From c513867561eeb07d24a0bdda1a18a8f91921a301 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 14 Oct 2008 18:08:48 -0400 Subject: ftrace: do not enclose logic in WARN_ON In ftrace, logic is defined in the WARN_ON_ONCE, which can become a nop with some configs. This patch fixes it. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index d073d981a73..8821ceabf51 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -62,6 +62,7 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, unsigned char *new_code) { unsigned char replaced[MCOUNT_INSN_SIZE]; + int ret; /* * Note: Due to modules and __init, code can @@ -77,8 +78,9 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) return 2; - WARN_ON_ONCE(__copy_to_user_inatomic((char __user *)ip, new_code, - MCOUNT_INSN_SIZE)); + ret = __copy_to_user_inatomic((char __user *)ip, new_code, + MCOUNT_INSN_SIZE); + WARN_ON_ONCE(ret); sync_core(); -- cgit v1.2.3 From 606576ce816603d9fe1fb453a88bc6eea16ca709 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 6 Oct 2008 19:06:12 -0400 Subject: ftrace: rename FTRACE to FUNCTION_TRACER Due to confusion between the ftrace infrastructure and the gcc profiling tracer "ftrace", this patch renames the config options from FTRACE to FUNCTION_TRACER. The other two names that are offspring from FTRACE DYNAMIC_FTRACE and FTRACE_MCOUNT_RECORD will stay the same. This patch was generated mostly by script, and partially by hand. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 2 +- arch/x86/kernel/Makefile | 2 +- arch/x86/kernel/entry_32.S | 4 ++-- arch/x86/kernel/entry_64.S | 4 ++-- arch/x86/kernel/i386_ksyms_32.c | 2 +- arch/x86/kernel/x8664_ksyms_64.c | 2 +- arch/x86/xen/Makefile | 2 +- 7 files changed, 9 insertions(+), 9 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 40ee8080956..290e21aa774 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -28,7 +28,7 @@ config X86 select HAVE_KRETPROBES select HAVE_FTRACE_MCOUNT_RECORD select HAVE_DYNAMIC_FTRACE - select HAVE_FTRACE + select HAVE_FUNCTION_TRACER select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64) select HAVE_ARCH_KGDB if !X86_VOYAGER select HAVE_ARCH_TRACEHOOK diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 0d41f0343dc..ec3d30136bf 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -6,7 +6,7 @@ extra-y := head_$(BITS).o head$(BITS).o head.o init_task.o vmlinu CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE) -ifdef CONFIG_FTRACE +ifdef CONFIG_FUNCTION_TRACER # Do not profile debug and lowlevel utilities CFLAGS_REMOVE_tsc.o = -pg CFLAGS_REMOVE_rtc.o = -pg diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 4e4269c73bb..9d49facc21f 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -1149,7 +1149,7 @@ ENDPROC(xen_failsafe_callback) #endif /* CONFIG_XEN */ -#ifdef CONFIG_FTRACE +#ifdef CONFIG_FUNCTION_TRACER #ifdef CONFIG_DYNAMIC_FTRACE ENTRY(mcount) @@ -1204,7 +1204,7 @@ trace: jmp ftrace_stub END(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ -#endif /* CONFIG_FTRACE */ +#endif /* CONFIG_FUNCTION_TRACER */ .section .rodata,"a" #include "syscall_table_32.S" diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 09e7145484c..b86f332c96a 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -61,7 +61,7 @@ .code64 -#ifdef CONFIG_FTRACE +#ifdef CONFIG_FUNCTION_TRACER #ifdef CONFIG_DYNAMIC_FTRACE ENTRY(mcount) retq @@ -138,7 +138,7 @@ trace: jmp ftrace_stub END(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ -#endif /* CONFIG_FTRACE */ +#endif /* CONFIG_FUNCTION_TRACER */ #ifndef CONFIG_PREEMPT #define retint_kernel retint_restore_args diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c index dd7ebee446a..43cec6bdda6 100644 --- a/arch/x86/kernel/i386_ksyms_32.c +++ b/arch/x86/kernel/i386_ksyms_32.c @@ -5,7 +5,7 @@ #include #include -#ifdef CONFIG_FTRACE +#ifdef CONFIG_FUNCTION_TRACER /* mcount is defined in assembly */ EXPORT_SYMBOL(mcount); #endif diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c index b545f371b5f..695e426aa35 100644 --- a/arch/x86/kernel/x8664_ksyms_64.c +++ b/arch/x86/kernel/x8664_ksyms_64.c @@ -12,7 +12,7 @@ #include #include -#ifdef CONFIG_FTRACE +#ifdef CONFIG_FUNCTION_TRACER /* mcount is defined in assembly */ EXPORT_SYMBOL(mcount); #endif diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile index 313947940a1..6dcefba7836 100644 --- a/arch/x86/xen/Makefile +++ b/arch/x86/xen/Makefile @@ -1,4 +1,4 @@ -ifdef CONFIG_FTRACE +ifdef CONFIG_FUNCTION_TRACER # Do not profile debug and lowlevel utilities CFLAGS_REMOVE_spinlock.o = -pg CFLAGS_REMOVE_time.o = -pg -- cgit v1.2.3 From 593eb8a2d63e95772a5f22d746f18a997c5ee463 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 23 Oct 2008 09:32:59 -0400 Subject: ftrace: return error on failed modified text. Have the ftrace_modify_code return error values: -EFAULT on error of reading the address -EINVAL if what is read does not match what it expected -EPERM if the write fails to update after a successful match. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 8821ceabf51..428291581cb 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -62,7 +62,6 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, unsigned char *new_code) { unsigned char replaced[MCOUNT_INSN_SIZE]; - int ret; /* * Note: Due to modules and __init, code can @@ -72,15 +71,16 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, * No real locking needed, this code is run through * kstop_machine, or before SMP starts. */ - if (__copy_from_user_inatomic(replaced, (char __user *)ip, MCOUNT_INSN_SIZE)) - return 1; + if (__copy_from_user_inatomic(replaced, (char __user *)ip, + MCOUNT_INSN_SIZE)) + return -EFAULT; if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) - return 2; + return -EINVAL; - ret = __copy_to_user_inatomic((char __user *)ip, new_code, - MCOUNT_INSN_SIZE); - WARN_ON_ONCE(ret); + if (__copy_to_user_inatomic((char __user *)ip, new_code, + MCOUNT_INSN_SIZE)) + return -EPERM; sync_core(); -- cgit v1.2.3 From 76aefee57657428fb77cbd8624119c1a440bee44 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 23 Oct 2008 09:33:00 -0400 Subject: ftrace: comment arch ftrace code Add comments to explain what is happening in the x86 arch ftrace code. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 428291581cb..783455454d7 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -66,18 +66,23 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, /* * Note: Due to modules and __init, code can * disappear and change, we need to protect against faulting - * as well as code changing. + * as well as code changing. We do this by using the + * __copy_*_user functions. * * No real locking needed, this code is run through * kstop_machine, or before SMP starts. */ + + /* read the text we want to modify */ if (__copy_from_user_inatomic(replaced, (char __user *)ip, MCOUNT_INSN_SIZE)) return -EFAULT; + /* Make sure it is what we expect it to be */ if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) return -EINVAL; + /* replace the text with the new text */ if (__copy_to_user_inatomic((char __user *)ip, new_code, MCOUNT_INSN_SIZE)) return -EPERM; -- cgit v1.2.3 From ab9a0918cbf0fa8883301838df8dbc8fc085ff50 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 23 Oct 2008 09:33:01 -0400 Subject: ftrace: use probe_kernel Andrew Morton suggested using the proper API for reading and writing kernel areas that might fault. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 783455454d7..da4fb0deecf 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -67,15 +67,14 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, * Note: Due to modules and __init, code can * disappear and change, we need to protect against faulting * as well as code changing. We do this by using the - * __copy_*_user functions. + * probe_kernel_* functions. * * No real locking needed, this code is run through * kstop_machine, or before SMP starts. */ /* read the text we want to modify */ - if (__copy_from_user_inatomic(replaced, (char __user *)ip, - MCOUNT_INSN_SIZE)) + if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) return -EFAULT; /* Make sure it is what we expect it to be */ @@ -83,8 +82,7 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, return -EINVAL; /* replace the text with the new text */ - if (__copy_to_user_inatomic((char __user *)ip, new_code, - MCOUNT_INSN_SIZE)) + if (probe_kernel_write((void *)ip, new_code, MCOUNT_INSN_SIZE)) return -EPERM; sync_core(); -- cgit v1.2.3 From 4d296c24326783bff1282ac72f310d8bac8df413 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 23 Oct 2008 09:33:06 -0400 Subject: ftrace: remove mcount set The arch dependent function ftrace_mcount_set was only used by the daemon start up code. This patch removes it. