->lcount = regs->lcount; + elfregs->sar = regs->sar; + elfregs->syscall = regs->syscall; + + /* Copy register file. + * The layout looks like this: + * + * | a0 ... a15 | Z ... Z | arX ... arY | + * current window unused saved frames + */ + + memset (elfregs->ar, 0, sizeof(elfregs->ar)); + + wb_offset = regs->windowbase * 4; + n = (regs->wmask&1)? 4 : (regs->wmask&2)? 8 : (regs->wmask&4)? 12 : 16; + + for (i = 0; i < n; i++) + elfregs->ar[(wb_offset + i) % XCHAL_NUM_AREGS] = regs->areg[i]; + + n = (regs->wmask >> 4) * 4; + + for (i = XCHAL_NUM_AREGS - n; n > 0; i++, n--) + elfregs->ar[(wb_offset + i) % XCHAL_NUM_AREGS] = regs->areg[i]; +} + +void xtensa_elf_core_copy_regs (xtensa_gregset_t *elfregs, struct pt_regs *regs) +{ + do_copy_regs ((xtensa_gregset_t *)elfregs, regs, current); +} + + +/* The inverse of do_copy_regs(). No error or sanity checking. */ + +void do_restore_regs (xtensa_gregset_t *elfregs, struct pt_regs *regs, + struct task_struct *tsk) +{ + int i, n, wb_offset; + + /* Note: PS.EXCM is not set while user task is running; it + * needs to be set in regs->ps is for exception handling convenience. + */ + + regs->pc = elfregs->pc; + regs->ps = (elfregs->ps | XCHAL_PS_EXCM_MASK); + regs->exccause = elfregs->exccause; + regs->excvaddr = elfregs->excvaddr; + regs->windowbase = elfregs->windowbase; + regs->windowstart = elfregs->windowstart; + regs->lbeg = elfregs->lbeg; + regs->lend = elfregs->lend; + regs->lcount = elfregs->lcount; + regs->sar = elfregs->sar; + regs->syscall = elfregs->syscall; + + /* Clear everything. */ + + memset (regs->areg, 0, sizeof(regs->areg)); + + /* Copy regs from live window frame. */ + + wb_offset = regs->windowbase * 4; + n = (regs->wmask&1)? 4 : (regs->wmask&2)? 8 : (regs->wmask&4)? 12 : 16; + + for (i = 0; i < n; i++) + regs->areg[(wb_offset+i) % XCHAL_NUM_AREGS] = elfregs->ar[i]; + + n = (regs->wmask >> 4) * 4; + + for (i = XCHAL_NUM_AREGS - n; n > 0; i++, n--) + regs->areg[(wb_offset+i) % XCHAL_NUM_AREGS] = elfregs->ar[i]; +} + +/* + * do_save_fpregs() gathers information from 'struct pt_regs' and + * 'current->thread' to fill in the elf_fpregset_t structure. + * + * Core files and ptrace use elf_fpregset_t. + */ + +void do_save_fpregs (elf_fpregset_t *fpregs, struct pt_regs *regs, + struct task_struct *tsk) +{ +#if XCHAL_HAVE_CP + + extern unsigned char _xtensa_reginfo_tables[]; + extern unsigned _xtensa_reginfo_table_size; + int i; + unsigned long flags; + + /* Before dumping coprocessor state from memory, + * ensure any live coprocessor contents for this + * task are first saved to memory: + */ + local_irq_save(flags); + + for (i = 0; i < XCHAL_CP_MAX; i++) { + if (tsk == coprocessor_info[i].owner) { + enable_coprocessor(i); + save_coprocessor_registers( + tsk->thread.cp_save+coprocessor_info[i].offset,i); + disable_coprocessor(i); + } + } + + local_irq_restore(flags); + + /* Now dump coprocessor & extra state: */ + memcpy((unsigned char*)fpregs, + _xtensa_reginfo_tables, _xtensa_reginfo_table_size); + memcpy((unsigned char*)fpregs + _xtensa_reginfo_table_size, + tsk->thread.cp_save, XTENSA_CP_EXTRA_SIZE); +#endif +} + +/* + * The inverse of do_save_fpregs(). + * Copies coprocessor and extra state from fpregs into regs and tsk->thread. + * Returns 0 on success, non-zero if layout doesn't match. + */ + +int do_restore_fpregs (elf_fpregset_t *fpregs, struct pt_regs *regs, + struct task_struct *tsk) +{ +#if XCHAL_HAVE_CP + + extern unsigned char _xtensa_reginfo_tables[]; + extern unsigned _xtensa_reginfo_table_size; + int i; + unsigned long flags; + + /* Make sure save area layouts match. + * FIXME: in the future we could allow restoring from + * a different layout of the same registers, by comparing + * fpregs' table with _xtensa_reginfo_tables and matching + * entries and copying registers one at a time. + * Not too sure yet whether that's very useful. + */ + + if( memcmp((unsigned char*)fpregs, + _xtensa_reginfo_tables, _xtensa_reginfo_table_size) ) { + return -1; + } + + /* Before restoring coprocessor state from memory, + * ensure any live coprocessor contents for this + * task are first invalidated. + */ + + local_irq_save(flags); + + for (i = 0; i < XCHAL_CP_MAX; i++) { + if (tsk == coprocessor_info[i].owner) { + enable_coprocessor(i); + save_coprocessor_registers( + tsk->thread.cp_save+coprocessor_info[i].offset,i); + coprocessor_info[i].owner = 0; + disable_coprocessor(i); + } + } + + local_irq_restore(flags); + + /* Now restore coprocessor & extra state: */ + + memcpy(tsk->thread.cp_save, + (unsigned char*)fpregs + _xtensa_reginfo_table_size, + XTENSA_CP_EXTRA_SIZE); +#endif + return 0; +} +/* + * Fill in the CP structure for a core dump for a particular task. + */ + +int +dump_task_fpu(struct pt_regs *regs, struct task_struct *task, elf_fpregset_t *r) +{ +/* see asm/coprocessor.h for this magic number 16 */ +#if TOTAL_CPEXTRA_SIZE > 16 + do_save_fpregs (r, regs, task); + + /* For now, bit 16 means some extra state may be present: */ +// FIXME!! need to track to return more accurate mask + return 0x10000 | XCHAL_CP_MASK; +#else + return 0; /* no coprocessors active on this processor */ +#endif +} + +/* + * Fill in the CP structure for a core dump. + * This includes any FPU coprocessor. + * Here, we dump all coprocessors, and other ("extra") custom state. + * + * This function is called by elf_core_dump() in fs/binfmt_elf.c + * (in which case 'regs' comes from calls to do_coredump, see signals.c). + */ +int dump_fpu(struct pt_regs *regs, elf_fpregset_t *r) +{ + return dump_task_fpu(regs, current, r); +} diff --git a/arch/xtensa/kernel/ptrace.c b/arch/xtensa/kernel/ptrace.c new file mode 100644 index 00000000000..9ef07a4dd2a --- /dev/null +++ b/arch/xtensa/kernel/ptrace.c @@ -0,0 +1,407 @@ +// TODO some minor issues +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Chris Zankel <chris@zankel.net> + * Scott Foehner<sfoehner@yahoo.com>, + * Kevin Chea + * Marc Gauthier<marc@tensilica.com> <marc@alumni.uwaterloo.ca> + */ + +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/mm.h> +#include <linux/errno.h> +#include <linux/ptrace.h> +#include <linux/smp.h> +#include <linux/smp_lock.h> +#include <linux/security.h> + +#include <asm/pgtable.h> +#include <asm/page.h> +#include <asm/system.h> +#include <asm/uaccess.h> +#include <asm/ptrace.h> +#include <asm/elf.h> + +#define TEST_KERNEL // verify kernel operations FIXME: remove + + +/* + * Called by kernel/ptrace.c when detaching.. + * + * Make sure single step bits etc are not set. + */ + +void ptrace_disable(struct task_struct *child) +{ + /* Nothing to do.. */ +} + +int sys_ptrace(long request, long pid, long addr, long data) +{ + struct task_struct *child; + int ret = -EPERM; + + lock_kernel(); + +#if 0 + if ((int)request != 1) + printk("ptrace(r=%d,pid=%d,addr=%08lx,data=%08lx)\n", + (int) request, (int) pid, (unsigned long) addr, + (unsigned long) data); +#endif + + if (request == PTRACE_TRACEME) { + + /* Are we already being traced? */ + + if (current->ptrace & PT_PTRACED) + goto out; + + if ((ret = security_ptrace(current->parent, current))) + goto out; + + /* Set the ptrace bit in the process flags. */ + + current->ptrace |= PT_PTRACED; + ret = 0; + goto out; + } + + ret = -ESRCH; + read_lock(&tasklist_lock); + child = find_task_by_pid(pid); + if (child) + get_task_struct(child); + read_unlock(&tasklist_lock); + if (!child) + goto out; + + ret = -EPERM; + if (pid == 1) /* you may not mess with init */ + goto out; + + if (request == PTRACE_ATTACH) { + ret = ptrace_attach(child); + goto out_tsk; + } + + if ((ret = ptrace_check_attach(child, request == PTRACE_KILL)) < 0) + goto out_tsk; + + switch (request) { + case PTRACE_PEEKTEXT: /* read word at location addr. */ + case PTRACE_PEEKDATA: + { + unsigned long tmp; + int copied; + + copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 0); + ret = -EIO; + if (copied != sizeof(tmp)) + break; + ret = put_user(tmp,(unsigned long *) data); + + goto out; + } + + /* Read the word at location addr in the USER area. */ + + case PTRACE_PEEKUSR: + { + struct pt_regs *regs; + unsigned long tmp; + + regs = xtensa_pt_regs(child); + tmp = 0; /* Default return value. */ + + switch(addr) { + + case REG_AR_BASE ... REG_AR_BASE + XCHAL_NUM_AREGS - 1: + { + int ar = addr - REG_AR_BASE - regs->windowbase * 4; + ar &= (XCHAL_NUM_AREGS - 1); + if (ar < 16 && ar + (regs->wmask >> 4) * 4 >= 0) + tmp = regs->areg[ar]; + else + ret = -EIO; + break; + } + case REG_A_BASE ... REG_A_BASE + 15: + tmp = regs->areg[addr - REG_A_BASE]; + break; + case REG_PC: + tmp = regs->pc; + break; + case REG_PS: + /* Note: PS.EXCM is not set while user task is running; + * its being set in regs is for exception handling + * convenience. */ + tmp = (regs->ps & ~XCHAL_PS_EXCM_MASK); + break; + case REG_WB: + tmp = regs->windowbase; + break; + case REG_WS: + tmp = regs->windowstart; + break; + case REG_LBEG: + tmp = regs->lbeg; + break; + case REG_LEND: + tmp = regs->lend; + break; + case REG_LCOUNT: + tmp = regs->lcount; + break; + case REG_SAR: + tmp = regs->sar; + break; + case REG_DEPC: + tmp = regs->depc; + break; + case REG_EXCCAUSE: + tmp = regs->exccause; + break; + case REG_EXCVADDR: + tmp = regs->excvaddr; + break; + case SYSCALL_NR: + tmp = regs->syscall; + break; + default: + tmp = 0; + ret = -EIO; + goto out; + } + ret = put_user(tmp, (unsigned long *) data); + goto out; + } + + case PTRACE_POKETEXT: /* write the word at location addr. */ + case PTRACE_POKEDATA: + if (access_process_vm(child, addr, &data, sizeof(data), 1) + == sizeof(data)) + break; + ret = -EIO; + goto out; + + case PTRACE_POKEUSR: + { + struct pt_regs *regs; + regs = xtensa_pt_regs(child); + + switch (addr) { + case REG_AR_BASE ... REG_AR_BASE + XCHAL_NUM_AREGS - 1: + { + int ar = addr - REG_AR_BASE - regs->windowbase * 4; + if (ar < 16 && ar + (regs->wmask >> 4) * 4 >= 0) + regs->areg[ar & (XCHAL_NUM_AREGS - 1)] = data; + else + ret = -EIO; + break; + } + case REG_A_BASE ... REG_A_BASE + 15: + regs->areg[addr - REG_A_BASE] = data; + break; + case REG_PC: + regs->pc = data; + break; + case SYSCALL_NR: + regs->syscall = data; + break; +#ifdef TEST_KERNEL + case REG_WB: + regs->windowbase = data; + break; + case REG_WS: + regs->windowstart = data; + break; +#endif + + default: + /* The rest are not allowed. */ + ret = -EIO; + break; + } + break; + } + + /* continue and stop at next (return from) syscall */ + case PTRACE_SYSCALL: + case PTRACE_CONT: /* restart after signal. */ + { + ret = -EIO; + if ((unsigned long) data > _NSIG) + break; + if (request == PTRACE_SYSCALL) + set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); + else + clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); + child->exit_code = data; + /* Make sure the single step bit is not set. */ + child->ptrace &= ~PT_SINGLESTEP; + wake_up_process(child); + ret = 0; + break; + } + + /* + * make the child exit. Best I can do is send it a sigkill. + * perhaps it should be put in the status that it wants to + * exit. + */ + case PTRACE_KILL: + ret = 0; + if (child->state == EXIT_ZOMBIE) /* already dead */ + break; + child->exit_code = SIGKILL; + child->ptrace &= ~PT_SINGLESTEP; + wake_up_process(child); + break; + + case PTRACE_SINGLESTEP: + ret = -EIO; + if ((unsigned long) data > _NSIG) + break; + clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); + child->ptrace |= PT_SINGLESTEP; + child->exit_code = data; + wake_up_process(child); + ret = 0; + break; + + case PTRACE_GETREGS: + { + /* 'data' points to user memory in which to write. + * Mainly due to the non-live register values, we + * reformat the register values into something more + * standard. For convenience, we use the handy + * elf_gregset_t format. */ + + xtensa_gregset_t format; + struct pt_regs *regs = xtensa_pt_regs(child); + + do_copy_regs (&format, regs, child); + + /* Now, copy to user space nice and easy... */ + ret = 0; + if (copy_to_user((void *)data, &format, sizeof(elf_gregset_t))) + ret = -EFAULT; + break; + } + + case PTRACE_SETREGS: + { + /* 'data' points to user memory that contains the new + * values in the elf_gregset_t format. */ + + xtensa_gregset_t format; + struct pt_regs *regs = xtensa_pt_regs(child); + + if (copy_from_user(&format,(void *)data,sizeof(elf_gregset_t))){ + ret = -EFAULT; + break; + } + + /* FIXME: Perhaps we want some sanity checks on + * these user-space values? See ARM version. Are + * debuggers a security concern? */ + + do_restore_regs (&format, regs, child); + + ret = 0; + break; + } + + case PTRACE_GETFPREGS: + { + /* 'data' points to user memory in which to write. + * For convenience, we use the handy + * elf_fpregset_t format. */ + + elf_fpregset_t fpregs; + struct pt_regs *regs = xtensa_pt_regs(child); + + do_save_fpregs (&fpregs, regs, child); + + /* Now, copy to user space nice and easy... */ + ret = 0; + if (copy_to_user((void *)data, &fpregs, sizeof(elf_fpregset_t))) + ret = -EFAULT; + + break; + } + + case PTRACE_SETFPREGS: + { + /* 'data' points to user memory that contains the new + * values in the elf_fpregset_t format. + */ + elf_fpregset_t fpregs; + struct pt_regs *regs = xtensa_pt_regs(child); + + ret = 0; + if (copy_from_user(&fpregs, (void *)data, sizeof(elf_fpregset_t))) { + ret = -EFAULT; + break; + } + + if (do_restore_fpregs (&fpregs, regs, child)) + ret = -EIO; + break; + } + + case PTRACE_GETFPREGSIZE: + /* 'data' points to 'unsigned long' set to the size + * of elf_fpregset_t + */ + ret = put_user(sizeof(elf_fpregset_t), (unsigned long *) data); + break; + + case PTRACE_DETACH: /* detach a process that was attached. */ + ret = ptrace_detach(child, data); + break; + + default: + ret = ptrace_request(child, request, addr, data); + goto out; + } +out_tsk: + put_task_struct(child); +out: + unlock_kernel(); + return ret; +} + +void do_syscall_trace(void) +{ + if (!test_thread_flag(TIF_SYSCALL_TRACE)) + return; + + if (!(current->ptrace & PT_PTRACED)) + return; + + /* + * The 0x80 provides a way for the tracing parent to distinguish + * between a syscall stop and SIGTRAP delivery + */ + ptrace_notify(SIGTRAP|((current->ptrace & PT_TRACESYSGOOD) ? 0x80 : 0)); + + /* + * this isn't the same as continuing with a signal, but it will do + * for normal use. strace only continues with a signal if the + * stopping signal is not SIGTRAP. -brl + */ + if (current->exit_code) { + send_sig(current->exit_code, current, 1); + current->exit_code = 0; + } +} diff --git a/arch/xtensa/kernel/semaphore.c b/arch/xtensa/kernel/semaphore.c new file mode 100644 index 00000000000..d40f4b1b75a --- /dev/null +++ b/arch/xtensa/kernel/semaphore.c @@ -0,0 +1,226 @@ +/* + * arch/xtensa/kernel/semaphore.c + * + * Generic semaphore code. Buyer beware. Do your own specific changes + * in <asm/semaphore-helper.h> + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Chris Zankel <chris@zankel.net> + * Marc Gauthier<marc@tensilica.com, marc@alumni.uwaterloo.ca> + * Kevin Chea + */ + +#include <linux/sched.h> +#include <linux/wait.h> +#include <linux/init.h> +#include <asm/semaphore.h> +#include <asm/errno.h> + +/* + * These two _must_ execute atomically wrt each other. + */ + +static __inline__ void wake_one_more(struct semaphore * sem) +{ + atomic_inc((atomic_t *)&sem->sleepers); +} + +static __inline__ int waking_non_zero(struct semaphore *sem) +{ + unsigned long flags; + int ret = 0; + + spin_lock_irqsave(&semaphore_wake_lock, flags); + if (sem->sleepers > 0) { + sem->sleepers--; + ret = 1; + } + spin_unlock_irqrestore(&semaphore_wake_lock, flags); + return ret; +} + +/* + * waking_non_zero_interruptible: + * 1 got the lock + * 0 go to sleep + * -EINTR interrupted + * + * We must undo the sem->count down_interruptible() increment while we are + * protected by the spinlock in order to make atomic this atomic_inc() with the + * atomic_read() in wake_one_more(), otherwise we can race. -arca + */ + +static __inline__ int waking_non_zero_interruptible(struct semaphore *sem, + struct task_struct *tsk) +{ + unsigned long flags; + int ret = 0; + + spin_lock_irqsave(&semaphore_wake_lock, flags); + if (sem->sleepers > 0) { + sem->sleepers--; + ret = 1; + } else if (signal_pending(tsk)) { + atomic_inc(&sem->count); + ret = -EINTR; + } + spin_unlock_irqrestore(&semaphore_wake_lock, flags); + return ret; +} + +/* + * waking_non_zero_trylock: + * 1 failed to lock + * 0 got the lock + * + * We must undo the sem->count down_trylock() increment while we are + * protected by the spinlock in order to make atomic this atomic_inc() with the + * atomic_read() in wake_one_more(), otherwise we can race. -arca + */ + +static __inline__ int waking_non_zero_trylock(struct semaphore *sem) +{ + unsigned long flags; + int ret = 1; + + spin_lock_irqsave(&semaphore_wake_lock, flags); + if (sem->sleepers <= 0) + atomic_inc(&sem->count); + else { + sem->sleepers--; + ret = 0; + } + spin_unlock_irqrestore(&semaphore_wake_lock, flags); + return ret; +} + +spinlock_t semaphore_wake_lock; + +/* + * Semaphores are implemented using a two-way counter: + * The "count" variable is decremented for each process + * that tries to sleep, while the "waking" variable is + * incremented when the "up()" code goes to wake up waiting + * processes. + * + * Notably, the inline "up()" and "down()" functions can + * efficiently test if they need to do any extra work (up + * needs to do something only if count was negative before + * the increment operation. + * + * waking_non_zero() (from asm/semaphore.h) must execute + * atomically. + * + * When __up() is called, the count was negative before + * incrementing it, and we need to wake up somebody. + * + * This routine adds one to the count of processes that need to + * wake up and exit. ALL waiting processes actually wake up but + * only the one that gets to the "waking" field first will gate + * through and acquire the semaphore. The others will go back + * to sleep. + * + * Note that these functions are only called when there is + * contention on the lock, and as such all this is the + * "non-critical" part of the whole semaphore business. The + * critical part is the inline stuff in <asm/semaphore.h> + * where we want to avoid any extra jumps and calls. + */ + +void __up(struct semaphore *sem) +{ + wake_one_more(sem); + wake_up(&sem->wait); +} + +/* + * Perform the "down" function. Return zero for semaphore acquired, + * return negative for signalled out of the function. + * + * If called from __down, the return is ignored and the wait loop is + * not interruptible. This means that a task waiting on a semaphore + * using "down()" cannot be killed until someone does an "up()" on + * the semaphore. + * + * If called from __down_interruptible, the return value gets checked + * upon return. If the return value is negative then the task continues + * with the negative value in the return register (it can be tested by + * the caller). + * + * Either form may be used in conjunction with "up()". + * + */ + +#define DOWN_VAR \ + struct task_struct *tsk = current; \ + wait_queue_t wait; \ + init_waitqueue_entry(&wait, tsk); + +#define DOWN_HEAD(task_state) \ + \ + \ + tsk->state = (task_state); \ + add_wait_queue(&sem->wait, &wait); \ + \ + /* \ + * Ok, we're set up. sem->count is known to be less than zero \ + * so we must wait. \ + * \ + * We can let go the lock for purposes of waiting. \ + * We re-acquire it after awaking so as to protect \ + * all semaphore operations. \ + * \ + * If "up()" is called before we call waking_non_zero() then \ + * we will catch it right away. If it is called later then \ + * we will have to go through a wakeup cycle to catch it. \ + * \ + * Multiple waiters contend for the semaphore lock to see \ + * who gets to gate through and who has to wait some more. \ + */ \ + for (;;) { + +#define DOWN_TAIL(task_state) \ + tsk->state = (task_state); \ + } \ + tsk->state = TASK_RUNNING; \ + remove_wait_queue(&sem->wait, &wait); + +void __sched __down(struct semaphore * sem) +{ + DOWN_VAR + DOWN_HEAD(TASK_UNINTERRUPTIBLE) + if (waking_non_zero(sem)) + break; + schedule(); + DOWN_TAIL(TASK_UNINTERRUPTIBLE) +} + +int __sched __down_interruptible(struct semaphore * sem) +{ + int ret = 0; + DOWN_VAR + DOWN_HEAD(TASK_INTERRUPTIBLE) + + ret = waking_non_zero_interruptible(sem, tsk); + if (ret) + { + if (ret == 1) + /* ret != 0 only if we get interrupted -arca */ + ret = 0; + break; + } + schedule(); + DOWN_TAIL(TASK_INTERRUPTIBLE) + return ret; +} + +int __down_trylock(struct semaphore * sem) +{ + return waking_non_zero_trylock(sem); +} diff --git a/arch/xtensa/kernel/setup.c b/arch/xtensa/kernel/setup.c new file mode 100644 index 00000000000..1f5bf5d624e --- /dev/null +++ b/arch/xtensa/kernel/setup.c @@ -0,0 +1,520 @@ +/* + * arch/xtensa/setup.c + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 1995 Linus Torvalds + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Kevin Chea + * Marc Gauthier<marc@tensilica.com> <marc@alumni.uwaterloo.ca> + */ + +#include <linux/config.h> +#include <linux/errno.h> +#include <linux/init.h> +#include <linux/proc_fs.h> +#include <linux/tty.h> +#include <linux/bootmem.h> +#include <linux/kernel.h> + +#if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_DUMMY_CONSOLE) +# include <linux/console.h> +#endif + +#ifdef CONFIG_RTC +# include <linux/timex.h> +#endif + +#ifdef CONFIG_PROC_FS +# include <linux/seq_file.h> +#endif + +#include <asm/system.h> +#include <asm/bootparam.h> +#include <asm/pgtable.h> +#include <asm/processor.h> +#include <asm/timex.h> +#include <asm/platform.h> +#include <asm/page.h> +#include <asm/setup.h> + +#include <xtensa/config/system.h> + +#if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_DUMMY_CONSOLE) +struct screen_info screen_info = { 0, 24, 0, 0, 0, 80, 0, 0, 0, 24, 1, 16}; +#endif + +#ifdef CONFIG_BLK_DEV_FD +extern struct fd_ops no_fd_ops; +struct fd_ops *fd_ops; +#endif + +#if defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_IDE_MODULE) +extern struct ide_ops no_ide_ops; +struct ide_ops *ide_ops; +#endif + +extern struct rtc_ops no_rtc_ops; +struct rtc_ops *rtc_ops; + +#ifdef CONFIG_PC_KEYB +extern struct kbd_ops no_kbd_ops; +struct kbd_ops *kbd_ops; +#endif + +#ifdef CONFIG_BLK_DEV_INITRD +extern void *initrd_start; +extern void *initrd_end; +extern void *__initrd_start; +extern void *__initrd_end; +int initrd_is_mapped = 0; +extern int initrd_below_start_ok; +#endif + +unsigned char aux_device_present; +extern unsigned long loops_per_jiffy; + +/* Command line specified as configuration option. */ + +static char command_line[COMMAND_LINE_SIZE]; + +#ifdef CONFIG_CMDLINE_BOOL +static char default_command_line[COMMAND_LINE_SIZE] __initdata = CONFIG_CMDLINE; +#endif + +sysmem_info_t __initdata sysmem; + +#ifdef CONFIG_BLK_DEV_INITRD +int initrd_is_mapped; +#endif + +extern void init_mmu(void); + +/* + * Boot parameter parsing. + * + * The Xtensa port uses a list of variable-sized tags to pass data to + * the kernel. The first tag must be a BP_TAG_FIRST tag for the list + * to be recognised. The list is terminated with a zero-sized + * BP_TAG_LAST tag. + */ + +typedef struct tagtable { + u32 tag; + int (*parse)(const bp_tag_t*); +} tagtable_t; + +#define __tagtable(tag, fn) static tagtable_t __tagtable_##fn \ + __attribute__((unused, __section__(".taglist"))) = { tag, fn } + +/* parse current tag */ + +static int __init parse_tag_mem(const bp_tag_t *tag) +{ + meminfo_t *mi = (meminfo_t*)(tag->data); + + if (mi->type != MEMORY_TYPE_CONVENTIONAL) + return -1; + + if (sysmem.nr_banks >= SYSMEM_BANKS_MAX) { + printk(KERN_WARNING + "Ignoring memory bank 0x%08lx size %ldKB\n", + (unsigned long)mi->start, + (unsigned long)mi->end - (unsigned long)mi->start); + return -EINVAL; + } + sysmem.bank[sysmem.nr_banks].type = mi->type; + sysmem.bank[sysmem.nr_banks].start = PAGE_ALIGN(mi->start); + sysmem.bank[sysmem.nr_banks].end = mi->end & PAGE_SIZE; + sysmem.nr_banks++; + + return 0; +} + +__tagtable(BP_TAG_MEMORY, parse_tag_mem); + +#ifdef CONFIG_BLK_DEV_INITRD + +static int __init parse_tag_initrd(const bp_tag_t* tag) +{ + meminfo_t* mi; + mi = (meminfo_t*)(tag->data); + initrd_start = (void*)(mi->start); + initrd_end = (void*)(mi->end); + + return 0; +} + +__tagtable(BP_TAG_INITRD, parse_tag_initrd); + +#endif /* CONFIG_BLK_DEV_INITRD */ + +static int __init parse_tag_cmdline(const bp_tag_t* tag) +{ + strncpy(command_line, (char*)(tag->data), COMMAND_LINE_SIZE); + command_line[COMMAND_LINE_SIZE - 1] = '\0'; + return 0; +} + +__tagtable(BP_TAG_COMMAND_LINE, parse_tag_cmdline); + +static int __init parse_bootparam(const bp_tag_t* tag) +{ + extern tagtable_t __tagtable_begin, __tagtable_end; + tagtable_t *t; + + /* Boot parameters must start with a BP_TAG_FIRST tag. */ + + if (tag->id != BP_TAG_FIRST) { + printk(KERN_WARNING "Invalid boot parameters!\n"); + return 0; + } + + tag = (bp_tag_t*)((unsigned long)tag + sizeof(bp_tag_t) + tag->size); + + /* Parse all tags. */ + + while (tag != NULL && tag->id != BP_TAG_LAST) { + for (t = &__tagtable_begin; t < &__tagtable_end; t++) { + if (tag->id == t->tag) { + t->parse(tag); + break; + } + } + if (t == &__tagtable_end) + printk(KERN_WARNING "Ignoring tag " + "0x%08x\n", tag->id); + tag = (bp_tag_t*)((unsigned long)(tag + 1) + tag->size); + } + + return 0; +} + +/* + * Initialize architecture. (Early stage) + */ + +void __init init_arch(bp_tag_t *bp_start) +{ + +#ifdef CONFIG_BLK_DEV_INITRD + initrd_start = &__initrd_start; + initrd_end = &__initrd_end; +#endif + + sysmem.nr_banks = 0; + +#ifdef CONFIG_CMDLINE_BOOL + strcpy(command_line, default_command_line); +#endif + + /* Parse boot parameters */ + + if (bp_start) + parse_bootparam(bp_start); + + if (sysmem.nr_banks == 0) { + sysmem.nr_banks = 1; + sysmem.bank[0].start = PLATFORM_DEFAULT_MEM_START; + sysmem.bank[0].end = PLATFORM_DEFAULT_MEM_START + + PLATFORM_DEFAULT_MEM_SIZE; + } + + /* Early hook for platforms */ + + platform_init(bp_start); + + /* Initialize MMU. */ + + init_mmu(); +} + +/* + * Initialize system. Setup memory and reserve regions. + */ + +extern char _end; +extern char _stext; +extern char _WindowVectors_text_start; +extern char _WindowVectors_text_end; +extern char _DebugInterruptVector_literal_start; +extern char _DebugInterruptVector_text_end; +extern char _KernelExceptionVector_literal_start; +extern char _KernelExceptionVector_text_end; +extern char _UserExceptionVector_literal_start; +extern char _UserExceptionVector_text_end; +extern char _DoubleExceptionVector_literal_start; +extern char _DoubleExceptionVector_text_end; + +void __init setup_arch(char **cmdline_p) +{ + extern int mem_reserve(unsigned long, unsigned long, int); + extern void bootmem_init(void); + + memcpy(saved_command_line, command_line, COMMAND_LINE_SIZE); + saved_command_line[COMMAND_LINE_SIZE-1] = '\0'; + *cmdline_p = command_line; + + /* Reserve some memory regions */ + +#ifdef CONFIG_BLK_DEV_INITRD + if (initrd_start < initrd_end) { + initrd_is_mapped = mem_reserve(__pa(initrd_start), + __pa(initrd_end), 0); + initrd_below_start_ok = 1; + } else { + initrd_start = 0; + } +#endif + + mem_reserve(__pa(&_stext),__pa(&_end), 1); + + mem_reserve(__pa(&_WindowVectors_text_start), + __pa(&_WindowVectors_text_end), 0); + + mem_reserve(__pa(&_DebugInterruptVector_literal_start), + __pa(&_DebugInterruptVector_text_end), 0); + + mem_reserve(__pa(&_KernelExceptionVector_literal_start), + __pa(&_KernelExceptionVector_text_end), 0); + + mem_reserve(__pa(&_UserExceptionVector_literal_start), + __pa(&_UserExceptionVector_text_end), 0); + + mem_reserve(__pa(&_DoubleExceptionVector_literal_start), + __pa(&_DoubleExceptionVector_text_end), 0); + + bootmem_init(); + + platform_setup(cmdline_p); + + + paging_init(); + +#ifdef CONFIG_VT +# if defined(CONFIG_VGA_CONSOLE) + conswitchp = &vga_con; +# elif defined(CONFIG_DUMMY_CONSOLE) + conswitchp = &dummy_con; +# endif +#endif + +#if CONFIG_PCI + platform_pcibios_init(); +#endif +} + +void machine_restart(char * cmd) +{ + platform_restart(); +} + +void machine_halt(void) +{ + platform_halt(); + while (1); +} + +void machine_power_off(void) +{ + platform_power_off(); + while (1); +} +#ifdef CONFIG_PROC_FS + +/* + * Display some core information through /proc/cpuinfo. + */ + +static int +c_show(struct seq_file *f, void *slot) +{ + /* high-level stuff */ + seq_printf(f,"processor\t: 0\n" + "vendor_id\t: Tensilica\n" + "model\t\t: Xtensa " XCHAL_HW_RELEASE_NAME "\n" + "core ID\t\t: " XCHAL_CORE_ID "\n" + "build ID\t: 0x%x\n" + "byte order\t: %s\n" + "cpu MHz\t\t: %lu.%02lu\n" + "bogomips\t: %lu.