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 7 ------- 1 file changed, 7 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index da4fb0deecf..b399eed2353 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -103,13 +103,6 @@ notrace int ftrace_update_ftrace_func(ftrace_func_t func) return ret; } -notrace int ftrace_mcount_set(unsigned long *data) -{ - /* mcount is initialized as a nop */ - *data = 0; - return 0; -} - int __init ftrace_dyn_arch_init(void *data) { extern const unsigned char ftrace_test_p6nop[]; -- cgit v1.2.3 From 15adc048986f6b54b6044f2b6fc4b48f49413e2f Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 23 Oct 2008 09:33:08 -0400 Subject: ftrace, powerpc, sparc64, x86: remove notrace from arch ftrace file The entire file of ftrace.c in the arch code needs to be marked as notrace. It is much cleaner to do this from the Makefile with CFLAGS_REMOVE_ftrace.o. [ powerpc already had this in its Makefile. ] Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/Makefile | 1 + arch/x86/kernel/ftrace.c | 10 +++++----- 2 files changed, 6 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index f1283fe6072..e489ff9cb3e 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -11,6 +11,7 @@ ifdef CONFIG_FUNCTION_TRACER CFLAGS_REMOVE_tsc.o = -pg CFLAGS_REMOVE_rtc.o = -pg CFLAGS_REMOVE_paravirt-spinlocks.o = -pg +CFLAGS_REMOVE_ftrace.o = -pg endif # diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index b399eed2353..b1e5e2244ec 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -33,17 +33,17 @@ union ftrace_code_union { }; -static int notrace ftrace_calc_offset(long ip, long addr) +static int ftrace_calc_offset(long ip, long addr) { return (int)(addr - ip); } -notrace unsigned char *ftrace_nop_replace(void) +unsigned char *ftrace_nop_replace(void) { return (char *)ftrace_nop; } -notrace unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) +unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) { static union ftrace_code_union calc; @@ -57,7 +57,7 @@ notrace unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) return calc.code; } -notrace int +int ftrace_modify_code(unsigned long ip, unsigned char *old_code, unsigned char *new_code) { @@ -90,7 +90,7 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, return 0; } -notrace int ftrace_update_ftrace_func(ftrace_func_t func) +int ftrace_update_ftrace_func(ftrace_func_t func) { unsigned long ip = (unsigned long)(&ftrace_call); unsigned char old[MCOUNT_INSN_SIZE], *new; -- cgit v1.2.3 From 75bebb7f0c2a709812cccb4d3151a21b012c5cad Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Thu, 23 Oct 2008 20:46:55 +0900 Subject: x86: use GFP_DMA for 24bit coherent_dma_mask dma_alloc_coherent (include/asm-x86/dma-mapping.h) avoids GFP_DMA allocation first and if the allocated address is not fit for the device's coherent_dma_mask, then dma_alloc_coherent does GFP_DMA allocation. This is because dma_alloc_coherent avoids precious GFP_DMA zone if possible. This is also how the old dma_alloc_coherent (arch/x86/kernel/pci-dma.c) works. However, if the coherent_dma_mask of a device is 24bit, there is no point to go into the above GFP_DMA retry mechanism. We had better use GFP_DMA in the first place. Signed-off-by: FUJITA Tomonori Tested-by: Takashi Iwai Signed-off-by: Ingo Molnar --- arch/x86/include/asm/dma-mapping.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/dma-mapping.h b/arch/x86/include/asm/dma-mapping.h index 4a5397bfce2..7f225a4b2a2 100644 --- a/arch/x86/include/asm/dma-mapping.h +++ b/arch/x86/include/asm/dma-mapping.h @@ -255,9 +255,11 @@ static inline unsigned long dma_alloc_coherent_mask(struct device *dev, static inline gfp_t dma_alloc_coherent_gfp_flags(struct device *dev, gfp_t gfp) { -#ifdef CONFIG_X86_64 unsigned long dma_mask = dma_alloc_coherent_mask(dev, gfp); + if (dma_mask <= DMA_24BIT_MASK) + gfp |= GFP_DMA; +#ifdef CONFIG_X86_64 if (dma_mask <= DMA_32BIT_MASK && !(gfp & GFP_DMA)) gfp |= GFP_DMA32; #endif -- cgit v1.2.3 From 03967c5267b0e7312d1d55dc814d94cf190ca573 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Thu, 23 Oct 2008 23:14:29 +0900 Subject: x86: restore the old swiotlb alloc_coherent behavior This restores the old swiotlb alloc_coherent behavior (before the alloc_coherent rewrite): http://lkml.org/lkml/2008/8/12/200 The old alloc_coherent avoids GFP_DMA allocation first and if the allocated address is not fit for the device's coherent_dma_mask, then dma_alloc_coherent does GFP_DMA allocation. If it fails, alloc_coherent calls swiotlb_alloc_coherent (in short, we rarely used swiotlb_alloc_coherent). After the alloc_coherent rewrite, dma_alloc_coherent (include/asm-x86/dma-mapping.h) directly calls swiotlb_alloc_coherent. It means that we possibly can't handle a device having dma_masks > 24bit < 32bits since swiotlb_alloc_coherent doesn't have the above GFP_DMA retry mechanism. This patch fixes x86's swiotlb alloc_coherent to use the GFP_DMA retry mechanism, which dma_generic_alloc_coherent() provides now (pci-nommu.c and GART IOMMU driver also use dma_generic_alloc_coherent). Signed-off-by: FUJITA Tomonori Signed-off-by: Ingo Molnar --- arch/x86/kernel/pci-swiotlb_64.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-swiotlb_64.c b/arch/x86/kernel/pci-swiotlb_64.c index c4ce0332759..3c539d111ab 100644 --- a/arch/x86/kernel/pci-swiotlb_64.c +++ b/arch/x86/kernel/pci-swiotlb_64.c @@ -18,9 +18,21 @@ swiotlb_map_single_phys(struct device *hwdev, phys_addr_t paddr, size_t size, return swiotlb_map_single(hwdev, phys_to_virt(paddr), size, direction); } +static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size, + dma_addr_t *dma_handle, gfp_t flags) +{ + void *vaddr; + + vaddr = dma_generic_alloc_coherent(hwdev, size, dma_handle, flags); + if (vaddr) + return vaddr; + + return swiotlb_alloc_coherent(hwdev, size, dma_handle, flags); +} + struct dma_mapping_ops swiotlb_dma_ops = { .mapping_error = swiotlb_dma_mapping_error, - .alloc_coherent = swiotlb_alloc_coherent, + .alloc_coherent = x86_swiotlb_alloc_coherent, .free_coherent = swiotlb_free_coherent, .map_single = swiotlb_map_single_phys, .unmap_single = swiotlb_unmap_single, -- cgit v1.2.3 From 3b15e581981b3ad35809f56d8131d5c19b6da1bd Mon Sep 17 00:00:00 2001 From: Fenghua Yu Date: Thu, 23 Oct 2008 16:51:00 -0700 Subject: x86/PCI: build failure at x86/kernel/pci-dma.c with !CONFIG_PCI On Thu, Oct 23, 2008 at 04:09:52PM -0700, Alexander Beregalov wrote: > arch/x86/kernel/built-in.o: In function `iommu_setup': > pci-dma.c:(.init.text+0x36ad): undefined reference to `forbid_dac' > pci-dma.c:(.init.text+0x36cc): undefined reference to `forbid_dac' > pci-dma.c:(.init.text+0x3711): undefined reference to `forbid_dac This patch partially reverts a patch to add IOMMU support to ia64. The forbid_dac variable was incorrectly moved to quirks.c, which isn't built when PCI is disabled. Tested-by: "Alexander Beregalov" Acked-by: FUJITA Tomonori Signed-off-by: Fenghua Yu Signed-off-by: Jesse Barnes --- arch/x86/include/asm/iommu.h | 1 - arch/x86/kernel/pci-dma.c | 16 ++++++++++++++++ 2 files changed, 16 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/iommu.h b/arch/x86/include/asm/iommu.h index 98e28ea8cd1..e4a552d4446 100644 --- a/arch/x86/include/asm/iommu.h +++ b/arch/x86/include/asm/iommu.h @@ -7,7 +7,6 @@ extern struct dma_mapping_ops nommu_dma_ops; extern int force_iommu, no_iommu; extern int iommu_detected; extern int dmar_disabled; -extern int forbid_dac; extern unsigned long iommu_nr_pages(unsigned long addr, unsigned long len); diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index 1972266e8ba..19262482021 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -9,6 +9,8 @@ #include #include +static int forbid_dac __read_mostly; + struct dma_mapping_ops *dma_ops; EXPORT_SYMBOL(dma_ops); @@ -291,3 +293,17 @@ void pci_iommu_shutdown(void) } /* Must execute after PCI subsystem */ fs_initcall(pci_iommu_init); + +#ifdef CONFIG_PCI +/* Many VIA bridges seem to corrupt data for DAC. Disable it here */ + +static __devinit void via_no_dac(struct pci_dev *dev) +{ + if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI && forbid_dac == 0) { + printk(KERN_INFO "PCI: VIA PCI bridge detected." + "Disabling DAC.