%02lu\n", + XCHAL_BUILD_UNIQUE_ID, + XCHAL_HAVE_BE ? "big" : "little", + CCOUNT_PER_JIFFY/(1000000/HZ), + (CCOUNT_PER_JIFFY/(10000/HZ)) % 100, + loops_per_jiffy/(500000/HZ), + (loops_per_jiffy/(5000/HZ)) % 100); + + seq_printf(f,"flags\t\t: " +#if XCHAL_HAVE_NMI + "nmi " +#endif +#if XCHAL_HAVE_DEBUG + "debug " +# if XCHAL_HAVE_OCD + "ocd " +# endif +#endif +#if XCHAL_HAVE_DENSITY + "density " +#endif +#if XCHAL_HAVE_BOOLEANS + "boolean " +#endif +#if XCHAL_HAVE_LOOPS + "loop " +#endif +#if XCHAL_HAVE_NSA + "nsa " +#endif +#if XCHAL_HAVE_MINMAX + "minmax " +#endif +#if XCHAL_HAVE_SEXT + "sext " +#endif +#if XCHAL_HAVE_CLAMPS + "clamps " +#endif +#if XCHAL_HAVE_MAC16 + "mac16 " +#endif +#if XCHAL_HAVE_MUL16 + "mul16 " +#endif +#if XCHAL_HAVE_MUL32 + "mul32 " +#endif +#if XCHAL_HAVE_MUL32_HIGH + "mul32h " +#endif +#if XCHAL_HAVE_FP + "fpu " +#endif + "\n"); + + /* Registers. */ + seq_printf(f,"physical aregs\t: %d\n" + "misc regs\t: %d\n" + "ibreak\t\t: %d\n" + "dbreak\t\t: %d\n", + XCHAL_NUM_AREGS, + XCHAL_NUM_MISC_REGS, + XCHAL_NUM_IBREAK, + XCHAL_NUM_DBREAK); + + + /* Interrupt. */ + seq_printf(f,"num ints\t: %d\n" + "ext ints\t: %d\n" + "int levels\t: %d\n" + "timers\t\t: %d\n" + "debug level\t: %d\n", + XCHAL_NUM_INTERRUPTS, + XCHAL_NUM_EXTINTERRUPTS, + XCHAL_NUM_INTLEVELS, + XCHAL_NUM_TIMERS, + XCHAL_DEBUGLEVEL); + + /* Coprocessors */ +#if XCHAL_HAVE_CP + seq_printf(f, "coprocessors\t: %d\n", XCHAL_CP_NUM); +#else + seq_printf(f, "coprocessors\t: none\n"); +#endif + + /* {I,D}{RAM,ROM} and XLMI */ + seq_printf(f,"inst ROMs\t: %d\n" + "inst RAMs\t: %d\n" + "data ROMs\t: %d\n" + "data RAMs\t: %d\n" + "XLMI ports\t: %d\n", + XCHAL_NUM_IROM, + XCHAL_NUM_IRAM, + XCHAL_NUM_DROM, + XCHAL_NUM_DRAM, + XCHAL_NUM_XLMI); + + /* Cache */ + seq_printf(f,"icache line size: %d\n" + "icache ways\t: %d\n" + "icache size\t: %d\n" + "icache flags\t: " +#if XCHAL_ICACHE_LINE_LOCKABLE + "lock" +#endif + "\n" + "dcache line size: %d\n" + "dcache ways\t: %d\n" + "dcache size\t: %d\n" + "dcache flags\t: " +#if XCHAL_DCACHE_IS_WRITEBACK + "writeback" +#endif +#if XCHAL_DCACHE_LINE_LOCKABLE + "lock" +#endif + "\n", + XCHAL_ICACHE_LINESIZE, + XCHAL_ICACHE_WAYS, + XCHAL_ICACHE_SIZE, + XCHAL_DCACHE_LINESIZE, + XCHAL_DCACHE_WAYS, + XCHAL_DCACHE_SIZE); + + /* MMU */ + seq_printf(f,"ASID bits\t: %d\n" + "ASID invalid\t: %d\n" + "ASID kernel\t: %d\n" + "rings\t\t: %d\n" + "itlb ways\t: %d\n" + "itlb AR ways\t: %d\n" + "dtlb ways\t: %d\n" + "dtlb AR ways\t: %d\n", + XCHAL_MMU_ASID_BITS, + XCHAL_MMU_ASID_INVALID, + XCHAL_MMU_ASID_KERNEL, + XCHAL_MMU_RINGS, + XCHAL_ITLB_WAYS, + XCHAL_ITLB_ARF_WAYS, + XCHAL_DTLB_WAYS, + XCHAL_DTLB_ARF_WAYS); + + return 0; +} + +/* + * We show only CPU #0 info. + */ +static void * +c_start(struct seq_file *f, loff_t *pos) +{ + return (void *) ((*pos == 0) ? (void *)1 : NULL); +} + +static void * +c_next(struct seq_file *f, void *v, loff_t *pos) +{ + return NULL; +} + +static void +c_stop(struct seq_file *f, void *v) +{ +} + +struct seq_operations cpuinfo_op = +{ + start: c_start, + next: c_next, + stop: c_stop, + show: c_show +}; + +#endif /* CONFIG_PROC_FS */ + diff --git a/arch/xtensa/kernel/signal.c b/arch/xtensa/kernel/signal.c new file mode 100644 index 00000000000..df6e1e17b09 --- /dev/null +++ b/arch/xtensa/kernel/signal.c @@ -0,0 +1,713 @@ +// TODO coprocessor stuff +/* + * linux/arch/xtensa/kernel/signal.c + * + * Copyright (C) 1991, 1992 Linus Torvalds + * 1997-11-28 Modified for POSIX.1b signals by Richard Henderson + * + * Joe Taylor <joe@tensilica.com> + * Chris Zankel <chris@zankel.net> + * + * + * + */ + +#include <xtensa/config/core.h> +#include <xtensa/hal.h> +#include <linux/sched.h> +#include <linux/mm.h> +#include <linux/smp.h> +#include <linux/smp_lock.h> +#include <linux/kernel.h> +#include <linux/signal.h> +#include <linux/errno.h> +#include <linux/wait.h> +#include <linux/ptrace.h> +#include <linux/unistd.h> +#include <linux/stddef.h> +#include <linux/personality.h> +#include <asm/ucontext.h> +#include <asm/uaccess.h> +#include <asm/pgtable.h> +#include <asm/cacheflush.h> + +#define DEBUG_SIG 0 + +#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) + +asmlinkage long sys_wait4(pid_t pid,unsigned int * stat_addr, int options, + struct rusage * ru); +asmlinkage int do_signal(struct pt_regs *regs, sigset_t *oldset); + +extern struct task_struct *coproc_owners[]; + + +/* + * Atomically swap in the new signal mask, and wait for a signal. + */ + +int sys_sigsuspend(struct pt_regs *regs) +{ + old_sigset_t mask = (old_sigset_t) regs->areg[3]; + sigset_t saveset; + + mask &= _BLOCKABLE; + spin_lock_irq(¤t->sighand->siglock); + saveset = current->blocked; + siginitset(¤t->blocked, mask); + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + regs->areg[2] = -EINTR; + while (1) { + current->state = TASK_INTERRUPTIBLE; + schedule(); + if (do_signal(regs, &saveset)) + return -EINTR; + } +} + +asmlinkage int +sys_rt_sigsuspend(struct pt_regs *regs) +{ + sigset_t *unewset = (sigset_t *) regs->areg[4]; + size_t sigsetsize = (size_t) regs->areg[3]; + sigset_t saveset, newset; + /* XXX: Don't preclude handling different sized sigset_t's. */ + if (sigsetsize != sizeof(sigset_t)) + return -EINVAL; + + if (copy_from_user(&newset, unewset, sizeof(newset))) + return -EFAULT; + sigdelsetmask(&newset, ~_BLOCKABLE); + spin_lock_irq(¤t->sighand->siglock); + saveset = current->blocked; + current->blocked = newset; + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + regs->areg[2] = -EINTR; + while (1) { + current->state = TASK_INTERRUPTIBLE; + schedule(); + if (do_signal(regs, &saveset)) + return -EINTR; + } +} + +asmlinkage int +sys_sigaction(int sig, const struct old_sigaction *act, + struct old_sigaction *oact) +{ + struct k_sigaction new_ka, old_ka; + int ret; + + if (act) { + old_sigset_t mask; + if (verify_area(VERIFY_READ, act, sizeof(*act)) || + __get_user(new_ka.sa.sa_handler, &act->sa_handler) || + __get_user(new_ka.sa.sa_restorer, &act->sa_restorer)) + return -EFAULT; + __get_user(new_ka.sa.sa_flags, &act->sa_flags); + __get_user(mask, &act->sa_mask); + siginitset(&new_ka.sa.sa_mask, mask); + } + + ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL); + + if (!ret && oact) { + if (verify_area(VERIFY_WRITE, oact, sizeof(*oact)) || + __put_user(old_ka.sa.sa_handler, &oact->sa_handler) || + __put_user(old_ka.sa.sa_restorer, &oact->sa_restorer)) + return -EFAULT; + __put_user(old_ka.sa.sa_flags, &oact->sa_flags); + __put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask); + } + + return ret; +} + +asmlinkage int +sys_sigaltstack(struct pt_regs *regs) +{ + const stack_t *uss = (stack_t *) regs->areg[4]; + stack_t *uoss = (stack_t *) regs->areg[3]; + + if (regs->depc > 64) + panic ("Double exception sys_sigreturn\n"); + + + return do_sigaltstack(uss, uoss, regs->areg[1]); +} + + +/* + * Do a signal return; undo the signal stack. + */ + +struct sigframe +{ + struct sigcontext sc; + struct _cpstate cpstate; + unsigned long extramask[_NSIG_WORDS-1]; + unsigned char retcode[6]; + unsigned int reserved[4]; /* Reserved area for chaining */ + unsigned int window[4]; /* Window of 4 registers for initial context */ +}; + +struct rt_sigframe +{ + struct siginfo info; + struct ucontext uc; + struct _cpstate cpstate; + unsigned char retcode[6]; + unsigned int reserved[4]; /* Reserved area for chaining */ + unsigned int window[4]; /* Window of 4 registers for initial context */ +}; + +extern void release_all_cp (struct task_struct *); + + +// FIXME restore_cpextra +static inline int +restore_cpextra (struct _cpstate *buf) +{ +#if 0 + /* The signal handler may have used coprocessors in which + * case they are still enabled. We disable them to force a + * reloading of the original task's CP state by the lazy + * context-switching mechanisms of CP exception handling. + * Also, we essentially discard any coprocessor state that the + * signal handler created. */ + + struct task_struct *tsk = current; + release_all_cp(tsk); + return __copy_from_user(tsk->thread.cpextra, buf, TOTAL_CPEXTRA_SIZE); +#endif + return 0; +} + +/* Note: We don't copy double exception 'tregs', we have to finish double exc. first before we return to signal handler! This dbl.exc.handler might cause another double exception, but I think we are fine as the situation is the same as if we had returned to the signal handerl and got an interrupt immediately... + */ + + +static int +restore_sigcontext(struct pt_regs *regs, struct sigcontext *sc) +{ + struct thread_struct *thread; + unsigned int err = 0; + unsigned long ps; + struct _cpstate *buf; + +#define COPY(x) err |= __get_user(regs->x, &sc->sc_##x) + COPY(pc); + COPY(depc); + COPY(wmask); + COPY(lbeg); + COPY(lend); + COPY(lcount); + COPY(sar); + COPY(windowbase); + COPY(windowstart); +#undef COPY + + /* For PS, restore only PS.CALLINC. + * Assume that all other bits are either the same as for the signal + * handler, or the user mode value doesn't matter (e.g. PS.OWB). + */ + err |= __get_user(ps, &sc->sc_ps); + regs->ps = (regs->ps & ~XCHAL_PS_CALLINC_MASK) + | (ps & XCHAL_PS_CALLINC_MASK); + + /* Additional corruption checks */ + + if ((regs->windowbase >= (XCHAL_NUM_AREGS/4)) + || ((regs->windowstart & ~((1<<(XCHAL_NUM_AREGS/4)) - 1)) != 0) ) + err = 1; + if ((regs->lcount > 0) + && ((regs->lbeg > TASK_SIZE) || (regs->lend > TASK_SIZE)) ) + err = 1; + + /* Restore extended register state. + * See struct thread_struct in processor.h. + */ + thread = ¤t->thread; + + err |= __copy_from_user (regs->areg, sc->sc_areg, XCHAL_NUM_AREGS*4); + err |= __get_user(buf, &sc->sc_cpstate); + if (buf) { + if (verify_area(VERIFY_READ, buf, sizeof(*buf))) + goto badframe; + err |= restore_cpextra(buf); + } + + regs->syscall = -1; /* disable syscall checks */ + return err; + +badframe: + return 1; +} + +static inline void +flush_my_cpstate(struct task_struct *tsk) +{ + unsigned long flags; + local_irq_save(flags); + +#if 0 // FIXME + for (i = 0; i < XCHAL_CP_NUM; i++) { + if (tsk == coproc_owners[i]) { + xthal_validate_cp(i); + xthal_save_cpregs(tsk->thread.cpregs_ptr[i], i); + + /* Invalidate and "disown" the cp to allow + * callers the chance to reset cp state in the + * task_struct. */ + + xthal_invalidate_cp(i); + coproc_owners[i] = 0; + } + } +#endif + local_irq_restore(flags); +} + +/* Return codes: + 0: nothing saved + 1: stuff to save, successful + -1: stuff to save, error happened +*/ +static int +save_cpextra (struct _cpstate *buf) +{ +#if (XCHAL_EXTRA_SA_SIZE == 0) && (XCHAL_CP_NUM == 0) + return 0; +#else + + /* FIXME: If a task has never used a coprocessor, there is + * no need to save and restore anything. Tracking this + * information would allow us to optimize this section. + * Perhaps we can use current->used_math or (current->flags & + * PF_USEDFPU) or define a new field in the thread + * structure. */ + + /* We flush any live, task-owned cp state to the task_struct, + * then copy it all to the sigframe. Then we clear all + * cp/extra state in the task_struct, effectively + * clearing/resetting all cp/extra state for the signal + * handler (cp-exception handling will load these new values + * into the cp/extra registers.) This step is important for + * things like a floating-point cp, where the OS must reset + * the FCR to the default rounding mode. */ + + int err = 0; + struct task_struct *tsk = current; + + flush_my_cpstate(tsk); + /* Note that we just copy everything: 'extra' and 'cp' state together.*/ + err |= __copy_to_user(buf, tsk->thread.cp_save, XTENSA_CP_EXTRA_SIZE); + memset(tsk->thread.cp_save, 0, XTENSA_CP_EXTRA_SIZE); + +#if (XTENSA_CP_EXTRA_SIZE == 0) +#error Sanity check on memset above, cpextra_size should not be zero. +#endif + + return err ? -1 : 1; +#endif +} + +static int +setup_sigcontext(struct sigcontext *sc, struct _cpstate *cpstate, + struct pt_regs *regs, unsigned long mask) +{ + struct thread_struct *thread; + int err = 0; + +//printk("setup_sigcontext\n"); +#define COPY(x) err |= __put_user(regs->x, &sc->sc_##x) + COPY(pc); + COPY(ps); + COPY(depc); + COPY(wmask); + COPY(lbeg); + COPY(lend); + COPY(lcount); + COPY(sar); + COPY(windowbase); + COPY(windowstart); +#undef COPY + + /* Save extended register state. + * See struct thread_struct in processor.h. + */ + thread = ¤t->thread; + err |= __copy_to_user (sc->sc_areg, regs->areg, XCHAL_NUM_AREGS * 4); + err |= save_cpextra(cpstate); + err |= __put_user(err ? NULL : cpstate, &sc->sc_cpstate); + /* non-iBCS2 extensions.. */ + err |= __put_user(mask, &sc->oldmask); + + return err; +} + +asmlinkage int sys_sigreturn(struct pt_regs *regs) +{ + struct sigframe *frame = (struct sigframe *)regs->areg[1]; + sigset_t set; + if (regs->depc > 64) + panic ("Double exception sys_sigreturn\n"); + + if (verify_area(VERIFY_READ, frame, sizeof(*frame))) + goto badframe; + + if (__get_user(set.sig[0], &frame->sc.oldmask) + || (_NSIG_WORDS > 1 + && __copy_from_user(&set.sig[1], &frame->extramask, + sizeof(frame->extramask)))) + goto badframe; + + sigdelsetmask(&set, ~_BLOCKABLE); + + spin_lock_irq(¤t->sighand->siglock); + current->blocked = set; + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + if (restore_sigcontext(regs, &frame->sc)) + goto badframe; + return regs->areg[2]; + +badframe: + force_sig(SIGSEGV, current); + return 0; +} + +asmlinkage int sys_rt_sigreturn(struct pt_regs *regs) +{ + struct rt_sigframe *frame = (struct rt_sigframe *)regs->areg[1]; + sigset_t set; + stack_t st; + int ret; + if (regs->depc > 64) + { + printk("!!!!!!! DEPC !!!!!!!\n"); + return 0; + } + + if (verify_area(VERIFY_READ, frame, sizeof(*frame))) + goto badframe; + + if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) + goto badframe; + + sigdelsetmask(&set, ~_BLOCKABLE); + spin_lock_irq(¤t->sighand->siglock); + current->blocked = set; + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + if (restore_sigcontext(regs, &frame->uc.uc_mcontext)) + goto badframe; + ret = regs->areg[2]; + + if (__copy_from_user(&st, &frame->uc.uc_stack, sizeof(st))) + goto badframe; + /* It is more difficult to avoid calling this function than to + call it and ignore errors. */ + do_sigaltstack(&st, NULL, regs->areg[1]); + + return ret; + +badframe: + force_sig(SIGSEGV, current); + return 0; +} + +/* + * Set up a signal frame. + */ + +/* + * Determine which stack to use.. + */ +static inline void * +get_sigframe(struct k_sigaction *ka, unsigned long sp, size_t frame_size) +{ + if ((ka->sa.sa_flags & SA_ONSTACK) != 0 && ! on_sig_stack(sp)) + sp = current->sas_ss_sp + current->sas_ss_size; + + return (void *)((sp - frame_size) & -16ul); +} + +#define USE_SIGRETURN 0 +#define USE_RT_SIGRETURN 1 + +static int +gen_return_code(unsigned char *codemem, unsigned int use_rt_sigreturn) +{ + unsigned int retcall; + int err = 0; + +#if 0 + /* Ignoring SA_RESTORER for now; it's supposed to be obsolete, + * and the xtensa glibc doesn't use it. + */ + if (ka->sa.sa_flags & SA_RESTORER) { + regs->pr = (unsigned long) ka->sa.sa_restorer; + } else +#endif /* 0 */ + { + +#if (__NR_sigreturn > 255) || (__NR_rt_sigreturn > 255) + +/* The 12-bit immediate is really split up within the 24-bit MOVI + * instruction. As long as the above system call numbers fit within + * 8-bits, the following code works fine. See the Xtensa ISA for + * details. + */ + +#error Generating the MOVI instruction below breaks! +#endif + + retcall = use_rt_sigreturn ? __NR_rt_sigreturn : __NR_sigreturn; + +#ifdef __XTENSA_EB__ /* Big Endian version */ + /* Generate instruction: MOVI a2, retcall */ + err |= __put_user(0x22, &codemem[0]); + err |= __put_user(0x0a, &codemem[1]); + err |= __put_user(retcall, &codemem[2]); + /* Generate instruction: SYSCALL */ + err |= __put_user(0x00, &codemem[3]); + err |= __put_user(0x05, &codemem[4]); + err |= __put_user(0x00, &codemem[5]); + +#elif defined __XTENSA_EL__ /* Little Endian version */ + /* Generate instruction: MOVI a2, retcall */ + err |= __put_user(0x22, &codemem[0]); + err |= __put_user(0xa0, &codemem[1]); + err |= __put_user(retcall, &codemem[2]); + /* Generate instruction: SYSCALL */ + err |= __put_user(0x00, &codemem[3]); + err |= __put_user(0x50, &codemem[4]); + err |= __put_user(0x00, &codemem[5]); +#else +#error Must use compiler for Xtensa processors. +#endif + } + + /* Flush generated code out of the data cache */ + + if (err == 0) + __flush_invalidate_cache_range((unsigned long)codemem, 6UL); + + return err; +} + +static void +set_thread_state(struct pt_regs *regs, void *stack, unsigned char *retaddr, + void *handler, unsigned long arg1, void *arg2, void *arg3) +{ + /* Set up registers for signal handler */ + start_thread(regs, (unsigned long) handler, (unsigned long) stack); + + /* Set up a stack frame for a call4 + * Note: PS.CALLINC is set to one by start_thread + */ + regs->areg[4] = (((unsigned long) retaddr) & 0x3fffffff) | 0x40000000; + regs->areg[6] = arg1; + regs->areg[7] = (unsigned long) arg2; + regs->areg[8] = (unsigned long) arg3; +} + +static void setup_frame(int sig, struct k_sigaction *ka, + sigset_t *set, struct pt_regs *regs) +{ + struct sigframe *frame; + int err = 0; + int signal; + + frame = get_sigframe(ka, regs->areg[1], sizeof(*frame)); + if (regs->depc > 64) + { + printk("!!!!!!! DEPC !!!!!!!\n"); + return; + } + + + if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) + goto give_sigsegv; + + signal = current_thread_info()->exec_domain + && current_thread_info()->exec_domain->signal_invmap + && sig < 32 + ? current_thread_info()->exec_domain->signal_invmap[sig] + : sig; + + err |= setup_sigcontext(&frame->sc, &frame->cpstate, regs, set->sig[0]); + + if (_NSIG_WORDS > 1) { + err |= __copy_to_user(frame->extramask, &set->sig[1], + sizeof(frame->extramask)); + } + + /* Create sys_sigreturn syscall in stack frame */ + err |= gen_return_code(frame->retcode, USE_SIGRETURN); + + if (err) + goto give_sigsegv; + + /* Create signal handler execution context. + * Return context not modified until this point. + */ + set_thread_state(regs, frame, frame->retcode, + ka->sa.sa_handler, signal, &frame->sc, NULL); + + /* Set access mode to USER_DS. Nomenclature is outdated, but + * functionality is used in uaccess.h + */ + set_fs(USER_DS); + + +#if DEBUG_SIG + printk("SIG deliver (%s:%d): signal=%d sp=%p pc=%08x\n", + current->comm, current->pid, signal, frame, regs->pc); +#endif + + return; + +give_sigsegv: + if (sig == SIGSEGV) + ka->sa.sa_handler = SIG_DFL; + force_sig(SIGSEGV, current); +} + +static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, + sigset_t *set, struct pt_regs *regs) +{ + struct rt_sigframe *frame; + int err = 0; + int signal; + + frame = get_sigframe(ka, regs->areg[1], sizeof(*frame)); + if (regs->depc > 64) + panic ("Double exception sys_sigreturn\n"); + + if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) + goto give_sigsegv; + + signal = current_thread_info()->exec_domain + && current_thread_info()->exec_domain->signal_invmap + && sig < 32 + ? current_thread_info()->exec_domain->signal_invmap[sig] + : sig; + + err |= copy_siginfo_to_user(&frame->info, info); + + /* Create the ucontext. */ + err |= __put_user(0, &frame->uc.uc_flags); + err |= __put_user(0, &frame->uc.uc_link); + err |= __put_user((void *)current->sas_ss_sp, + &frame->uc.uc_stack.ss_sp); + err |= __put_user(sas_ss_flags(regs->areg[1]), + &frame->uc.uc_stack.ss_flags); + err |= __put_user(current->sas_ss_size, &frame->uc.uc_stack.ss_size); + err |= setup_sigcontext(&frame->uc.uc_mcontext, &frame->cpstate, + regs, set->sig[0]); + err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); + + /* Create sys_rt_sigreturn syscall in stack frame */ + err |= gen_return_code(frame->retcode, USE_RT_SIGRETURN); + + if (err) + goto give_sigsegv; + + /* Create signal handler execution context. + * Return context not modified until this point. + */ + set_thread_state(regs, frame, frame->retcode, + ka->sa.sa_handler, signal, &frame->info, &frame->uc); + + /* Set access mode to USER_DS. Nomenclature is outdated, but + * functionality is used in uaccess.h + */ + set_fs(USER_DS); + +#if DEBUG_SIG + printk("SIG rt deliver (%s:%d): signal=%d sp=%p pc=%08x\n", + current->comm, current->pid, signal, frame, regs->pc); +#endif + + return; + +give_sigsegv: + if (sig == SIGSEGV) + ka->sa.sa_handler = SIG_DFL; + force_sig(SIGSEGV, current); +} + + + +/* + * Note that 'init' is a special process: it doesn't get signals it doesn't + * want to handle. Thus you cannot kill init even with a SIGKILL even by + * mistake. + * + * Note that we go through the signals twice: once to check the signals that + * the kernel can handle, and then we build all the user-level signal handling + * stack-frames in one go after that. + */ +int do_signal(struct pt_regs *regs, sigset_t *oldset) +{ + siginfo_t info; + int signr; + struct k_sigaction ka; + + if (!oldset) + oldset = ¤t->blocked; + + signr = get_signal_to_deliver(&info, &ka, regs, NULL); + + /* Are we from a system call? */ + if (regs->syscall >= 0) { + /* If so, check system call restarting.. */ + switch (regs->areg[2]) { + case ERESTARTNOHAND: + case ERESTART_RESTARTBLOCK: + regs->areg[2] = -EINTR; + break; + + case ERESTARTSYS: + if (!(ka.sa.sa_flags & SA_RESTART)) { + regs->areg[2] = -EINTR; + break; + } + /* fallthrough */ + case ERESTARTNOINTR: + regs->areg[2] = regs->syscall; + regs->pc -= 3; + } + } + + if (signr == 0) + return 0; /* no signals delivered */ + + /* Whee! Actually deliver the signal. */ + + /* Set up the stack frame */ + if (ka.sa.sa_flags & SA_SIGINFO) + setup_rt_frame(signr, &ka, &info, oldset, regs); + else + setup_frame(signr, &ka, oldset, regs); + + if (ka.sa.sa_flags & SA_ONESHOT) + ka.sa.sa_handler = SIG_DFL; + + if (!(ka.sa.sa_flags & SA_NODEFER)) { + spin_lock_irq(¤t->sighand->siglock); + sigorsets(¤t->blocked, ¤t->blocked, &ka.sa.sa_mask); + sigaddset(¤t->blocked, signr); + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + } + return 1; +} diff --git a/arch/xtensa/kernel/syscalls.c b/arch/xtensa/kernel/syscalls.c new file mode 100644 index 00000000000..abc8ed6c702 --- /dev/null +++ b/arch/xtensa/kernel/syscalls.c @@ -0,0 +1,418 @@ +/* + * arch/xtensa/kernel/syscall.c + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * Copyright (C) 2000 Silicon Graphics, Inc. + * Copyright (C) 1995 - 2000 by Ralf Baechle + * + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca> + * Chris Zankel <chris@zankel.net> + * Kevin Chea + * + */ + +#define DEBUG 0 + +#include <linux/config.h> +#include <linux/linkage.h> +#include <linux/mm.h> +#include <linux/smp.h> +#include <linux/smp_lock.h> +#include <linux/mman.h> +#include <linux/sched.h> +#include <linux/file.h> +#include <linux/slab.h> +#include <linux/utsname.h> +#include <linux/unistd.h> +#include <linux/stringify.h> +#include <linux/syscalls.h> +#include <linux/sem.h> +#include <linux/msg.h> +#include <linux/shm.h> +#include <linux/errno.h> +#include <asm/ptrace.h> +#include <asm/signal.h> +#include <asm/uaccess.h> +#include <asm/hardirq.h> +#include <asm/mman.h> +#include <asm/shmparam.h> +#include <asm/page.h> +#include <asm/ipc.h> + +extern void do_syscall_trace(void); +typedef int (*syscall_t)(void *a0,...); +extern int (*do_syscalls)(struct pt_regs *regs, syscall_t fun, + int narg); +extern syscall_t sys_call_table[]; +extern unsigned char sys_narg_table[]; + +/* + * sys_pipe() is the normal C calling standard for creating a pipe. It's not + * the way unix traditional does this, though. + */ + +int sys_pipe(int __user *userfds) +{ + int fd[2]; + int error; + + error = do_pipe(fd); + if (!error) { + if (copy_to_user(userfds, fd, 2 * sizeof(int))) + error = -EFAULT; + } + return error; +} + +/* + * Common code for old and new mmaps. + */ + +static inline long do_mmap2(unsigned long addr, unsigned long len, + unsigned long prot, unsigned long flags, + unsigned long fd, unsigned long pgoff) +{ + int error = -EBADF; + struct file * file = NULL; + + flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE); + if (!(flags & MAP_ANONYMOUS)) { + file = fget(fd); + if (!file) + goto out; + } + + down_write(¤t->mm->mmap_sem); + error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff); + up_write(¤t->mm->mmap_sem); + + if (file) + fput(file); +out: + return error; +} + +unsigned long old_mmap(unsigned long addr, size_t len, int prot, + int flags, int fd, off_t offset) +{ + return do_mmap2(addr, len, prot, flags, fd, offset >> PAGE_SHIFT); +} + +long sys_mmap2(unsigned long addr, unsigned long len, unsigned long prot, + unsigned long flags, unsigned long fd, unsigned long pgoff) +{ + return do_mmap2(addr, len, prot, flags, fd, pgoff); +} + +int sys_fork(struct pt_regs *regs) +{ + return do_fork(SIGCHLD, regs->areg[1], regs, 0, NULL, NULL); +} + +int sys_vfork(struct pt_regs *regs) +{ + return do_fork(CLONE_VFORK|CLONE_VM|SIGCHLD, regs->areg[1], + regs, 0, NULL, NULL); +} + +int sys_clone(struct pt_regs *regs) +{ + unsigned long clone_flags; + unsigned long newsp; + int __user *parent_tidptr, *child_tidptr; + clone_flags = regs->areg[4]; + newsp = regs->areg[3]; + parent_tidptr = (int __user *)regs->areg[5]; + child_tidptr = (int __user *)regs->areg[6]; + if (!newsp) + newsp = regs->areg[1]; + return do_fork(clone_flags,newsp,regs,0,parent_tidptr,child_tidptr); +} + +/* + * sys_execve() executes a new program. + */ + +int sys_execve(struct pt_regs *regs) +{ + int error; + char * filename; + + filename = getname((char *) (long)regs->areg[5]); + error = PTR_ERR(filename); + if (IS_ERR(filename)) + goto out; + error = do_execve(filename, (char **) (long)regs->areg[3], + (char **) (long)regs->areg[4], regs); + putname(filename); + +out: + return error; +} + +int sys_uname(struct old_utsname * name) +{ + if (name && !copy_to_user(name, &system_utsname, sizeof (*name))) + return 0; + return -EFAULT; +} + +int sys_olduname(struct oldold_utsname * name) +{ + int error; + + if (!name) + return -EFAULT; + if (!access_ok(VERIFY_WRITE,name,sizeof(struct oldold_utsname))) + return -EFAULT; + + error = __copy_to_user(&name->sysname,&system_utsname.sysname,__OLD_UTS_LEN); + error -= __put_user(0,name->sysname+__OLD_UTS_LEN); + error -= __copy_to_user(&name->nodename,&system_utsname.nodename,__OLD_UTS_LEN); + error -= __put_user(0,name->nodename+__OLD_UTS_LEN); + error -= __copy_to_user(&name->release,&system_utsname.release,__OLD_UTS_LEN); + error -= __put_user(0,name->release+__OLD_UTS_LEN); + error -= __copy_to_user(&name->version,&system_utsname.version,__OLD_UTS_LEN); + error -= __put_user(0,name->version+__OLD_UTS_LEN); + error -= __copy_to_user(&name->machine,&system_utsname.machine,__OLD_UTS_LEN); + error -= __put_user(0,name->machine+__OLD_UTS_LEN); + + return error ? -EFAULT : 0; +} + + +/* + * Build the string table for the builtin "poor man's strace". + */ + +#if DEBUG +#define SYSCALL(fun, narg) #fun, +static char *sfnames[] = { +#include "syscalls.h" +}; +#undef SYS +#endif + +void system_call (struct pt_regs *regs) +{ + syscall_t syscall; + unsigned long parm0, parm1, parm2, parm3, parm4, parm5; + int nargs, res; + unsigned int syscallnr; + int ps; + +#if DEBUG + int i; + unsigned long parms[6]; + char *sysname; +#endif + + regs->syscall = regs->areg[2]; + + do_syscall_trace(); + + /* Have to load after syscall_trace because strace + * sometimes changes regs->syscall. + */ + syscallnr = regs->syscall; + + parm0 = parm1 = parm2 = parm3 = parm4 = parm5 = 0; + + /* Restore interrupt level to syscall invoker's. + * If this were in assembly, we wouldn't disable + * interrupts in the first place: + */ + local_save_flags (ps); + local_irq_restore((ps & ~XCHAL_PS_INTLEVEL_MASK) | + (regs->ps & XCHAL_PS_INTLEVEL_MASK) ); + + if (syscallnr > __NR_Linux_syscalls) { + regs->areg[2] = -ENOSYS; + return; + } + + syscall = sys_call_table[syscallnr]; + nargs = sys_narg_table[syscallnr]; + + if (syscall == NULL) { + regs->areg[2] = -ENOSYS; + return; + } + + /* There shouldn't be more than six arguments in the table! */ + + if (nargs > 6) + panic("Internal error - too many syscall arguments (%d)!\n", + nargs); + + /* Linux takes system-call arguments in registers. The ABI + * and Xtensa software conventions require the system-call + * number in a2. If an argument exists in a2, we move it to + * the next available register. Note that for improved + * efficiency, we do NOT shift all parameters down one + * register to maintain the original order. + * + * At best case (zero arguments), we just write the syscall + * number to a2. At worst case (1 to 6 arguments), we move + * the argument in a2 to the next available register, then + * write the syscall number to a2. + * + * For clarity, the following truth table enumerates all + * possibilities. + * + * arguments syscall number arg0, arg1, arg2, arg3, arg4, arg5 + * --------- -------------- ---------------------------------- + * 0 a2 + * 1 a2 a3 + * 2 a2 a4, a3 + * 3 a2 a5, a3, a4 + * 4 a2 a6, a3, a4, a5 + * 5 a2 a7, a3, a4, a5, a6 + * 6 a2 a8, a3, a4, a5, a6, a7 + */ + if (nargs) { + parm0 = regs->areg[nargs+2]; + parm1 = regs->areg[3]; + parm2 = regs->areg[4]; + parm3 = regs->areg[5]; + parm4 = regs->areg[6]; + parm5 = regs->areg[7]; + } else /* nargs == 0 */ + parm0 = (unsigned long) regs; + +#if DEBUG + parms[0] = parm0; + parms[1] = parm1; + parms[2] = parm2; + parms[3] = parm3; + parms[4] = parm4; + parms[5] = parm5; + + sysname = sfnames[syscallnr]; + if (strncmp(sysname, "sys_", 4) == 0) + sysname = sysname + 4; + + printk("\017SYSCALL:I:%x:%d:%s %s(", regs->pc, current->pid, + current->comm, sysname); + for (i = 0; i < nargs; i++) + printk((i>0) ? ", %#lx" : "%#lx", parms[i]); + printk(")\n"); +#endif + + res = syscall((void *)parm0, parm1, parm2, parm3, parm4, parm5); + +#if DEBUG + printk("\017SYSCALL:O:%d:%s %s(",current->pid, current->comm, sysname); + for (i = 0; i < nargs; i++) + printk((i>0) ? ", %#lx" : "%#lx", parms[i]); + if (res < 4096) + printk(") = %d\n", res); + else + printk(") = %#x\n", res); +#endif /* DEBUG */ + + regs->areg[2] = res; + do_syscall_trace(); +} + +/* + * sys_ipc() is the de-multiplexer for the SysV IPC calls.. + * + * This is really horribly ugly. + */ + +int sys_ipc (uint call, int first, int second, + int third, void __user *ptr, long fifth) +{ + int version, ret; + + version = call >> 16; /* hack for backward compatibility */ + call &= 0xffff; + ret = -ENOSYS; + + switch (call) { + case SEMOP: + ret = sys_semtimedop (first, (struct sembuf __user *)ptr, + second, NULL); + break; + + case SEMTIMEDOP: + ret = sys_semtimedop (first, (struct sembuf __user *)ptr, + second, (const struct timespec *) fifth); + break; + + case SEMGET: + ret = sys_semget (first, second, third); + break; + + case SEMCTL: { + union semun fourth; + + if (ptr && !get_user(fourth.