\n"); + forbid_dac = 1; + } +} +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_ANY_ID, via_no_dac); +#endif -- cgit v1.2.3 From 9f32d21c981bb638d0991ce5675a20337312066b Mon Sep 17 00:00:00 2001 From: Chris Lalancette Date: Thu, 23 Oct 2008 17:40:25 -0700 Subject: xen: fix Xen domU boot with batched mprotect Impact: fix guest kernel boot crash on certain configs Recent i686 2.6.27 kernels with a certain amount of memory (between 736 and 855MB) have a problem booting under a hypervisor that supports batched mprotect (this includes the RHEL-5 Xen hypervisor as well as any 3.3 or later Xen hypervisor). The problem ends up being that xen_ptep_modify_prot_commit() is using virt_to_machine to calculate which pfn to update. However, this only works for pages that are in the p2m list, and the pages coming from change_pte_range() in mm/mprotect.c are kmap_atomic pages. Because of this, we can run into the situation where the lookup in the p2m table returns an INVALID_MFN, which we then try to pass to the hypervisor, which then (correctly) denies the request to a totally bogus pfn. The right thing to do is to use arbitrary_virt_to_machine, so that we can be sure we are modifying the right pfn. This unfortunately introduces a performance penalty because of a full page-table-walk, but we can avoid that penalty for pages in the p2m list by checking if virt_addr_valid is true, and if so, just doing the lookup in the p2m table. The attached patch implements this, and allows my 2.6.27 i686 based guest with 768MB of memory to boot on a RHEL-5 hypervisor again. Thanks to Jeremy for the suggestions about how to fix this particular issue. Signed-off-by: Chris Lalancette Signed-off-by: Jeremy Fitzhardinge Cc: Chris Lalancette Signed-off-by: Ingo Molnar --- arch/x86/xen/mmu.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index d4d52f5a1cf..aba77b2b7d1 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -246,11 +246,21 @@ xmaddr_t arbitrary_virt_to_machine(void *vaddr) { unsigned long address = (unsigned long)vaddr; unsigned int level; - pte_t *pte = lookup_address(address, &level); - unsigned offset = address & ~PAGE_MASK; + pte_t *pte; + unsigned offset; - BUG_ON(pte == NULL); + /* + * if the PFN is in the linear mapped vaddr range, we can just use + * the (quick) virt_to_machine() p2m lookup + */ + if (virt_addr_valid(vaddr)) + return virt_to_machine(vaddr); + /* otherwise we have to do a (slower) full page-table walk */ + + pte = lookup_address(address, &level); + BUG_ON(pte == NULL); + offset = address & ~PAGE_MASK; return XMADDR(((phys_addr_t)pte_mfn(*pte) << PAGE_SHIFT) + offset); } @@ -410,7 +420,7 @@ void xen_ptep_modify_prot_commit(struct mm_struct *mm, unsigned long addr, xen_mc_batch(); - u.ptr = virt_to_machine(ptep).maddr | MMU_PT_UPDATE_PRESERVE_AD; + u.ptr = arbitrary_virt_to_machine(ptep).maddr | MMU_PT_UPDATE_PRESERVE_AD; u.val = pte_val_ma(pte); xen_extend_mmu_update(&u); -- cgit v1.2.3 From ef020ab0109aa5cd6eac2e93519b7641c9862828 Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Thu, 23 Oct 2008 17:54:05 -0500 Subject: x86/uv: memory allocation at initialization Impact: on SGI UV platforms, fix boot crash UV initialization is currently called too late to call alloc_bootmem_pages(). The current sequence is: start_kernel() mem_init() free_all_bootmem() <--- discard of bootmem rest_init() kernel_init() smp_prepare_cpus() native_smp_prepare_cpus() uv_system_init() <--- uses alloc_bootmem_pages() It should be calling kmalloc(). Signed-off-by: Cliff Wickman Signed-off-by: Ingo Molnar --- arch/x86/kernel/genx2apic_uv_x.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/genx2apic_uv_x.c b/arch/x86/kernel/genx2apic_uv_x.c index 680a06557c5..2c7dbdb9827 100644 --- a/arch/x86/kernel/genx2apic_uv_x.c +++ b/arch/x86/kernel/genx2apic_uv_x.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include @@ -398,16 +397,16 @@ void __init uv_system_init(void) printk(KERN_DEBUG "UV: Found %d blades\n", uv_num_possible_blades()); bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades(); - uv_blade_info = alloc_bootmem_pages(bytes); + uv_blade_info = kmalloc(bytes, GFP_KERNEL); get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes(); - uv_node_to_blade = alloc_bootmem_pages(bytes); + uv_node_to_blade = kmalloc(bytes, GFP_KERNEL); memset(uv_node_to_blade, 255, bytes); bytes = sizeof(uv_cpu_to_blade[0]) * num_possible_cpus(); - uv_cpu_to_blade = alloc_bootmem_pages(bytes); + uv_cpu_to_blade = kmalloc(bytes, GFP_KERNEL); memset(uv_cpu_to_blade, 255, bytes); blade = 0; -- cgit v1.2.3 From 8115f3f0c939c5db0fe3c6c6c58911fd3a205b1e Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 24 Oct 2008 09:12:17 -0400 Subject: ftrace: use a real variable for ftrace_nop in x86 Impact: avoid section mismatch warning, clean up The dynamic ftrace determines which nop is safe to use at start up. When it finds a safe nop for patching, it sets a pointer called ftrace_nop to point to the code. All call sites are then patched to this nop. Later, when tracing is turned on, this ftrace_nop variable is again used to compare the location to make sure it is a nop before we update it to an mcount call. If this fails just once, a warning is printed and ftrace is disabled. Rakib Mullick noted that the code that sets up the nop is a .init section where as the nop itself is in the .text section. This is needed because the nop is used later on after boot up. The problem is that the test of the nop jumps back to the setup code and causes a "section mismatch" warning. Rakib first recommended to convert the nop to .init.text, but as stated above, this would fail since that text is used later. The real solution is to extend Rabik's patch, and to make the ftrace_nop into an array, and just save the code from the assembly to this array. Now the section can stay as an init section, and we have a nop to use later on. Reported-by: Rakib Mullick Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index b1e5e2244ec..50ea0ac8c9b 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -21,8 +21,7 @@ #include -/* Long is fine, even if it is only 4 bytes ;-) */ -static unsigned long *ftrace_nop; +static unsigned char ftrace_nop[MCOUNT_INSN_SIZE]; union ftrace_code_union { char code[MCOUNT_INSN_SIZE]; @@ -40,7 +39,7 @@ static int ftrace_calc_offset(long ip, long addr) unsigned char *ftrace_nop_replace(void) { - return (char *)ftrace_nop; + return ftrace_nop; } unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) @@ -125,9 +124,6 @@ int __init ftrace_dyn_arch_init(void *data) * TODO: check the cpuid to determine the best nop. */ asm volatile ( - "jmp ftrace_test_jmp\n" - /* This code needs to stay around */ - ".section .text, \"ax\"\n" "ftrace_test_jmp:" "jmp ftrace_test_p6nop\n" "nop\n" @@ -138,8 +134,6 @@ int __init ftrace_dyn_arch_init(void *data) "jmp 1f\n" "ftrace_test_nop5:" ".byte 0x66,0x66,0x66,0x66,0x90\n" - "jmp 1f\n" - ".previous\n" "1:" ".section .fixup, \"ax\"\n" "2: movl $1, %0\n" @@ -154,15 +148,15 @@ int __init ftrace_dyn_arch_init(void *data) switch (faulted) { case 0: pr_info("ftrace: converting mcount calls to 0f 1f 44 00 00\n"); - ftrace_nop = (unsigned long *)ftrace_test_p6nop; + memcpy(ftrace_nop, ftrace_test_p6nop, MCOUNT_INSN_SIZE); break; case 1: pr_info("ftrace: converting mcount calls to 66 66 66 66 90\n"); - ftrace_nop = (unsigned long *)ftrace_test_nop5; + memcpy(ftrace_nop, ftrace_test_nop5, MCOUNT_INSN_SIZE); break; case 2: pr_info("ftrace: converting mcount calls to jmp . + 