__pad, (void *__user *) ptr)) + ret = sys_semctl (first, second, third, fourth); + break; + } + + case MSGSND: + ret = sys_msgsnd (first, (struct msgbuf __user*) ptr, + second, third); + break; + + case MSGRCV: + switch (version) { + case 0: { + struct ipc_kludge tmp; + + if (ptr && !copy_from_user(&tmp, + (struct ipc_kludge *) ptr, + sizeof (tmp))) + ret = sys_msgrcv (first, tmp.msgp, second, + tmp.msgtyp, third); + break; + } + + default: + ret = sys_msgrcv (first, (struct msgbuf __user *) ptr, + second, 0, third); + break; + } + break; + + case MSGGET: + ret = sys_msgget ((key_t) first, second); + break; + + case MSGCTL: + ret = sys_msgctl (first, second, (struct msqid_ds __user*) ptr); + break; + + case SHMAT: { + ulong raddr; + ret = do_shmat (first, (char __user *) ptr, second, &raddr); + + if (!ret) + ret = put_user (raddr, (ulong __user *) third); + + break; + } + + case SHMDT: + ret = sys_shmdt ((char __user *)ptr); + break; + + case SHMGET: + ret = sys_shmget (first, second, third); + break; + + case SHMCTL: + ret = sys_shmctl (first, second, (struct shmid_ds __user*) ptr); + break; + } + return ret; +} + diff --git a/arch/xtensa/kernel/syscalls.h b/arch/xtensa/kernel/syscalls.h new file mode 100644 index 00000000000..5b3f75f50fe --- /dev/null +++ b/arch/xtensa/kernel/syscalls.h @@ -0,0 +1,248 @@ +/* + * arch/xtensa/kernel/syscalls.h + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 1995, 1996, 1997, 1998 by Ralf Baechle + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Changes by Joe Taylor <joe@tensilica.com> + */ + +/* + * This file is being included twice - once to build a list of all + * syscalls and once to build a table of how many arguments each syscall + * accepts. Syscalls that receive a pointer to the saved registers are + * marked as having zero arguments. + * + * The binary compatibility calls are in a separate list. + * + * Entry '0' used to be system_call. It's removed to disable indirect + * system calls for now so user tasks can't recurse. See mips' + * sys_syscall for a comparable example. + */ + +SYSCALL(0, 0) /* 00 */ + +SYSCALL(sys_exit, 1) +SYSCALL(sys_fork, 0) +SYSCALL(sys_read, 3) +SYSCALL(sys_write, 3) +SYSCALL(sys_open, 3) /* 05 */ +SYSCALL(sys_close, 1) +SYSCALL(sys_waitpid, 3) +SYSCALL(sys_creat, 2) +SYSCALL(sys_link, 2) +SYSCALL(sys_unlink, 1) /* 10 */ +SYSCALL(sys_execve, 0) +SYSCALL(sys_chdir, 1) +SYSCALL(sys_time, 1) +SYSCALL(sys_mknod, 3) +SYSCALL(sys_chmod, 2) /* 15 */ +SYSCALL(sys_lchown, 3) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_stat, 2) +SYSCALL(sys_lseek, 3) +SYSCALL(sys_getpid, 0) /* 20 */ +SYSCALL(sys_mount, 5) +SYSCALL(sys_oldumount, 1) +SYSCALL(sys_setuid, 1) +SYSCALL(sys_getuid, 0) +SYSCALL(sys_stime, 1) /* 25 */ +SYSCALL(sys_ptrace, 4) +SYSCALL(sys_alarm, 1) +SYSCALL(sys_fstat, 2) +SYSCALL(sys_pause, 0) +SYSCALL(sys_utime, 2) /* 30 */ +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_access, 2) +SYSCALL(sys_nice, 1) +SYSCALL(sys_ni_syscall, 0) /* 35 */ +SYSCALL(sys_sync, 0) +SYSCALL(sys_kill, 2) +SYSCALL(sys_rename, 2) +SYSCALL(sys_mkdir, 2) +SYSCALL(sys_rmdir, 1) /* 40 */ +SYSCALL(sys_dup, 1) +SYSCALL(sys_pipe, 1) +SYSCALL(sys_times, 1) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_brk, 1) /* 45 */ +SYSCALL(sys_setgid, 1) +SYSCALL(sys_getgid, 0) +SYSCALL(sys_ni_syscall, 0) /* was signal(2) */ +SYSCALL(sys_geteuid, 0) +SYSCALL(sys_getegid, 0) /* 50 */ +SYSCALL(sys_acct, 1) +SYSCALL(sys_umount, 2) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_ioctl, 3) +SYSCALL(sys_fcntl, 3) /* 55 */ +SYSCALL(sys_ni_syscall, 2) +SYSCALL(sys_setpgid, 2) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_olduname, 1) +SYSCALL(sys_umask, 1) /* 60 */ +SYSCALL(sys_chroot, 1) +SYSCALL(sys_ustat, 2) +SYSCALL(sys_dup2, 2) +SYSCALL(sys_getppid, 0) +SYSCALL(sys_getpgrp, 0) /* 65 */ +SYSCALL(sys_setsid, 0) +SYSCALL(sys_sigaction, 3) +SYSCALL(sys_sgetmask, 0) +SYSCALL(sys_ssetmask, 1) +SYSCALL(sys_setreuid, 2) /* 70 */ +SYSCALL(sys_setregid, 2) +SYSCALL(sys_sigsuspend, 0) +SYSCALL(sys_sigpending, 1) +SYSCALL(sys_sethostname, 2) +SYSCALL(sys_setrlimit, 2) /* 75 */ +SYSCALL(sys_getrlimit, 2) +SYSCALL(sys_getrusage, 2) +SYSCALL(sys_gettimeofday, 2) +SYSCALL(sys_settimeofday, 2) +SYSCALL(sys_getgroups, 2) /* 80 */ +SYSCALL(sys_setgroups, 2) +SYSCALL(sys_ni_syscall, 0) /* old_select */ +SYSCALL(sys_symlink, 2) +SYSCALL(sys_lstat, 2) +SYSCALL(sys_readlink, 3) /* 85 */ +SYSCALL(sys_uselib, 1) +SYSCALL(sys_swapon, 2) +SYSCALL(sys_reboot, 3) +SYSCALL(old_readdir, 3) +SYSCALL(old_mmap, 6) /* 90 */ +SYSCALL(sys_munmap, 2) +SYSCALL(sys_truncate, 2) +SYSCALL(sys_ftruncate, 2) +SYSCALL(sys_fchmod, 2) +SYSCALL(sys_fchown, 3) /* 95 */ +SYSCALL(sys_getpriority, 2) +SYSCALL(sys_setpriority, 3) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_statfs, 2) +SYSCALL(sys_fstatfs, 2) /* 100 */ +SYSCALL(sys_ni_syscall, 3) +SYSCALL(sys_socketcall, 2) +SYSCALL(sys_syslog, 3) +SYSCALL(sys_setitimer, 3) +SYSCALL(sys_getitimer, 2) /* 105 */ +SYSCALL(sys_newstat, 2) +SYSCALL(sys_newlstat, 2) +SYSCALL(sys_newfstat, 2) +SYSCALL(sys_uname, 1) +SYSCALL(sys_ni_syscall, 0) /* 110 */ +SYSCALL(sys_vhangup, 0) +SYSCALL(sys_ni_syscall, 0) /* was sys_idle() */ +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_wait4, 4) +SYSCALL(sys_swapoff, 1) /* 115 */ +SYSCALL(sys_sysinfo, 1) +SYSCALL(sys_ipc, 5) /* 6 really, but glibc uses only 5) */ +SYSCALL(sys_fsync, 1) +SYSCALL(sys_sigreturn, 0) +SYSCALL(sys_clone, 0) /* 120 */ +SYSCALL(sys_setdomainname, 2) +SYSCALL(sys_newuname, 1) +SYSCALL(sys_ni_syscall, 0) /* sys_modify_ldt */ +SYSCALL(sys_adjtimex, 1) +SYSCALL(sys_mprotect, 3) /* 125 */ +SYSCALL(sys_sigprocmask, 3) +SYSCALL(sys_ni_syscall, 2) /* old sys_create_module */ +SYSCALL(sys_init_module, 2) +SYSCALL(sys_delete_module, 1) +SYSCALL(sys_ni_syscall, 1) /* old sys_get_kernel_sysm */ /* 130 */ +SYSCALL(sys_quotactl, 0) +SYSCALL(sys_getpgid, 1) +SYSCALL(sys_fchdir, 1) +SYSCALL(sys_bdflush, 2) +SYSCALL(sys_sysfs, 3) /* 135 */ +SYSCALL(sys_personality, 1) +SYSCALL(sys_ni_syscall, 0) /* for afs_syscall */ +SYSCALL(sys_setfsuid, 1) +SYSCALL(sys_setfsgid, 1) +SYSCALL(sys_llseek, 5) /* 140 */ +SYSCALL(sys_getdents, 3) +SYSCALL(sys_select, 5) +SYSCALL(sys_flock, 2) +SYSCALL(sys_msync, 3) +SYSCALL(sys_readv, 3) /* 145 */ +SYSCALL(sys_writev, 3) +SYSCALL(sys_ni_syscall, 3) +SYSCALL(sys_ni_syscall, 3) +SYSCALL(sys_ni_syscall, 4) /* handled in fast syscall handler. */ +SYSCALL(sys_ni_syscall, 0) /* 150 */ +SYSCALL(sys_getsid, 1) +SYSCALL(sys_fdatasync, 1) +SYSCALL(sys_sysctl, 1) +SYSCALL(sys_mlock, 2) +SYSCALL(sys_munlock, 2) /* 155 */ +SYSCALL(sys_mlockall, 1) +SYSCALL(sys_munlockall, 0) +SYSCALL(sys_sched_setparam,2) +SYSCALL(sys_sched_getparam,2) +SYSCALL(sys_sched_setscheduler,3) /* 160 */ +SYSCALL(sys_sched_getscheduler,1) +SYSCALL(sys_sched_yield,0) +SYSCALL(sys_sched_get_priority_max,1) +SYSCALL(sys_sched_get_priority_min,1) +SYSCALL(sys_sched_rr_get_interval,2) /* 165 */ +SYSCALL(sys_nanosleep,2) +SYSCALL(sys_mremap,4) +SYSCALL(sys_accept, 3) +SYSCALL(sys_bind, 3) +SYSCALL(sys_connect, 3) /* 170 */ +SYSCALL(sys_getpeername, 3) +SYSCALL(sys_getsockname, 3) +SYSCALL(sys_getsockopt, 5) +SYSCALL(sys_listen, 2) +SYSCALL(sys_recv, 4) /* 175 */ +SYSCALL(sys_recvfrom, 6) +SYSCALL(sys_recvmsg, 3) +SYSCALL(sys_send, 4) +SYSCALL(sys_sendmsg, 3) +SYSCALL(sys_sendto, 6) /* 180 */ +SYSCALL(sys_setsockopt, 5) +SYSCALL(sys_shutdown, 2) +SYSCALL(sys_socket, 3) +SYSCALL(sys_socketpair, 4) +SYSCALL(sys_setresuid, 3) /* 185 */ +SYSCALL(sys_getresuid, 3) +SYSCALL(sys_ni_syscall, 5) /* old sys_query_module */ +SYSCALL(sys_poll, 3) +SYSCALL(sys_nfsservctl, 3) +SYSCALL(sys_setresgid, 3) /* 190 */ +SYSCALL(sys_getresgid, 3) +SYSCALL(sys_prctl, 5) +SYSCALL(sys_rt_sigreturn, 0) +SYSCALL(sys_rt_sigaction, 4) +SYSCALL(sys_rt_sigprocmask, 4) /* 195 */ +SYSCALL(sys_rt_sigpending, 2) +SYSCALL(sys_rt_sigtimedwait, 4) +SYSCALL(sys_rt_sigqueueinfo, 3) +SYSCALL(sys_rt_sigsuspend, 0) +SYSCALL(sys_pread64, 5) /* 200 */ +SYSCALL(sys_pwrite64, 5) +SYSCALL(sys_chown, 3) +SYSCALL(sys_getcwd, 2) +SYSCALL(sys_capget, 2) +SYSCALL(sys_capset, 2) /* 205 */ +SYSCALL(sys_sigaltstack, 0) +SYSCALL(sys_sendfile, 4) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_ni_syscall, 0) +SYSCALL(sys_mmap2, 6) /* 210 */ +SYSCALL(sys_truncate64, 2) +SYSCALL(sys_ftruncate64, 2) +SYSCALL(sys_stat64, 2) +SYSCALL(sys_lstat64, 2) +SYSCALL(sys_fstat64, 2) /* 215 */ +SYSCALL(sys_pivot_root, 2) +SYSCALL(sys_mincore, 3) +SYSCALL(sys_madvise, 3) +SYSCALL(sys_getdents64, 3) +SYSCALL(sys_vfork, 0) /* 220 */ diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c new file mode 100644 index 00000000000..e07287db5a4 --- /dev/null +++ b/arch/xtensa/kernel/time.c @@ -0,0 +1,227 @@ +/* + * arch/xtensa/kernel/time.c + * + * Timer and clock support. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + */ + +#include <linux/config.h> +#include <linux/errno.h> +#include <linux/time.h> +#include <linux/timex.h> +#include <linux/interrupt.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/irq.h> +#include <linux/profile.h> +#include <linux/delay.h> + +#include <asm/timex.h> +#include <asm/platform.h> + + +extern volatile unsigned long wall_jiffies; + +u64 jiffies_64 = INITIAL_JIFFIES; +EXPORT_SYMBOL(jiffies_64); + +spinlock_t rtc_lock = SPIN_LOCK_UNLOCKED; +EXPORT_SYMBOL(rtc_lock); + + +#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT +unsigned long ccount_per_jiffy; /* per 1/HZ */ +unsigned long ccount_nsec; /* nsec per ccount increment */ +#endif + +unsigned int last_ccount_stamp; +static long last_rtc_update = 0; + +/* + * Scheduler clock - returns current tim in nanosec units. + */ + +unsigned long long sched_clock(void) +{ + return (unsigned long long)jiffies * (1000000000 / HZ); +} + +static irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs); +static struct irqaction timer_irqaction = { + .handler = timer_interrupt, + .flags = SA_INTERRUPT, + .name = "timer", +}; + +void __init time_init(void) +{ + time_t sec_o, sec_n = 0; + + /* The platform must provide a function to calibrate the processor + * speed for the CALIBRATE. + */ + +#if CONFIG_XTENSA_CALIBRATE_CCOUNT + printk("Calibrating CPU frequency "); + platform_calibrate_ccount(); + printk("%d.%02d MHz\n", (int)ccount_per_jiffy/(1000000/HZ), + (int)(ccount_per_jiffy/(10000/HZ))%100); +#endif + + /* Set time from RTC (if provided) */ + + if (platform_get_rtc_time(&sec_o) == 0) + while (platform_get_rtc_time(&sec_n)) + if (sec_o != sec_n) + break; + + xtime.tv_nsec = 0; + last_rtc_update = xtime.tv_sec = sec_n; + last_ccount_stamp = get_ccount(); + + set_normalized_timespec(&wall_to_monotonic, + -xtime.tv_sec, -xtime.tv_nsec); + + /* Initialize the linux timer interrupt. */ + + setup_irq(LINUX_TIMER_INT, &timer_irqaction); + set_linux_timer(get_ccount() + CCOUNT_PER_JIFFY); +} + + +int do_settimeofday(struct timespec *tv) +{ + time_t wtm_sec, sec = tv->tv_sec; + long wtm_nsec, nsec = tv->tv_nsec; + unsigned long ccount; + + if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC) + return -EINVAL; + + write_seqlock_irq(&xtime_lock); + + /* This is revolting. We need to set "xtime" correctly. However, the + * value in this location is the value at the most recent update of + * wall time. Discover what correction gettimeofday() would have + * made, and then undo it! + */ + ccount = get_ccount(); + nsec -= (ccount - last_ccount_stamp) * CCOUNT_NSEC; + nsec -= (jiffies - wall_jiffies) * CCOUNT_PER_JIFFY * CCOUNT_NSEC; + + wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - sec); + wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - nsec); + + set_normalized_timespec(&xtime, sec, nsec); + set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec); + + time_adjust = 0; /* stop active adjtime() */ + time_status |= STA_UNSYNC; + time_maxerror = NTP_PHASE_LIMIT; + time_esterror = NTP_PHASE_LIMIT; + write_sequnlock_irq(&xtime_lock); + return 0; +} + +EXPORT_SYMBOL(do_settimeofday); + + +void do_gettimeofday(struct timeval *tv) +{ + unsigned long flags; + unsigned long sec, usec, delta, lost, seq; + + do { + seq = read_seqbegin_irqsave(&xtime_lock, flags); + + delta = get_ccount() - last_ccount_stamp; + sec = xtime.tv_sec; + usec = (xtime.tv_nsec / NSEC_PER_USEC); + + lost = jiffies - wall_jiffies; + + } while (read_seqretry_irqrestore(&xtime_lock, seq, flags)); + + usec += lost * (1000000UL/HZ) + (delta * CCOUNT_NSEC) / NSEC_PER_USEC; + for (; usec >= 1000000; sec++, usec -= 1000000) + ; + + tv->tv_sec = sec; + tv->tv_usec = usec; +} + +EXPORT_SYMBOL(do_gettimeofday); + +/* + * The timer interrupt is called HZ times per second. + */ + +irqreturn_t timer_interrupt (int irq, void *dev_id, struct pt_regs *regs) +{ + + unsigned long next; + + next = get_linux_timer(); + +again: + while ((signed long)(get_ccount() - next) > 0) { + + profile_tick(CPU_PROFILING, regs); +#ifndef CONFIG_SMP + update_process_times(user_mode(regs)); +#endif + + write_seqlock(&xtime_lock); + + last_ccount_stamp = next; + next += CCOUNT_PER_JIFFY; + do_timer (regs); /* Linux handler in kernel/timer.c */ + + if ((time_status & STA_UNSYNC) == 0 && + xtime.tv_sec - last_rtc_update >= 659 && + abs((xtime.tv_nsec/1000)-(1000000-1000000/HZ))<5000000/HZ && + jiffies - wall_jiffies == 1) { + + if (platform_set_rtc_time(xtime.tv_sec+1) == 0) + last_rtc_update = xtime.tv_sec+1; + else + /* Do it again in 60 s */ + last_rtc_update += 60; + } + write_sequnlock(&xtime_lock); + } + + /* NOTE: writing CCOMPAREn clears the interrupt. */ + + set_linux_timer (next); + + /* Make sure we didn't miss any tick... */ + + if ((signed long)(get_ccount() - next) > 0) + goto again; + + /* Allow platform to do something usefull (Wdog). */ + + platform_heartbeat(); + + return IRQ_HANDLED; +} + +#ifndef CONFIG_GENERIC_CALIBRATE_DELAY +void __devinit calibrate_delay(void) +{ + loops_per_jiffy = CCOUNT_PER_JIFFY; + printk("Calibrating delay loop (skipped)... " + "%lu.%02lu BogoMIPS preset\n", + loops_per_jiffy/(1000000/HZ), + (loops_per_jiffy/(10000/HZ)) % 100); +} +#endif + diff --git a/arch/xtensa/kernel/traps.c b/arch/xtensa/kernel/traps.c new file mode 100644 index 00000000000..804246e743b --- /dev/null +++ b/arch/xtensa/kernel/traps.c @@ -0,0 +1,498 @@ +/* + * arch/xtensa/kernel/traps.c + * + * Exception handling. + * + * Derived from code with the following copyrights: + * Copyright (C) 1994 - 1999 by Ralf Baechle + * Modified for R3000 by Paul M. Antoine, 1995, 1996 + * Complete output from die() by Ulf Carlsson, 1998 + * Copyright (C) 1999 Silicon Graphics, Inc. + * + * Essentially rewritten for the Xtensa architecture port. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Chris Zankel <chris@zankel.net> + * Marc Gauthier<marc@tensilica.com, marc@alumni.uwaterloo.ca> + * Kevin Chea + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ + +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/stringify.h> +#include <linux/kallsyms.h> + +#include <asm/ptrace.h> +#include <asm/timex.h> +#include <asm/uaccess.h> +#include <asm/pgtable.h> +#include <asm/processor.h> + +#ifdef CONFIG_KGDB +extern int gdb_enter; +extern int return_from_debug_flag; +#endif + +/* + * Machine specific interrupt handlers + */ + +extern void kernel_exception(void); +extern void user_exception(void); + +extern void fast_syscall_kernel(void); +extern void fast_syscall_user(void); +extern void fast_alloca(void); +extern void fast_unaligned(void); +extern void fast_second_level_miss(void); +extern void fast_store_prohibited(void); +extern void fast_coprocessor(void); + +extern void do_illegal_instruction (struct pt_regs*); +extern void do_interrupt (struct pt_regs*); +extern void do_unaligned_user (struct pt_regs*); +extern void do_multihit (struct pt_regs*, unsigned long); +extern void do_page_fault (struct pt_regs*, unsigned long); +extern void do_debug (struct pt_regs*); +extern void system_call (struct pt_regs*); + +/* + * The vector table must be preceded by a save area (which + * implies it must be in RAM, unless one places RAM immediately + * before a ROM and puts the vector at the start of the ROM (!)) + */ + +#define KRNL 0x01 +#define USER 0x02 + +#define COPROCESSOR(x) \ +{ XCHAL_EXCCAUSE_COPROCESSOR ## x ## _DISABLED, USER, fast_coprocessor } + +typedef struct { + int cause; + int fast; + void* handler; +} dispatch_init_table_t; + +dispatch_init_table_t __init dispatch_init_table[] = { + +{ XCHAL_EXCCAUSE_ILLEGAL_INSTRUCTION, 0, do_illegal_instruction}, +{ XCHAL_EXCCAUSE_SYSTEM_CALL, KRNL, fast_syscall_kernel }, +{ XCHAL_EXCCAUSE_SYSTEM_CALL, USER, fast_syscall_user }, +{ XCHAL_EXCCAUSE_SYSTEM_CALL, 0, system_call }, +/* XCHAL_EXCCAUSE_INSTRUCTION_FETCH unhandled */ +/* XCHAL_EXCCAUSE_LOAD_STORE_ERROR unhandled*/ +{ XCHAL_EXCCAUSE_LEVEL1_INTERRUPT, 0, do_interrupt }, +{ XCHAL_EXCCAUSE_ALLOCA, USER|KRNL, fast_alloca }, +/* XCHAL_EXCCAUSE_INTEGER_DIVIDE_BY_ZERO unhandled */ +/* XCHAL_EXCCAUSE_PRIVILEGED unhandled */ +#if XCHAL_UNALIGNED_LOAD_EXCEPTION || XCHAL_UNALIGNED_STORE_EXCEPTION +#ifdef CONFIG_UNALIGNED_USER +{ XCHAL_EXCCAUSE_UNALIGNED, USER, fast_unaligned }, +#else +{ XCHAL_EXCCAUSE_UNALIGNED, 0, do_unaligned_user }, +#endif +{ XCHAL_EXCCAUSE_UNALIGNED, KRNL, fast_unaligned }, +#endif +{ XCHAL_EXCCAUSE_ITLB_MISS, 0, do_page_fault }, +{ XCHAL_EXCCAUSE_ITLB_MISS, USER|KRNL, fast_second_level_miss}, +{ XCHAL_EXCCAUSE_ITLB_MULTIHIT, 0, do_multihit }, +{ XCHAL_EXCCAUSE_ITLB_PRIVILEGE, 0, do_page_fault }, +/* XCHAL_EXCCAUSE_SIZE_RESTRICTION unhandled */ +{ XCHAL_EXCCAUSE_FETCH_CACHE_ATTRIBUTE, 0, do_page_fault }, +{ XCHAL_EXCCAUSE_DTLB_MISS, USER|KRNL, fast_second_level_miss}, +{ XCHAL_EXCCAUSE_DTLB_MISS, 0, do_page_fault }, +{ XCHAL_EXCCAUSE_DTLB_MULTIHIT, 0, do_multihit }, +{ XCHAL_EXCCAUSE_DTLB_PRIVILEGE, 0, do_page_fault }, +/* XCHAL_EXCCAUSE_DTLB_SIZE_RESTRICTION unhandled */ +{ XCHAL_EXCCAUSE_STORE_CACHE_ATTRIBUTE, USER|KRNL, fast_store_prohibited }, +{ XCHAL_EXCCAUSE_STORE_CACHE_ATTRIBUTE, 0, do_page_fault }, +{ XCHAL_EXCCAUSE_LOAD_CACHE_ATTRIBUTE, 0, do_page_fault }, +/* XCCHAL_EXCCAUSE_FLOATING_POINT unhandled */ +#if (XCHAL_CP_MASK & 1) +COPROCESSOR(0), +#endif +#if (XCHAL_CP_MASK & 2) +COPROCESSOR(1), +#endif +#if (XCHAL_CP_MASK & 4) +COPROCESSOR(2), +#endif +#if (XCHAL_CP_MASK & 8) +COPROCESSOR(3), +#endif +#if (XCHAL_CP_MASK & 16) +COPROCESSOR(4), +#endif +#if (XCHAL_CP_MASK & 32) +COPROCESSOR(5), +#endif +#if (XCHAL_CP_MASK & 64) +COPROCESSOR(6), +#endif +#if (XCHAL_CP_MASK & 128) +COPROCESSOR(7), +#endif +{ EXCCAUSE_MAPPED_DEBUG, 0, do_debug }, +{ -1, -1, 0 } + +}; + +/* The exception table <exc_table> serves two functions: + * 1. it contains three dispatch tables (fast_user, fast_kernel, default-c) + * 2. it is a temporary memory buffer for the exception handlers. + */ + +unsigned long exc_table[EXC_TABLE_SIZE/4]; + +void die(const char*, struct pt_regs*, long); + +static inline void +__die_if_kernel(const char *str, struct pt_regs *regs, long err) +{ + if (!user_mode(regs)) + die(str, regs, err); +} + +/* + * Unhandled Exceptions. Kill user task or panic if in kernel space. + */ + +void do_unhandled(struct pt_regs *regs, unsigned long exccause) +{ + __die_if_kernel("Caught unhandled exception - should not happen", + regs, SIGKILL); + + /* If in user mode, send SIGILL signal to current process */ + printk("Caught unhandled exception in '%s' " + "(pid = %d, pc = %#010lx) - should not happen\n" + "\tEXCCAUSE is %ld\n", + current->comm, current->pid, regs->pc, exccause); + force_sig(SIGILL, current); +} + +/* + * Multi-hit exception. This if fatal! + */ + +void do_multihit(struct pt_regs *regs, unsigned long exccause) +{ + die("Caught multihit exception", regs, SIGKILL); +} + +/* + * Level-1 interrupt. + * We currently have no priority encoding. + */ + +unsigned long ignored_level1_interrupts; +extern void do_IRQ(int, struct pt_regs *); + +void do_interrupt (struct pt_regs *regs) +{ + unsigned long intread = get_sr (INTREAD); + unsigned long intenable = get_sr (INTENABLE); + int i, mask; + + /* Handle all interrupts (no priorities). + * (Clear the interrupt before processing, in case it's + * edge-triggered or software-generated) + */ + + for (i=0, mask = 1; i < XCHAL_NUM_INTERRUPTS; i++, mask <<= 1) { + if (mask & (intread & intenable)) { + set_sr (mask, INTCLEAR); + do_IRQ (i,regs); + } + } +} + +/* + * Illegal instruction. Fatal if in kernel space. + */ + +void +do_illegal_instruction(struct pt_regs *regs) +{ + __die_if_kernel("Illegal instruction in kernel", regs, SIGKILL); + + /* If in user mode, send SIGILL signal to current process. */ + + printk("Illegal Instruction in '%s' (pid = %d, pc = %#010lx)\n", + current->comm, current->pid, regs->pc); + force_sig(SIGILL, current); +} + + +/* + * Handle unaligned memory accesses from user space. Kill task. + * + * If CONFIG_UNALIGNED_USER is not set, we don't allow unaligned memory + * accesses causes from user space. + */ + +#if XCHAL_UNALIGNED_LOAD_EXCEPTION || XCHAL_UNALIGNED_STORE_EXCEPTION +#ifndef CONFIG_UNALIGNED_USER +void +do_unaligned_user (struct pt_regs *regs) +{ + siginfo_t info; + + __die_if_kernel("Unhandled unaligned exception in kernel", + regs, SIGKILL); + + current->thread.bad_vaddr = regs->excvaddr; + current->thread.error_code = -3; + printk("Unaligned memory access to %08lx in '%s' " + "(pid = %d, pc = %#010lx)\n", + regs->excvaddr, current->comm, current->pid, regs->pc); + info.si_signo = SIGBUS; + info.si_errno = 0; + info.si_code = BUS_ADRALN; + info.si_addr = (void *) regs->excvaddr; + force_sig_info(SIGSEGV, &info, current); + +} +#endif +#endif + +void +do_debug(struct pt_regs *regs) +{ +#ifdef CONFIG_KGDB + /* If remote debugging is configured AND enabled, we give control to + * kgdb. Otherwise, we fall through, perhaps giving control to the + * native debugger. + */ + + if (gdb_enter) { + extern void gdb_handle_exception(struct pt_regs *); + gdb_handle_exception(regs); + return_from_debug_flag = 1; + return; + } +#endif + + __die_if_kernel("Breakpoint in kernel", regs, SIGKILL); + + /* If in user mode, send SIGTRAP signal to current process */ + + force_sig(SIGTRAP, current); +} + + +/* + * Initialize dispatch tables. + * + * The exception vectors are stored compressed the __init section in the + * dispatch_init_table. This function initializes the following three tables + * from that compressed table: + * - fast user first dispatch table for user exceptions + * - fast kernel first dispatch table for kernel exceptions + * - default C-handler C-handler called by the default fast handler. + * + * See vectors.S for more details. + */ + +#define set_handler(idx,handler) (exc_table[idx] = (unsigned long) (handler)) + +void trap_init(void) +{ + int i; + + /* Setup default vectors. */ + + for(i = 0; i < 64; i++) { + set_handler(EXC_TABLE_FAST_USER/4 + i, user_exception); + set_handler(EXC_TABLE_FAST_KERNEL/4 + i, kernel_exception); + set_handler(EXC_TABLE_DEFAULT/4 + i, do_unhandled); + } + + /* Setup specific handlers. */ + + for(i = 0; dispatch_init_table[i].cause >= 0; i++) { + + int fast = dispatch_init_table[i].fast; + int cause = dispatch_init_table[i].cause; + void *handler = dispatch_init_table[i].handler; + + if (fast == 0) + set_handler (EXC_TABLE_DEFAULT/4 + cause, handler); + if (fast && fast & USER) + set_handler (EXC_TABLE_FAST_USER/4 + cause, handler); + if (fast && fast & KRNL) + set_handler (EXC_TABLE_FAST_KERNEL/4 + cause, handler); + } + + /* Initialize EXCSAVE_1 to hold the address of the exception table. */ + + i = (unsigned long)exc_table; + __asm__ __volatile__("wsr %0, "__stringify(EXCSAVE_1)"\n" : : "a" (i)); +} + +/* + * This function dumps the current valid window frame and other base registers. + */ + +void show_regs(struct pt_regs * regs) +{ + int i, wmask; + + wmask = regs->wmask & ~1; + + for (i = 0; i < 32; i++) { + if (wmask & (1 << (i / 4))) + break; + if ((i % 8) == 0) + printk ("\n" KERN_INFO "a%02d: ", i); + printk("%08lx ", regs->areg[i]); + } + printk("\n"); + + printk("pc: %08lx, ps: %08lx, depc: %08lx, excvaddr: %08lx\n", + regs->pc, regs->ps, regs->depc, regs->excvaddr); + printk("lbeg: %08lx, lend: %08lx lcount: %08lx, sar: %08lx\n", + regs->lbeg, regs->lend, regs->lcount, regs->sar); + if (user_mode(regs)) + printk("wb: %08lx, ws: %08lx, wmask: %08lx, syscall: %ld\n", + regs->windowbase, regs->windowstart, regs->wmask, + regs->syscall); +} + +void show_trace(struct task_struct *task, unsigned long *sp) +{ + unsigned long a0, a1, pc; + unsigned long sp_start, sp_end; + + a1 = (unsigned long)sp; + + if (a1 == 0) + __asm__ __volatile__ ("mov %0, a1\n" : "=a"(a1)); + + + sp_start = a1 & ~(THREAD_SIZE-1); + sp_end = sp_start + THREAD_SIZE; + + printk("Call Trace:"); +#ifdef CONFIG_KALLSYMS + printk("\n"); +#endif + spill_registers(); + + while (a1 > sp_start && a1 < sp_end) { + sp = (unsigned long*)a1; + + a0 = *(sp - 4); + a1 = *(sp - 3); + + if (a1 <= (unsigned long) sp) + break; + + pc = MAKE_PC_FROM_RA(a0, a1); + + if (kernel_text_address(pc)) { + printk(" [<%08lx>] ", pc); + print_symbol("%s\n", pc); + } + } + printk("\n"); +} + +/* + * This routine abuses get_user()/put_user() to reference pointers + * with at least a bit of error checking ... + */ + +static int kstack_depth_to_print = 24; + +void show_stack(struct task_struct *task, unsigned long *sp) +{ + int i = 0; + unsigned long *stack; + + if (sp == 0) + __asm__ __volatile__ ("mov %0, a1\n" : "=a"(sp)); + + stack = sp; + + printk("\nStack: "); + + for (i = 0; i < kstack_depth_to_print; i++) { + if (kstack_end(sp)) + break; + if (i && ((i % 8) == 0)) + printk("\n "); + printk("%08lx ", *sp++); + } + printk("\n"); + show_trace(task, stack); +} + +void dump_stack(void) +{ + show_stack(current, NULL); +} + +EXPORT_SYMBOL(dump_stack); + + +void show_code(unsigned int *pc) +{ + long i; + + printk("\nCode:"); + + for(i = -3 ; i < 6 ; i++) { + unsigned long insn; + if (__get_user(insn, pc + i)) { + printk(" (Bad address in pc)\n"); + break; + } + printk("%c%08lx%c",(i?' ':'<'),insn,(i?' ':'>')); + } +} + +spinlock_t die_lock = SPIN_LOCK_UNLOCKED; + +void die(const char * str, struct pt_regs * regs, long err) +{ + static int die_counter; + int nl = 0; + + console_verbose(); + spin_lock_irq(&die_lock); + + printk("%s: sig: %ld [#%d]\n", str, err, ++die_counter); +#ifdef CONFIG_PREEMPT + printk("PREEMPT "); + nl = 1; +#endif + if (nl) + printk("\n"); + show_regs(regs); + if (!user_mode(regs)) + show_stack(NULL, (unsigned long*)regs->areg[1]); + + spin_unlock_irq(&die_lock); + + if (in_interrupt()) + panic("Fatal exception in interrupt"); + + if (panic_on_oops) { + printk(KERN_EMERG "Fatal exception: panic in 5 seconds\n"); + set_current_state(TASK_UNINTERRUPTIBLE); + schedule_timeout(5 * HZ); + panic("Fatal exception"); + } + do_exit(err); +} + + diff --git a/arch/xtensa/kernel/vectors.S b/arch/xtensa/kernel/vectors.S new file mode 100644 index 00000000000..81808f0c674 --- /dev/null +++ b/arch/xtensa/kernel/vectors.S @@ -0,0 +1,464 @@ +/* + * arch/xtensa/kernel/vectors.S + * + * This file contains all exception vectors (user, kernel, and double), + * as well as the window vectors (overflow and underflow), and the debug + * vector. These are the primary vectors executed by the processor if an + * exception occurs. + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Copyright (C) 2005 Tensilica, Inc. + * + * Chris Zankel <chris@zankel.net> + * + */ + +/* + * We use a two-level table approach. The user and kernel exception vectors + * use a first-level dispatch table to dispatch the exception to a registered + * fast handler or the default handler, if no fast handler was registered. + * The default handler sets up a C-stack and dispatches the exception to a + * registerd C handler in the second-level dispatch table. + * + * Fast handler entry condition: + * + * a0: trashed, original value saved on stack (PT_AREG0) + * a1: a1 + * a2: new stack pointer, original value in depc + * a3: dispatch table + * depc: a2, original value saved on stack (PT_DEPC) + * excsave_1: a3 + * + * The value for PT_DEPC saved to stack also functions as a boolean to + * indicate that the exception is either a double or a regular exception: + * + * PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception + * < VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception + * + * Note: Neither the kernel nor the user exception handler generate literals. + * + */ + +#include <linux/linkage.h> +#include <asm/ptrace.h> +#include <asm/ptrace.h> +#include <asm/current.h> +#include <asm/offsets.h> +#include <asm/pgtable.h> +#include <asm/processor.h> +#include <asm/page.h> +#include <asm/thread_info.h> +#include <asm/processor.h> + + +/* + * User exception vector. (Exceptions with PS.UM == 1, PS.EXCM == 0) + * + * We get here when an exception occurred while we were in userland. + * We switch to