5\n"); - ftrace_nop = (unsigned long *)ftrace_test_jmp; + memcpy(ftrace_nop, ftrace_test_jmp, MCOUNT_INSN_SIZE); break; } -- cgit v1.2.3 From 3afa39493de510c33c56ddc76e6e1af7f87c5392 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sat, 25 Oct 2008 22:58:21 -0700 Subject: x86: keep the /proc/meminfo page count correct Impact: get correct page count in /proc/meminfo found page count in /proc/meminfo is nor correct on 1G system in VirtualBox 2.0.4 # cat /proc/meminfo MemTotal: 1017508 kB MemFree: 822700 kB Buffers: 1456 kB Cached: 26632 kB SwapCached: 0 kB ... Hugepagesize: 2048 kB DirectMap4k: 4032 kB DirectMap2M: 18446744073709549568 kB with this patch get: ... DirectMap4k: 4032 kB DirectMap2M: 1044480 kB which is consistent to kernel_page_tables ---[ Low Kernel Mapping ]--- 0xffff880000000000-0xffff880000001000 4K RW PCD GLB x pte 0xffff880000001000-0xffff88000009f000 632K RW GLB x pte 0xffff88000009f000-0xffff8800000a0000 4K RW PCD GLB x pte 0xffff8800000a0000-0xffff880000200000 1408K RW GLB x pte 0xffff880000200000-0xffff88003fe00000 1020M RW PSE GLB x pmd 0xffff88003fe00000-0xffff88003fff0000 1984K RW GLB NX pte 0xffff88003fff0000-0xffff880040000000 64K pte 0xffff880040000000-0xffff888000000000 511G pud 0xffff888000000000-0xffffc20000000000 58880G pgd Signed-off-by: Yinghai Lu Acked-by: Suresh Siddha Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index b8e461d4941..c7a4c5a9a21 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -350,8 +350,10 @@ phys_pte_init(pte_t *pte_page, unsigned long addr, unsigned long end, * pagetable pages as RO. So assume someone who pre-setup * these mappings are more intelligent. */ - if (pte_val(*pte)) + if (pte_val(*pte)) { + pages++; continue; + } if (0) printk(" pte=%p addr=%lx pte=%016lx\n", @@ -418,8 +420,10 @@ phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end, * not differ with respect to page frame and * attributes. */ - if (page_size_mask & (1 << PG_LEVEL_2M)) + if (page_size_mask & (1 << PG_LEVEL_2M)) { + pages++; continue; + } new_prot = pte_pgprot(pte_clrhuge(*(pte_t *)pmd)); } @@ -499,8 +503,10 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end, * not differ with respect to page frame and * attributes. */ - if (page_size_mask & (1 << PG_LEVEL_1G)) + if (page_size_mask & (1 << PG_LEVEL_1G)) { + pages++; continue; + } prot = pte_pgprot(pte_clrhuge(*(pte_t *)pud)); } -- cgit v1.2.3 From 60817c9b31ef7897d60bca2f384cbc316a3fdd8b Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Mon, 27 Oct 2008 13:03:18 -0700 Subject: x86, memory hotplug: remove wrong -1 in calling init_memory_mapping() Impact: fix crash with memory hotplug Shuahua Li found: | I just did some experiments on a desktop for memory hotplug and this bug | triggered a crash in my test. | | Yinghai's suggestion also fixed the bug. We don't need to round it, just remove that extra -1 Signed-off-by: Yinghai Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index c7a4c5a9a21..f79a02f64d1 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -837,7 +837,7 @@ int arch_add_memory(int nid, u64 start, u64 size) unsigned long nr_pages = size >> PAGE_SHIFT; int ret; - last_mapped_pfn = init_memory_mapping(start, start + size-1); + last_mapped_pfn = init_memory_mapping(start, start + size); if (last_mapped_pfn > max_pfn_mapped) max_pfn_mapped = last_mapped_pfn; -- cgit v1.2.3 From 6ad9f15c94822c3f067a7d443f3b414e08b34460 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 15 Oct 2008 07:45:08 -0200 Subject: KVM: MMU: sync root on paravirt TLB flush The pvmmu TLB flush handler should request a root sync, similarly to a native read-write CR3. Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- arch/x86/kvm/mmu.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 99c239c5c0a..2a5e64881d9 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -2634,6 +2634,7 @@ static int kvm_pv_mmu_write(struct kvm_vcpu *vcpu, static int kvm_pv_mmu_flush_tlb(struct kvm_vcpu *vcpu) { kvm_x86_ops->tlb_flush(vcpu); + set_bit(KVM_REQ_MMU_SYNC, &vcpu->requests); return 1; } -- cgit v1.2.3 From 5550af4df179e52753d3a43a788a113ad8cd95cd Mon Sep 17 00:00:00 2001 From: Sheng Yang Date: Wed, 15 Oct 2008 20:15:06 +0800 Subject: KVM: Fix guest shared interrupt with in-kernel irqchip Every call of kvm_set_irq() should offer an irq_source_id, which is allocated by kvm_request_irq_source_id(). Based on irq_source_id, we identify the irq source and implement logical OR for shared level interrupts. The allocated irq_source_id can be freed by kvm_free_irq_source_id(). Currently, we support at most sizeof(unsigned long) different irq sources. [Amit: - rebase to kvm.git HEAD - move definition of KVM_USERSPACE_IRQ_SOURCE_ID to common file - move kvm_request_irq_source_id to the update_irq ioctl] [Xiantao: - Add kvm/ia64 stuff and make it work for kvm/ia64 guests] Signed-off-by: Sheng Yang Signed-off-by: Amit Shah Signed-off-by: Xiantao Zhang Signed-off-by: Avi Kivity --- arch/x86/include/asm/kvm_host.h | 3 +++ arch/x86/kvm/i8254.c | 11 +++++++++-- arch/x86/kvm/i8254.h | 1 + arch/x86/kvm/x86.c | 6 +++++- 4 files changed, 18 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 65679d00633..8346be87cfa 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -364,6 +364,9 @@ struct kvm_arch{ struct page *ept_identity_pagetable; bool ept_identity_pagetable_done; + + unsigned long irq_sources_bitmap; + unsigned long irq_states[KVM_IOAPIC_NUM_PINS]; }; struct kvm_vm_stat { diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c index 11c6725fb79..8772dc94682 100644 --- a/arch/x86/kvm/i8254.c +++ b/arch/x86/kvm/i8254.c @@ -545,6 +545,12 @@ struct kvm_pit *kvm_create_pit(struct kvm *kvm) if (!pit) return NULL; + mutex_lock(&kvm->lock); + pit->irq_source_id = kvm_request_irq_source_id(kvm); + mutex_unlock(&kvm->lock); + if (pit->irq_source_id < 0) + return NULL; + mutex_init(&pit->pit_state.lock); mutex_lock(&pit->pit_state.lock); spin_lock_init(&pit->pit_state.inject_lock); @@ -587,6 +593,7 @@ void kvm_free_pit(struct kvm *kvm) mutex_lock(&kvm->arch.vpit->pit_state.lock); timer = &kvm->arch.vpit->pit_state.pit_timer.timer; hrtimer_cancel(timer); + kvm_free_irq_source_id(kvm, kvm->arch.vpit->irq_source_id); mutex_unlock(&kvm->arch.vpit->pit_state.lock); kfree(kvm->arch.vpit); } @@ -595,8 +602,8 @@ void kvm_free_pit(struct kvm *kvm) static void __inject_pit_timer_intr(struct kvm *kvm) { mutex_lock(&kvm->lock); - kvm_set_irq(kvm, 0, 1); - kvm_set_irq(kvm, 0, 0); + kvm_set_irq(kvm, kvm->arch.vpit->irq_source_id, 0, 1); + kvm_set_irq(kvm, kvm->arch.vpit->irq_source_id, 0, 0); mutex_unlock(&kvm->lock); } diff --git a/arch/x86/kvm/i8254.h b/arch/x86/kvm/i8254.h index e436d4983aa..4178022b97a 100644 --- a/arch/x86/kvm/i8254.h +++ b/arch/x86/kvm/i8254.h @@ -44,6 +44,7 @@ struct kvm_pit { struct kvm_io_device speaker_dev; struct kvm *kvm; struct kvm_kpit_state pit_state; + int irq_source_id; }; #define KVM_PIT_BASE_ADDRESS 0x40 diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 4f0677d1eae..f1f8ff2f1fa 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1742,7 +1742,8 @@ long kvm_arch_vm_ioctl(struct file *filp, goto out; if (irqchip_in_kernel(kvm)) { mutex_lock(&kvm->lock); - kvm_set_irq(kvm, irq_event.irq, irq_event.level); + kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID, + irq_event.irq, irq_event.level); mutex_unlock(&kvm->lock); r = 0; } @@ -4013,6 +4014,9 @@ struct kvm *kvm_arch_create_vm(void) INIT_LIST_HEAD(&kvm->arch.active_mmu_pages); INIT_LIST_HEAD(&kvm->arch.assigned_dev_head); + /* Reserve bit 0 of irq_sources_bitmap for userspace irq source */ + set_bit(KVM_USERSPACE_IRQ_SOURCE_ID, &kvm->arch.irq_sources_bitmap); + return kvm; } -- cgit v1.2.3 From 531f6ed7de911e975352fbb2b228367121da630a Mon Sep 17 00:00:00 2001 From: Markus Metzger Date: Fri, 17 Oct 2008 09:09:27 +0200 Subject: x86, bts: improve help text for BTS config Improve the help text of the X86_PTRACE_BTS config. Make X86_DS invisible and depend on X86_PTRACE_BTS. Reported-by: Roland Dreier Signed-off-by: Markus Metzger Signed-off-by: Ingo Molnar --- arch/x86/Kconfig.cpu | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index 0b7c4a3f065..b815664fe37 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -513,19 +513,19 @@ config CPU_SUP_UMC_32 If unsure, say N. config X86_DS - bool "Debug Store support" - default y - help - Add support for Debug Store. - This allows the kernel to provide a memory buffer to the hardware - to store various profiling and tracing events. + def_bool X86_PTRACE_BTS + depends on X86_DEBUGCTLMSR config X86_PTRACE_BTS - bool "ptrace interface to Branch Trace Store" + bool "Branch Trace Store" default y - depends on (X86_DS && X86_DEBUGCTLMSR) + depends on X86_DEBUGCTLMSR help - Add a ptrace interface to allow collecting an execution trace - of the traced task. - This collects control flow changes in a (cyclic) buffer and allows - debuggers to fill in the gaps and show an execution trace of the debuggee. + This adds a ptrace interface to the hardware's branch trace store. + + Debuggers may use it to collect an execution trace of the debugged + application in order to answer the question 'how did I get here?'