the kernel stack and jump to the first level handler + * associated to the exception cause. + * + * Note: the saved kernel stack pointer (EXC_TABLE_KSTK) is already + * decremented by PT_USER_SIZE. + */ + + .section .UserExceptionVector.text, "ax" + +ENTRY(_UserExceptionVector) + + xsr a3, EXCSAVE_1 # save a3 and get dispatch table + wsr a2, DEPC # save a2 + l32i a2, a3, EXC_TABLE_KSTK # load kernel stack to a2 + s32i a0, a2, PT_AREG0 # save a0 to ESF + rsr a0, EXCCAUSE # retrieve exception cause + s32i a0, a2, PT_DEPC # mark it as a regular exception + addx4 a0, a0, a3 # find entry in table + l32i a0, a0, EXC_TABLE_FAST_USER # load handler + jx a0 + +/* + * Kernel exception vector. (Exceptions with PS.UM == 0, PS.EXCM == 0) + * + * We get this exception when we were already in kernel space. + * We decrement the current stack pointer (kernel) by PT_SIZE and + * jump to the first-level handler associated with the exception cause. + * + * Note: we need to preserve space for the spill region. + */ + + .section .KernelExceptionVector.text, "ax" + +ENTRY(_KernelExceptionVector) + + xsr a3, EXCSAVE_1 # save a3, and get dispatch table + wsr a2, DEPC # save a2 + addi a2, a1, -16-PT_SIZE # adjust stack pointer + s32i a0, a2, PT_AREG0 # save a0 to ESF + rsr a0, EXCCAUSE # retrieve exception cause + s32i a0, a2, PT_DEPC # mark it as a regular exception + addx4 a0, a0, a3 # find entry in table + l32i a0, a0, EXC_TABLE_FAST_KERNEL # load handler address + jx a0 + + +/* + * Double exception vector (Exceptions with PS.EXCM == 1) + * We get this exception when another exception occurs while were are + * already in an exception, such as window overflow/underflow exception, + * or 'expected' exceptions, for example memory exception when we were trying + * to read data from an invalid address in user space. + * + * Note that this vector is never invoked for level-1 interrupts, because such + * interrupts are disabled (masked) when PS.EXCM is set. + * + * We decode the exception and take the appropriate action. However, the + * double exception vector is much more careful, because a lot more error + * cases go through the double exception vector than through the user and + * kernel exception vectors. + * + * Occasionally, the kernel expects a double exception to occur. This usually + * happens when accessing user-space memory with the user's permissions + * (l32e/s32e instructions). The kernel state, though, is not always suitable + * for immediate transfer of control to handle_double, where "normal" exception + * processing occurs. Also in kernel mode, TLB misses can occur if accessing + * vmalloc memory, possibly requiring repair in a double exception handler. + * + * The variable at TABLE_FIXUP offset from the pointer in EXCSAVE_1 doubles as + * a boolean variable and a pointer to a fixup routine. If the variable + * EXC_TABLE_FIXUP is non-zero, this handler jumps to that address. A value of + * zero indicates to use the default kernel/user exception handler. + * There is only one exception, when the value is identical to the exc_table + * label, the kernel is in trouble. This mechanism is used to protect critical + * sections, mainly when the handler writes to the stack to assert the stack + * pointer is valid. Once the fixup/default handler leaves that area, the + * EXC_TABLE_FIXUP variable is reset to the fixup handler or zero. + * + * Procedures wishing to use this mechanism should set EXC_TABLE_FIXUP to the + * nonzero address of a fixup routine before it could cause a double exception + * and reset it before it returns. + * + * Some other things to take care of when a fast exception handler doesn't + * specify a particular fixup handler but wants to use the default handlers: + * + * - The original stack pointer (in a1) must not be modified. The fast + * exception handler should only use a2 as the stack pointer. + * + * - If the fast handler manipulates the stack pointer (in a2), it has to + * register a valid fixup handler and cannot use the default handlers. + * + * - The handler can use any other generic register from a3 to a15, but it + * must save the content of these registers to stack (PT_AREG3...PT_AREGx) + * + * - These registers must be saved before a double exception can occur. + * + * - If we ever implement handling signals while in double exceptions, the + * number of registers a fast handler has saved (excluding a0 and a1) must + * be written to PT_AREG1. (1 if only a3 is used, 2 for a3 and a4, etc. ) + * + * The fixup handlers are special handlers: + * + * - Fixup entry conditions differ from regular exceptions: + * + * a0: DEPC + * a1: a1 + * a2: trashed, original value in EXC_TABLE_DOUBLE_A2 + * a3: exctable + * depc: a0 + * excsave_1: a3 + * + * - When the kernel enters the fixup handler, it still assumes it is in a + * critical section, so EXC_TABLE_FIXUP variable is set to exc_table. + * The fixup handler, therefore, has to re-register itself as the fixup + * handler before it returns from the double exception. + * + * - Fixup handler can share the same exception frame with the fast handler. + * The kernel stack pointer is not changed when entering the fixup handler. + * + * - Fixup handlers can jump to the default kernel and user exception + * handlers. Before it jumps, though, it has to setup a exception frame + * on stack. Because the default handler resets the register fixup handler + * the fixup handler must make sure that the default handler returns to + * it instead of the exception address, so it can re-register itself as + * the fixup handler. + * + * In case of a critical condition where the kernel cannot recover, we jump + * to unrecoverable_exception with the following entry conditions. + * All registers a0...a15 are unchanged from the last exception, except: + * + * a0: last address before we jumped to the unrecoverable_exception. + * excsave_1: a0 + * + * + * See the handle_alloca_user and spill_registers routines for example clients. + * + * FIXME: Note: we currently don't allow signal handling coming from a double + * exception, so the item markt with (*) is not required. + */ + + .section .DoubleExceptionVector.text, "ax" + .begin literal_prefix .DoubleExceptionVector + +ENTRY(_DoubleExceptionVector) + + /* Deliberately destroy excsave (don't assume it's value was valid). */ + + wsr a3, EXCSAVE_1 # save a3 + + /* Check for kernel double exception (usually fatal). */ + + rsr a3, PS + _bbci.l a3, PS_UM_SHIFT, .Lksp + + /* Check if we are currently handling a window exception. */ + /* Note: We don't need to indicate that we enter a critical section. */ + + xsr a0, DEPC # get DEPC, save a0 + + movi a3, XCHAL_WINDOW_VECTORS_VADDR + _bltu a0, a3, .Lfixup + addi a3, a3, XSHAL_WINDOW_VECTORS_SIZE + _bgeu a0, a3, .Lfixup + + /* Window overflow/underflow exception. Get stack pointer. */ + + mov a3, a2 + movi a2, exc_table + l32i a2, a2, EXC_TABLE_KSTK + + /* Check for overflow/underflow exception, jump if overflow. */ + + _bbci.l a0, 6, .Lovfl + + /* a0: depc, a1: a1, a2: kstk, a3: a2, depc: a0, excsave: a3 */ + + /* Restart window underflow exception. + * We return to the instruction in user space that caused the window + * underflow exception. Therefore, we change window base to the value + * before we entered the window underflow exception and prepare the + * registers to return as if we were coming from a regular exception + * by changing depc (in a0). + * Note: We can trash the current window frame (a0...a3) and depc! + */ + + wsr a2, DEPC # save stack pointer temporarily + rsr a0, PS + extui a0, a0, XCHAL_PS_OWB_SHIFT, XCHAL_PS_OWB_BITS + wsr a0, WINDOWBASE + rsync + + /* We are now in the previous window frame. Save registers again. */ + + xsr a2, DEPC # save a2 and get stack pointer + s32i a0, a2, PT_AREG0 + + wsr a3, EXCSAVE_1 # save a3 + movi a3, exc_table + + rsr a0, EXCCAUSE + s32i a0, a2, PT_DEPC # mark it as a regular exception + addx4 a0, a0, a3 + l32i a0, a0, EXC_TABLE_FAST_USER + jx a0 + +.Lfixup:/* Check for a fixup handler or if we were in a critical section. */ + + /* a0: depc, a1: a1, a2: a2, a3: trashed, depc: a0, excsave1: a3 */ + + movi a3, exc_table + s32i a2, a3, EXC_TABLE_DOUBLE_SAVE # temporary variable + + /* Enter critical section. */ + + l32i a2, a3, EXC_TABLE_FIXUP + s32i a3, a3, EXC_TABLE_FIXUP + beq a2, a3, .Lunrecoverable_fixup # critical! + beqz a2, .Ldflt # no handler was registered + + /* a0: depc, a1: a1, a2: trash, a3: exctable, depc: a0, excsave: a3 */ + + jx a2 + +.Ldflt: /* Get stack pointer. */ + + l32i a3, a3, EXC_TABLE_DOUBLE_SAVE + addi a2, a3, -PT_USER_SIZE + +.Lovfl: /* Jump to default handlers. */ + + /* a0: depc, a1: a1, a2: kstk, a3: a2, depc: a0, excsave: a3 */ + + xsr a3, DEPC + s32i a0, a2, PT_DEPC + s32i a3, a2, PT_AREG0 + + /* a0: avail, a1: a1, a2: kstk, a3: avail, depc: a2, excsave: a3 */ + + movi a3, exc_table + rsr a0, EXCCAUSE + addx4 a0, a0, a3 + l32i a0, a0, EXC_TABLE_FAST_USER + jx a0 + + /* + * We only allow the ITLB miss exception if we are in kernel space. + * All other exceptions are unexpected and thus unrecoverable! + */ + + .extern fast_second_level_miss_double_kernel + +.Lksp: /* a0: a0, a1: a1, a2: a2, a3: trashed, depc: depc, excsave: a3 */ + + rsr a3, EXCCAUSE + beqi a3, XCHAL_EXCCAUSE_ITLB_MISS, 1f + addi a3, a3, -XCHAL_EXCCAUSE_DTLB_MISS + bnez a3, .Lunrecoverable +1: movi a3, fast_second_level_miss_double_kernel + jx a3 + + /* Critical! We can't handle this situation. PANIC! */ + + .extern unrecoverable_exception + +.Lunrecoverable_fixup: + l32i a2, a3, EXC_TABLE_DOUBLE_SAVE + xsr a0, DEPC + +.Lunrecoverable: + rsr a3, EXCSAVE_1 + wsr a0, EXCSAVE_1 + movi a0, unrecoverable_exception + callx0 a0 + + .end literal_prefix + + +/* + * Debug interrupt vector + * + * There is not much space here, so simply jump to another handler. + * EXCSAVE[DEBUGLEVEL] has been set to that handler. + */ + + .section .DebugInterruptVector.text, "ax" + +ENTRY(_DebugInterruptVector) + xsr a0, EXCSAVE + XCHAL_DEBUGLEVEL + jx a0 + + + +/* Window overflow and underflow handlers. + * The handlers must be 64 bytes apart, first starting with the underflow + * handlers underflow-4 to underflow-12, then the overflow handlers + * overflow-4 to overflow-12. + * + * Note: We rerun the underflow handlers if we hit an exception, so + * we try to access any page that would cause a page fault early. + */ + + .section .WindowVectors.text, "ax" + + +/* 4-Register Window Overflow Vector (Handler) */ + + .align 64 +.global _WindowOverflow4 +_WindowOverflow4: + s32e a0, a5, -16 + s32e a1, a5, -12 + s32e a2, a5, -8 + s32e a3, a5, -4 + rfwo + + +/* 4-Register Window Underflow Vector (Handler) */ + + .align 64 +.global _WindowUnderflow4 +_WindowUnderflow4: + l32e a0, a5, -16 + l32e a1, a5, -12 + l32e a2, a5, -8 + l32e a3, a5, -4 + rfwu + + +/* 8-Register Window Overflow Vector (Handler) */ + + .align 64 +.global _WindowOverflow8 +_WindowOverflow8: + s32e a0, a9, -16 + l32e a0, a1, -12 + s32e a2, a9, -8 + s32e a1, a9, -12 + s32e a3, a9, -4 + s32e a4, a0, -32 + s32e a5, a0, -28 + s32e a6, a0, -24 + s32e a7, a0, -20 + rfwo + +/* 8-Register Window Underflow Vector (Handler) */ + + .align 64 +.global _WindowUnderflow8 +_WindowUnderflow8: + l32e a1, a9, -12 + l32e a0, a9, -16 + l32e a7, a1, -12 + l32e a2, a9, -8 + l32e a4, a7, -32 + l32e a3, a9, -4 + l32e a5, a7, -28 + l32e a6, a7, -24 + l32e a7, a7, -20 + rfwu + + +/* 12-Register Window Overflow Vector (Handler) */ + + .align 64 +.global _WindowOverflow12 +_WindowOverflow12: + s32e a0, a13, -16 + l32e a0, a1, -12 + s32e a1, a13, -12 + s32e a2, a13, -8 + s32e a3, a13, -4 + s32e a4, a0, -48 + s32e a5, a0, -44 + s32e a6, a0, -40 + s32e a7, a0, -36 + s32e a8, a0, -32 + s32e a9, a0, -28 + s32e a10, a0, -24 + s32e a11, a0, -20 + rfwo + +/* 12-Register Window Underflow Vector (Handler) */ + + .align 64 +.global _WindowUnderflow12 +_WindowUnderflow12: + l32e a1, a13, -12 + l32e a0, a13, -16 + l32e a11, a1, -12 + l32e a2, a13, -8 + l32e a4, a11, -48 + l32e a8, a11, -32 + l32e a3, a13, -4 + l32e a5, a11, -44 + l32e a6, a11, -40 + l32e a7, a11, -36 + l32e a9, a11, -28 + l32e a10, a11, -24 + l32e a11, a11, -20 + rfwu + + .text + + diff --git a/arch/xtensa/kernel/vmlinux.lds.S b/arch/xtensa/kernel/vmlinux.lds.S new file mode 100644 index 00000000000..476b2b53cd0 --- /dev/null +++ b/arch/xtensa/kernel/vmlinux.lds.S @@ -0,0 +1,341 @@ +/* + * arch/xtensa/kernel/vmlinux.lds.S + * + * Xtensa linker script + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca> + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + */ + +#include <asm-generic/vmlinux.lds.h> + +#include <linux/config.h> +#define _NOCLANGUAGE +#include <xtensa/config/core.h> +#include <xtensa/config/system.h> +OUTPUT_ARCH(xtensa) +ENTRY(_start) + +#if XCHAL_MEMORY_ORDER == XTHAL_BIGENDIAN +jiffies = jiffies_64 + 4; +#else +jiffies = jiffies_64; +#endif + +#define KERNELOFFSET 0x1000 + +/* Note: In the following macros, it would be nice to specify only the + vector name and section kind and construct "sym" and "section" using + CPP concatenation, but that does not work reliably. Concatenating a + string with "." produces an invalid token. CPP will not print a + warning because it thinks this is an assembly file, but it leaves + them as multiple tokens and there may or may not be whitespace + between them. */ + +/* Macro for a relocation entry */ + +#define RELOCATE_ENTRY(sym, section) \ + LONG(sym ## _start); \ + LONG(sym ## _end); \ + LONG(LOADADDR(section)) + +/* Macro to define a section for a vector. + * + * Use of the MIN function catches the types of errors illustrated in + * the following example: + * + * Assume the section .DoubleExceptionVector.literal is completely + * full. Then a programmer adds code to .DoubleExceptionVector.text + * that produces another literal. The final literal position will + * overlay onto the first word of the adjacent code section + * .DoubleExceptionVector.text. (In practice, the literals will + * overwrite the code, and the first few instructions will be + * garbage.) + */ + +#define SECTION_VECTOR(sym, section, addr, max_prevsec_size, prevsec) \ + section addr : AT((MIN(LOADADDR(prevsec) + max_prevsec_size, \ + LOADADDR(prevsec) + SIZEOF(prevsec)) + 3) & ~ 3) \ + { \ + . = ALIGN(4); \ + sym ## _start = ABSOLUTE(.); \ + *(section) \ + sym ## _end = ABSOLUTE(.); \ + } + +/* + * Mapping of input sections to output sections when linking. + */ + +SECTIONS +{ + . = XCHAL_KSEG_CACHED_VADDR + KERNELOFFSET; + /* .text section */ + + _text = .; + _stext = .; + _ftext = .; + + .text : + { + /* The .head.text section must be the first section! */ + *(.head.text) + *(.literal .text) + *(.srom.text) + VMLINUX_SYMBOL(__sched_text_start) = .; + *(.sched.text.literal .sched.text) + VMLINUX_SYMBOL(__sched_text_end) = .; + VMLINUX_SYMBOL(__lock_text_start) = .; + *(.spinlock.text.literal .spinlock.text) + VMLINUX_SYMBOL(__lock_text_end) = .; + + } + _etext = .; + + . = ALIGN(16); + + RODATA + + /* Relocation table */ + + . = ALIGN(16); + __boot_reloc_table_start = ABSOLUTE(.); + + __relocate : { + + RELOCATE_ENTRY(_WindowVectors_text, + .WindowVectors.text); +#if 0 + RELOCATE_ENTRY(_KernelExceptionVector_literal, + .KernelExceptionVector.literal); +#endif + RELOCATE_ENTRY(_KernelExceptionVector_text, + .KernelExceptionVector.text); +#if 0 + RELOCATE_ENTRY(_UserExceptionVector_literal, + .UserExceptionVector.literal); +#endif + RELOCATE_ENTRY(_UserExceptionVector_text, + .UserExceptionVector.text); + RELOCATE_ENTRY(_DoubleExceptionVector_literal, + .DoubleExceptionVector.literal); + RELOCATE_ENTRY(_DoubleExceptionVector_text, + .DoubleExceptionVector.text); + } + __boot_reloc_table_end = ABSOLUTE(.) ; + + .fixup : { *(.fixup) } + + . = ALIGN(16); + + __ex_table : { + __start___ex_table = .; + *(__ex_table) + __stop___ex_table = .; + } + + /* Data section */ + + . = ALIGN(XCHAL_ICACHE_LINESIZE); + _fdata = .; + .data : + { + *(.data) CONSTRUCTORS + . = ALIGN(XCHAL_ICACHE_LINESIZE); + *(.data.cacheline_aligned) + } + + _edata = .; + + /* The initial task */ + . = ALIGN(8192); + .data.init_task : { *(.data.init_task) } + + /* Initialization code and data: */ + + . = ALIGN(1<<XCHAL_MMU_MIN_PTE_PAGE_SIZE); + __init_begin = .; + .init.text : { + _sinittext = .; + *(.init.text.literal) *(.init.text) + _einittext = .; + } + + .init.data : + { + *(.init.data) + . = ALIGN(0x4); + __tagtable_begin = .; + *(.taglist) + __tagtable_end = .; + } + + . = ALIGN(XCHAL_ICACHE_LINESIZE); + + __setup_start = .; + .init.setup : { *(.init.setup) } + __setup_end = .; + + __initcall_start = .; + .initcall.init : { + *(.initcall1.init) + *(.initcall2.init) + *(.initcall3.init) + *(.initcall4.init) + *(.initcall5.init) + *(.initcall6.init) + *(.initcall7.init) + } + __initcall_end = .; + + __con_initcall_start = .; + .con_initcall.init : { *(.con_initcall.init) } + __con_initcall_end = .; + + SECURITY_INIT + + . = ALIGN(4); + + __start___ftr_fixup = .; + __ftr_fixup : { *(__ftr_fixup) } + __stop___ftr_fixup = .; + + . = ALIGN(32); + __per_cpu_start = .; + .data.percpu : { *(.data.percpu) } + __per_cpu_end = .; + + . = ALIGN(4096); + __initramfs_start =.; + .init.ramfs : { *(.init.ramfs) } + __initramfs_end = .; + + /* We need this dummy segment here */ + + . = ALIGN(4); + .dummy : { LONG(0) } + + /* The vectors are relocated to the real position at startup time */ + + SECTION_VECTOR (_WindowVectors_text, + .WindowVectors.text, + XCHAL_WINDOW_VECTORS_VADDR, 4, + .dummy) + SECTION_VECTOR (_DebugInterruptVector_literal, + .DebugInterruptVector.literal, + XCHAL_INTLEVEL_VECTOR_VADDR(XCHAL_DEBUGLEVEL) - 4, + SIZEOF(.WindowVectors.text), + .WindowVectors.text) + SECTION_VECTOR (_DebugInterruptVector_text, + .DebugInterruptVector.text, + XCHAL_INTLEVEL_VECTOR_VADDR(XCHAL_DEBUGLEVEL), + 4, + .DebugInterruptVector.literal) + SECTION_VECTOR (_KernelExceptionVector_literal, + .KernelExceptionVector.literal, + XCHAL_KERNELEXC_VECTOR_VADDR - 4, + SIZEOF(.DebugInterruptVector.text), + .DebugInterruptVector.text) + SECTION_VECTOR (_KernelExceptionVector_text, + .KernelExceptionVector.text, + XCHAL_KERNELEXC_VECTOR_VADDR, + 4, + .KernelExceptionVector.literal) + SECTION_VECTOR (_UserExceptionVector_literal, + .UserExceptionVector.literal, + XCHAL_USEREXC_VECTOR_VADDR - 4, + SIZEOF(.KernelExceptionVector.text), + .KernelExceptionVector.text) + SECTION_VECTOR (_UserExceptionVector_text, + .UserExceptionVector.text, + XCHAL_USEREXC_VECTOR_VADDR, + 4, + .UserExceptionVector.literal) + SECTION_VECTOR (_DoubleExceptionVector_literal, + .DoubleExceptionVector.literal, + XCHAL_DOUBLEEXC_VECTOR_VADDR - 16, + SIZEOF(.UserExceptionVector.text), + .UserExceptionVector.text) + SECTION_VECTOR (_DoubleExceptionVector_text, + .DoubleExceptionVector.text, + XCHAL_DOUBLEEXC_VECTOR_VADDR, + 32, + .DoubleExceptionVector.literal) + + . = (LOADADDR( .DoubleExceptionVector.text ) + SIZEOF( .DoubleExceptionVector.text ) + 3) & ~ 3; + . = ALIGN(1<<XCHAL_MMU_MIN_PTE_PAGE_SIZE); + + __init_end = .; + + . = ALIGN(8192); + + /* BSS section */ + _bss_start = .; + .sbss : { *(.sbss) *(.scommon) } + .bss : { *(COMMON) *(.bss) } + _bss_end = .; + _end = .; + + /* only used by the boot loader */ + + . = ALIGN(0x10); + .bootstrap : { *(.bootstrap.literal .bootstrap.text .bootstrap.data) } + + . = ALIGN(0x1000); + __initrd_start = .; + .initrd : { *(.initrd) } + __initrd_end = .; + + .ResetVector.text XCHAL_RESET_VECTOR_VADDR : + { + *(.ResetVector.text) + } + + + /* Sections to be discarded */ + /DISCARD/ : + { + *(.text.exit) + *(.text.exit.literal) + *(.data.exit) + *(.exitcall.exit) + } + + + .debug 0 : { *(.debug) } + .line 0 : { *(.line) } + .debug_srcinfo 0 : { *(.debug_srcinfo) } + .debug_sfnames 0 : { *(.debug_sfnames) } + .debug_aranges 0 : { *(.debug_aranges) } + .debug_pubnames 0 : { *(.debug_pubnames) } + .debug_info 0 : { *(.debug_info) } + .debug_abbrev 0 : { *(.debug_abbrev) } + .debug_line 0 : { *(.debug_line) } + .debug_frame 0 : { *(.debug_frame) } + .debug_str 0 : { *(.debug_str) } + .debug_loc 0 : { *(.debug_loc) } + .debug_macinfo 0 : { *(.debug_macinfo) } + .debug_weaknames 0 : { *(.debug_weaknames) } + .debug_funcnames 0 : { *(.debug_funcnames) } + .debug_typenames 0 : { *(.debug_typenames) } + .debug_varnames 0 : { *(.debug_varnames) } + + .xt.insn 0 : + { + *(.xt.insn) + *(.gnu.linkonce.x*) + } + + .xt.lit 0 : + { + *(.xt.lit) + *(.gnu.linkonce.p*) + } +} diff --git a/arch/xtensa/kernel/xtensa_ksyms.c b/arch/xtensa/kernel/xtensa_ksyms.c new file mode 100644 index 00000000000..efae56a5147 --- /dev/null +++ b/arch/xtensa/kernel/xtensa_ksyms.c @@ -0,0 +1,123 @@ +/* + * arch/xtensa/kernel/xtensa_ksyms.c + * + * Export Xtensa-specific functions for loadable modules. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Joe Taylor <joe@tensilica.com> + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/string.h> +#include <linux/mm.h> +#include <linux/interrupt.h> +#include <asm/irq.h> +#include <linux/in6.h> +#include <linux/pci.h> +#include <linux/ide.h> + +#include <asm/uaccess.h> +#include <asm/checksum.h> +#include <asm/dma.h> +#include <asm/io.h> +#include <asm/page.h> +#include <asm/pgalloc.h> +#include <asm/semaphore.h> +#ifdef CONFIG_BLK_DEV_FD +#include <asm/floppy.h> +#endif +#ifdef CONFIG_NET +#include <net/checksum.h> +#endif /* CONFIG_NET */ + + +/* + * String functions + */ +EXPORT_SYMBOL(memcmp); +EXPORT_SYMBOL(memset); +EXPORT_SYMBOL(memcpy); +EXPORT_SYMBOL(memmove); +EXPORT_SYMBOL(memchr); +EXPORT_SYMBOL(strcat); +EXPORT_SYMBOL(strchr); +EXPORT_SYMBOL(strlen); +EXPORT_SYMBOL(strpbrk); +EXPORT_SYMBOL(strncat); +EXPORT_SYMBOL(strnlen); +EXPORT_SYMBOL(strrchr); +EXPORT_SYMBOL(strstr); + +EXPORT_SYMBOL(enable_irq); +EXPORT_SYMBOL(disable_irq); +EXPORT_SYMBOL(kernel_thread); + +/* + * gcc internal math functions + */ +extern long long __ashrdi3(long long, int); +extern long long __ashldi3(long long, int); +extern long long __lshrdi3(long long, int); +extern int __divsi3(int, int); +extern int __modsi3(int, int); +extern long long __muldi3(long long, long long); +extern int __mulsi3(int, int); +extern unsigned int __udivsi3(unsigned int, unsigned int); +extern unsigned int __umodsi3(unsigned int, unsigned int); +extern unsigned long long __umoddi3(unsigned long long, unsigned long long); +extern unsigned long long __udivdi3(unsigned long long, unsigned long long); + +EXPORT_SYMBOL(__ashldi3); +EXPORT_SYMBOL(__ashrdi3); +EXPORT_SYMBOL(__lshrdi3); +EXPORT_SYMBOL(__divsi3); +EXPORT_SYMBOL(__modsi3); +EXPORT_SYMBOL(__muldi3); +EXPORT_SYMBOL(__mulsi3); +EXPORT_SYMBOL(__udivsi3); +EXPORT_SYMBOL(__umodsi3); +EXPORT_SYMBOL(__udivdi3); +EXPORT_SYMBOL(__umoddi3); + +/* + * Semaphore operations + */ +EXPORT_SYMBOL(__down); +EXPORT_SYMBOL(__down_interruptible); +EXPORT_SYMBOL(__down_trylock); +EXPORT_SYMBOL(__up); + +#ifdef CONFIG_NET +/* + * Networking support + */ +EXPORT_SYMBOL(csum_partial_copy_generic); +#endif /* CONFIG_NET */ + +/* + * Architecture-specific symbols + */ +EXPORT_SYMBOL(__xtensa_copy_user); + +/* + * Kernel hacking ... + */ + +#if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_DUMMY_CONSOLE) +// FIXME EXPORT_SYMBOL(screen_info); +#endif + +EXPORT_SYMBOL(get_wchan); + +EXPORT_SYMBOL(outsb); +EXPORT_SYMBOL(outsw); +EXPORT_SYMBOL(outsl); +EXPORT_SYMBOL(insb); +EXPORT_SYMBOL(insw); +EXPORT_SYMBOL(insl); diff --git a/arch/xtensa/lib/Makefile b/arch/xtensa/lib/Makefile new file mode 100644 index 00000000000..ed935b58e8a --- /dev/null +++ b/arch/xtensa/lib/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for Xtensa-specific library files. +# + +lib-y += memcopy.o memset.o checksum.o strcasecmp.o \ + usercopy.o strncpy_user.o strnlen_user.o +lib-$(CONFIG_PCI) += pci-auto.o diff --git a/arch/xtensa/lib/checksum.S b/arch/xtensa/lib/checksum.S new file mode 100644 index 00000000000..e2d64dfd530 --- /dev/null +++ b/arch/xtensa/lib/checksum.S @@ -0,0 +1,410 @@ +/* + * INET An implementation of the TCP/IP protocol suite for the LINUX + * operating system. INET is implemented using the BSD Socket + * interface as the means of communication with the user level. + * + * IP/TCP/UDP checksumming routines + * + * Xtensa version: Copyright (C) 2001 Tensilica, Inc. by Kevin Chea + * Optimized by Joe Taylor + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <asm/errno.h> +#include <linux/linkage.h> +#define _ASMLANGUAGE +#include <xtensa/config/core.h> + +/* + * computes a partial checksum, e.g. for TCP/UDP fragments + */ + +/* + * unsigned int csum_partial(const unsigned char *buf, int len, + * unsigned int sum); + * a2 = buf + * a3 = len + * a4 = sum + * + * This function assumes 2- or 4-byte alignment. Other alignments will fail! + */ + +/* ONES_ADD converts twos-complement math to ones-complement. */ +#define ONES_ADD(sum, val) \ + add sum, sum, val ; \ + bgeu sum, val, 99f ; \ + addi sum, sum, 1 ; \ +99: ; + +.text +ENTRY(csum_partial) + /* + * Experiments with Ethernet and SLIP connections show that buf + * is aligned on either a 2-byte or 4-byte boundary. + */ + entry sp, 32 + extui a5, a2, 0, 2 + bnez a5, 8f /* branch if 2-byte aligned */ + /* Fall-through on common case, 4-byte alignment */ +1: + srli a5, a3, 5 /* 32-byte chunks */ +#if XCHAL_HAVE_LOOPS + loopgtz a5, 2f +#else + beqz a5, 2f + slli a5, a5, 5 + add a5, a5, a2 /* a5 = end of last 32-byte chunk */ +.Loop1: +#endif + l32i a6, a2, 0 + l32i a7, a2, 4 + ONES_ADD(a4, a6) + ONES_ADD(a4, a7) + l32i a6, a2, 8 + l32i a7, a2, 12 + ONES_ADD(a4, a6) + ONES_ADD(a4, a7) + l32i a6, a2, 16 + l32i a7, a2, 20 + ONES_ADD(a4, a6) + ONES_ADD(a4, a7) + l32i a6, a2, 24 + l32i a7, a2, 28 + ONES_ADD(a4, a6) + ONES_ADD(a4, a7) + addi a2, a2, 4*8 +#if !XCHAL_HAVE_LOOPS + blt a2, a5, .Loop1 +#endif +2: + extui a5, a3, 2, 3 /* remaining 4-byte chunks */ +#if XCHAL_HAVE_LOOPS + loopgtz a5, 3f +#else + beqz a5, 3f + slli a5, a5, 2 + add a5, a5, a2 /* a5 = end of last 4-byte chunk */ +.Loop2: +#endif + l32i a6, a2, 0 + ONES_ADD(a4, a6) + addi a2, a2, 4 +#if !XCHAL_HAVE_LOOPS + blt a2, a5, .Loop2 +#endif +3: + _bbci.l a3, 1, 5f /* remaining 2-byte chunk */ + l16ui a6, a2, 0 + ONES_ADD(a4, a6) + addi a2, a2, 2 +5: + _bbci.l a3, 0, 7f /* remaining 1-byte chunk */ +6: l8ui a6, a2, 0 +#ifdef __XTENSA_EB__ + slli a6, a6, 8 /* load byte into bits 8..15 */ +#endif + ONES_ADD(a4, a6) +7: + mov a2, a4 + retw + + /* uncommon case, buf is 2-byte aligned */ +8: + beqz a3, 7b /* branch if len == 0 */ + beqi a3, 1, 6b /* branch if len == 1 */ + + extui a5, a2, 0, 1 + bnez a5, 8f /* branch if 1-byte aligned */ + + l16ui a6, a2, 0 /* common case, len >= 2 */ + ONES_ADD(a4, a6) + addi a2, a2, 2 /* adjust buf */ + addi a3, a3, -2 /* adjust len */ + j 1b /* now buf is 4-byte aligned */ + + /* case: odd-byte aligned, len > 1 + * This case is dog slow, so don't give us an odd address. + * (I don't think this ever happens, but just in case.) + */ +8: + srli a5, a3, 2 /* 4-byte chunks */ +#if XCHAL_HAVE_LOOPS + loopgtz a5, 2f +#else + beqz a5, 2f + slli a5, a5, 2 + add a5, a5, a2 /* a5 = end of last 4-byte chunk */ +.Loop3: +#endif + l8ui a6, a2, 0 /* bits 24..31 */ + l16ui a7, a2, 1 /* bits 8..23 */ + l8ui a8, a2, 3 /* bits 0.. 8 */ +#ifdef __XTENSA_EB__ + slli a6, a6, 24 +#else + slli a8, a8, 24 +#endif + slli a7, a7, 8 + or a7, a7, a6 + or a7, a7, a8 + ONES_ADD(a4, a7) + addi a2, a2, 4 +#if !XCHAL_HAVE_LOOPS + blt a2, a5, .Loop3 +#endif +2: + _bbci.l a3, 1, 3f /* remaining 2-byte chunk, still odd addr */ + l8ui a6, a2, 0 + l8ui a7, a2, 1 +#ifdef __XTENSA_EB__ + slli a6, a6, 8 +#else + slli a7, a7, 8 +#endif + or a7, a7, a6 + ONES_ADD(a4, a7) + addi a2, a2, 2 +3: + j 5b /* branch to handle the remaining byte */ + + + +/* + * Copy from ds while checksumming, otherwise like csum_partial + * + * The macros SRC and DST specify the type of access for the instruction. + * thus we can call a custom exception handler for each access type. + */ + +#define SRC(y...) \ + 9999: y; \ + .section __ex_table, "a"; \ + .long 9999b, 6001f ; \ + .previous + +#define DST(y...) \ + 9999: y; \ + .section __ex_table, "a"; \ + .long 9999b, 6002f ; \ + .previous + +/* +unsigned int csum_partial_copy_generic (const char *src, char *dst, int len, + int sum, int *src_err_ptr, int *dst_err_ptr) + a2 = src + a3 = dst + a4 = len + a5 = sum + a6 = src_err_ptr + a7 = dst_err_ptr + a8 = temp + a9 = temp + a10 = temp + a11 = original len for exception handling + a12 = original dst for exception handling + + This function is optimized for 4-byte aligned addresses. Other + alignments work, but not nearly as efficiently. + */ + +ENTRY(csum_partial_copy_generic) + entry sp, 32 + mov a12, a3 + mov a11, a4 + or a10, a2, a3 + + /* We optimize the following alignment tests for the 4-byte + aligned case. Two bbsi.l instructions might seem more optimal + (commented out below). However, both labels 5: and 3: are out + of the imm8 range, so the assembler relaxes them into + equivalent bbci.l, j combinations, which is actually + slower. */ + + extui a9, a10, 0, 2 + beqz a9, 1f /* branch if both are 4-byte aligned */ + bbsi.l a10, 0, 5f /* branch if one address is odd */ + j 3f /* one address is 2-byte aligned */ + +/* _bbsi.l a10, 0, 5f */ /* branch if odd address */ +/* _bbsi.l a10, 1, 3f */ /* branch if 2-byte-aligned address */ + +1: + /* src and dst are both 4-byte aligned */ + srli a10, a4, 5 /* 32-byte chunks */ +#if XCHAL_HAVE_LOOPS + loopgtz a10, 2f +#else + beqz a10, 2f + slli a10, a10, 5 + add a10, a10, a2 /* a10 = end of last 32-byte src chunk */ +.Loop5: +#endif +SRC( l32i a9, a2, 0 ) +SRC( l32i a8, a2, 4 ) +DST( s32i a9, a3, 0 ) +DST( s32i a8, a3, 4 ) + ONES_ADD(a5, a9) + ONES_ADD(a5, a8) +SRC( l32i a9, a2, 8 ) +SRC( l32i a8, a2, 12 ) +DST( s32i a9, a3, 8 ) +DST( s32i a8, a3, 12 ) + ONES_ADD(a5, a9) + ONES_ADD(a5, a8) +SRC( l32i a9, a2, 16 ) +SRC( l32i a8, a2, 20 ) +DST( s32i a9, a3, 16 ) +DST( s32i a8, a3, 20 ) + ONES_ADD(a5, a9) + ONES_ADD(a5, a8) +SRC( l32i a9, a2, 24 ) +SRC( l32i a8, a2, 28 ) +DST( s32i a9, a3, 24 ) +DST( s32i a8, a3, 28 ) + ONES_ADD(a5, a9) + ONES_ADD(a5, a8) + addi a2, a2, 32 + addi a3, a3, 32 +#if !XCHAL_HAVE_LOOPS + blt a2, a10, .Loop5 +#endif +2: + extui a10, a4, 2, 3 /* remaining 4-byte chunks */ + extui a4, a4, 0, 2 /* reset len for general-case, 2-byte chunks */ +#if XCHAL_HAVE_LOOPS + loopgtz a10, 3f +#else + beqz a10, 3f + slli a10, a10, 2 + add a10, a10, a2 /* a10 = end of last 4-byte src chunk */ +.Loop6: +#endif +SRC( l32i a9, a2, 0 ) +DST( s32i a9, a3, 0 ) + ONES_ADD(a5, a9) + addi a2, a2, 4 + addi a3, a3, 4 +#if !XCHAL_HAVE_LOOPS + blt a2, a10, .Loop6 +#endif +3: + /* + Control comes to here in two cases: (1) It may fall through + to here from the 4-byte alignment case to process, at most, + one 2-byte chunk. (2) It branches to here from above if + either src or dst is 2-byte aligned, and we process all bytes + here, except for perhaps a trailing odd byte. It's + inefficient, so align your addresses to 4-byte boundaries. + + a2 = src + a3 = dst + a4 = len + a5 = sum + */ + srli a10, a4, 1 /* 2-byte chunks */ +#if XCHAL_HAVE_LOOPS + loopgtz a10, 4f +#else + beqz a10, 4f + slli a10, a10, 1 + add a10, a10, a2 /* a10 = end of last 2-byte src chunk */ +.Loop7: +#endif +SRC( l16ui a9, a2, 0 ) +DST( s16i a9, a3, 0 ) + ONES_ADD(a5, a9) + addi a2, a2, 2 + addi a3, a3, 2 +#if !XCHAL_HAVE_LOOPS + blt a2, a10, .Loop7 +#endif +4: + /* This section processes a possible trailing odd byte. */ + _bbci.l a4, 0, 8f /* 1-byte chunk */ +SRC( l8ui a9, a2, 0 ) +DST( s8i a9, a3, 0 ) +#ifdef __XTENSA_EB__ + slli a9, a9, 8 /* shift byte to bits 8..15 */ +#endif + ONES_ADD(a5, a9) +8: + mov a2, a5 + retw + +5: + /* Control branch to here when either src or dst is odd. We + process all bytes using 8-bit accesses. Grossly inefficient, + so don't feed us an odd address. */ + + srli a10, a4, 1 /* handle in pairs for 16-bit csum */ +#if XCHAL_HAVE_LOOPS + loopgtz a10, 6f +#else + beqz a10, 6f + slli a10, a10, 1 + add a10, a10, a2 /* a10 = end of last odd-aligned, 2-byte src chunk */ +.Loop8: +#endif +SRC( l8ui a9, a2, 0 ) +SRC( l8ui a8, a2, 1 ) +DST( s8i a9, a3, 0 ) +DST( s8i a8, a3, 1 ) +#ifdef __XTENSA_EB__ + slli a9, a9, 8 /* combine into a single 16-bit value */ +#else /* for checksum computation */ + slli a8, a8, 8 +#endif + or a9, a9, a8 + ONES_ADD(a5, a9) + addi a2, a2, 2 + addi a3, a3, 2 +#if !XCHAL_HAVE_LOOPS + blt a2, a10, .Loop8 +#endif +6: + j 4b /* process the possible trailing odd byte */ + + +# Exception handler: +.section .fixup, "ax" +/* + a6 = src_err_ptr + a7 = dst_err_ptr + a11 = original len for exception handling + a12 = original dst for exception handling +*/ + +6001: + _movi a2, -EFAULT + s32i a2, a6, 0 /* src_err_ptr */ + + # clear the complete destination - computing the rest + # is too much work + movi a2, 0 +#if XCHAL_HAVE_LOOPS + loopgtz a11, 2f +#else + beqz a11, 2f + add a11, a11, a12 /* a11 = ending address */ +.Leloop: +#endif + s8i a2, a12, 0 + addi a12, a12, 1 +#if !XCHAL_HAVE_LOOPS + blt a12, a11, .Leloop +#endif +2: + retw + +6002: + movi a2, -EFAULT + s32i a2, a7, 0 /* dst_err_ptr */ + movi a2, 0 + retw + +.previous + diff --git a/arch/xtensa/lib/memcopy.S b/arch/xtensa/lib/memcopy.S new file mode 100644 index 00000000000..e8f6d7eb722 --- /dev/null +++ b/arch/xtensa/lib/memcopy.S @@ -0,0 +1,315 @@ +/* + * arch/xtensa/lib/hal/memcopy.S -- Core HAL library functions + * xthal_memcpy and xthal_bcopy + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2002 - 2005 Tensilica Inc. + */ + +#include <xtensa/coreasm.h> + + .macro src_b r, w0, w1 +#ifdef __XTENSA_EB__ + src \r, \w0, \w1 +#else + src \r, \w1, \w0 +#endif + .endm + + .macro ssa8 r +#ifdef __XTENSA_EB__ + ssa8b \r +#else + ssa8l \r +#endif + .endm + + +/* + * void *memcpy(void *dst, const void *src, size_t len); + * void *memmove(void *dst, const void *src, size_t len); + * void *bcopy(const void *src, void *dst, size_t len); + * + * This function is intended to do the same thing as the standard + * library function memcpy() (or bcopy()) for most cases. + * However, where the source and/or destination references + * an instruction RAM or ROM or a data RAM or ROM, that + * source and/or destination will always be accessed with + * 32-bit load and store instructions (as required for these + * types of devices). + * + * !!!!!!! XTFIXME: + * !!!!!!! Handling of IRAM/IROM has not yet + * !!!!!!! been implemented. + * + * The bcopy version is provided here to avoid the overhead + * of an extra call, for callers that require this convention. + * + * The (general case) algorithm is as follows: + * If destination is unaligned, align it by conditionally + * copying 1 and 2 bytes. + * If source is aligned, + * do 16 bytes with a loop, and then finish up with + * 8, 4, 2, and 1 byte copies conditional on the length; + * else (if source is unaligned), + * do the same, but use SRC to align the source data. + * This code tries to use fall-through branches for the common + * case of aligned source and destination and multiple + * of 4 (or 8) length. + * + * Register use: + * a0/ return address + * a1/ stack pointer + * a2/ return value + * a3/ src + * a4/ length + * a5/ dst + * a6/ tmp + * a7/ tmp + * a8/ tmp + * a9/ tmp + * a10/ tmp + * a11/ tmp + */ + + .text + .align 4 + .global bcopy + .type bcopy,@function +bcopy: + entry sp, 16 # minimal stack frame + # a2=src, a3=dst, a4=len + mov a5, a3 # copy dst so that a2 is return value + mov a3, a2 + mov a2, a5 + j .Lcommon # go to common code for memcpy+bcopy + + +/* + * Byte by byte copy + */ + .align 4 + .byte 0 # 1 mod 4 alignment for LOOPNEZ + # (0 mod 4 alignment for LBEG) +.Lbytecopy: +#if XCHAL_HAVE_LOOPS + loopnez a4, .Lbytecopydone +#else /* !XCHAL_HAVE_LOOPS */ + beqz a4, .Lbytecopydone + add a7, a3, a4 # a7 = end address for source +#endif /* !XCHAL_HAVE_LOOPS */ +.Lnextbyte: + l8ui a6, a3, 0 + addi a3, a3, 1 + s8i a6, a5, 0 + addi a5, a5, 1 +#if !XCHAL_HAVE_LOOPS + blt a3, a7, .Lnextbyte +#endif /* !XCHAL_HAVE_LOOPS */ +.Lbytecopydone: + retw + +/* + * Destination is unaligned + */ + + .align 4 +.Ldst1mod2: # dst is only byte aligned + _bltui a4, 7, .Lbytecopy # do short copies byte by byte + + # copy 1 byte + l8ui a6, a3, 0 + addi a3, a3, 1 + addi a4, a4, -1 + s8i a6, a5, 0 + addi a5, a5, 1 + _bbci.l a5, 1, .Ldstaligned # if dst is now aligned, then + # return to main algorithm +.Ldst2mod4: # dst 16-bit aligned + # copy 2 bytes + _bltui a4, 6, .Lbytecopy # do short copies byte by byte + l8ui a6, a3, 0 + l8ui a7, a3, 1 + addi a3, a3, 2 + addi a4, a4, -2 + s8i a6, a5, 0 + s8i a7, a5, 1 + addi a5, a5, 2 + j .Ldstaligned # dst is now aligned, return to main algorithm + + .align 4 + .global memcpy + .type memcpy,@function +memcpy: + .global memmove + .type memmove,@function +memmove: + + entry sp, 16 # minimal stack frame + # a2/ dst, a3/ src, a4/ len + mov a5, a2 # copy dst so that a2 is return value +.Lcommon: + _bbsi.l a2, 0, .Ldst1mod2 # if dst is 1 mod 2 + _bbsi.l a2, 1, .Ldst2mod4 # if dst is 2 mod 4 +.Ldstaligned: # return here from .Ldst?mod? once dst is aligned + srli a7, a4, 4 # number of loop iterations with 16B + # per iteration + movi a8, 3 # if source is not aligned, + _bany a3, a8, .Lsrcunaligned # then use shifting copy + /* + * Destination and source are word-aligned, use word copy. + */ + # copy 16 bytes per iteration for word-aligned dst and word-aligned src +#if XCHAL_HAVE_LOOPS + loopnez a7, .Loop1done +#else /* !XCHAL_HAVE_LOOPS */ + beqz a7, .Loop1done + slli a8, a7, 4 + add a8, a8, a3 # a8 = end of last 16B source chunk +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop1: + l32i a6, a3, 0 + l32i a7, a3, 4 + s32i a6, a5, 0 + l32i a6, a3, 8 + s32i a7, a5, 4 + l32i a7, a3, 12 + s32i a6, a5, 8 + addi a3, a3, 16 + s32i a7, a5, 12 + addi a5, a5, 16 +#if !XCHAL_HAVE_LOOPS + blt a3, a8, .Loop1 +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop1done: + bbci.l a4, 3, .L2 + # copy 8 bytes + l32i a6, a3, 0 + l32i a7, a3, 4 + addi a3, a3, 8 + s32i a6, a5, 0 + s32i a7, a5, 4 + addi a5, a5, 8 +.L2: + bbsi.l a4, 2, .L3 + bbsi.l a4, 1, .L4 + bbsi.l a4, 0, .L5 + retw +.L3: + # copy 4 bytes + l32i a6, a3, 0 + addi a3, a3, 4 + s32i a6, a5, 0 + addi a5, a5, 4 + bbsi.l a4, 1, .L4 + bbsi.l a4, 0, .L5 + retw +.L4: + # copy 2 bytes + l16ui a6, a3, 0 + addi a3, a3, 2 + s16i a6, a5, 0 + addi a5, a5, 2 + bbsi.l a4, 0, .L5 + retw +.L5: + # copy 1 byte + l8ui a6, a3, 0 + s8i a6, a5, 0 + retw + +/* + * Destination is aligned, Source is unaligned + */ + + .align 4 +.Lsrcunaligned: + _beqz a4, .Ldone # avoid loading anything for zero-length copies + # copy 16 bytes per iteration for word-aligned dst and unaligned src + ssa8 a3 # set shift amount from byte offset +#define SIM_CHECKS_ALIGNMENT 1 /* set to 1 when running on ISS (simulator) with the + lint or ferret client, or 0 to save a few cycles */ +#if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT + and a11, a3, a8 # save unalignment offset for below + sub a3, a3, a11 # align a3 +#endif + l32i a6, a3, 0 # load first word +#if XCHAL_HAVE_LOOPS + loopnez a7, .Loop2done +#else /* !XCHAL_HAVE_LOOPS */ + beqz a7, .Loop2done + slli a10, a7, 4 + add a10, a10, a3 # a10 = end of last 16B source chunk +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop2: + l32i a7, a3, 4 + l32i a8, a3, 8 + src_b a6, a6, a7 + s32i a6, a5, 0 + l32i a9, a3, 12 + src_b a7, a7, a8 + s32i a7, a5, 4 + l32i a6, a3, 16 + src_b a8, a8, a9 + s32i a8, a5, 8 + addi a3, a3, 16 + src_b a9, a9, a6 + s32i a9, a5, 12 + addi a5, a5, 16 +#if !XCHAL_HAVE_LOOPS + blt a3, a10, .Loop2 +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop2done: + bbci.l a4, 3, .L12 + # copy 8 bytes + l32i a7, a3, 4 + l32i a8, a3, 8 + src_b a6, a6, a7 + s32i a6, a5, 0 + addi a3, a3, 8 + src_b a7, a7, a8 + s32i a7, a5, 4 + addi a5, a5, 8 + mov a6, a8 +.L12: + bbci.l a4, 2, .L13 + # copy 4 bytes + l32i a7, a3, 4 + addi a3, a3, 4 + src_b a6, a6, a7 + s32i a6, a5, 0 + addi a5, a5, 4 + mov a6, a7 +.L13: +#if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT + add a3, a3, a11 # readjust a3 with correct misalignment +#endif + bbsi.l a4, 1, .L14 + bbsi.l a4, 0, .L15 +.Ldone: retw +.L14: + # copy 2 bytes + l8ui a6, a3, 0 + l8ui a7, a3, 1 + addi a3, a3, 2 + s8i a6, a5, 0 + s8i a7, a5, 1 + addi a5, a5, 2 + bbsi.l a4, 0, .L15 + retw +.L15: + # copy 1 byte + l8ui a6, a3, 0 + s8i a6, a5, 0 + retw + +/* + * Local Variables: + * mode:fundamental + * comment-start: "# " + * comment-start-skip: "# *" + * End: + */ diff --git a/arch/xtensa/lib/memset.S b/arch/xtensa/lib/memset.S new file mode 100644 index 00000000000..4de25134bc6 --- /dev/null +++ b/arch/xtensa/lib/memset.S @@ -0,0 +1,160 @@ +/* + * arch/xtensa/lib/memset.S + * + * ANSI C standard library function memset + * (Well, almost. .fixup code might return zero.) + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Copyright (C) 2002 Tensilica Inc. + */ + +#include <xtensa/coreasm.h> + +/* + * void *memset(void *dst, int c, size_t length) + * + * The algorithm is as follows: + * Create a word with c in all byte positions + * If the destination is aligned, + * do 16B chucks with a loop, and then finish up with + * 8B, 4B, 2B, and 1B stores conditional on the length. + * If destination is unaligned, align it by conditionally + * setting 1B and 2B and then go to aligned case. + * This code tries to use fall-through branches for the common + * case of an aligned destination (except for the branches to + * the alignment labels). + */ + +/* Load or store instructions that may cause exceptions use the EX macro. */ + +#define EX(insn,reg1,reg2,offset,handler) \ +9: insn reg1, reg2, offset; \ + .section __ex_table, "a"; \ + .word 9b, handler; \ + .previous + + +.text +.align 4 +.global memset +.type memset,@function +memset: + entry sp, 16 # minimal stack frame + # a2/ dst, a3/ c, a4/ length + extui a3, a3, 0, 8 # mask to just 8 bits + slli a7, a3, 8 # duplicate character in all bytes of word + or a3, a3, a7 # ... + slli a7, a3, 16 # ... + or a3, a3, a7 # ... + mov a5, a2 # copy dst so that a2 is return value + movi a6, 3 # for alignment tests + bany a2, a6, .Ldstunaligned # if dst is unaligned +.L0: # return here from .Ldstunaligned when dst is aligned + srli a7, a4, 4 # number of loop iterations with 16B + # per iteration + bnez a4, .Laligned + retw + +/* + * Destination is word-aligned. + */ + # set 16 bytes per iteration for word-aligned dst + .align 4 # 1 mod 4 alignment for LOOPNEZ + .byte 0 # (0 mod 4 alignment for LBEG) +.Laligned: +#if XCHAL_HAVE_LOOPS + loopnez a7, .Loop1done +#else /* !XCHAL_HAVE_LOOPS */ + beqz a7, .Loop1done + slli a6, a7, 4 + add a6, a6, a5 # a6 = end of last 16B chunk +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop1: + EX(s32i, a3, a5, 0, memset_fixup) + EX(s32i, a3, a5, 4, memset_fixup) + EX(s32i, a3, a5, 8, memset_fixup) + EX(s32i, a3, a5, 12, memset_fixup) + addi a5, a5, 16 +#if !XCHAL_HAVE_LOOPS + blt a5, a6, .Loop1 +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop1done: + bbci.l a4, 3, .L2 + # set 8 bytes + EX(s32i, a3, a5, 0, memset_fixup) + EX(s32i, a3, a5, 4, memset_fixup) + addi a5, a5, 8 +.L2: + bbci.l a4, 2, .L3 + # set 4 bytes + EX(s32i, a3, a5, 0, memset_fixup) + addi a5, a5, 4 +.L3: + bbci.l a4, 1, .L4 + # set 2 bytes + EX(s16i, a3, a5, 0, memset_fixup) + addi a5, a5, 2 +.L4: + bbci.l a4, 0, .L5 + # set 1 byte + EX(s8i, a3, a5, 0, memset_fixup) +.L5: +.Lret1: + retw + +/* + * Destination is unaligned + */ + +.Ldstunaligned: + bltui a4, 8, .Lbyteset # do short copies byte by byte + bbci.l a5, 0, .L20 # branch if dst alignment half-aligned + # dst is only byte aligned + # set 1 byte + EX(s8i, a3, a5, 0, memset_fixup) + addi a5, a5, 1 + addi a4, a4, -1 + # now retest if dst aligned + bbci.l a5, 1, .L0 # if now aligned, return to main algorithm +.L20: + # dst half-aligned + # set 2 bytes + EX(s16i, a3, a5, 0, memset_fixup) + addi a5, a5, 2 + addi a4, a4, -2 + j .L0 # dst is now aligned, return to main algorithm + +/* + * Byte by byte set + */ + .align 4 + .byte 0 # 1 mod 4 alignment for LOOPNEZ + # (0 mod 4 alignment for LBEG) +.Lbyteset: +#if XCHAL_HAVE_LOOPS + loopnez a4, .Lbytesetdone +#else /* !XCHAL_HAVE_LOOPS */ + beqz a4, .Lbytesetdone + add a6, a5, a4 # a6 = ending address +#endif /* !XCHAL_HAVE_LOOPS */ +.Lbyteloop: + EX(s8i, a3, a5, 0, memset_fixup) + addi a5, a5, 1 +#if !XCHAL_HAVE_LOOPS + blt a5, a6, .Lbyteloop +#endif /* !XCHAL_HAVE_LOOPS */ +.Lbytesetdone: + retw + + + .section .fixup, "ax" + .align 4 + +/* We return zero if a failure occurred. */ + +memset_fixup: + movi a2, 0 + retw diff --git a/arch/xtensa/lib/pci-auto.c b/arch/xtensa/lib/pci-auto.c new file mode 100644 index 00000000000..90c790f6123 --- /dev/null +++ b/arch/xtensa/lib/pci-auto.c @@ -0,0 +1,352 @@ +/* + * arch/xtensa/kernel/pci-auto.c + * + * PCI autoconfiguration library + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <zankel@tensilica.com, cez@zankel.net> + * + * Based on work from Matt Porter <mporter@mvista.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/pci.h> + +#include <asm/pci-bridge.h> + + +/* + * + * Setting up a PCI + * + * pci_ctrl->first_busno = <first bus number (0)> + * pci_ctrl->last_busno = <last bus number (0xff)> + * pci_ctrl->ops = <PCI config operations> + * pci_ctrl->map_irq = <function to return the interrupt number for a device> + * + * pci_ctrl->io_space.start = <IO space start address (PCI view)> + * pci_ctrl->io_space.end = <IO space end address (PCI view)> + * pci_ctrl->io_space.base = <IO space offset: address 0 from CPU space> + * pci_ctrl->mem_space.start = <MEM space start address (PCI view)> + * pci_ctrl->mem_space.end = <MEM space end address (PCI view)> + * pci_ctrl->mem_space.base = <MEM space offset: address 0 from CPU space> + * + * pcibios_init_resource(&pci_ctrl->io_resource, <IO space start>, + * <IO space end>, IORESOURCE_IO, "PCI host bridge"); + * pcibios_init_resource(&pci_ctrl->mem_resources[0], <MEM space start>, + * <MEM space end>, IORESOURCE_MEM, "PCI host bridge"); + * + * pci_ctrl->last_busno = pciauto_bus_scan(pci_ctrl,pci_ctrl->first_busno); + * + * int __init pciauto_bus_scan(struct pci_controller *pci_ctrl, int current_bus) + * + */ + + +/* define DEBUG to print some debugging messages. */ + +#undef DEBUG + +#ifdef DEBUG +# define DBG(x...) printk(x) +#else +# define DBG(x...) +#endif + +static int pciauto_upper_iospc; +static int pciauto_upper_memspc; + +static struct pci_dev pciauto_dev; +static struct pci_bus pciauto_bus; + +/* + * Helper functions + */ + +/* Initialize the bars of a PCI device. */ + +static void __init +pciauto_setup_bars(struct pci_dev *dev, int bar_limit) +{ + int bar_size; + int bar, bar_nr; + int *upper_limit; + int found_mem64 = 0; + + for (bar = PCI_BASE_ADDRESS_0, bar_nr = 0; + bar <= bar_limit; + bar+=4, bar_nr++) + { + /* Tickle the BAR and get the size */ + pci_write_config_dword(dev, bar, 0xffffffff); + pci_read_config_dword(dev, bar, &bar_size); + + /* If BAR is not implemented go to the next BAR */ + if (!bar_size) + continue; + + /* Check the BAR type and set our address mask */ + if (bar_size & PCI_BASE_ADDRESS_SPACE_IO) + { + bar_size &= PCI_BASE_ADDRESS_IO_MASK; + upper_limit = &pciauto_upper_iospc; + DBG("PCI Autoconfig: BAR %d, I/O, ", bar_nr); + } + else + { + if ((bar_size & PCI_BASE_ADDRESS_MEM_TYPE_MASK) == + PCI_BASE_ADDRESS_MEM_TYPE_64) + found_mem64 = 1; + + bar_size &= PCI_BASE_ADDRESS_MEM_MASK; + upper_limit = &pciauto_upper_memspc; + DBG("PCI Autoconfig: BAR %d, Mem, ", bar_nr); + } + + /* Allocate a base address (bar_size is negative!) */ + *upper_limit = (*upper_limit + bar_size) & bar_size; + + /* Write it out and update our limit */ + pci_write_config_dword(dev, bar, *upper_limit); + + /* + * If we are a 64-bit decoder then increment to the + * upper 32 bits of the bar and force it to locate + * in the lower 4GB of memory. + */ + + if (found_mem64) + pci_write_config_dword(dev, (bar+=4), 0x00000000); + + DBG("size=0x%x, address=0x%x\n", ~bar_size + 1, *upper_limit); + } +} + +/* Initialize the interrupt number. */ + +static void __init +pciauto_setup_irq(struct pci_controller* pci_ctrl,struct pci_dev *dev,int devfn) +{ + u8 pin; + int irq = 0; + + pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin); + + /* Fix illegal pin numbers. */ + + if (pin == 0 || pin > 4) + pin = 1; + + if (pci_ctrl->map_irq) + irq = pci_ctrl->map_irq(dev, PCI_SLOT(devfn), pin); + + if (irq == -1) + irq = 0; + + DBG("PCI Autoconfig: Interrupt %d, pin %d\n", irq, pin); + + pci_write_config_byte(dev, PCI_INTERRUPT_LINE, irq); +} + + +static void __init +pciauto_prescan_setup_bridge(struct pci_dev *dev, int current_bus, + int sub_bus, int *iosave, int *memsave) +{ + /* Configure bus number registers */ + pci_write_config_byte(dev, PCI_PRIMARY_BUS, current_bus); + pci_write_config_byte(dev, PCI_SECONDARY_BUS, sub_bus + 1); + pci_write_config_byte(dev, PCI_SUBORDINATE_BUS, 0xff); + + /* Round memory allocator to 1MB boundary */ + pciauto_upper_memspc &= ~(0x100000 - 1); + *memsave = pciauto_upper_memspc; + + /* Round I/O allocator to 4KB boundary */ + pciauto_upper_iospc &= ~(0x1000 - 1); + *iosave = pciauto_upper_iospc; + + /* Set up memory and I/O filter limits, assume 32-bit I/O space */ + pci_write_config_word(dev, PCI_MEMORY_LIMIT, + ((pciauto_upper_memspc - 1) & 0xfff00000) >> 16); + pci_write_config_byte(dev, PCI_IO_LIMIT, + ((pciauto_upper_iospc - 1) & 0x0000f000) >> 8); + pci_write_config_word(dev, PCI_IO_LIMIT_UPPER16, + ((pciauto_upper_iospc - 1) & 0xffff0000) >> 16); +} + +static void __init +pciauto_postscan_setup_bridge(struct pci_dev *dev, int current_bus, int sub_bus, + int *iosave, int *memsave) +{ + int cmdstat; + + /* Configure bus number registers */ + pci_write_config_byte(dev, PCI_SUBORDINATE_BUS, sub_bus); + + /* + * Round memory allocator to 1MB boundary. + * If no space used, allocate minimum. + */ + pciauto_upper_memspc &= ~(0x100000 - 1); + if (*memsave == pciauto_upper_memspc) + pciauto_upper_memspc -= 0x00100000; + + pci_write_config_word(dev, PCI_MEMORY_BASE, pciauto_upper_memspc >> 16); + + /* Allocate 1MB for pre-fretch */ + pci_write_config_word(dev, PCI_PREF_MEMORY_LIMIT, + ((pciauto_upper_memspc - 1) & 0xfff00000) >> 16); + + pciauto_upper_memspc -= 0x100000; + + pci_write_config_word(dev, PCI_PREF_MEMORY_BASE, + pciauto_upper_memspc >> 16); + + /* Round I/O allocator to 4KB boundary */ + pciauto_upper_iospc &= ~(0x1000 - 1); + if (*iosave == pciauto_upper_iospc) + pciauto_upper_iospc -= 0x1000; + + pci_write_config_byte(dev, PCI_IO_BASE, + (pciauto_upper_iospc & 0x0000f000) >> 8); + pci_write_config_word(dev, PCI_IO_BASE_UPPER16, + pciauto_upper_iospc >> 16); + + /* Enable memory and I/O accesses, enable bus master */ + pci_read_config_dword(dev, PCI_COMMAND, &cmdstat); + pci_write_config_dword(dev, PCI_COMMAND, + cmdstat | + PCI_COMMAND_IO | + PCI_COMMAND_MEMORY | + PCI_COMMAND_MASTER); +} + +/* + * Scan the current PCI bus. + */ + + +int __init pciauto_bus_scan(struct pci_controller *pci_ctrl, int current_bus) +{ + int sub_bus, pci_devfn, pci_class, cmdstat, found_multi=0; + unsigned short vid; + unsigned char header_type; + struct pci_dev *dev = &pciauto_dev; + + pciauto_dev.bus = &pciauto_bus; + pciauto_dev.sysdata = pci_ctrl; + pciauto_bus.ops = pci_ctrl->ops; + + /* + * Fetch our I/O and memory space upper boundaries used + * to allocated base addresses on this pci_controller. + */ + + if (current_bus == pci_ctrl->first_busno) + { + pciauto_upper_iospc = pci_ctrl->io_resource.end + 1; + pciauto_upper_memspc = pci_ctrl->mem_resources[0].end + 1; + } + + sub_bus = current_bus; + + for (pci_devfn = 0; pci_devfn < 0xff; pci_devfn++) + { + /* Skip our host bridge */ + if ((current_bus == pci_ctrl->first_busno) && (pci_devfn == 0)) + continue; + + if (PCI_FUNC(pci_devfn) && !found_multi) + continue; + + pciauto_bus.number = current_bus; + pciauto_dev.devfn = pci_devfn; + + /* If config space read fails from this device, move on */ + if (pci_read_config_byte(dev, PCI_HEADER_TYPE, &header_type)) + continue; + + if (!PCI_FUNC(pci_devfn)) + found_multi = header_type & 0x80; + pci_read_config_word(dev, PCI_VENDOR_ID, &vid); + + if (vid == 0xffff || vid == 0x0000) { + found_multi = 0; + continue; + } + + pci_read_config_dword(dev, PCI_CLASS_REVISION, &pci_class); + + if ((pci_class >> 16) == PCI_CLASS_BRIDGE_PCI) { + + int iosave, memsave; + + DBG("PCI Autoconfig: Found P2P bridge, device %d\n", + PCI_SLOT(pci_devfn)); + + /* Allocate PCI I/O and/or memory space */ + pciauto_setup_bars(dev, PCI_BASE_ADDRESS_1); + + pciauto_prescan_setup_bridge(dev, current_bus, sub_bus, + &iosave, &memsave); + sub_bus = pciauto_bus_scan(pci_ctrl, sub_bus+1); + pciauto_postscan_setup_bridge(dev, current_bus, sub_bus, + &iosave, &memsave); + pciauto_bus.number = current_bus; + + continue; + + } + + +#if 0 + /* Skip legacy mode IDE controller */ + + if ((pci_class >> 16) == PCI_CLASS_STORAGE_IDE) { + + unsigned char prg_iface; + pci_read_config_byte(dev, PCI_CLASS_PROG, &prg_iface); + + if (!(prg_iface & PCIAUTO_IDE_MODE_MASK)) { + DBG("PCI Autoconfig: Skipping legacy mode " + "IDE controller\n"); + continue; + } + } +#endif + + /* + * Found a peripheral, enable some standard + * settings + */ + + pci_read_config_dword(dev, PCI_COMMAND, &cmdstat); + pci_write_config_dword(dev, PCI_COMMAND, + cmdstat | + PCI_COMMAND_IO | + PCI_COMMAND_MEMORY | + PCI_COMMAND_MASTER); + pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0x80); + + /* Allocate PCI I/O and/or memory space */ + DBG("PCI Autoconfig: Found Bus %d, Device %d, Function %d\n", + current_bus, PCI_SLOT(pci_devfn), PCI_FUNC(pci_devfn) ); + + pciauto_setup_bars(dev, PCI_BASE_ADDRESS_5); + pciauto_setup_irq(pci_ctrl, dev, pci_devfn); + } + return sub_bus; +} + + + + + diff --git a/arch/xtensa/lib/strcasecmp.c b/arch/xtensa/lib/strcasecmp.c new file mode 100644 index 00000000000..165b2d6effa --- /dev/null +++ b/arch/xtensa/lib/strcasecmp.c @@ -0,0 +1,32 @@ +/* + * linux/arch/xtensa/lib/strcasecmp.c + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Copyright (C) 2002 Tensilica Inc. + */ + +#include <linux/string.h> + + +/* We handle nothing here except the C locale. Since this is used in + only one place, on strings known to contain only 7 bit ASCII, this + is ok. */ + +int strcasecmp(const char *a, const char *b) +{ + int ca, cb; + + do { + ca = *a++ & 0xff; + cb = *b++ & 0xff; + if (ca >= 'A' && ca <= 'Z') + ca += 'a' - 'A'; + if (cb >= 'A' && cb <= 'Z') + cb += 'a' - 'A'; + } while (ca == cb && ca != '\0'); + + return ca - cb; +} diff --git a/arch/xtensa/lib/strncpy_user.S b/arch/xtensa/lib/strncpy_user.S new file mode 100644 index 00000000000..71d55df4389 --- /dev/null +++ b/arch/xtensa/lib/strncpy_user.S @@ -0,0 +1,224 @@ +/* + * arch/xtensa/lib/strncpy_user.S + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Returns: -EFAULT if exception before terminator, N if the entire + * buffer filled, else strlen. + * + * Copyright (C) 2002 Tensilica Inc. + */ + +#include <xtensa/coreasm.h> +#include <linux/errno.h> + +/* Load or store instructions that may cause exceptions use the EX macro. */ + +#define EX(insn,reg1,reg2,offset,handler) \ +9: insn reg1, reg2, offset; \ + .section __ex_table, "a"; \ + .word 9b, handler; \ + .previous + +/* + * char *__strncpy_user(char *dst, const char *src, size_t len) + */ +.text +.begin literal +.align 4 +.Lmask0: + .byte 0xff, 0x00, 0x00, 0x00 +.Lmask1: + .byte 0x00, 0xff, 0x00, 0x00 +.Lmask2: + .byte 0x00, 0x00, 0xff, 0x00 +.Lmask3: + .byte 0x00, 0x00, 0x00, 0xff +.end literal + +# Register use +# a0/ return address +# a1/ stack pointer +# a2/ return value +# a3/ src +# a4/ len +# a5/ mask0 +# a6/ mask1 +# a7/ mask2 +# a8/ mask3 +# a9/ tmp +# a10/ tmp +# a11/ dst +# a12/ tmp + +.align 4 +.global __strncpy_user +.type __strncpy_user,@function +__strncpy_user: + entry sp, 16 # minimal stack frame + # a2/ dst, a3/ src, a4/ len + mov a11, a2 # leave dst in return value register + beqz a4, .Lret # if len is zero + l32r a5, .Lmask0 # mask for byte 0 + l32r a6, .Lmask1 # mask for byte 1 + l32r a7, .Lmask2 # mask for byte 2 + l32r a8, .Lmask3 # mask for byte 3 + bbsi.l a3, 0, .Lsrc1mod2 # if only 8-bit aligned + bbsi.l a3, 1, .Lsrc2mod4 # if only 16-bit aligned +.Lsrcaligned: # return here when src is word-aligned + srli a12, a4, 2 # number of loop iterations with 4B per loop + movi a9, 3 + bnone a11, a9, .Laligned + j .Ldstunaligned + +.Lsrc1mod2: # src address is odd + EX(l8ui, a9, a3, 0, fixup_l) # get byte 0 + addi a3, a3, 1 # advance src pointer + EX(s8i, a9, a11, 0, fixup_s) # store byte 0 + beqz a9, .Lret # if byte 0 is zero + addi a11, a11, 1 # advance dst pointer + addi a4, a4, -1 # decrement len + beqz a4, .Lret # if len is zero + bbci.l a3, 1, .Lsrcaligned # if src is now word-aligned + +.Lsrc2mod4: # src address is 2 mod 4 + EX(l8ui, a9, a3, 0, fixup_l) # get byte 0 + /* 1-cycle interlock */ + EX(s8i, a9, a11, 0, fixup_s) # store byte 0 + beqz a9, .Lret # if byte 0 is zero + addi a11, a11, 1 # advance dst pointer + addi a4, a4, -1 # decrement len + beqz a4, .Lret # if len is zero + EX(l8ui, a9, a3, 1, fixup_l) # get byte 0 + addi a3, a3, 2 # advance src pointer + EX(s8i, a9, a11, 0, fixup_s) # store byte 0 + beqz a9, .Lret # if byte 0 is zero + addi a11, a11, 1 # advance dst pointer + addi a4, a4, -1 # decrement len + bnez a4, .Lsrcaligned # if len is nonzero +.Lret: + sub a2, a11, a2 # compute strlen + retw + +/* + * dst is word-aligned, src is word-aligned + */ + .align 4 # 1 mod 4 alignment for LOOPNEZ + .byte 0 # (0 mod 4 alignment for LBEG) +.Laligned: +#if XCHAL_HAVE_LOOPS + loopnez a12, .Loop1done +#else + beqz a12, .Loop1done + slli a12, a12, 2 + add a12, a12, a11 # a12 = end of last 4B chunck +#endif +.Loop1: + EX(l32i, a9, a3, 0, fixup_l) # get word from src + addi a3, a3, 4 # advance src pointer + bnone a9, a5, .Lz0 # if byte 0 is zero + bnone a9, a6, .Lz1 # if byte 1 is zero + bnone a9, a7, .Lz2 # if byte 2 is zero + EX(s32i, a9, a11, 0, fixup_s) # store word to dst + bnone a9, a8, .Lz3 # if byte 3 is zero + addi a11, a11, 4 # advance dst pointer +#if !XCHAL_HAVE_LOOPS + blt a11, a12, .Loop1 +#endif + +.Loop1done: + bbci.l a4, 1, .L100 + # copy 2 bytes + EX(l16ui, a9, a3, 0, fixup_l) + addi a3, a3, 2 # advance src pointer +#ifdef __XTENSA_EB__ + bnone a9, a7, .Lz0 # if byte 2 is zero + bnone a9, a8, .Lz1 # if byte 3 is zero +#else + bnone a9, a5, .Lz0 # if byte 0 is zero + bnone a9, a6, .Lz1 # if byte 1 is zero +#endif + EX(s16i, a9, a11, 0, fixup_s) + addi a11, a11, 2 # advance dst pointer +.L100: + bbci.l a4, 0, .Lret + EX(l8ui, a9, a3, 0, fixup_l) + /* slot */ + EX(s8i, a9, a11, 0, fixup_s) + beqz a9, .Lret # if byte is zero + addi a11, a11, 1-3 # advance dst ptr 1, but also cancel + # the effect of adding 3 in .Lz3 code + /* fall thru to .Lz3 and "retw" */ + +.Lz3: # byte 3 is zero + addi a11, a11, 3 # advance dst pointer + sub a2, a11, a2 # compute strlen + retw +.Lz0: # byte 0 is zero +#ifdef __XTENSA_EB__ + movi a9, 0 +#endif /* __XTENSA_EB__ */ + EX(s8i, a9, a11, 0, fixup_s) + sub a2, a11, a2 # compute strlen + retw +.Lz1: # byte 1 is zero +#ifdef __XTENSA_EB__ + extui a9, a9, 16, 16 +#endif /* __XTENSA_EB__ */ + EX(s16i, a9, a11, 0, fixup_s) + addi a11, a11, 1 # advance dst pointer + sub a2, a11, a2 # compute strlen + retw +.Lz2: # byte 2 is zero +#ifdef __XTENSA_EB__ + extui a9, a9, 16, 16 +#endif /* __XTENSA_EB__ */ + EX(s16i, a9, a11, 0, fixup_s) + movi a9, 0 + EX(s8i, a9, a11, 2, fixup_s) + addi a11, a11, 2 # advance dst pointer + sub a2, a11, a2 # compute strlen + retw + + .align 4 # 1 mod 4 alignment for LOOPNEZ + .byte 0 # (0 mod 4 alignment for LBEG) +.Ldstunaligned: +/* + * for now just use byte copy loop + */ +#if XCHAL_HAVE_LOOPS + loopnez a4, .Lunalignedend +#else + beqz a4, .Lunalignedend + add a12, a11, a4 # a12 = ending address +#endif /* XCHAL_HAVE_LOOPS */ +.Lnextbyte: + EX(l8ui, a9, a3, 0, fixup_l) + addi a3, a3, 1 + EX(s8i, a9, a11, 0, fixup_s) + beqz a9, .Lunalignedend + addi a11, a11, 1 +#if !XCHAL_HAVE_LOOPS + blt a11, a12, .Lnextbyte +#endif + +.Lunalignedend: + sub a2, a11, a2 # compute strlen + retw + + + .section .fixup, "ax" + .align 4 + + /* For now, just return -EFAULT. Future implementations might + * like to clear remaining kernel space, like the fixup + * implementation in memset(). Thus, we differentiate between + * load/store fixups. */ + +fixup_s: +fixup_l: + movi a2, -EFAULT + retw + diff --git a/arch/xtensa/lib/strnlen_user.S b/arch/xtensa/lib/strnlen_user.S new file mode 100644 index 00000000000..cdff4d670f3 --- /dev/null +++ b/arch/xtensa/lib/strnlen_user.S @@ -0,0 +1,147 @@ +/* + * arch/xtensa/lib/strnlen_user.S + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Returns strnlen, including trailing zero terminator. + * Zero indicates error. + * + * Copyright (C) 2002 Tensilica Inc. + */ + +#include <xtensa/coreasm.h> + +/* Load or store instructions that may cause exceptions use the EX macro. */ + +#define EX(insn,reg1,reg2,offset,handler) \ +9: insn reg1, reg2, offset; \ + .section __ex_table, "a"; \ + .word 9b, handler; \ + .previous + +/* + * size_t __strnlen_user(const char *s, size_t len) + */ +.text +.begin literal +.align 4 +.Lmask0: + .byte 0xff, 0x00, 0x00, 0x00 +.Lmask1: + .byte 0x00, 0xff, 0x00, 0x00 +.Lmask2: + .byte 0x00, 0x00, 0xff, 0x00 +.Lmask3: + .byte 0x00, 0x00, 0x00, 0xff +.end literal + +# Register use: +# a2/ src +# a3/ len +# a4/ tmp +# a5/ mask0 +# a6/ mask1 +# a7/ mask2 +# a8/ mask3 +# a9/ tmp +# a10/ tmp + +.align 4 +.global __strnlen_user +.type __strnlen_user,@function +__strnlen_user: + entry sp, 16 # minimal stack frame + # a2/ s, a3/ len + addi a4, a2, -4 # because we overincrement at the end; + # we compensate with load offsets of 4 + l32r a5, .Lmask0 # mask for byte 0 + l32r a6, .Lmask1 # mask for byte 1 + l32r a7, .Lmask2 # mask for byte 2 + l32r a8, .Lmask3 # mask for byte 3 + bbsi.l a2, 0, .L1mod2 # if only 8-bit aligned + bbsi.l a2, 1, .L2mod4 # if only 16-bit aligned + +/* + * String is word-aligned. + */ +.Laligned: + srli a10, a3, 2 # number of loop iterations with 4B per loop +#if XCHAL_HAVE_LOOPS + loopnez a10, .Ldone +#else + beqz a10, .Ldone + slli a10, a10, 2 + add a10, a10, a4 # a10 = end of last 4B chunk +#endif /* XCHAL_HAVE_LOOPS */ +.Loop: + EX(l32i, a9, a4, 4, lenfixup) # get next word of string + addi a4, a4, 4 # advance string pointer + bnone a9, a5, .Lz0 # if byte 0 is zero + bnone a9, a6, .Lz1 # if byte 1 is zero + bnone a9, a7, .Lz2 # if byte 2 is zero + bnone a9, a8, .Lz3 # if byte 3 is zero +#if !XCHAL_HAVE_LOOPS + blt a4, a10, .Loop +#endif + +.Ldone: + EX(l32i, a9, a4, 4, lenfixup) # load 4 bytes for remaining checks + + bbci.l a3, 1, .L100 + # check two more bytes (bytes 0, 1 of word) + addi a4, a4, 2 # advance string pointer + bnone a9, a5, .Lz0 # if byte 0 is zero + bnone a9, a6, .Lz1 # if byte 1 is zero +.L100: + bbci.l a3, 0, .L101 + # check one more byte (byte 2 of word) + # Actually, we don't need to check. Zero or nonzero, we'll add one. + # Do not add an extra one for the NULL terminator since we have + # exhausted the original len parameter. + addi a4, a4, 1 # advance string pointer +.L101: + sub a2, a4, a2 # compute length + retw + +# NOTE that in several places below, we point to the byte just after +# the zero byte in order to include the NULL terminator in the count. + +.Lz3: # byte 3 is zero + addi a4, a4, 3 # point to zero byte +.Lz0: # byte 0 is zero + addi a4, a4, 1 # point just beyond zero byte + sub a2, a4, a2 # subtract to get length + retw +.Lz1: # byte 1 is zero + addi a4, a4, 1+1 # point just beyond zero byte + sub a2, a4, a2 # subtract to get length + retw +.Lz2: # byte 2 is zero + addi a4, a4, 2+1 # point just beyond zero byte + sub a2, a4, a2 # subtract to get length + retw + +.L1mod2: # address is odd + EX(l8ui, a9, a4, 4, lenfixup) # get byte 0 + addi a4, a4, 1 # advance string pointer + beqz a9, .Lz3 # if byte 0 is zero + bbci.l a4, 1, .Laligned # if string pointer is now word-aligned + +.L2mod4: # address is 2 mod 4 + addi a4, a4, 2 # advance ptr for aligned access + EX(l32i, a9, a4, 0, lenfixup) # get word with first two bytes of string + bnone a9, a7, .Lz2 # if byte 2 (of word, not string) is zero + bany a9, a8, .Laligned # if byte 3 (of word, not string) is nonzero + # byte 3 is zero + addi a4, a4, 3+1 # point just beyond zero byte + sub a2, a4, a2 # subtract to get length + retw + + .section .fixup, "ax" + .align 4 +lenfixup: + movi a2, 0 + retw + diff --git a/arch/xtensa/lib/usercopy.S b/arch/xtensa/lib/usercopy.S new file mode 100644 index 00000000000..265db2693cb --- /dev/null +++ b/arch/xtensa/lib/usercopy.S @@ -0,0 +1,321 @@ +/* + * arch/xtensa/lib/usercopy.S + * + * Copy to/from user space (derived from arch/xtensa/lib/hal/memcopy.S) + * + * DO NOT COMBINE this function with <arch/xtensa/lib/hal/memcopy.S>. + * It needs to remain separate and distinct. The hal files are part + * of the the Xtensa link-time HAL, and those files may differ per + * processor configuration. Patching the kernel for another + * processor configuration includes replacing the hal files, and we + * could loose the special functionality for accessing user-space + * memory during such a patch. We sacrifice a little code space here + * in favor to simplify code maintenance. + * + * This file is subject to the terms and conditions of the GNU General + * Public License. See the file "COPYING" in the main directory of + * this archive for more details. + * + * Copyright (C) 2002 Tensilica Inc. + */ + + +/* + * size_t __xtensa_copy_user (void *dst, const void *src, size_t len); + * + * The returned value is the number of bytes not copied. Implies zero + * is success. + * + * The general case algorithm is as follows: + * If the destination and source are both aligned, + * do 16B chunks with a loop, and then finish up with + * 8B, 4B, 2B, and 1B copies conditional on the length. + * If destination is aligned and source unaligned, + * do the same, but use SRC to align the source data. + * If destination is unaligned, align it by conditionally + * copying 1B and 2B and then retest. + * This code tries to use fall-through braches for the common + * case of aligned destinations (except for the branches to + * the alignment label). + * + * Register use: + * a0/ return address + * a1/ stack pointer + * a2/ return value + * a3/ src + * a4/ length + * a5/ dst + * a6/ tmp + * a7/ tmp + * a8/ tmp + * a9/ tmp + * a10/ tmp + * a11/ original length + */ + +#include <xtensa/coreasm.h> + +#ifdef __XTENSA_EB__ +#define ALIGN(R, W0, W1) src R, W0, W1 +#define SSA8(R) ssa8b R +#else +#define ALIGN(R, W0, W1) src R, W1, W0 +#define SSA8(R) ssa8l R +#endif + +/* Load or store instructions that may cause exceptions use the EX macro. */ + +#define EX(insn,reg1,reg2,offset,handler) \ +9: insn reg1, reg2, offset; \ + .section __ex_table, "a"; \ + .word 9b, handler; \ + .previous + + + .text + .align 4 + .global __xtensa_copy_user + .type __xtensa_copy_user,@function +__xtensa_copy_user: + entry sp, 16 # minimal stack frame + # a2/ dst, a3/ src, a4/ len + mov a5, a2 # copy dst so that a2 is return value + mov a11, a4 # preserve original len for error case +.Lcommon: + bbsi.l a2, 0, .Ldst1mod2 # if dst is 1 mod 2 + bbsi.l a2, 1, .Ldst2mod4 # if dst is 2 mod 4 +.Ldstaligned: # return here from .Ldstunaligned when dst is aligned + srli a7, a4, 4 # number of loop iterations with 16B + # per iteration + movi a8, 3 # if source is also aligned, + bnone a3, a8, .Laligned # then use word copy + SSA8( a3) # set shift amount from byte offset + bnez a4, .Lsrcunaligned + movi a2, 0 # return success for len==0 + retw + +/* + * Destination is unaligned + */ + +.Ldst1mod2: # dst is only byte aligned + bltui a4, 7, .Lbytecopy # do short copies byte by byte + + # copy 1 byte + EX(l8ui, a6, a3, 0, l_fixup) + addi a3, a3, 1 + EX(s8i, a6, a5, 0, s_fixup) + addi a5, a5, 1 + addi a4, a4, -1 + bbci.l a5, 1, .Ldstaligned # if dst is now aligned, then + # return to main algorithm +.Ldst2mod4: # dst 16-bit aligned + # copy 2 bytes + bltui a4, 6, .Lbytecopy # do short copies byte by byte + EX(l8ui, a6, a3, 0, l_fixup) + EX(l8ui, a7, a3, 1, l_fixup) + addi a3, a3, 2 + EX(s8i, a6, a5, 0, s_fixup) + EX(s8i, a7, a5, 1, s_fixup) + addi a5, a5, 2 + addi a4, a4, -2 + j .Ldstaligned # dst is now aligned, return to main algorithm + +/* + * Byte by byte copy + */ + .align 4 + .byte 0 # 1 mod 4 alignment for LOOPNEZ + # (0 mod 4 alignment for LBEG) +.Lbytecopy: +#if XCHAL_HAVE_LOOPS + loopnez a4, .Lbytecopydone +#else /* !XCHAL_HAVE_LOOPS */ + beqz a4, .Lbytecopydone + add a7, a3, a4 # a7 = end address for source +#endif /* !XCHAL_HAVE_LOOPS */ +.Lnextbyte: + EX(l8ui, a6, a3, 0, l_fixup) + addi a3, a3, 1 + EX(s8i, a6, a5, 0, s_fixup) + addi a5, a5, 1 +#if !XCHAL_HAVE_LOOPS + blt a3, a7, .Lnextbyte +#endif /* !XCHAL_HAVE_LOOPS */ +.Lbytecopydone: + movi a2, 0 # return success for len bytes copied + retw + +/* + * Destination and source are word-aligned. + */ + # copy 16 bytes per iteration for word-aligned dst and word-aligned src + .align 4 # 1 mod 4 alignment for LOOPNEZ + .byte 0 # (0 mod 4 alignment for LBEG) +.Laligned: +#if XCHAL_HAVE_LOOPS + loopnez a7, .Loop1done +#else /* !XCHAL_HAVE_LOOPS */ + beqz a7, .Loop1done + slli a8, a7, 4 + add a8, a8, a3 # a8 = end of last 16B source chunk +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop1: + EX(l32i, a6, a3, 0, l_fixup) + EX(l32i, a7, a3, 4, l_fixup) + EX(s32i, a6, a5, 0, s_fixup) + EX(l32i, a6, a3, 8, l_fixup) + EX(s32i, a7, a5, 4, s_fixup) + EX(l32i, a7, a3, 12, l_fixup) + EX(s32i, a6, a5, 8, s_fixup) + addi a3, a3, 16 + EX(s32i, a7, a5, 12, s_fixup) + addi a5, a5, 16 +#if !XCHAL_HAVE_LOOPS + blt a3, a8, .Loop1 +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop1done: + bbci.l a4, 3, .L2 + # copy 8 bytes + EX(l32i, a6, a3, 0, l_fixup) + EX(l32i, a7, a3, 4, l_fixup) + addi a3, a3, 8 + EX(s32i, a6, a5, 0, s_fixup) + EX(s32i, a7, a5, 4, s_fixup) + addi a5, a5, 8 +.L2: + bbci.l a4, 2, .L3 + # copy 4 bytes + EX(l32i, a6, a3, 0, l_fixup) + addi a3, a3, 4 + EX(s32i, a6, a5, 0, s_fixup) + addi a5, a5, 4 +.L3: + bbci.l a4, 1, .L4 + # copy 2 bytes + EX(l16ui, a6, a3, 0, l_fixup) + addi a3, a3, 2 + EX(s16i, a6, a5, 0, s_fixup) + addi a5, a5, 2 +.L4: + bbci.l a4, 0, .L5 + # copy 1 byte + EX(l8ui, a6, a3, 0, l_fixup) + EX(s8i, a6, a5, 0, s_fixup) +.L5: + movi a2, 0 # return success for len bytes copied + retw + +/* + * Destination is aligned, Source is unaligned + */ + + .align 4 + .byte 0 # 1 mod 4 alignement for LOOPNEZ + # (0 mod 4 alignment for LBEG) +.Lsrcunaligned: + # copy 16 bytes per iteration for word-aligned dst and unaligned src + and a10, a3, a8 # save unalignment offset for below + sub a3, a3, a10 # align a3 (to avoid sim warnings only; not needed for hardware) + EX(l32i, a6, a3, 0, l_fixup) # load first word +#if XCHAL_HAVE_LOOPS + loopnez a7, .Loop2done +#else /* !XCHAL_HAVE_LOOPS */ + beqz a7, .Loop2done + slli a10, a7, 4 + add a10, a10, a3 # a10 = end of last 16B source chunk +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop2: + EX(l32i, a7, a3, 4, l_fixup) + EX(l32i, a8, a3, 8, l_fixup) + ALIGN( a6, a6, a7) + EX(s32i, a6, a5, 0, s_fixup) + EX(l32i, a9, a3, 12, l_fixup) + ALIGN( a7, a7, a8) + EX(s32i, a7, a5, 4, s_fixup) + EX(l32i, a6, a3, 16, l_fixup) + ALIGN( a8, a8, a9) + EX(s32i, a8, a5, 8, s_fixup) + addi a3, a3, 16 + ALIGN( a9, a9, a6) + EX(s32i, a9, a5, 12, s_fixup) + addi a5, a5, 16 +#if !XCHAL_HAVE_LOOPS + blt a3, a10, .Loop2 +#endif /* !XCHAL_HAVE_LOOPS */ +.Loop2done: + bbci.l a4, 3, .L12 + # copy 8 bytes + EX(l32i, a7, a3, 4, l_fixup) + EX(l32i, a8, a3, 8, l_fixup) + ALIGN( a6, a6, a7) + EX(s32i, a6, a5, 0, s_fixup) + addi a3, a3, 8 + ALIGN( a7, a7, a8) + EX(s32i, a7, a5, 4, s_fixup) + addi a5, a5, 8 + mov a6, a8 +.L12: + bbci.l a4, 2, .L13 + # copy 4 bytes + EX(l32i, a7, a3, 4, l_fixup) + addi a3, a3, 4 + ALIGN( a6, a6, a7) + EX(s32i, a6, a5, 0, s_fixup) + addi a5, a5, 4 + mov a6, a7 +.L13: + add a3, a3, a10 # readjust a3 with correct misalignment + bbci.l a4, 1, .L14 + # copy 2 bytes + EX(l8ui, a6, a3, 0, l_fixup) + EX(l8ui, a7, a3, 1, l_fixup) + addi a3, a3, 2 + EX(s8i, a6, a5, 0, s_fixup) + EX(s8i, a7, a5, 1, s_fixup) + addi a5, a5, 2 +.L14: + bbci.l a4, 0, .L15 + # copy 1 byte + EX(l8ui, a6, a3, 0, l_fixup) + EX(s8i, a6, a5, 0, s_fixup) +.L15: + movi a2, 0 # return success for len bytes copied + retw + + + .section .fixup, "ax" + .align 4 + +/* a2 = original dst; a5 = current dst; a11= original len + * bytes_copied = a5 - a2 + * retval = bytes_not_copied = original len - bytes_copied + * retval = a11 - (a5 - a2) + * + * Clearing the remaining pieces of kernel memory plugs security + * holes. This functionality is the equivalent of the *_zeroing + * functions that some architectures provide. + */ + +.Lmemset: + .word memset + +s_fixup: + sub a2, a5, a2 /* a2 <-- bytes copied */ + sub a2, a11, a2 /* a2 <-- bytes not copied */ + retw + +l_fixup: + sub a2, a5, a2 /* a2 <-- bytes copied */ + sub a2, a11, a2 /* a2 <-- bytes not copied == return value */ + + /* void *memset(void *s, int c, size_t n); */ + mov a6, a5 /* s */ + movi a7, 0 /* c */ + mov a8, a2 /* n */ + l32r a4, .Lmemset + callx4 a4 + /* Ignore memset return value in a6. */ + /* a2 still contains bytes not copied. */ + retw + diff --git a/arch/xtensa/mm/Makefile b/arch/xtensa/mm/Makefile new file mode 100644 index 00000000000..a5aed5932d7 --- /dev/null +++ b/arch/xtensa/mm/Makefile @@ -0,0 +1,13 @@ +# +# Makefile for the Linux/Xtensa-specific parts of the memory manager. +# +# Note! Dependencies are done automagically by 'make dep', which also +# removes any old dependencies. DON'T put your own dependencies here +# unless it's something special (ie not a .c file). +# +# Note 2! The CFLAGS definition is now in the main makefile... + +obj-y := init.o fault.o tlb.o misc.o +obj-m := +obj-n := +obj- := diff --git a/arch/xtensa/mm/fault.c b/arch/xtensa/mm/fault.c new file mode 100644 index 00000000000..a945a33e85a --- /dev/null +++ b/arch/xtensa/mm/fault.c @@ -0,0 +1,241 @@ +// TODO VM_EXEC flag work-around, cache aliasing +/* + * arch/xtensa/mm/fault.c + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + */ + +#include <linux/mm.h> +#include <linux/module.h> +#include <asm/mmu_context.h> +#include <asm/cacheflush.h> +#include <asm/hardirq.h> +#include <asm/uaccess.h> +#include <asm/system.h> +#include <asm/pgalloc.h> + +unsigned long asid_cache = ASID_FIRST_VERSION; +void bad_page_fault(struct pt_regs*, unsigned long, int); + +/* + * This routine handles page faults. It determines the address, + * and the problem, and then passes it off to one of the appropriate + * routines. + * + * Note: does not handle Miss and MultiHit. + */ + +void do_page_fault(struct pt_regs *regs) +{ + struct vm_area_struct * vma; + struct mm_struct *mm = current->mm; + unsigned int exccause = regs->exccause; + unsigned int address = regs->excvaddr; + siginfo_t info; + + int is_write, is_exec; + + info.si_code = SEGV_MAPERR; + + /* We fault-in kernel-space virtual memory on-demand. The + * 'reference' page table is init_mm.pgd. + */ + if (address >= TASK_SIZE && !user_mode(regs)) + goto vmalloc_fault; + + /* If we're in an interrupt or have no user + * context, we must not take the fault.. + */ + if (in_atomic() || !mm) { + bad_page_fault(regs, address, SIGSEGV); + return; + } + + is_write = (exccause == XCHAL_EXCCAUSE_STORE_CACHE_ATTRIBUTE) ? 1 : 0; + is_exec = (exccause == XCHAL_EXCCAUSE_ITLB_PRIVILEGE || + exccause == XCHAL_EXCCAUSE_ITLB_MISS || + exccause == XCHAL_EXCCAUSE_FETCH_CACHE_ATTRIBUTE) ? 1 : 0; + +#if 0 + printk("[%s:%d:%08x:%d:%08x:%s%s]\n", current->comm, current->pid, + address, exccause, regs->pc, is_write? "w":"", is_exec? "x":""); +#endif + + down_read(&mm->mmap_sem); + vma = find_vma(mm, address); + + if (!vma) + goto bad_area; + if (vma->vm_start <= address) + goto good_area; + if (!(vma->vm_flags & VM_GROWSDOWN)) + goto bad_area; + if (expand_stack(vma, address)) + goto bad_area; + + /* Ok, we have a good vm_area for this memory access, so + * we can handle it.. + */ + +good_area: + info.si_code = SEGV_ACCERR; + + if (is_write) { + if (!(vma->vm_flags & VM_WRITE)) + goto bad_area; + } else if (is_exec) { + if (!(vma->vm_flags & VM_EXEC)) + goto bad_area; + } else /* Allow read even from write-only pages. */ + if (!(vma->vm_flags & (VM_READ | VM_WRITE))) + goto bad_area; + + /* If for any reason at all we couldn't handle the fault, + * make sure we exit gracefully rather than endlessly redo + * the fault. + */ +survive: + switch (handle_mm_fault(mm, vma, address, is_write)) { + case VM_FAULT_MINOR: + current->min_flt++; + break; + case VM_FAULT_MAJOR: + current->maj_flt++; + break; + case VM_FAULT_SIGBUS: + goto do_sigbus; + case VM_FAULT_OOM: + goto out_of_memory; + default: + BUG(); + } + + up_read(&mm->mmap_sem); + return; + + /* Something tried to access memory that isn't in our memory map.. + * Fix it, but check if it's kernel or user first.. + */ +bad_area: + up_read(&mm->mmap_sem); + if (user_mode(regs)) { + current->thread.bad_vaddr = address; + current->thread.error_code = is_write; + info.si_signo = SIGSEGV; + info.si_errno = 0; + /* info.si_code has been set above */ + info.si_addr = (void *) address; + force_sig_info(SIGSEGV, &info, current); + return; + } + bad_page_fault(regs, address, SIGSEGV); + return; + + + /* We ran out of memory, or some other thing happened to us that made + * us unable to handle the page fault gracefully. + */ +out_of_memory: + up_read(&mm->mmap_sem); + if (current->pid == 1) { + yield(); + down_read(&mm->mmap_sem); + goto survive; + } + printk("VM: killing process %s\n", current->comm); + if (user_mode(regs)) + do_exit(SIGKILL); + bad_page_fault(regs, address, SIGKILL); + return; + +do_sigbus: + up_read(&mm->mmap_sem); + + /* Send a sigbus, regardless of whether we were in kernel + * or user mode. + */ + current->thread.bad_vaddr = address; + info.si_code = SIGBUS; + info.si_errno = 0; + info.si_code = BUS_ADRERR; + info.si_addr = (void *) address; + force_sig_info(SIGBUS, &info, current); + + /* Kernel mode? Handle exceptions or die */ + if (!user_mode(regs)) + bad_page_fault(regs, address, SIGBUS); + +vmalloc_fault: + { + /* Synchronize this task's top level page-table + * with the 'reference' page table. + */ + struct mm_struct *act_mm = current->active_mm; + int index = pgd_index(address); + pgd_t *pgd, *pgd_k; + pmd_t *pmd, *pmd_k; + pte_t *pte_k; + + if (act_mm == NULL) + goto bad_page_fault; + + pgd = act_mm->pgd + index; + pgd_k = init_mm.pgd + index; + + if (!pgd_present(*pgd_k)) + goto bad_page_fault; + + pgd_val(*pgd) = pgd_val(*pgd_k); + + pmd = pmd_offset(pgd, address); + pmd_k = pmd_offset(pgd_k, address); + if (!pmd_present(*pmd) || !pmd_present(*pmd_k)) + goto bad_page_fault; + + pmd_val(*pmd) = pmd_val(*pmd_k); + pte_k = pte_offset_kernel(pmd_k, address); + + if (!pte_present(*pte_k)) + goto bad_page_fault; + return; + } +bad_page_fault: + bad_page_fault(regs, address, SIGKILL); + return; +} + + +void +bad_page_fault(struct pt_regs *regs, unsigned long address, int sig) +{ + extern void die(const char*, struct pt_regs*, long); + const struct exception_table_entry *entry; + + /* Are we prepared to handle this kernel fault? */ + if ((entry = search_exception_tables(regs->pc)) != NULL) { +#if 1 + printk(KERN_DEBUG "%s: Exception at pc=%#010lx (%lx)\n", + current->comm, regs->pc, entry->fixup); +#endif + current->thread.bad_uaddr = address; + regs->pc = entry->fixup; + return; + } + + /* Oops. The kernel tried to access some bad page. We'll have to + * terminate things with extreme prejudice. + */ + printk(KERN_ALERT "Unable to handle kernel paging request at virtual " + "address %08lx\n pc = %08lx, ra = %08lx\n", + address, regs->pc, regs->areg[0]); + die("Oops", regs, sig); + do_exit(sig); +} + diff --git a/arch/xtensa/mm/init.c b/arch/xtensa/mm/init.c new file mode 100644 index 00000000000..56aace84aae --- /dev/null +++ b/arch/xtensa/mm/init.c @@ -0,0 +1,551 @@ +/* + * arch/xtensa/mm/init.c + * + * Derived from MIPS, PPC. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> + * Marc Gauthier + * Kevin Chea + */ + +#include <linux/config.h> +#include <linux/init.h> +#include <linux/signal.h> +#include <linux/sched.h> +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/string.h> +#include <linux/types.h> +#include <linux/ptrace.h> +#include <linux/bootmem.h> +#include <linux/swap.h> + +#include <asm/pgtable.h> +#include <asm/bootparam.h> +#include <asm/mmu_context.h> +#include <asm/tlb.h> +#include <asm/tlbflush.h> +#include <asm/page.h> +#include <asm/pgalloc.h> +#include <asm/pgtable.h> + + +#define DEBUG 0 + +DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); +//static DEFINE_SPINLOCK(tlb_lock); + +/* + * This flag is used to indicate that the page was mapped and modified in + * kernel space, so the cache is probably dirty at that address. + * If cache aliasing is enabled and the page color mismatches, update_mmu_cache + * synchronizes the caches if this bit is set. + */ + +#define PG_cache_clean PG_arch_1 + +/* References to section boundaries */ + +extern char _ftext, _etext, _fdata, _edata, _rodata_end; +extern char __init_begin, __init_end; + +/* + * mem_reserve(start, end, must_exist) + * + * Reserve some memory from the memory pool. + * + * Parameters: + * start Start of region, + * end End of region, + * must_exist Must exist in memory pool. + * + * Returns: + * 0 (memory area couldn't be mapped) + * -1 (success) + */ + +int __init mem_reserve(unsigned long start, unsigned long end, int must_exist) +{ + int i; + + if (start == end) + return 0; + + start = start & PAGE_MASK; + end = PAGE_ALIGN(end); + + for (i = 0; i < sysmem.nr_banks; i++) + if (start < sysmem.bank[i].end + && end >= sysmem.bank[i].start) + break; + + if (i == sysmem.nr_banks) { + if (must_exist) + printk (KERN_WARNING "mem_reserve: [0x%0lx, 0x%0lx) " + "not in any region!\n", start, end); + return 0; + } + + if (start > sysmem.bank[i].start) { + if (end < sysmem.bank[i].end) { + /* split entry */ + if (sysmem.nr_banks >= SYSMEM_BANKS_MAX) + panic("meminfo overflow\n"); + sysmem.bank[sysmem.nr_banks].start = end; + sysmem.bank[sysmem.nr_banks].end = sysmem.bank[i].end; + sysmem.nr_banks++; + } + sysmem.bank[i].end = start; + } else { + if (end < sysmem.bank[i].end) + sysmem.bank[i].start = end; + else { + /* remove entry */ + sysmem.nr_banks--; + sysmem.bank[i].start = sysmem.bank[sysmem.nr_banks].start; + sysmem.bank[i].end = sysmem.bank[sysmem.nr_banks].end; + } + } + return -1; +} + + +/* + * Initialize the bootmem system and give it all the memory we have available. + */ + +void __init bootmem_init(void) +{ + unsigned long pfn; + unsigned long bootmap_start, bootmap_size; + int i; + + max_low_pfn = max_pfn = 0; + min_low_pfn = ~0; + + for (i=0; i < sysmem.nr_banks; i++) { + pfn = PAGE_ALIGN(sysmem.bank[i].start) >> PAGE_SHIFT; + if (pfn < min_low_pfn) + min_low_pfn = pfn; + pfn = PAGE_ALIGN(sysmem.bank[i].end - 1) >> PAGE_SHIFT; + if (pfn > max_pfn) + max_pfn = pfn; + } + + if (min_low_pfn > max_pfn) + panic("No memory found!\n"); + + max_low_pfn = max_pfn < MAX_LOW_MEMORY >> PAGE_SHIFT ? + max_pfn : MAX_LOW_MEMORY >> PAGE_SHIFT; + + /* Find an area to use for the bootmem bitmap. */ + + bootmap_size = bootmem_bootmap_pages(max_low_pfn) << PAGE_SHIFT; + bootmap_start = ~0; + + for (i=0; i<sysmem.nr_banks; i++) + if (sysmem.bank[i].end - sysmem.bank[i].start >= bootmap_size) { + bootmap_start = sysmem.bank[i].start; + break; + } + + if (bootmap_start == ~0UL) + panic("Cannot find %ld bytes for bootmap\n", bootmap_size); + + /* Reserve the bootmem bitmap area */ + + mem_reserve(bootmap_start, bootmap_start + bootmap_size, 1); + bootmap_size = init_bootmem_node(NODE_DATA(0), min_low_pfn, + bootmap_start >> PAGE_SHIFT, + max_low_pfn); + + /* Add all remaining memory pieces into the bootmem map */ + + for (i=0; i<sysmem.nr_banks; i++) + free_bootmem(sysmem.bank[i].start, + sysmem.bank[i].end - sysmem.bank[i].start); + +} + + +void __init paging_init(void) +{ + unsigned long zones_size[MAX_NR_ZONES]; + int i; + + /* All pages are DMA-able, so we put them all in the DMA zone. */ + + zones_size[ZONE_DMA] = max_low_pfn; + for (i = 1; i < MAX_NR_ZONES; i++) + zones_size[i] = 0; + +#ifdef CONFIG_HIGHMEM + zones_size[ZONE_HIGHMEM] = max_pfn - max_low_pfn; +#endif + + /* Initialize the kernel's page tables. */ + + memset(swapper_pg_dir, 0, PAGE_SIZE); + + free_area_init(zones_size); +} + +/* + * Flush the mmu and reset associated register to default values. + */ + +void __init init_mmu (void) +{ + /* Writing zeros to the <t>TLBCFG special registers ensure + * that valid values exist in the register. For existing + * PGSZID<w> fields, zero selects the first element of the + * page-size array. For nonexistant PGSZID<w> fields, zero is + * the best value to write. Also, when changing PGSZID<w> + * fields, the corresponding TLB must be flushed. + */ + set_itlbcfg_register (0); + set_dtlbcfg_register (0); + flush_tlb_all (); + + /* Set rasid register to a known value. */ + + set_rasid_register (ASID_ALL_RESERVED); + + /* Set PTEVADDR special register to the start of the page + * table, which is in kernel mappable space (ie. not + * statically mapped). This register's value is undefined on + * reset. + */ + set_ptevaddr_register (PGTABLE_START); +} + +/* + * Initialize memory pages. + */ + +void __init mem_init(void) +{ + unsigned long codesize, reservedpages, datasize, initsize; + unsigned long highmemsize, tmp, ram; + + max_mapnr = num_physpages = max_low_pfn; + high_memory = (void *) __va(max_mapnr << PAGE_SHIFT); + highmemsize = 0; + +#if CONFIG_HIGHMEM +#error HIGHGMEM not implemented in init.c +#endif + + totalram_pages += free_all_bootmem(); + + reservedpages = ram = 0; + for (tmp = 0; tmp < max_low_pfn; tmp++) { + ram++; + if (PageReserved(mem_map+tmp)) + reservedpages++; + } + + codesize = (unsigned long) &_etext - (unsigned long) &_ftext; + datasize = (unsigned long) &_edata - (unsigned long) &_fdata; + initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin; + + printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, " + "%ldk data, %ldk init %ldk highmem)\n", + (unsigned long) nr_free_pages() << (PAGE_SHIFT-10), + ram << (PAGE_SHIFT-10), + codesize >> 10, + reservedpages << (PAGE_SHIFT-10), + datasize >> 10, + initsize >> 10, + highmemsize >> 10); +} + +void +free_reserved_mem(void *start, void *end) +{ + for (; start < end; start += PAGE_SIZE) { + ClearPageReserved(virt_to_page(start)); + set_page_count(virt_to_page(start), 1); + free_page((unsigned long)start); + totalram_pages++; + } +} + +#ifdef CONFIG_BLK_DEV_INITRD +extern int initrd_is_mapped; + +void free_initrd_mem(unsigned long start, unsigned long end) +{ + if (initrd_is_mapped) { + free_reserved_mem((void*)start, (void*)end); + printk ("Freeing initrd memory: %ldk freed\n",(end-start)>>10); + } +} +#endif + +void free_initmem(void) +{ + free_reserved_mem(&__init_begin, &__init_end); + printk("Freeing unused kernel memory: %dk freed\n", + (&__init_end - &__init_begin) >> 10); +} + +void show_mem(void) +{ + int i, free = 0, total = 0, reserved = 0; + int shared = 0, cached = 0; + + printk("Mem-info:\n"); + show_free_areas(); + printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); + i = max_mapnr; + while (i-- > 0) { + total++; + if (PageReserved(mem_map+i)) + reserved++; + else if (PageSwapCache(mem_map+i)) + cached++; + else if (!page_count(mem_map + i)) + free++; + else + shared += page_count(mem_map + i) - 1; + } + printk("%d pages of RAM\n", total); + printk("%d reserved pages\n", reserved); + printk("%d pages shared\n", shared); + printk("%d pages swap cached\n",cached); + printk("%d free pages\n", free); +} + +/* ------------------------------------------------------------------------- */ + +#if (DCACHE_WAY_SIZE > PAGE_SIZE) + +/* + * With cache aliasing, the page color of the page in kernel space and user + * space might mismatch. We temporarily map the page to a different virtual + * address with the same color and clear the page there. + */ + +void clear_user_page(void *kaddr, unsigned long vaddr, struct page* page) +{ + + /* There shouldn't be any entries for this page. */ + + __flush_invalidate_dcache_page_phys(__pa(page_address(page))); + + if (!PAGE_COLOR_EQ(vaddr, kaddr)) { + unsigned long v, p; + + /* Temporarily map page to DTLB_WAY_DCACHE_ALIAS0. */ + + spin_lock(&tlb_lock); + + p = (unsigned long)pte_val((mk_pte(page,PAGE_KERNEL))); + kaddr = (void*)PAGE_COLOR_MAP0(vaddr); + v = (unsigned long)kaddr | DTLB_WAY_DCACHE_ALIAS0; + __asm__ __volatile__("wdtlb %0,%1; dsync" : :"a" (p), "a" (v)); + + clear_page(kaddr); + + spin_unlock(&tlb_lock); + } else { + clear_page(kaddr); + } + + /* We need to make sure that i$ and d$ are coherent. */ + + clear_bit(PG_cache_clean, &page->flags); +} + +/* + * With cache aliasing, we have to make sure that the page color of the page + * in kernel space matches that of the virtual user address before we read + * the page. If the page color differ, we create a temporary DTLB entry with + * the corrent page color and use this 'temporary' address as the source. + * We then use the same approach as in clear_user_page and copy the data + * to the kernel space and clear the PG_cache_clean bit to synchronize caches + * later. + * + * Note: + * Instead of using another 'way' for the temporary DTLB entry, we could + * probably use the same entry that points to the kernel address (after + * saving the original value and restoring it when we are done). + */ + +void copy_user_page(void* to, void* from, unsigned long vaddr, + struct page* to_page) +{ + /* There shouldn't be any entries for the new page. */ + + __flush_invalidate_dcache_page_phys(__pa(page_address(to_page))); + + spin_lock(&tlb_lock); + + if (!PAGE_COLOR_EQ(vaddr, from)) { + unsigned long v, p, t; + + __asm__ __volatile__ ("pdtlb %1,%2; rdtlb1 %0,%1" + : "=a"(p), "=a"(t) : "a"(from)); + from = (void*)PAGE_COLOR_MAP0(vaddr); + v = (unsigned long)from | DTLB_WAY_DCACHE_ALIAS0; + __asm__ __volatile__ ("wdtlb %0,%1; dsync" ::"a" (p), "a" (v)); + } + + if (!PAGE_COLOR_EQ(vaddr, to)) { + unsigned long v, p; + + p = (unsigned long)pte_val((mk_pte(to_page,PAGE_KERNEL))); + to = (void*)PAGE_COLOR_MAP1(vaddr); + v = (unsigned long)to | DTLB_WAY_DCACHE_ALIAS1; + __asm__ __volatile__ ("wdtlb %0,%1; dsync" ::"a" (p), "a" (v)); + } + copy_page(to, from); + + spin_unlock(&tlb_lock); + + /* We need to make sure that i$ and d$ are coherent. */ + + clear_bit(PG_cache_clean, &to_page->flags); +} + + + +/* + * Any time the kernel writes to a user page cache page, or it is about to + * read from a page cache page this routine is called. + * + * Note: + * The kernel currently only provides one architecture bit in the page + * flags that we use for I$/D$ coherency. Maybe, in future, we can + * use a sepearte bit for deferred dcache aliasing: + * If the page is not mapped yet, we only need to set a flag, + * if mapped, we need to invalidate the page. + */ +// FIXME: we probably need this for WB caches not only for Page Coloring.. + +void flush_dcache_page(struct page *page) +{ + unsigned long addr = __pa(page_address(page)); + struct address_space *mapping = page_mapping(page); + + __flush_invalidate_dcache_page_phys(addr); + + if (!test_bit(PG_cache_clean, &page->flags)) + return; + + /* If this page hasn't been mapped, yet, handle I$/D$ coherency later.