. + Debuggers may trace user mode as well as kernel mode. + + Say Y unless there is no application development on this machine + and you want to save a small amount of code size. -- cgit v1.2.3 From 36b75da27bb51dc34e358d0b7487406132806c46 Mon Sep 17 00:00:00 2001 From: Peter Oruba Date: Fri, 17 Oct 2008 15:30:37 +0200 Subject: x86: microcode patch loader author update Removed one author's email address from module init message. Signed-off-by: Peter Oruba Signed-off-by: Ingo Molnar --- arch/x86/kernel/microcode_core.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/microcode_core.c b/arch/x86/kernel/microcode_core.c index 936d8d55f23..82fb2809ce3 100644 --- a/arch/x86/kernel/microcode_core.c +++ b/arch/x86/kernel/microcode_core.c @@ -480,8 +480,8 @@ static int __init microcode_init(void) printk(KERN_INFO "Microcode Update Driver: v" MICROCODE_VERSION - " " - " \n"); + " ," + " Peter Oruba\n"); return 0; } -- cgit v1.2.3 From 3c52204bb90834bca8e9e78a3628d886ad6d4db5 Mon Sep 17 00:00:00 2001 From: Peter Oruba Date: Fri, 17 Oct 2008 15:30:38 +0200 Subject: x86: AMD microcode patch loader author update Removed author's email address from MODULE_AUTHOR. Signed-off-by: Peter Oruba Signed-off-by: Ingo Molnar --- arch/x86/kernel/microcode_amd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c index 7a1f8eeac2c..5f8e5d75a25 100644 --- a/arch/x86/kernel/microcode_amd.c +++ b/arch/x86/kernel/microcode_amd.c @@ -39,7 +39,7 @@ #include MODULE_DESCRIPTION("AMD Microcode Update Driver"); -MODULE_AUTHOR("Peter Oruba "); +MODULE_AUTHOR("Peter Oruba"); MODULE_LICENSE("GPL v2"); #define UCODE_MAGIC 0x00414d44 -- cgit v1.2.3 From 1281675e9c0d4d42d993697f4daab45ef22d49da Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 14 Oct 2008 18:59:17 -0700 Subject: x86: fix APIC_DEBUG with inquire_remote_apic APIC_DEBUG is always 2. need to update inquire_remote_apic to check apic_verbosity with it instead. Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/include/asm/es7000/wakecpu.h | 9 ++++----- arch/x86/include/asm/mach-default/mach_wakecpu.h | 9 ++++----- 2 files changed, 8 insertions(+), 10 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/es7000/wakecpu.h b/arch/x86/include/asm/es7000/wakecpu.h index 3ffc5a7bf66..39849346191 100644 --- a/arch/x86/include/asm/es7000/wakecpu.h +++ b/arch/x86/include/asm/es7000/wakecpu.h @@ -50,10 +50,9 @@ static inline void restore_NMI_vector(unsigned short *high, unsigned short *low) { } -#if APIC_DEBUG - #define inquire_remote_apic(apicid) __inquire_remote_apic(apicid) -#else - #define inquire_remote_apic(apicid) {} -#endif +#define inquire_remote_apic(apicid) do { \ + if (apic_verbosity >= APIC_DEBUG) \ + __inquire_remote_apic(apicid); \ + } while (0) #endif /* __ASM_MACH_WAKECPU_H */ diff --git a/arch/x86/include/asm/mach-default/mach_wakecpu.h b/arch/x86/include/asm/mach-default/mach_wakecpu.h index d5c0b826a4f..9d80db91e99 100644 --- a/arch/x86/include/asm/mach-default/mach_wakecpu.h +++ b/arch/x86/include/asm/mach-default/mach_wakecpu.h @@ -33,10 +33,9 @@ static inline void restore_NMI_vector(unsigned short *high, unsigned short *low) { } -#if APIC_DEBUG - #define inquire_remote_apic(apicid) __inquire_remote_apic(apicid) -#else - #define inquire_remote_apic(apicid) {} -#endif +#define inquire_remote_apic(apicid) do { \ + if (apic_verbosity >= APIC_DEBUG) \ + __inquire_remote_apic(apicid); \ + } while (0) #endif /* _ASM_X86_MACH_DEFAULT_MACH_WAKECPU_H */ -- cgit v1.2.3 From e7706fc691513b0f06adb3de3d6ac04293180146 Mon Sep 17 00:00:00 2001 From: Ken'ichi Ohmichi Date: Mon, 20 Oct 2008 13:51:52 +0900 Subject: x86, kdump: fix invalid access on i386 sparsemem Impact: fix kdump crash on 32-bit sparsemem kernels Since linux-2.6.27, kdump has failed on i386 sparsemem kernel. 1st-kernel gets a panic just before switching to 2nd-kernel. The cause is that a kernel accesses invalid mem_section by page_to_pfn(image->swap_page) at machine_kexec(). image->swap_page is allocated if kexec for hibernation, but it is not allocated if kdump. So if kdump, a kernel should not access the mem_section corresponding to image->swap_page. The attached patch fixes this invalid access. Signed-off-by: Ken'ichi Ohmichi Cc: kexec-ml Cc: Andrew Morton Signed-off-by: Ingo Molnar --- arch/x86/kernel/machine_kexec_32.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index 0732adba05c..7a385746509 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -162,7 +162,10 @@ void machine_kexec(struct kimage *image) page_list[VA_PTE_0] = (unsigned long)kexec_pte0; page_list[PA_PTE_1] = __pa(kexec_pte1); page_list[VA_PTE_1] = (unsigned long)kexec_pte1; - page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page) << PAGE_SHIFT); + + if (image->type == KEXEC_TYPE_DEFAULT) + page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page) + << PAGE_SHIFT); /* The segment registers are funny things, they have both a * visible and an invisible part. Whenever the visible part is -- cgit v1.2.3 From 11a6b0c933b55654a58afd84f63a5dde1607d78f Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 14 Oct 2008 18:59:18 -0700 Subject: x86: 64 bit print out absent pages num too so users are not confused with memhole causing big total ram we don't need to worry about 32 bit, because memhole is always above max_low_pfn. Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index f79a02f64d1..ad38648bddb 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -884,6 +884,7 @@ static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel, void __init mem_init(void) { long codesize, reservedpages, datasize, initsize; + unsigned long absent_pages; start_periodic_check_for_corruption(); @@ -899,8 +900,9 @@ void __init mem_init(void) #else totalram_pages = free_all_bootmem(); #endif - reservedpages = max_pfn - totalram_pages - - absent_pages_in_range(0, max_pfn); + + absent_pages = absent_pages_in_range(0, max_pfn); + reservedpages = max_pfn - totalram_pages - absent_pages; after_bootmem = 1; codesize = (unsigned long) &_etext - (unsigned long) &_text; @@ -917,10 +919,11 @@ void __init mem_init(void) VSYSCALL_END - VSYSCALL_START); printk(KERN_INFO "Memory: %luk/%luk available (%ldk kernel code, " - "%ldk reserved, %ldk data, %ldk init)\n", + "%ldk absent, %ldk reserved, %ldk data, %ldk init)\n", (unsigned long) nr_free_pages() << (PAGE_SHIFT-10), max_pfn << (PAGE_SHIFT-10), codesize >> 10, + absent_pages << (PAGE_SHIFT-10), reservedpages << (PAGE_SHIFT-10), datasize >> 10, initsize >> 10); -- cgit v1.2.3 From 87c6f40128f92621698f97a62d2ead5184d1dd97 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Tue, 28 Oct 2008 16:13:54 +0100 Subject: x86, gart: fix gart detection for Fam11h CPUs Impact: fix AMD Family 11h boot hangs / USB device problems The AMD Fam11h CPUs have a K8 northbridge. This northbridge is different from other family's because it lacks GART support (as I just learned). But the kernel implicitly expects a GART if it finds an AMD northbridge. Fix this by removing the Fam11h northbridge id from the scan list of K8 northbridges. This patch also changes the message in the GART driver about missing K8 northbridges to tell that the GART is missing which is the correct information in this case. Reported-by: Jouni Malinen Signed-off-by: Joerg Roedel Signed-off-by: Ingo Molnar --- arch/x86/kernel/k8.c | 1 - arch/x86/kernel/pci-gart_64.c | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/k8.c b/arch/x86/kernel/k8.c index 304d8bad655..cbc4332a77b 100644 --- a/arch/x86/kernel/k8.c +++ b/arch/x86/kernel/k8.c @@ -18,7 +18,6 @@ static u32 *flush_words; struct pci_device_id k8_nb_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) }, - { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_11H_NB_MISC) }, {} }; EXPORT_SYMBOL(k8_nb_ids); diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index e3f75bbcede..a42b02b4df6 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -744,7 +744,7 @@ void __init gart_iommu_init(void) long i; if (cache_k8_northbridges() < 0 || num_k8_northbridges == 0) { - printk(KERN_INFO "PCI-GART: No AMD northbridge found.