*/ +#if 0 + if (mapping && !mapping_mapped(mapping)) + clear_bit(PG_cache_clean, &page->flags); + else +#endif + __invalidate_icache_page_phys(addr); +} + +void flush_cache_range(struct vm_area_struct* vma, unsigned long s, + unsigned long e) +{ + __flush_invalidate_cache_all(); +} + +void flush_cache_page(struct vm_area_struct* vma, unsigned long address, + unsigned long pfn) +{ + struct page *page = pfn_to_page(pfn); + + /* Remove any entry for the old mapping. */ + + if (current->active_mm == vma->vm_mm) { + unsigned long addr = __pa(page_address(page)); + __flush_invalidate_dcache_page_phys(addr); + if ((vma->vm_flags & VM_EXEC) != 0) + __invalidate_icache_page_phys(addr); + } else { + BUG(); + } +} + +#endif /* (DCACHE_WAY_SIZE > PAGE_SIZE) */ + + +pte_t* pte_alloc_one_kernel (struct mm_struct* mm, unsigned long addr) +{ + pte_t* pte = (pte_t*)__get_free_pages(GFP_KERNEL|__GFP_REPEAT, 0); + if (likely(pte)) { + pte_t* ptep = (pte_t*)(pte_val(*pte) + PAGE_OFFSET); + int i; + for (i = 0; i < 1024; i++, ptep++) + pte_clear(mm, addr, ptep); + } + return pte; +} + +struct page* pte_alloc_one(struct mm_struct *mm, unsigned long addr) +{ + struct page *page; + + page = alloc_pages(GFP_KERNEL | __GFP_REPEAT, 0); + + if (likely(page)) { + pte_t* ptep = kmap_atomic(page, KM_USER0); + int i; + + for (i = 0; i < 1024; i++, ptep++) + pte_clear(mm, addr, ptep); + + kunmap_atomic(ptep, KM_USER0); + } + return page; +} + + +/* + * Handle D$/I$ coherency. + * + * Note: + * We only have one architecture bit for the page flags, so we cannot handle + * cache aliasing, yet. + */ + +void +update_mmu_cache(struct vm_area_struct * vma, unsigned long addr, pte_t pte) +{ + unsigned long pfn = pte_pfn(pte); + struct page *page; + unsigned long vaddr = addr & PAGE_MASK; + + if (!pfn_valid(pfn)) + return; + + page = pfn_to_page(pfn); + + invalidate_itlb_mapping(addr); + invalidate_dtlb_mapping(addr); + + /* We have a new mapping. Use it. */ + + write_dtlb_entry(pte, dtlb_probe(addr)); + + /* If the processor can execute from this page, synchronize D$/I$. */ + + if ((vma->vm_flags & VM_EXEC) != 0) { + + write_itlb_entry(pte, itlb_probe(addr)); + + /* Synchronize caches, if not clean. */ + + if (!test_and_set_bit(PG_cache_clean, &page->flags)) { + __flush_dcache_page(vaddr); + __invalidate_icache_page(vaddr); + } + } +} + diff --git a/arch/xtensa/mm/misc.S b/arch/xtensa/mm/misc.S new file mode 100644 index 00000000000..327c0f17187 --- /dev/null +++ b/arch/xtensa/mm/misc.S @@ -0,0 +1,374 @@ +/* + * arch/xtensa/mm/misc.S + * + * Miscellaneous assembly functions. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + */ + +/* Note: we might want to implement some of the loops as zero-overhead-loops, + * where applicable and if supported by the processor. + */ + +#include <linux/linkage.h> +#include <asm/page.h> +#include <asm/pgtable.h> + +#include <xtensa/cacheasm.h> +#include <xtensa/cacheattrasm.h> + +/* clear_page (page) */ + +ENTRY(clear_page) + entry a1, 16 + addi a4, a2, PAGE_SIZE + movi a3, 0 + +1: s32i a3, a2, 0 + s32i a3, a2, 4 + s32i a3, a2, 8 + s32i a3, a2, 12 + s32i a3, a2, 16 + s32i a3, a2, 20 + s32i a3, a2, 24 + s32i a3, a2, 28 + addi a2, a2, 32 + blt a2, a4, 1b + + retw + +/* + * copy_page (void *to, void *from) + * a2 a3 + */ + +ENTRY(copy_page) + entry a1, 16 + addi a4, a2, PAGE_SIZE + +1: l32i a5, a3, 0 + l32i a6, a3, 4 + l32i a7, a3, 8 + s32i a5, a2, 0 + s32i a6, a2, 4 + s32i a7, a2, 8 + l32i a5, a3, 12 + l32i a6, a3, 16 + l32i a7, a3, 20 + s32i a5, a2, 12 + s32i a6, a2, 16 + s32i a7, a2, 20 + l32i a5, a3, 24 + l32i a6, a3, 28 + s32i a5, a2, 24 + s32i a6, a2, 28 + addi a2, a2, 32 + addi a3, a3, 32 + blt a2, a4, 1b + + retw + + +/* + * void __flush_invalidate_cache_all(void) + */ + +ENTRY(__flush_invalidate_cache_all) + entry sp, 16 + dcache_writeback_inv_all a2, a3 + icache_invalidate_all a2, a3 + retw + +/* + * void __invalidate_icache_all(void) + */ + +ENTRY(__invalidate_icache_all) + entry sp, 16 + icache_invalidate_all a2, a3 + retw + +/* + * void __flush_invalidate_dcache_all(void) + */ + +ENTRY(__flush_invalidate_dcache_all) + entry sp, 16 + dcache_writeback_inv_all a2, a3 + retw + + +/* + * void __flush_invalidate_cache_range(ulong start, ulong size) + */ + +ENTRY(__flush_invalidate_cache_range) + entry sp, 16 + mov a4, a2 + mov a5, a3 + dcache_writeback_inv_region a4, a5, a6 + icache_invalidate_region a2, a3, a4 + retw + +/* + * void __invalidate_icache_page(ulong start) + */ + +ENTRY(__invalidate_icache_page) + entry sp, 16 + movi a3, PAGE_SIZE + icache_invalidate_region a2, a3, a4 + retw + +/* + * void __invalidate_dcache_page(ulong start) + */ + +ENTRY(__invalidate_dcache_page) + entry sp, 16 + movi a3, PAGE_SIZE + dcache_invalidate_region a2, a3, a4 + retw + +/* + * void __invalidate_icache_range(ulong start, ulong size) + */ + +ENTRY(__invalidate_icache_range) + entry sp, 16 + icache_invalidate_region a2, a3, a4 + retw + +/* + * void __invalidate_dcache_range(ulong start, ulong size) + */ + +ENTRY(__invalidate_dcache_range) + entry sp, 16 + dcache_invalidate_region a2, a3, a4 + retw + +/* + * void __flush_dcache_page(ulong start) + */ + +ENTRY(__flush_dcache_page) + entry sp, 16 + movi a3, PAGE_SIZE + dcache_writeback_region a2, a3, a4 + retw + +/* + * void __flush_invalidate_dcache_page(ulong start) + */ + +ENTRY(__flush_invalidate_dcache_page) + entry sp, 16 + movi a3, PAGE_SIZE + dcache_writeback_inv_region a2, a3, a4 + retw + +/* + * void __flush_invalidate_dcache_range(ulong start, ulong size) + */ + +ENTRY(__flush_invalidate_dcache_range) + entry sp, 16 + dcache_writeback_inv_region a2, a3, a4 + retw + +/* + * void __invalidate_dcache_all(void) + */ + +ENTRY(__invalidate_dcache_all) + entry sp, 16 + dcache_invalidate_all a2, a3 + retw + +/* + * void __flush_invalidate_dcache_page_phys(ulong start) + */ + +ENTRY(__flush_invalidate_dcache_page_phys) + entry sp, 16 + + movi a3, XCHAL_DCACHE_SIZE + movi a4, PAGE_MASK | 1 + addi a2, a2, 1 + +1: addi a3, a3, -XCHAL_DCACHE_LINESIZE + + ldct a6, a3 + dsync + and a6, a6, a4 + beq a6, a2, 2f + bgeui a3, 2, 1b + retw + +2: diwbi a3, 0 + bgeui a3, 2, 1b + retw + +ENTRY(check_dcache_low0) + entry sp, 16 + + movi a3, XCHAL_DCACHE_SIZE / 4 + movi a4, PAGE_MASK | 1 + addi a2, a2, 1 + +1: addi a3, a3, -XCHAL_DCACHE_LINESIZE + + ldct a6, a3 + dsync + and a6, a6, a4 + beq a6, a2, 2f + bgeui a3, 2, 1b + retw + +2: j 2b + +ENTRY(check_dcache_high0) + entry sp, 16 + + movi a5, XCHAL_DCACHE_SIZE / 4 + movi a3, XCHAL_DCACHE_SIZE / 2 + movi a4, PAGE_MASK | 1 + addi a2, a2, 1 + +1: addi a3, a3, -XCHAL_DCACHE_LINESIZE + addi a5, a5, -XCHAL_DCACHE_LINESIZE + + ldct a6, a3 + dsync + and a6, a6, a4 + beq a6, a2, 2f + bgeui a5, 2, 1b + retw + +2: j 2b + +ENTRY(check_dcache_low1) + entry sp, 16 + + movi a5, XCHAL_DCACHE_SIZE / 4 + movi a3, XCHAL_DCACHE_SIZE * 3 / 4 + movi a4, PAGE_MASK | 1 + addi a2, a2, 1 + +1: addi a3, a3, -XCHAL_DCACHE_LINESIZE + addi a5, a5, -XCHAL_DCACHE_LINESIZE + + ldct a6, a3 + dsync + and a6, a6, a4 + beq a6, a2, 2f + bgeui a5, 2, 1b + retw + +2: j 2b + +ENTRY(check_dcache_high1) + entry sp, 16 + + movi a5, XCHAL_DCACHE_SIZE / 4 + movi a3, XCHAL_DCACHE_SIZE + movi a4, PAGE_MASK | 1 + addi a2, a2, 1 + +1: addi a3, a3, -XCHAL_DCACHE_LINESIZE + addi a5, a5, -XCHAL_DCACHE_LINESIZE + + ldct a6, a3 + dsync + and a6, a6, a4 + beq a6, a2, 2f + bgeui a5, 2, 1b + retw + +2: j 2b + + +/* + * void __invalidate_icache_page_phys(ulong start) + */ + +ENTRY(__invalidate_icache_page_phys) + entry sp, 16 + + movi a3, XCHAL_ICACHE_SIZE + movi a4, PAGE_MASK | 1 + addi a2, a2, 1 + +1: addi a3, a3, -XCHAL_ICACHE_LINESIZE + + lict a6, a3 + isync + and a6, a6, a4 + beq a6, a2, 2f + bgeui a3, 2, 1b + retw + +2: iii a3, 0 + bgeui a3, 2, 1b + retw + + +#if 0 + + movi a3, XCHAL_DCACHE_WAYS - 1 + movi a4, PAGE_SIZE + +1: mov a5, a2 + add a6, a2, a4 + +2: diwbi a5, 0 + diwbi a5, XCHAL_DCACHE_LINESIZE + diwbi a5, XCHAL_DCACHE_LINESIZE * 2 + diwbi a5, XCHAL_DCACHE_LINESIZE * 3 + + addi a5, a5, XCHAL_DCACHE_LINESIZE * 4 + blt a5, a6, 2b + + addi a3, a3, -1 + addi a2, a2, XCHAL_DCACHE_SIZE / XCHAL_DCACHE_WAYS + bgez a3, 1b + + retw + +ENTRY(__invalidate_icache_page_index) + entry sp, 16 + + movi a3, XCHAL_ICACHE_WAYS - 1 + movi a4, PAGE_SIZE + +1: mov a5, a2 + add a6, a2, a4 + +2: iii a5, 0 + iii a5, XCHAL_ICACHE_LINESIZE + iii a5, XCHAL_ICACHE_LINESIZE * 2 + iii a5, XCHAL_ICACHE_LINESIZE * 3 + + addi a5, a5, XCHAL_ICACHE_LINESIZE * 4 + blt a5, a6, 2b + + addi a3, a3, -1 + addi a2, a2, XCHAL_ICACHE_SIZE / XCHAL_ICACHE_WAYS + bgez a3, 2b + + retw + +#endif + + + + + + diff --git a/arch/xtensa/mm/pgtable.c b/arch/xtensa/mm/pgtable.c new file mode 100644 index 00000000000..e5e119c820e --- /dev/null +++ b/arch/xtensa/mm/pgtable.c @@ -0,0 +1,76 @@ +/* + * arch/xtensa/mm/fault.c + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2005 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + */ + +#if (DCACHE_SIZE > PAGE_SIZE) + +pte_t* pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address) +{ + pte_t *pte, p; + int color = ADDR_COLOR(address); + int i; + + p = (pte_t*) __get_free_pages(GFP_KERNEL|__GFP_REPEAT, COLOR_ORDER); + + if (likely(p)) { + struct page *page; + + for (i = 0; i < COLOR_SIZE; i++, p++) { + page = virt_to_page(pte); + + set_page_count(page, 1); + ClearPageCompound(page); + + if (ADDR_COLOR(p) == color) + pte = p; + else + free_page(p); + } + clear_page(pte); + } + return pte; +} + +#ifdef PROFILING + +int mask; +int hit; +int flush; + +#endif + +struct page* pte_alloc_one(struct mm_struct *mm, unsigned long address) +{ + struct page *page, p; + int color = ADDR_COLOR(address); + + p = alloc_pages(GFP_KERNEL | __GFP_REPEAT, PTE_ORDER); + + if (likely(p)) { + for (i = 0; i < PAGE_ORDER; i++) { + set_page_count(p, 1); + ClearPageCompound(p); + + if (PADDR_COLOR(page_address(pg)) == color) + page = p; + else + free_page(p); + } + clear_highpage(page); + } + + return page; +} + +#endif + + + diff --git a/arch/xtensa/mm/tlb.c b/arch/xtensa/mm/tlb.c new file mode 100644 index 00000000000..d3bd3bfc3b3 --- /dev/null +++ b/arch/xtensa/mm/tlb.c @@ -0,0 +1,545 @@ +/* + * arch/xtensa/mm/mmu.c + * + * Logic that manipulates the Xtensa MMU. Derived from MIPS. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001 - 2003 Tensilica Inc. + * + * Joe Taylor + * Chris Zankel <chris@zankel.net> + * Marc Gauthier + */ + +#include <linux/mm.h> +#include <asm/processor.h> +#include <asm/mmu_context.h> +#include <asm/tlbflush.h> +#include <asm/system.h> +#include <asm/cacheflush.h> + + +static inline void __flush_itlb_all (void) +{ + int way, index; + + for (way = 0; way < XCHAL_ITLB_ARF_WAYS; way++) { + for (index = 0; index < ITLB_ENTRIES_PER_ARF_WAY; index++) { + int entry = way + (index << PAGE_SHIFT); + invalidate_itlb_entry_no_isync (entry); + } + } + asm volatile ("isync\n"); +} + +static inline void __flush_dtlb_all (void) +{ + int way, index; + + for (way = 0; way < XCHAL_DTLB_ARF_WAYS; way++) { + for (index = 0; index < DTLB_ENTRIES_PER_ARF_WAY; index++) { + int entry = way + (index << PAGE_SHIFT); + invalidate_dtlb_entry_no_isync (entry); + } + } + asm volatile ("isync\n"); +} + + +void flush_tlb_all (void) +{ + __flush_itlb_all(); + __flush_dtlb_all(); +} + +/* If mm is current, we simply assign the current task a new ASID, thus, + * invalidating all previous tlb entries. If mm is someone else's user mapping, + * wie invalidate the context, thus, when that user mapping is swapped in, + * a new context will be assigned to it. + */ + +void flush_tlb_mm(struct mm_struct *mm) +{ +#if 0 + printk("[tlbmm<%lx>]\n", (unsigned long)mm->context); +#endif + + if (mm == current->active_mm) { + int flags; + local_save_flags(flags); + get_new_mmu_context(mm, asid_cache); + set_rasid_register(ASID_INSERT(mm->context)); + local_irq_restore(flags); + } + else + mm->context = 0; +} + +void flush_tlb_range (struct vm_area_struct *vma, + unsigned long start, unsigned long end) +{ + struct mm_struct *mm = vma->vm_mm; + unsigned long flags; + + if (mm->context == NO_CONTEXT) + return; + +#if 0 + printk("[tlbrange<%02lx,%08lx,%08lx>]\n", + (unsigned long)mm->context, start, end); +#endif + local_save_flags(flags); + + if (end-start + (PAGE_SIZE-1) <= SMALLEST_NTLB_ENTRIES << PAGE_SHIFT) { + int oldpid = get_rasid_register(); + set_rasid_register (ASID_INSERT(mm->context)); + start &= PAGE_MASK; + if (vma->vm_flags & VM_EXEC) + while(start < end) { + invalidate_itlb_mapping(start); + invalidate_dtlb_mapping(start); + start += PAGE_SIZE; + } + else + while(start < end) { + invalidate_dtlb_mapping(start); + start += PAGE_SIZE; + } + + set_rasid_register(oldpid); + } else { + get_new_mmu_context(mm, asid_cache); + if (mm == current->active_mm) + set_rasid_register(ASID_INSERT(mm->context)); + } + local_irq_restore(flags); +} + +void flush_tlb_page (struct vm_area_struct *vma, unsigned long page) +{ + struct mm_struct* mm = vma->vm_mm; + unsigned long flags; + int oldpid; +#if 0 + printk("[tlbpage<%02lx,%08lx>]\n", + (unsigned long)mm->context, page); +#endif + + if(mm->context == NO_CONTEXT) + return; + + local_save_flags(flags); + + oldpid = get_rasid_register(); + + if (vma->vm_flags & VM_EXEC) + invalidate_itlb_mapping(page); + invalidate_dtlb_mapping(page); + + set_rasid_register(oldpid); + + local_irq_restore(flags); + +#if 0 + flush_tlb_all(); + return; +#endif +} + + +#ifdef DEBUG_TLB + +#define USE_ITLB 0 +#define USE_DTLB 1 + +struct way_config_t { + int indicies; + int indicies_log2; + int pgsz_log2; + int arf; +}; + +static struct way_config_t itlb[XCHAL_ITLB_WAYS] = +{ + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY0_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY0_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY0_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY0_SET, ARF) + }, + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY1_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY1_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY1_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY1_SET, ARF) + }, + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY2_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY2_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY2_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY2_SET, ARF) + }, + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY3_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY3_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY3_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY3_SET, ARF) + }, + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY4_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY4_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY4_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY4_SET, ARF) + }, + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY5_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY5_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY5_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY5_SET, ARF) + }, + { XCHAL_ITLB_SET(XCHAL_ITLB_WAY6_SET, ENTRIES), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY6_SET, ENTRIES_LOG2), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY6_SET, PAGESZ_LOG2_MIN), + XCHAL_ITLB_SET(XCHAL_ITLB_WAY6_SET, ARF) + } +}; + +static struct way_config_t dtlb[XCHAL_DTLB_WAYS] = +{ + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY0_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY0_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY0_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY0_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY1_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY1_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY1_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY1_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY2_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY2_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY2_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY2_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY3_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY3_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY3_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY3_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY4_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY4_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY4_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY4_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY5_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY5_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY5_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY5_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY6_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY6_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY6_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY6_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY7_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY7_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY7_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY7_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY8_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY8_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY8_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY8_SET, ARF) + }, + { XCHAL_DTLB_SET(XCHAL_DTLB_WAY9_SET, ENTRIES), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY9_SET, ENTRIES_LOG2), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY9_SET, PAGESZ_LOG2_MIN), + XCHAL_DTLB_SET(XCHAL_DTLB_WAY9_SET, ARF) + } +}; + +/* Total number of entries: */ +#define ITLB_TOTAL_ENTRIES \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY0_SET, ENTRIES) + \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY1_SET, ENTRIES) + \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY2_SET, ENTRIES) + \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY3_SET, ENTRIES) + \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY4_SET, ENTRIES) + \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY5_SET, ENTRIES) + \ + XCHAL_ITLB_SET(XCHAL_ITLB_WAY6_SET, ENTRIES) +#define DTLB_TOTAL_ENTRIES \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY0_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY1_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY2_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY3_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY4_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY5_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY6_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY7_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY8_SET, ENTRIES) + \ + XCHAL_DTLB_SET(XCHAL_DTLB_WAY9_SET, ENTRIES) + + +typedef struct { + unsigned va; + unsigned pa; + unsigned char asid; + unsigned char ca; + unsigned char way; + unsigned char index; + unsigned char pgsz_log2; /* 0 .. 32 */ + unsigned char type; /* 0=ITLB 1=DTLB */ +} tlb_dump_entry_t; + +/* Return -1 if a precedes b, +1 if a follows b, 0 if same: */ +int cmp_tlb_dump_info( tlb_dump_entry_t *a, tlb_dump_entry_t *b ) +{ + if (a->asid < b->asid) return -1; + if (a->asid > b->asid) return 1; + if (a->va < b->va) return -1; + if (a->va > b->va) return 1; + if (a->pa < b->pa) return -1; + if (a->pa > b->pa) return 1; + if (a->ca < b->ca) return -1; + if (a->ca > b->ca) return 1; + if (a->way < b->way) return -1; + if (a->way > b->way) return 1; + if (a->index < b->index) return -1; + if (a->index > b->index) return 1; + return 0; +} + +void sort_tlb_dump_info( tlb_dump_entry_t *t, int n ) +{ + int i, j; + /* Simple O(n*n) sort: */ + for (i = 0; i < n-1; i++) + for (j = i+1; j < n; j++) + if (cmp_tlb_dump_info(t+i, t+j) > 0) { + tlb_dump_entry_t tmp = t[i]; + t[i] = t[j]; + t[j] = tmp; + } +} + + +static tlb_dump_entry_t itlb_dump_info[ITLB_TOTAL_ENTRIES]; +static tlb_dump_entry_t dtlb_dump_info[DTLB_TOTAL_ENTRIES]; + + +static inline char *way_type (int type) +{ + return type ? "autorefill" : "non-autorefill"; +} + +void print_entry (struct way_config_t *way_info, + unsigned int way, + unsigned int index, + unsigned int virtual, + unsigned int translation) +{ + char valid_chr; + unsigned int va, pa, asid, ca; + + va = virtual & + ~((1 << (way_info->pgsz_log2 + way_info->indicies_log2)) - 1); + asid = virtual & ((1 << XCHAL_MMU_ASID_BITS) - 1); + pa = translation & ~((1 << way_info->pgsz_log2) - 1); + ca = translation & ((1 << XCHAL_MMU_CA_BITS) - 1); + valid_chr = asid ? 'V' : 'I'; + + /* Compute and incorporate the effect of the index bits on the + * va. It's more useful for kernel debugging, since we always + * want to know the effective va anyway. */ + + va += index << way_info->pgsz_log2; + + printk ("\t[%d,%d] (%c) vpn 0x%.8x ppn 0x%.8x asid 0x%.2x am 0x%x\n", + way, index, valid_chr, va, pa, asid, ca); +} + +void print_itlb_entry (struct way_config_t *way_info, int way, int index) +{ + print_entry (way_info, way, index, + read_itlb_virtual (way + (index << way_info->pgsz_log2)), + read_itlb_translation (way + (index << way_info->pgsz_log2))); +} + +void print_dtlb_entry (struct way_config_t *way_info, int way, int index) +{ + print_entry (way_info, way, index, + read_dtlb_virtual (way + (index << way_info->pgsz_log2)), + read_dtlb_translation (way + (index << way_info->pgsz_log2))); +} + +void dump_itlb (void) +{ + int way, index; + + printk ("\nITLB: ways = %d\n", XCHAL_ITLB_WAYS); + + for (way = 0; way < XCHAL_ITLB_WAYS; way++) { + printk ("\nWay: %d, Entries: %d, MinPageSize: %d, Type: %s\n", + way, itlb[way].indicies, + itlb[way].pgsz_log2, way_type(itlb[way].arf)); + for (index = 0; index < itlb[way].indicies; index++) { + print_itlb_entry(&itlb[way], way, index); + } + } +} + +void dump_dtlb (void) +{ + int way, index; + + printk ("\nDTLB: ways = %d\n", XCHAL_DTLB_WAYS); + + for (way = 0; way < XCHAL_DTLB_WAYS; way++) { + printk ("\nWay: %d, Entries: %d, MinPageSize: %d, Type: %s\n", + way, dtlb[way].indicies, + dtlb[way].pgsz_log2, way_type(dtlb[way].arf)); + for (index = 0; index < dtlb[way].indicies; index++) { + print_dtlb_entry(&dtlb[way], way, index); + } + } +} + +void dump_tlb (tlb_dump_entry_t *tinfo, struct way_config_t *config, + int entries, int ways, int type, int show_invalid) +{ + tlb_dump_entry_t *e = tinfo; + int way, i; + + /* Gather all info: */ + for (way = 0; way < ways; way++) { + struct way_config_t *cfg = config + way; + for (i = 0; i < cfg->indicies; i++) { + unsigned wayindex = way + (i << cfg->pgsz_log2); + unsigned vv = (type ? read_dtlb_virtual (wayindex) + : read_itlb_virtual (wayindex)); + unsigned pp = (type ? read_dtlb_translation (wayindex) + : read_itlb_translation (wayindex)); + + /* Compute and incorporate the effect of the index bits on the + * va. It's more useful for kernel debugging, since we always + * want to know the effective va anyway. */ + + e->va = (vv & ~((1 << (cfg->pgsz_log2 + cfg->indicies_log2)) - 1)); + e->va += (i << cfg->pgsz_log2); + e->pa = (pp & ~((1 << cfg->pgsz_log2) - 1)); + e->asid = (vv & ((1 << XCHAL_MMU_ASID_BITS) - 1)); + e->ca = (pp & ((1 << XCHAL_MMU_CA_BITS) - 1)); + e->way = way; + e->index = i; + e->pgsz_log2 = cfg->pgsz_log2; + e->type = type; + e++; + } + } +#if 1 + /* Sort by ASID and VADDR: */ + sort_tlb_dump_info (tinfo, entries); +#endif + + /* Display all sorted info: */ + printk ("\n%cTLB dump:\n", (type ? 'D' : 'I')); + for (e = tinfo, i = 0; i < entries; i++, e++) { +#if 0 + if (e->asid == 0 && !show_invalid) + continue; +#endif + printk ("%c way=%d i=%d ASID=%02X V=%08X -> P=%08X CA=%X (%d %cB)\n", + (e->type ? 