\n"); + printk(KERN_INFO "PCI-GART: No AMD GART found.\n"); return; } -- cgit v1.2.3 From f96f57d91c2df75011d1e260c23edca429f37361 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 28 Oct 2008 12:39:23 -0700 Subject: x86: fix init_memory_mapping for [dc000000 - e0000000) - v2 Impact: change over-mapping to precise mapping, fix /proc/meminfo output v2: fix less than 1G ram system handling when gart aperture is 0xdc000000 - 0xe0000000 it return 0xc0000000 - 0xe0000000 that is not right. this patch fix that will get exact mapping on 256g sytem with that aperture after patch LBSuse:~ # cat /proc/meminfo MemTotal: 264742432 kB MemFree: 263920628 kB Buffers: 1416 kB Cached: 24468 kB ... DirectMap4k: 5760 kB DirectMap2M: 3205120 kB DirectMap1G: 265289728 kB it is consistent to LBSuse:~ # cat /sys/kernel/debug/kernel_page_tables .. ---[ Low Kernel Mapping ]--- 0xffff880000000000-0xffff880000200000 2M RW GLB x pte 0xffff880000200000-0xffff880040000000 1022M RW PSE GLB x pmd 0xffff880040000000-0xffff8800c0000000 2G RW PSE GLB NX pud 0xffff8800c0000000-0xffff8800d7e00000 382M RW PSE GLB NX pmd 0xffff8800d7e00000-0xffff8800d7fa0000 1664K RW GLB NX pte 0xffff8800d7fa0000-0xffff8800d8000000 384K pte 0xffff8800d8000000-0xffff8800dc000000 64M pmd 0xffff8800dc000000-0xffff8800e0000000 64M RW PSE GLB NX pmd 0xffff8800e0000000-0xffff880100000000 512M pmd 0xffff880100000000-0xffff880800000000 28G RW PSE GLB NX pud 0xffff880800000000-0xffff880824600000 582M RW PSE GLB NX pmd 0xffff880824600000-0xffff8808247f0000 1984K RW GLB NX pte 0xffff8808247f0000-0xffff880824800000 64K RW PCD GLB NX pte 0xffff880824800000-0xffff880840000000 440M RW PSE GLB NX pmd 0xffff880840000000-0xffff884000000000 223G RW PSE GLB NX pud 0xffff884000000000-0xffff884028000000 640M RW PSE GLB NX pmd 0xffff884028000000-0xffff884040000000 384M pmd 0xffff884040000000-0xffff888000000000 255G pud 0xffff888000000000-0xffffc20000000000 58880G pgd Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 50 +++++++++++++++++++++++++++++++++----------------- 1 file changed, 33 insertions(+), 17 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ad38648bddb..ebe1811e5b1 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -671,12 +671,13 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, unsigned long last_map_addr = 0; unsigned long page_size_mask = 0; unsigned long start_pfn, end_pfn; + unsigned long pos; struct map_range mr[NR_RANGE_MR]; int nr_range, i; int use_pse, use_gbpages; - printk(KERN_INFO "init_memory_mapping\n"); + printk(KERN_INFO "init_memory_mapping: %016lx-%016lx\n", start, end); /* * Find space for the kernel direct mapping tables. @@ -710,35 +711,50 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, /* head if not big page alignment ?*/ start_pfn = start >> PAGE_SHIFT; - end_pfn = ((start + (PMD_SIZE - 1)) >> PMD_SHIFT) + pos = start_pfn << PAGE_SHIFT; + end_pfn = ((pos + (PMD_SIZE - 1)) >> PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT); - nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0); + if (start_pfn < end_pfn) { + nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0); + pos = end_pfn << PAGE_SHIFT; + } /* big page (2M) range*/ - start_pfn = ((start + (PMD_SIZE - 1))>>PMD_SHIFT) + start_pfn = ((pos + (PMD_SIZE - 1))>>PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT); - end_pfn = ((start + (PUD_SIZE - 1))>>PUD_SHIFT) + end_pfn = ((pos + (PUD_SIZE - 1))>>PUD_SHIFT) << (PUD_SHIFT - PAGE_SHIFT); - if (end_pfn > ((end>>PUD_SHIFT)<<(PUD_SHIFT - PAGE_SHIFT))) - end_pfn = ((end>>PUD_SHIFT)<<(PUD_SHIFT - PAGE_SHIFT)); - nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, - page_size_mask & (1< ((end>>PMD_SHIFT)<<(PMD_SHIFT - PAGE_SHIFT))) + end_pfn = ((end>>PMD_SHIFT)<<(PMD_SHIFT - PAGE_SHIFT)); + if (start_pfn < end_pfn) { + nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, + page_size_mask & (1<>PUD_SHIFT) << (PUD_SHIFT - PAGE_SHIFT); - nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, + start_pfn = ((pos + (PUD_SIZE - 1))>>PUD_SHIFT) + << (PUD_SHIFT - PAGE_SHIFT); + end_pfn = (end >> PUD_SHIFT) << (PUD_SHIFT - PAGE_SHIFT); + if (start_pfn < end_pfn) { + nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, page_size_mask & ((1<>PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT); - nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, - page_size_mask & (1<>PMD_SHIFT) + << (PMD_SHIFT - PAGE_SHIFT); + end_pfn = (end >> PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT); + if (start_pfn < end_pfn) { + nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, + page_size_mask & (1<>PAGE_SHIFT; end_pfn = end>>PAGE_SHIFT; nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0); -- cgit v1.2.3 From 9352f5698db2c6d7f2789f6cd37e3996d49ac4b5 Mon Sep 17 00:00:00 2001 From: Harvey Harrison Date: Tue, 28 Oct 2008 23:05:22 -0700 Subject: x86: two trivial sparse annotations Impact: fewer sparse warnings, no functional changes arch/x86/kernel/vsmp_64.c:87:14: warning: incorrect type in argument 1 (different address spaces) arch/x86/kernel/vsmp_64.c:87:14: expected void const volatile [noderef] *addr arch/x86/kernel/vsmp_64.c:87:14: got void *[assigned] address arch/x86/kernel/vsmp_64.c:88:22: warning: incorrect type in argument 1 (different address spaces) arch/x86/kernel/vsmp_64.c:88:22: expected void const volatile [noderef] *addr arch/x86/kernel/vsmp_64.c:88:22: got void * arch/x86/kernel/vsmp_64.c:100:23: warning: incorrect type in argument 2 (different address spaces) arch/x86/kernel/vsmp_64.c:100:23: expected void volatile [noderef] *addr arch/x86/kernel/vsmp_64.c:100:23: got void * arch/x86/kernel/vsmp_64.c:101:23: warning: incorrect type in argument 1 (different address spaces) arch/x86/kernel/vsmp_64.c:101:23: expected void const volatile [noderef] *addr arch/x86/kernel/vsmp_64.c:101:23: got void * arch/x86/mm/gup.c:235:6: warning: incorrect type in argument 1 (different base types) arch/x86/mm/gup.c:235:6: expected void const volatile [noderef] * arch/x86/mm/gup.c:235:6: got unsigned long [unsigned] [assigned] start Signed-off-by: Harvey Harrison Signed-off-by: Ingo Molnar --- arch/x86/kernel/vsmp_64.c | 2 +- arch/x86/mm/gup.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index 7766d36983f..a688f3bfaec 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -78,7 +78,7 @@ static unsigned __init_or_module vsmp_patch(u8 type, u16 clobbers, void *ibuf, static void __init set_vsmp_pv_ops(void) { - void *address; + void __iomem *address; unsigned int cap, ctl, cfg; /* set vSMP magic bits to indicate vSMP capable kernel */ diff --git a/arch/x86/mm/gup.c b/arch/x86/mm/gup.c index 4ba373c5b8c..be54176e9eb 100644 --- a/arch/x86/mm/gup.c +++ b/arch/x86/mm/gup.c @@ -233,7 +233,7 @@ int get_user_pages_fast(unsigned long start, int nr_pages, int write, len = (unsigned long) nr_pages << PAGE_SHIFT; end = start + len; if (unlikely(!access_ok(write ? VERIFY_WRITE : VERIFY_READ, - start, len))) + (void __user *)start, len))) goto slow_irqon; /* -- cgit v1.2.3 From 1d6cf1feb854c53c6d59e0d879603692b379e208 Mon Sep 17 00:00:00 2001 From: Harvey Harrison Date: Tue, 28 Oct 2008 22:46:04 -0700 Subject: x86: start annotating early ioremap pointers with __iomem Impact: some new sparse warnings in e820.c etc, but no functional change. As with regular ioremap, iounmap etc, annotate with __iomem. Fixes the following sparse warnings, will produce some new ones elsewhere in arch/x86 that will get worked out over time. arch/x86/mm/ioremap.c:402:9: warning: cast removes address space of expression arch/x86/mm/ioremap.c:406:10: warning: cast adds address space to expression () arch/x86/mm/ioremap.c:782:19: warning: Using plain integer as NULL pointer Signed-off-by: Harvey Harrison Signed-off-by: Ingo Molnar --- arch/x86/include/asm/io.h | 6 +++--- arch/x86/mm/ioremap.c | 22 +++++++++++----------- 2 files changed, 14 insertions(+), 14 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/io.h b/arch/x86/include/asm/io.h index 5618a103f39..ac2abc88cd9 100644 --- a/arch/x86/include/asm/io.h +++ b/arch/x86/include/asm/io.h @@ -82,9 +82,9 @@ extern void __iomem *ioremap_wc(unsigned long offset, unsigned long size); extern void early_ioremap_init(void); extern void early_ioremap_clear(void); extern void early_ioremap_reset(void); -extern void *early_ioremap(unsigned long offset, unsigned long size); -extern void *early_memremap(unsigned long offset, unsigned long size); -extern void early_iounmap(void *addr, unsigned long size); +extern void __iomem *early_ioremap(unsigned long offset, unsigned long size); +extern void __iomem *early_memremap(unsigned long offset, unsigned long size); +extern void early_iounmap(void __iomem *addr, unsigned long size); extern void __iomem *fix_ioremap(unsigned idx, unsigned long phys); diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index ae71e11eb3e..d4c4307ff3e 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -387,7 +387,7 @@ static void __iomem *ioremap_default(resource_size_t phys_addr, unsigned long size) { unsigned long flags; - void *ret; + void __iomem *ret; int err; /* @@ -399,11 +399,11 @@ static void __iomem *ioremap_default(resource_size_t phys_addr, if (err < 0) return NULL; - ret = (void *) __ioremap_caller(phys_addr, size, flags, - __builtin_return_address(0)); + ret = __ioremap_caller(phys_addr, size, flags, + __builtin_return_address(0)); free_memtype(phys_addr, phys_addr + size); - return (void __iomem *)ret; + return ret; } void __iomem *ioremap_prot(resource_size_t phys_addr, unsigned long size, @@ -622,7 +622,7 @@ static inline void __init early_clear_fixmap(enum fixed_addresses idx) __early_set_fixmap(idx, 0, __pgprot(0)); } -static void *prev_map[FIX_BTMAPS_SLOTS] __initdata; +static void __iomem *prev_map[FIX_BTMAPS_SLOTS] __initdata; static unsigned long prev_size[FIX_BTMAPS_SLOTS] __initdata; static int __init check_early_ioremap_leak(void) { @@ -645,7 +645,7 @@ static int __init check_early_ioremap_leak(void) } late_initcall(check_early_ioremap_leak); -static void __init *__early_ioremap(unsigned long phys_addr, unsigned long size, pgprot_t prot) +static void __init __iomem *__early_ioremap(unsigned long phys_addr, unsigned long size, pgprot_t prot) { unsigned long offset, last_addr; unsigned int nrpages; @@ -713,23 +713,23 @@ static void __init *__early_ioremap(unsigned long phys_addr, unsigned long size, if (early_ioremap_debug) printk(KERN_CONT "%08lx + %08lx\n", offset, fix_to_virt(idx0)); - prev_map[slot] = (void *) (offset + fix_to_virt(idx0)); + prev_map[slot] = (void __iomem *)(offset + fix_to_virt(idx0)); return prev_map[slot]; } /* Remap an IO device */ -void __init *early_ioremap(unsigned long phys_addr, unsigned long size) +void __init __iomem *early_ioremap(unsigned long phys_addr, unsigned long size) { return __early_ioremap(phys_addr, size, PAGE_KERNEL_IO); } /* Remap memory */ -void __init *early_memremap(unsigned long phys_addr, unsigned long size) +void __init __iomem *early_memremap(unsigned long phys_addr, unsigned long size) { return __early_ioremap(phys_addr, size, PAGE_KERNEL); } -void __init early_iounmap(void *addr, unsigned long size) +void __init early_iounmap(void __iomem *addr, unsigned long size) { unsigned long virt_addr; unsigned long offset; @@ -779,7 +779,7 @@ void __init early_iounmap(void *addr, unsigned long size) --idx; --nrpages; } - prev_map[slot] = 0; + prev_map[slot] = NULL; } void __this_fixmap_does_not_exist(void) -- cgit v1.2.3 From fe8b868eccb9f85a0e231e35f0abac5b39bac801 Mon Sep 17 00:00:00 2001 From: Gary Hade Date: Tue, 28 Oct 2008 16:43:14 -0700 Subject: x86: remove debug code from arch_add_memory() Impact: remove incorrect WARN_ON(1) Gets rid of dmesg spam created during physical memory hot-add which will very likely confuse users. The change removes what appears to be debugging code which I assume was unintentionally included in: x86: arch/x86/mm/init_64.c printk fixes commit 10f22dde556d1ed41d55355d1fb8ad495f9810c8 Signed-off-by: Gary Hade Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ebe1811e5b1..9db01db6e3c 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -858,7 +858,7 @@ int arch_add_memory(int nid, u64 start, u64 size) max_pfn_mapped = last_mapped_pfn; ret = __add_pages(zone, start_pfn, nr_pages); - WARN_ON(1); + WARN_ON_ONCE(ret); return ret; } -- cgit v1.2.3 From ab00fee30cddf975200b3c97aef25bea144a0d89 Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Thu, 30 Oct 2008 10:37:21 +0000 Subject: i386/PAE: fix pud_page() Impact: cleanup To the unsuspecting user it is quite annoying that this broken and inconsistent with x86-64 definition still exists. Signed-off-by: Jan Beulich Signed-off-by: Ingo Molnar --- arch/x86/include/asm/pgtable-3level.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/pgtable-3level.h b/arch/x86/include/asm/pgtable-3level.h index fb16cec702e..52597aeadff 100644 --- a/arch/x86/include/asm/pgtable-3level.h +++ b/arch/x86/include/asm/pgtable-3level.h @@ -120,13 +120,13 @@ static inline void pud_clear(pud_t *pudp) write_cr3(pgd); } -#define pud_page(pud) ((struct page *) __va(pud_val(pud) & PTE_PFN_MASK)) +#define pud_page(pud) pfn_to_page(pud_val(pud) >> PAGE_SHIFT) #define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PTE_PFN_MASK)) /* Find an entry in the second-level page table.. */ -#define pmd_offset(pud, address) ((pmd_t *)pud_page(*(pud)) + \ +#define pmd_offset(pud, address) ((pmd_t *)pud_page_vaddr(*(pud)) + \ pmd_index(address)) #ifdef CONFIG_SMP -- cgit v1.2.3 From ae9b9403644f3ecc76867af042e7e1cfd5c099d0 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 30 Oct 2008 17:43:57 +0100 Subject: AMD IOMMU: fix detection of NP capable IOMMUs This patch changes the code to use IOMMU_CAP_NPCACHE as a shift and not as a mask. Signed-off-by: Joerg Roedel --- arch/x86/kernel/amd_iommu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 3b346c6f551..38e88d40ab1 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -50,7 +50,7 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, /* returns !0 if the IOMMU is caching non-present entries in its TLB */ static int iommu_has_npcache(struct amd_iommu *iommu) { - return iommu->cap & IOMMU_CAP_NPCACHE; + return iommu->cap & (1UL << IOMMU_CAP_NPCACHE); } /**************************************************************************** -- cgit v1.2.3 From c17dad6905fc82d8f523399e5c3f014e81d61df6 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Wed, 29 Oct 2008 14:00:50 -0700 Subject: .gitignore updates Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/boot/compressed/.gitignore | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/boot/compressed/.gitignore b/arch/x86/boot/compressed/.gitignore index be0ed065249..63eff3b04d0 100644 --- a/arch/x86/boot/compressed/.gitignore +++ b/arch/x86/boot/compressed/.gitignore @@ -1 +1,3 @@ relocs +vmlinux.bin.all +vmlinux.relocs -- cgit v1.2.3 From c08b6acc9b996ba6231105cb12a4125c957e0c97 Mon Sep 17 00:00:00 2001 From: Mike Travis Date: Thu, 30 Oct 2008 11:33:19 -0700 Subject: x86, uv: fix compile error in uv_hub.h Impact: include file dependency cleanup Fix compile errors of files that include asm/uv/uv_hub.h but do not include linux/timer.h. [ such files are not mainline right now. ] Signed-of-by: Mike Travis Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_hub.h | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h index c6ad93e315c..7a5782610b2 100644 --- a/arch/x86/include/asm/uv/uv_hub.h +++ b/arch/x86/include/asm/uv/uv_hub.h @@ -13,6 +13,7 @@ #include #include +#include #include #include -- cgit v1.2.3 From 08c33308575b370c89b4ed1198ece5f93145a2aa Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 30 Oct 2008 16:08:38 -0500 Subject: x86/voyager: fix boot breakage caused by x86: boot secondary cpus through initial_code Impact: boot up secondary CPUs as well on x86/Voyager systems This commit: | commit 3e9704739daf46a8ba6593d749c67b5f7cd633d2 | Author: Glauber Costa | Date: Wed May 28 13:01:54 2008 -0300 | | x86: boot secondary cpus through initial_code removed the use of initialize_secondary. However, it didn't update voyager, so the secondary cpus no longer boot. Fix this by adding the initial_code switch to voyager as well. Signed-off-by: James Bottomley Cc: Glauber Costa Signed-off-by: Ingo Molnar --- arch/x86/mach-voyager/voyager_smp.