'D' : 'I'), e->way, e->index, + e->asid, e->va, e->pa, e->ca, + (1 << (e->pgsz_log2 % 10)), + " kMG"[e->pgsz_log2 / 10] + ); + } +} + +void dump_tlbs2 (int showinv) +{ + dump_tlb (itlb_dump_info, itlb, ITLB_TOTAL_ENTRIES, XCHAL_ITLB_WAYS, 0, showinv); + dump_tlb (dtlb_dump_info, dtlb, DTLB_TOTAL_ENTRIES, XCHAL_DTLB_WAYS, 1, showinv); +} + +void dump_all_tlbs (void) +{ + dump_tlbs2 (1); +} + +void dump_valid_tlbs (void) +{ + dump_tlbs2 (0); +} + + +void dump_tlbs (void) +{ + dump_itlb(); + dump_dtlb(); +} + +void dump_cache_tag(int dcache, int idx) +{ + int w, i, s, e; + unsigned long tag, index; + unsigned long num_lines, num_ways, cache_size, line_size; + + num_ways = dcache ? XCHAL_DCACHE_WAYS : XCHAL_ICACHE_WAYS; + cache_size = dcache ? XCHAL_DCACHE_SIZE : XCHAL_ICACHE_SIZE; + line_size = dcache ? XCHAL_DCACHE_LINESIZE : XCHAL_ICACHE_LINESIZE; + + num_lines = cache_size / num_ways; + + s = 0; e = num_lines; + + if (idx >= 0) + e = (s = idx * line_size) + 1; + + for (i = s; i < e; i+= line_size) { + printk("\nline %#08x:", i); + for (w = 0; w < num_ways; w++) { + index = w * num_lines + i; + if (dcache) + __asm__ __volatile__("ldct %0, %1\n\t" + : "=a"(tag) : "a"(index)); + else + __asm__ __volatile__("lict %0, %1\n\t" + : "=a"(tag) : "a"(index)); + + printk(" %#010lx", tag); + } + } + printk ("\n"); +} + +void dump_icache(int index) +{ + unsigned long data, addr; + int w, i; + + const unsigned long num_ways = XCHAL_ICACHE_WAYS; + const unsigned long cache_size = XCHAL_ICACHE_SIZE; + const unsigned long line_size = XCHAL_ICACHE_LINESIZE; + const unsigned long num_lines = cache_size / num_ways / line_size; + + for (w = 0; w < num_ways; w++) { + printk ("\nWay %d", w); + + for (i = 0; i < line_size; i+= 4) { + addr = w * num_lines + index * line_size + i; + __asm__ __volatile__("licw %0, %1\n\t" + : "=a"(data) : "a"(addr)); + printk(" %#010lx", data); + } + } + printk ("\n"); +} + +void dump_cache_tags(void) +{ + printk("Instruction cache\n"); + dump_cache_tag(0, -1); + printk("Data cache\n"); + dump_cache_tag(1, -1); +} + +#endif diff --git a/arch/xtensa/platform-iss/Makefile b/arch/xtensa/platform-iss/Makefile new file mode 100644 index 00000000000..5b394e9620e --- /dev/null +++ b/arch/xtensa/platform-iss/Makefile @@ -0,0 +1,13 @@ +# $Id: Makefile,v 1.1.1.1 2002/08/28 16:10:14 aroll Exp $ +# +# Makefile for the Xtensa Instruction Set Simulator (ISS) +# "prom monitor" library routines under Linux. +# +# Note! Dependencies are done automagically by 'make dep', which also +# removes any old dependencies. DON'T put your own dependencies here +# unless it's something special (ie not a .c file). +# +# Note 2! The CFLAGS definitions are in the main makefile... + +obj-y = io.o console.o setup.o network.o + diff --git a/arch/xtensa/platform-iss/console.c b/arch/xtensa/platform-iss/console.c new file mode 100644 index 00000000000..4fbddf92a92 --- /dev/null +++ b/arch/xtensa/platform-iss/console.c @@ -0,0 +1,300 @@ +/* + * arch/xtensa/platform-iss/console.c + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2001-2005 Tensilica Inc. + * Authors Christian Zankel, Joe Taylor + */ + +#include <linux/module.h> +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/console.h> +#include <linux/init.h> +#include <linux/slab.h> +#include <linux/mm.h> +#include <linux/major.h> +#include <linux/param.h> +#include <linux/serial.h> +#include <linux/serialP.h> +#include <linux/console.h> + +#include <asm/uaccess.h> +#include <asm/irq.h> + +#include <xtensa/simcall.h> + +#include <linux/tty.h> +#include <linux/tty_flip.h> + +#ifdef SERIAL_INLINE +#define _INLINE_ inline +#endif + +#define SERIAL_MAX_NUM_LINES 1 +#define SERIAL_TIMER_VALUE (20 * HZ) + +static struct tty_driver *serial_driver; +static struct timer_list serial_timer; + +static DEFINE_SPINLOCK(timer_lock); + +int errno; + +static int __simc (int a, int b, int c, int d, int e, int f) +{ + int ret; + __asm__ __volatile__ ("simcall\n" + "mov %0, a2\n" + "mov %1, a3\n" : "=a" (ret), "=a" (errno) + : : "a2", "a3"); + return ret; +} + +static char *serial_version = "0.1"; +static char *serial_name = "ISS serial driver"; + +/* + * This routine is called whenever a serial port is opened. It + * enables interrupts for a serial port, linking in its async structure into + * the IRQ chain. It also performs the serial-specific + * initialization for the tty structure. + */ + +static void rs_poll(unsigned long); + +static int rs_open(struct tty_struct *tty, struct file * filp) +{ + int line = tty->index; + + if ((line < 0) || (line >= SERIAL_MAX_NUM_LINES)) + return -ENODEV; + + spin_lock(&timer_lock); + + if (tty->count == 1) { + init_timer(&serial_timer); + serial_timer.data = (unsigned long) tty; + serial_timer.function = rs_poll; + mod_timer(&serial_timer, jiffies + SERIAL_TIMER_VALUE); + } + spin_unlock(&timer_lock); + + return 0; +} + + +/* + * ------------------------------------------------------------ + * iss_serial_close() + * + * This routine is called when the serial port gets closed. First, we + * wait for the last remaining data to be sent. Then, we unlink its + * async structure from the interrupt chain if necessary, and we free + * that IRQ if nothing is left in the chain. + * ------------------------------------------------------------ + */ +static void rs_close(struct tty_struct *tty, struct file * filp) +{ + spin_lock(&timer_lock); + if (tty->count == 1) + del_timer_sync(&serial_timer); + spin_unlock(&timer_lock); +} + + +static int rs_write(struct tty_struct * tty, + const unsigned char *buf, int count) +{ + /* see drivers/char/serialX.c to reference original version */ + + __simc (SYS_write, 1, (unsigned long)buf, count, 0, 0); + return count; +} + +static void rs_poll(unsigned long priv) +{ + struct tty_struct* tty = (struct tty_struct*) priv; + + struct timeval tv = { .tv_sec = 0, .tv_usec = 0 }; + int i = 0; + unsigned char c; + + spin_lock(&timer_lock); + + while (__simc(SYS_select_one, 0, XTISS_SELECT_ONE_READ, (int)&tv,0,0)){ + __simc (SYS_read, 0, (unsigned long)&c, 1, 0, 0); + tty->flip.count++; + *tty->flip.char_buf_ptr++ = c; + *tty->flip.flag_buf_ptr++ = TTY_NORMAL; + i++; + } + + if (i) + tty_flip_buffer_push(tty); + + + mod_timer(&serial_timer, jiffies + SERIAL_TIMER_VALUE); + spin_unlock(&timer_lock); +} + + +static void rs_put_char(struct tty_struct *tty, unsigned char ch) +{ + char buf[2]; + + if (!tty) + return; + + buf[0] = ch; + buf[1] = '\0'; /* Is this NULL necessary? */ + __simc (SYS_write, 1, (unsigned long) buf, 1, 0, 0); +} + +static void rs_flush_chars(struct tty_struct *tty) +{ +} + +static int rs_write_room(struct tty_struct *tty) +{ + /* Let's say iss can always accept 2K characters.. */ + return 2 * 1024; +} + +static int rs_chars_in_buffer(struct tty_struct *tty) +{ + /* the iss doesn't buffer characters */ + return 0; +} + +static void rs_hangup(struct tty_struct *tty) +{ + /* Stub, once again.. */ +} + +static void rs_wait_until_sent(struct tty_struct *tty, int timeout) +{ + /* Stub, once again.. */ +} + +static int rs_read_proc(char *page, char **start, off_t off, int count, + int *eof, void *data) +{ + int len = 0; + off_t begin = 0; + + len += sprintf(page, "serinfo:1.0 driver:%s\n", serial_version); + *eof = 1; + + if (off >= len + begin) + return 0; + + *start = page + (off - begin); + return ((count < begin + len - off) ? count : begin + len - off); +} + + +static struct tty_operations serial_ops = { + .open = rs_open, + .close = rs_close, + .write = rs_write, + .put_char = rs_put_char, + .flush_chars = rs_flush_chars, + .write_room = rs_write_room, + .chars_in_buffer = rs_chars_in_buffer, + .hangup = rs_hangup, + .wait_until_sent = rs_wait_until_sent, + .read_proc = rs_read_proc +}; + +int __init rs_init(void) +{ + serial_driver = alloc_tty_driver(1); + + printk ("%s %s\n", serial_name, serial_version); + + /* Initialize the tty_driver structure */ + + serial_driver->owner = THIS_MODULE; + serial_driver->driver_name = "iss_serial"; + serial_driver->name = "ttyS"; + serial_driver->major = TTY_MAJOR; + serial_driver->minor_start = 64; + serial_driver->type = TTY_DRIVER_TYPE_SERIAL; + serial_driver->subtype = SERIAL_TYPE_NORMAL; + serial_driver->init_termios = tty_std_termios; + serial_driver->init_termios.c_cflag = + B9600 | CS8 | CREAD | HUPCL | CLOCAL; + serial_driver->flags = TTY_DRIVER_REAL_RAW; + + tty_set_operations(serial_driver, &serial_ops); + + if (tty_register_driver(serial_driver)) + panic("Couldn't register serial driver\n"); + return 0; +} + + +static __exit void rs_exit(void) +{ + int error; + + if ((error = tty_unregister_driver(serial_driver))) + printk("ISS_SERIAL: failed to unregister serial driver (%d)\n", + error); + put_tty_driver(serial_driver); +} + + +/* We use `late_initcall' instead of just `__initcall' as a workaround for + * the fact that (1) simcons_tty_init can't be called before tty_init, + * (2) tty_init is called via `module_init', (3) if statically linked, + * module_init == device_init, and (4) there's no ordering of init lists. + * We can do this easily because simcons is always statically linked, but + * other tty drivers that depend on tty_init and which must use + * `module_init' to declare their init routines are likely to be broken. + */ + +late_initcall(rs_init); + + +#ifdef CONFIG_SERIAL_CONSOLE + +static void iss_console_write(struct console *co, const char *s, unsigned count) +{ + int len = strlen(s); + + if (s != 0 && *s != 0) + __simc (SYS_write, 1, (unsigned long)s, + count < len ? count : len,0,0); +} + +static struct tty_driver* iss_console_device(struct console *c, int *index) +{ + *index = c->index; + return serial_driver; +} + + +static struct console sercons = { + .name = "ttyS", + .write = iss_console_write, + .device = iss_console_device, + .flags = CON_PRINTBUFFER, + .index = -1 +}; + +static int __init iss_console_init(void) +{ + register_console(&sercons); + return 0; +} + +console_initcall(iss_console_init); + +#endif /* CONFIG_SERIAL_CONSOLE */ + diff --git a/arch/xtensa/platform-iss/io.c b/arch/xtensa/platform-iss/io.c new file mode 100644 index 00000000000..5b161a5cb65 --- /dev/null +++ b/arch/xtensa/platform-iss/io.c @@ -0,0 +1,32 @@ +/* This file isn't really needed right now. */ + +#if 0 + +#include <asm/io.h> +#include <xtensa/simcall.h> + +extern int __simc (); + + +char iss_serial_getc() +{ + char c; + __simc( SYS_read, 0, &c, 1 ); + return c; +} + +void iss_serial_putc( char c ) +{ + __simc( SYS_write, 1, &c, 1 ); +} + +void iss_serial_puts( char *s ) +{ + if( s != 0 && *s != 0 ) + __simc( SYS_write, 1, s, strlen(s) ); +} + +/*#error Need I/O ports to specific hardware!*/ + +#endif + diff --git a/arch/xtensa/platform-iss/network.c b/arch/xtensa/platform-iss/network.c new file mode 100644 index 00000000000..498d7dced1f --- /dev/null +++ b/arch/xtensa/platform-iss/network.c @@ -0,0 +1,855 @@ +/* + * + * arch/xtensa/platform-iss/network.c + * + * Platform specific initialization. + * + * Authors: Chris Zankel <chris@zankel.net> + * Based on work form the UML team. + * + * Copyright 2005 Tensilica Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + */ + +#include <linux/config.h> +#include <linux/list.h> +#include <linux/irq.h> +#include <linux/spinlock.h> +#include <linux/slab.h> +#include <linux/timer.h> +#include <linux/if_ether.h> +#include <linux/inetdevice.h> +#include <linux/init.h> +#include <linux/if_tun.h> +#include <linux/etherdevice.h> +#include <linux/interrupt.h> +#include <linux/ioctl.h> +#include <linux/bootmem.h> +#include <linux/ethtool.h> +#include <linux/rtnetlink.h> +#include <linux/timer.h> + +#include <xtensa/simcall.h> + +#define DRIVER_NAME "iss-netdev" +#define ETH_MAX_PACKET 1500 +#define ETH_HEADER_OTHER 14 +#define ISS_NET_TIMER_VALUE (2 * HZ) + + +static DEFINE_SPINLOCK(opened_lock); +static LIST_HEAD(opened); + +static DEFINE_SPINLOCK(devices_lock); +static LIST_HEAD(devices); + +/* ------------------------------------------------------------------------- */ + +/* We currently only support the TUNTAP transport protocol. */ + +#define TRANSPORT_TUNTAP_NAME "tuntap" +#define TRANSPORT_TUNTAP_MTU ETH_MAX_PACKET + +struct tuntap_info { + char dev_name[IFNAMSIZ]; + int fixed_config; + unsigned char gw[ETH_ALEN]; + int fd; +}; + +/* ------------------------------------------------------------------------- */ + + +/* This structure contains out private information for the driver. */ + +struct iss_net_private { + + struct list_head device_list; + struct list_head opened_list; + + spinlock_t lock; + struct net_device *dev; + struct platform_device pdev; + struct timer_list tl; + struct net_device_stats stats; + + struct timer_list timer; + unsigned int timer_val; + + int index; + int mtu; + + unsigned char mac[ETH_ALEN]; + int have_mac; + + struct { + union { + struct tuntap_info tuntap; + } info; + + int (*open)(struct iss_net_private *lp); + void (*close)(struct iss_net_private *lp); + int (*read)(struct iss_net_private *lp, struct sk_buff **skb); + int (*write)(struct iss_net_private *lp, struct sk_buff **skb); + unsigned short (*protocol)(struct sk_buff *skb); + int (*poll)(struct iss_net_private *lp); + } tp; + +}; + +/* ======================= ISS SIMCALL INTERFACE =========================== */ + +/* Note: __simc must _not_ be declared inline! */ + +static int errno; + +static int __simc (int a, int b, int c, int d, int e, int f) +{ + int ret; + __asm__ __volatile__ ("simcall\n" + "mov %0, a2\n" + "mov %1, a3\n" : "=a" (ret), "=a" (errno) + : : "a2", "a3"); + return ret; +} + +static int inline simc_open(char *file, int flags, int mode) +{ + return __simc(SYS_open, (int) file, flags, mode, 0, 0); +} + +static int inline simc_close(int fd) +{ + return __simc(SYS_close, fd, 0, 0, 0, 0); +} + +static int inline simc_ioctl(int fd, int request, void *arg) +{ + return __simc(SYS_ioctl, fd, request, (int) arg, 0, 0); +} + +static int inline simc_read(int fd, void *buf, size_t count) +{ + return __simc(SYS_read, fd, (int) buf, count, 0, 0); +} + +static int inline simc_write(int fd, void *buf, size_t count) +{ + return __simc(SYS_write, fd, (int) buf, count, 0, 0); +} + +static int inline simc_poll(int fd) +{ + struct timeval tv = { .tv_sec = 0, .tv_usec = 0 }; + + return __simc(SYS_select_one, fd, XTISS_SELECT_ONE_READ, (int)&tv,0,0); +} + +/* ================================ HELPERS ================================ */ + + +static char *split_if_spec(char *str, ...) +{ + char **arg, *end; + va_list ap; + + va_start(ap, str); + while ((arg = va_arg(ap, char**)) != NULL) { + if (*str == '\0') + return NULL; + end = strchr(str, ','); + if (end != str) + *arg = str; + if (end == NULL) + return NULL; + *end ++ = '\0'; + str = end; + } + va_end(ap); + return str; +} + + +#if 0 +/* Adjust SKB. */ + +struct sk_buff *ether_adjust_skb(struct sk_buff *skb, int extra) +{ + if ((skb != NULL) && (skb_tailroom(skb) < extra)) { + struct sk_buff *skb2; + + skb2 = skb_copy_expand(skb, 0, extra, GFP_ATOMIC); + dev_kfree_skb(skb); + skb = skb2; + } + if (skb != NULL) + skb_put(skb, extra); + + return skb; +} +#endif + +/* Return the IP address as a string for a given device. */ + +static void dev_ip_addr(void *d, char *buf, char *bin_buf) +{ + struct net_device *dev = d; + struct in_device *ip = dev->ip_ptr; + struct in_ifaddr *in; + u32 addr; + + if ((ip == NULL) || ((in = ip->ifa_list) == NULL)) { + printk(KERN_WARNING "Device not assigned an IP address!\n"); + return; + } + + addr = in->ifa_address; + sprintf(buf, "%d.%d.%d.%d", addr & 0xff, (addr >> 8) & 0xff, + (addr >> 16) & 0xff, addr >> 24); + + if (bin_buf) { + bin_buf[0] = addr & 0xff; + bin_buf[1] = (addr >> 8) & 0xff; + bin_buf[2] = (addr >> 16) & 0xff; + bin_buf[3] = addr >> 24; + } +} + +/* Set Ethernet address of the specified device. */ + +static void inline set_ether_mac(void *d, unsigned char *addr) +{ + struct net_device *dev = d; + memcpy(dev->dev_addr, addr, ETH_ALEN); +} + + +/* ======================= TUNTAP TRANSPORT INTERFACE ====================== */ + +static int tuntap_open(struct iss_net_private *lp) +{ + struct ifreq ifr; + char *dev_name = lp->tp.info.tuntap.dev_name; + int err = -EINVAL; + int fd; + + /* We currently only support a fixed configuration. */ + + if (!lp->tp.info.tuntap.fixed_config) + return -EINVAL; + + if ((fd = simc_open("/dev/net/tun", 02, 0)) < 0) { /* O_RDWR */ + printk("Failed to open /dev/net/tun, returned %d " + "(errno = %d)\n", fd, errno); + return fd; + } + + memset(&ifr, 0, sizeof ifr); + ifr.ifr_flags = IFF_TAP | IFF_NO_PI; + strlcpy(ifr.ifr_name, dev_name, sizeof ifr.ifr_name - 1); + + if ((err = simc_ioctl(fd, TUNSETIFF, (void*) &ifr)) < 0) { + printk("Failed to set interface, returned %d " + "(errno = %d)\n", err, errno); + simc_close(fd); + return err; + } + + lp->tp.info.tuntap.fd = fd; + return err; +} + +static void tuntap_close(struct iss_net_private *lp) +{ +#if 0 + if (lp->tp.info.tuntap.fixed_config) + iter_addresses(lp->tp.info.tuntap.dev, close_addr, lp->host.dev_name); +#endif + simc_close(lp->tp.info.tuntap.fd); + lp->tp.info.tuntap.fd = -1; +} + +static int tuntap_read (struct iss_net_private *lp, struct sk_buff **skb) +{ +#if 0 + *skb = ether_adjust_skb(*skb, ETH_HEADER_OTHER); + if (*skb == NULL) + return -ENOMEM; +#endif + + return simc_read(lp->tp.info.tuntap.fd, + (*skb)->data, (*skb)->dev->mtu + ETH_HEADER_OTHER); +} + +static int tuntap_write (struct iss_net_private *lp, struct sk_buff **skb) +{ + return simc_write(lp->tp.info.tuntap.fd, (*skb)->data, (*skb)->len); +} + +unsigned short tuntap_protocol(struct sk_buff *skb) +{ + return eth_type_trans(skb, skb->dev); +} + +static int tuntap_poll(struct iss_net_private *lp) +{ + return simc_poll(lp->tp.info.tuntap.fd); +} + +/* + * Currently only a device name is supported. + * ethX=tuntap[,[mac address][,[device name]]] + */ + +static int tuntap_probe(struct iss_net_private *lp, int index, char *init) +{ + const int len = strlen(TRANSPORT_TUNTAP_NAME); + char *dev_name = NULL, *mac_str = NULL, *rem = NULL; + + /* Transport should be 'tuntap': ethX=tuntap,mac,dev_name */ + + if (strncmp(init, TRANSPORT_TUNTAP_NAME, len)) + return 0; + + if (*(init += strlen(TRANSPORT_TUNTAP_NAME)) == ',') { + if ((rem=split_if_spec(init+1, &mac_str, &dev_name)) != NULL) { + printk("Extra garbage on specification : '%s'\n", rem); + return 0; + } + } else if (*init != '\0') { + printk("Invalid argument: %s. Skipping device!\n", init); + return 0; + } + + if (dev_name) { + strncpy(lp->tp.info.tuntap.dev_name, dev_name, + sizeof lp->tp.info.tuntap.dev_name); + lp->tp.info.tuntap.fixed_config = 1; + } else + strcpy(lp->tp.info.tuntap.dev_name, TRANSPORT_TUNTAP_NAME); + + +#if 0 + if (setup_etheraddr(mac_str, lp->mac)) + lp->have_mac = 1; +#endif + lp->mtu = TRANSPORT_TUNTAP_MTU; + + //lp->info.tuntap.gate_addr = gate_addr; + + lp->tp.info.tuntap.fd = -1; + + lp->tp.open = tuntap_open; + lp->tp.close = tuntap_close; + lp->tp.read = tuntap_read; + lp->tp.write = tuntap_write; + lp->tp.protocol = tuntap_protocol; + lp->tp.poll = tuntap_poll; + + printk("TUN/TAP backend - "); +#if 0 + if (lp->host.gate_addr != NULL) + printk("IP = %s", lp->host.gate_addr); +#endif + printk("\n"); + + return 1; +} + +/* ================================ ISS NET ================================ */ + +static int iss_net_rx(struct net_device *dev) +{ + struct iss_net_private *lp = dev->priv; + int pkt_len; + struct sk_buff *skb; + + /* Check if there is any new data. */ + + if (lp->tp.poll(lp) == 0) + return 0; + + /* Try to allocate memory, if it fails, try again next round. */ + + if ((skb = dev_alloc_skb(dev->mtu + 2 + ETH_HEADER_OTHER)) == NULL) { + lp->stats.rx_dropped++; + return 0; + } + + skb_reserve(skb, 2); + + /* Setup skb */ + + skb->dev = dev; + skb->mac.raw = skb->data; + pkt_len = lp->tp.read(lp, &skb); + skb_put(skb, pkt_len); + + if (pkt_len > 0) { + skb_trim(skb, pkt_len); + skb->protocol = lp->tp.protocol(skb); + // netif_rx(skb); + netif_rx_ni(skb); + + lp->stats.rx_bytes += skb->len; + lp->stats.rx_packets++; + return pkt_len; + } + kfree_skb(skb); + return pkt_len; +} + +static int iss_net_poll(void) +{ + struct list_head *ele; + int err, ret = 0; + + spin_lock(&opened_lock); + + list_for_each(ele, &opened) { + struct iss_net_private *lp; + + lp = list_entry(ele, struct iss_net_private, opened_list); + + if (!netif_running(lp->dev)) + break; + + spin_lock(&lp->lock); + + while ((err = iss_net_rx(lp->dev)) > 0) + ret++; + + spin_unlock(&lp->lock); + + if (err < 0) { + printk(KERN_ERR "Device '%s' read returned %d, " + "shutting it down\n", lp->dev->name, err); + dev_close(lp->dev); + } else { + // FIXME reactivate_fd(lp->fd, ISS_ETH_IRQ); + } + } + + spin_unlock(&opened_lock); + return ret; +} + + +static void iss_net_timer(unsigned long priv) +{ + struct iss_net_private* lp = (struct iss_net_private*) priv; + + spin_lock(&lp->lock); + + iss_net_poll(); + + mod_timer(&lp->timer, jiffies + lp->timer_val); + + spin_unlock(&lp->lock); +} + + +static int iss_net_open(struct net_device *dev) +{ + struct iss_net_private *lp = dev->priv; + char addr[sizeof "255.255.255.255\0"]; + int err; + + spin_lock(&lp->lock); + + if ((err = lp->tp.open(lp)) < 0) + goto out; + + if (!lp->have_mac) { + dev_ip_addr(dev, addr, &lp->mac[2]); + set_ether_mac(dev, lp->mac); + } + + netif_start_queue(dev); + + /* clear buffer - it can happen that the host side of the interface + * is full when we gethere. In this case, new data is never queued, + * SIGIOs never arrive, and the net never works. + */ + while ((err = iss_net_rx(dev)) > 0) + ; + + spin_lock(&opened_lock); + list_add(&lp->opened_list, &opened); + spin_unlock(&opened_lock); + + init_timer(&lp->timer); + lp->timer_val = ISS_NET_TIMER_VALUE; + lp->timer.data = (unsigned long) lp; + lp->timer.function = iss_net_timer; + mod_timer(&lp->timer, jiffies + lp->timer_val); + +out: + spin_unlock(&lp->lock); + return err; +} + +static int iss_net_close(struct net_device *dev) +{ + struct iss_net_private *lp = dev->priv; +printk("iss_net_close!\n"); + netif_stop_queue(dev); + spin_lock(&lp->lock); + + spin_lock(&opened_lock); + list_del(&opened); + spin_unlock(&opened_lock); + + del_timer_sync(&lp->timer); + + lp->tp.close(lp); + + spin_unlock(&lp->lock); + return 0; +} + +static int iss_net_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct iss_net_private *lp = dev->priv; + unsigned long flags; + int len; + + netif_stop_queue(dev); + spin_lock_irqsave(&lp->lock, flags); + + len = lp->tp.write(lp, &skb); + + if (len == skb->len) { + lp->stats.tx_packets++; + lp->stats.tx_bytes += skb->len; + dev->trans_start = jiffies; + netif_start_queue(dev); + + /* this is normally done in the interrupt when tx finishes */ + netif_wake_queue(dev); + + } else if (len == 0) { + netif_start_queue(dev); + lp->stats.tx_dropped++; + + } else { + netif_start_queue(dev); + printk(KERN_ERR "iss_net_start_xmit: failed(%d)\n", len); + } + + spin_unlock_irqrestore(&lp->lock, flags); + + dev_kfree_skb(skb); + return 0; +} + + +static struct net_device_stats *iss_net_get_stats(struct net_device *dev) +{ + struct iss_net_private *lp = dev->priv; + return &lp->stats; +} + +static void iss_net_set_multicast_list(struct net_device *dev) +{ +#if 0 + if (dev->flags & IFF_PROMISC) + return; + else if (dev->mc_count) + dev->flags |= IFF_ALLMULTI; + else + dev->flags &= ~IFF_ALLMULTI; +#endif +} + +static void iss_net_tx_timeout(struct net_device *dev) +{ +#if 0 + dev->trans_start = jiffies; + netif_wake_queue(dev); +#endif +} + +static int iss_net_set_mac(struct net_device *dev, void *addr) +{ +#if 0 + struct iss_net_private *lp = dev->priv; + struct sockaddr *hwaddr = addr; + + spin_lock(&lp->lock); + memcpy(dev->dev_addr, hwaddr->sa_data, ETH_ALEN); + spin_unlock(&lp->lock); +#endif + + return 0; +} + +static int iss_net_change_mtu(struct net_device *dev, int new_mtu) +{ +#if 0 + struct iss_net_private *lp = dev->priv; + int err = 0; + + spin_lock(&lp->lock); + + // FIXME not needed new_mtu = transport_set_mtu(new_mtu, &lp->user); + + if (new_mtu < 0) + err = new_mtu; + else + dev->mtu = new_mtu; + + spin_unlock(&lp->lock); + return err; +#endif + return -EINVAL; +} + +static int iss_net_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +{ +#if 0 + static const struct ethtool_drvinfo info = { + .cmd = ETHTOOL_GDRVINFO, + .driver = DRIVER_NAME, + .version = "42", + }; + void *useraddr; + u32 ethcmd; + + switch (cmd) { + case SIOCETHTOOL: + useraddr = ifr->ifr_data; + if (copy_from_user(ðcmd, useraddr, sizeof(ethcmd))) + return -EFAULT; + + switch (ethcmd) { + case ETHTOOL_GDRVINFO: + if (copy_to_user(useraddr, &info, sizeof(info))) + return -EFAULT; + return 0; + default: + return -EOPNOTSUPP; + } + default: + return -EINVAL; + } +#endif + return -EINVAL; +} + +void iss_net_user_timer_expire(unsigned long _conn) +{ +} + + +static struct device_driver iss_net_driver = { + .name = DRIVER_NAME, + .bus = &platform_bus_type, +}; + +static int driver_registered; + +static int iss_net_configure(int index, char *init) +{ + struct net_device *dev; + struct iss_net_private *lp; + int err; + + if ((dev = alloc_etherdev(sizeof *lp)) == NULL) { + printk(KERN_ERR "eth_configure: failed to allocate device\n"); + return 1; + } + + /* Initialize private element. */ + + lp = dev->priv; + *lp = ((struct iss_net_private) { + .device_list = LIST_HEAD_INIT(lp->device_list), + .opened_list = LIST_HEAD_INIT(lp->opened_list), + .lock = SPIN_LOCK_UNLOCKED, + .dev = dev, + .index = index, + //.fd = -1, + .mac = { 0xfe, 0xfd, 0x0, 0x0, 0x0, 0x0 }, + .have_mac = 0, + }); + + /* + * Try all transport protocols. + * Note: more protocols can be added by adding '&& !X_init(lp, eth)'. + */ + + if (!tuntap_probe(lp, index, init)) { + printk("Invalid arguments. Skipping device!\n"); + goto errout; + } + + printk(KERN_INFO "Netdevice %d ", index); + if (lp->have_mac) + printk("(%02x:%02x:%02x:%02x:%02x:%02x) ", + lp->mac[0], lp->mac[1], + lp->mac[2], lp->mac[3], + lp->mac[4], lp->mac[5]); + printk(": "); + + /* sysfs register */ + + if (!driver_registered) { + driver_register(&iss_net_driver); + driver_registered = 1; + } + + spin_lock(&devices_lock); + list_add(&lp->device_list, &devices); + spin_unlock(&devices_lock); + + lp->pdev.id = index; + lp->pdev.name = DRIVER_NAME; + platform_device_register(&lp->pdev); + SET_NETDEV_DEV(dev,&lp->pdev.dev); + + /* + * If this name ends up conflicting with an existing registered + * netdevice, that is OK, register_netdev{,ice}() will notice this + * and fail. + */ + snprintf(dev->name, sizeof dev->name, "eth%d", index); + + dev->mtu = lp->mtu; + dev->open = iss_net_open; + dev->hard_start_xmit = iss_net_start_xmit; + dev->stop = iss_net_close; + dev->get_stats = iss_net_get_stats; + dev->set_multicast_list = iss_net_set_multicast_list; + dev->tx_timeout = iss_net_tx_timeout; + dev->set_mac_address = iss_net_set_mac; + dev->change_mtu = iss_net_change_mtu; + dev->do_ioctl = iss_net_ioctl; + dev->watchdog_timeo = (HZ >> 1); + dev->irq = -1; + + rtnl_lock(); + err = register_netdevice(dev); + rtnl_unlock(); + + if (err) { + printk("Error registering net device!\n"); + /* XXX: should we call ->remove() here? */ + free_netdev(dev); + return 1; + } + + init_timer(&lp->tl); + lp->tl.function = iss_net_user_timer_expire; + +#if 0 + if (lp->have_mac) + set_ether_mac(dev, lp->mac); +#endif + return 0; + +errout: + // FIXME: unregister; free, etc.. + return -EIO; + +} + +/* ------------------------------------------------------------------------- */ + +/* Filled in during early boot */ + +struct list_head eth_cmd_line = LIST_HEAD_INIT(eth_cmd_line); + +struct iss_net_init { + struct list_head list; + char *init; /* init string */ + int index; +}; + +/* + * Parse the command line and look for 'ethX=...' fields, and register all + * those fields. They will be later initialized in iss_net_init. + */ + +#define ERR KERN_ERR "iss_net_setup: " + +static int iss_net_setup(char *str) +{ + struct iss_net_private *device = NULL; + struct iss_net_init *new; + struct list_head *ele; + char *end; + int n; + + n = simple_strtoul(str, &end, 0); + if (end == str) { + printk(ERR "Failed to parse '%s'\n", str); + return 1; + } + if (n < 0) { + printk(ERR "Device %d is negative\n", n); + return 1; + } + if (*(str = end) != '=') { + printk(ERR "Expected '=' after device number\n"); + return 1; + } + + spin_lock(&devices_lock); + + list_for_each(ele, &devices) { + device = list_entry(ele, struct iss_net_private, device_list); + if (device->index == n) + break; + } + + spin_unlock(&devices_lock); + + if (device && device->index == n) { + printk(ERR "Device %d already configured\n", n); + return 1; + } + + if ((new = alloc_bootmem(sizeof new)) == NULL) { + printk("Alloc_bootmem failed\n"); + return 1; + } + + INIT_LIST_HEAD(&new->list); + new->index = n; + new->init = str + 1; + + list_add_tail(&new->list, ð_cmd_line); + return 1; +} + +#undef ERR + +__setup("eth", iss_net_setup); + +/* + * Initialize all ISS Ethernet devices previously registered in iss_net_setup. + */ + +static int iss_net_init(void) +{ + struct list_head *ele, *next; + + /* Walk through all Ethernet devices specified in the command line. */ + + list_for_each_safe(ele, next, ð_cmd_line) { + struct iss_net_init *eth; + eth = list_entry(ele, struct iss_net_init, list); + iss_net_configure(eth->index, eth->init); + } + + return 1; +} + +module_init(iss_net_init); + diff --git a/arch/xtensa/platform-iss/setup.c b/arch/xtensa/platform-iss/setup.c new file mode 100644 index 00000000000..2e6dcbf0cc0 --- /dev/null +++ b/arch/xtensa/platform-iss/setup.c @@ -0,0 +1,112 @@ +/* + * + * arch/xtensa/platform-iss/setup.c + * + * Platform specific initialization. + * + * Authors: Chris Zankel <chris@zankel.net> + * Joe Taylor <joe@tensilica.com> + * + * Copyright 2001 - 2005 Tensilica Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + */ +#include <linux/config.h> +#include <linux/stddef.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/errno.h> +#include <linux/reboot.h> +#include <linux/pci.h> +#include <linux/kdev_t.h> +#include <linux/types.h> +#include <linux/major.h> +#include <linux/blkdev.h> +#include <linux/console.h> +#include <linux/delay.h> +#include <linux/stringify.h> +#include <linux/notifier.h> + +#include <asm/platform.h> +#include <asm/bootparam.h> + + +void __init platform_init(bp_tag_t* bootparam) +{ + +} + +void platform_halt(void) +{ + printk (" ** Called platform_halt(), looping forever! **\n"); + while (1); +} + +void platform_power_off(void) +{ + printk (" ** Called platform_power_off(), looping forever! **\n"); + while (1); +} +void platform_restart(void) +{ + /* Flush and reset the mmu, simulate a processor reset, and + * jump to the reset vector. */ + + __asm__ __volatile__("movi a2, 15\n\t" + "wsr a2, " __stringify(ICOUNTLEVEL) "\n\t" + "movi a2, 0\n\t" + "wsr a2, " __stringify(ICOUNT) "\n\t" + "wsr a2, " __stringify(IBREAKENABLE) "\n\t" + "wsr a2, " __stringify(LCOUNT) "\n\t" + "movi a2, 0x1f\n\t" + "wsr a2, " __stringify(PS) "\n\t" + "isync\n\t" + "jx %0\n\t" + : + : "a" (XCHAL_RESET_VECTOR_VADDR) + : "a2"); + + /* control never gets here */ +} + +extern void iss_net_poll(void); + +const char twirl[]="|/-\\|/-\\"; + +void platform_heartbeat(void) +{ +#if 0 + static int i = 0, j = 0; + + if (--i < 0) { + i = 99; + printk("\r%c\r", twirl[j++]); + if (j == 8) + j = 0; + } +#endif +} + + + +static int +iss_panic_event(struct notifier_block *this, unsigned long event, void *ptr) +{ + __asm__ __volatile__("movi a2, -1; simcall\n"); + return NOTIFY_DONE; +} + +static struct notifier_block iss_panic_block = { + iss_panic_event, + NULL, + 0 +}; + +void __init platform_setup(char **p_cmdline) +{ + notifier_chain_register(&panic_notifier_list, &iss_panic_block); +} |