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c index 0f6e8a6523a..9cd327a278a 100644 --- a/arch/x86/mach-voyager/voyager_smp.c +++ b/arch/x86/mach-voyager/voyager_smp.c @@ -90,6 +90,7 @@ static void ack_vic_irq(unsigned int irq); static void vic_enable_cpi(void); static void do_boot_cpu(__u8 cpuid); static void do_quad_bootstrap(void); +static void initialize_secondary(void); int hard_smp_processor_id(void); int safe_smp_processor_id(void); @@ -650,6 +651,8 @@ void __init smp_boot_cpus(void) smp_tune_scheduling(); */ smp_store_cpu_info(boot_cpu_id); + /* setup the jump vector */ + initial_code = (unsigned long)initialize_secondary; printk("CPU%d: ", boot_cpu_id); print_cpu_info(&cpu_data(boot_cpu_id)); @@ -702,7 +705,7 @@ void __init smp_boot_cpus(void) /* Reload the secondary CPUs task structure (this function does not * return ) */ -void __init initialize_secondary(void) +static void __init initialize_secondary(void) { #if 0 // AC kernels only -- cgit v1.2.3 From 017d9d20d88cacb0a6a29f343b23c95e203f6645 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 30 Oct 2008 16:05:39 -0500 Subject: x86: use CONFIG_X86_SMP instead of CONFIG_SMP Impact: fix x86/Voyager boot CONFIG_SMP is used for features which work on *all* x86 boxes. CONFIG_X86_SMP is used for standard PC like x86 boxes (for things like multi core and apics) Signed-off-by: James Bottomley Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/addon_cpuid_features.c | 2 +- arch/x86/kernel/tsc.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c index 0d9c993aa93..ef8f831af82 100644 --- a/arch/x86/kernel/cpu/addon_cpuid_features.c +++ b/arch/x86/kernel/cpu/addon_cpuid_features.c @@ -69,7 +69,7 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c) */ void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c) { -#ifdef CONFIG_SMP +#ifdef CONFIG_X86_SMP unsigned int eax, ebx, ecx, edx, sub_index; unsigned int ht_mask_width, core_plus_mask_width; unsigned int core_select_mask, core_level_siblings; diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 161bb850fc4..62348e4fd8d 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -759,7 +759,7 @@ __cpuinit int unsynchronized_tsc(void) if (!cpu_has_tsc || tsc_unstable) return 1; -#ifdef CONFIG_SMP +#ifdef CONFIG_X86_SMP if (apic_is_clustered_box()) return 1; #endif -- cgit v1.2.3 From ee477524b461324ed8fc950f451c3671dc79f12e Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 30 Oct 2008 16:28:35 -0500 Subject: x86/voyager: fix compile breakage casued by x86: move prefill_possible_map calling early Impact: fix build failure on x86/Voyager Before: | commit 329513a35d1a2b6b28d54f5c2c0dde4face8200b | Author: Yinghai Lu | Date: Wed Jul 2 18:54:40 2008 -0700 | | x86: move prefill_possible_map calling early prefill_possible_mask() was hidden under CONFIG_HOTPLUG_CPU rendering it invisitble to voyager. Since this commit it's exposed, but not provided by the voyager subarch, so add a dummy stub to fix the link breakage. Signed-off-by: James Bottomley Cc: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/mach-voyager/voyager_smp.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c index 9cd327a278a..01285af5782 100644 --- a/arch/x86/mach-voyager/voyager_smp.c +++ b/arch/x86/mach-voyager/voyager_smp.c @@ -345,6 +345,12 @@ static void do_quad_bootstrap(void) } } +void prefill_possible_map(void) +{ + /* This is empty on voyager because we need a much + * earlier detection which is done in find_smp_config */ +} + /* Set up all the basic stuff: read the SMP config and make all the * SMP information reflect only the boot cpu. All others will be * brought on-line later. */ -- cgit v1.2.3 From 9e41bff2708e420e61e6b89a54c15232857069b1 Mon Sep 17 00:00:00 2001 From: Ravikiran G Thirumalai Date: Thu, 30 Oct 2008 13:59:21 -0700 Subject: x86: fix /dev/mem mmap breakage when PAT is disabled Impact: allow /dev/mem mmaps on non-PAT CPUs/platforms Fix mmap to /dev/mem when CONFIG_X86_PAT is off and CONFIG_STRICT_DEVMEM is off mmap to /dev/mem on kernel memory has been failing since the introduction of PAT (CONFIG_STRICT_DEVMEM=n case). Seems like the check to avoid cache aliasing with PAT is kicking in even when PAT is disabled. The bug seems to have crept in 2.6.26. This patch makes sure that the mmap to regular kernel memory succeeds if CONFIG_STRICT_DEVMEM=n and PAT is disabled, and the checks to avoid cache aliasing still happens if PAT is enabled. Signed-off-by: Ravikiran Thirumalai Tested-by: Tim Sirianni Cc: Acked-by: H. Peter Anvin Signed-off-by: Ingo Molnar --- arch/x86/mm/pat.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index 738fd0f2495..eb1bf000d12 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -481,12 +481,16 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size) return 1; } #else +/* This check is needed to avoid cache aliasing when PAT is enabled */ static inline int range_is_allowed(unsigned long pfn, unsigned long size) { u64 from = ((u64)pfn) << PAGE_SHIFT; u64 to = from + size; u64 cursor = from; + if (!pat_enabled) + return 1; + while (cursor < to) { if (!devmem_is_allowed(pfn)) { printk(KERN_INFO -- cgit v1.2.3 From b3572e361b6b2ac5e724bc4bb932b7774b720b95 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 30 Oct 2008 16:00:59 -0500 Subject: x86/voyager: fix compile breakage caused by dc1e35c6e95e8923cf1d3510438b63c600fee1e2 Impact: build fix on x86/Voyager Given commits like this: | Author: Suresh Siddha | Date: Tue Jul 29 10:29:19 2008 -0700 | | x86, xsave: enable xsave/xrstor on cpus with xsave support Which deliberately expose boot cpu dependence to pieces of the system, I think it's time to explicitly have a variable for it to prevent this continual misassumption that the boot CPU is zero. Signed-off-by: James Bottomley Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 4 ++++ arch/x86/include/asm/smp.h | 6 ++++++ arch/x86/kernel/cpu/common.c | 2 +- 3 files changed, 11 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 350bee1d54d..2a40c4c6dd7 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -231,6 +231,10 @@ config SMP If you don't know what to do here, say N. +config X86_HAS_BOOT_CPU_ID + def_bool y + depends on X86_VOYAGER + config X86_FIND_SMP_CONFIG def_bool y depends on X86_MPPARSE || X86_VOYAGER diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h index 2766021aef8..d12811ce51d 100644 --- a/arch/x86/include/asm/smp.h +++ b/arch/x86/include/asm/smp.h @@ -225,5 +225,11 @@ static inline int hard_smp_processor_id(void) #endif /* CONFIG_X86_LOCAL_APIC */ +#ifdef CONFIG_X86_HAS_BOOT_CPU_ID +extern unsigned char boot_cpu_id; +#else +#define boot_cpu_id 0 +#endif + #endif /* __ASSEMBLY__ */ #endif /* _ASM_X86_SMP_H */ diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 25581dcb280..93e9393ea64 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1134,7 +1134,7 @@ void __cpuinit cpu_init(void) /* * Boot processor to setup the FP and extended state context info. */ - if (!smp_processor_id()) + if (smp_processor_id() == boot_cpu_id) init_thread_xstate(); xsave_init(); -- cgit v1.2.3 From bfcb4c1becf93b1592f4a03a4d6e00a3ab89d5ec Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 30 Oct 2008 16:13:37 -0500 Subject: x86/voyager: fix missing cpu_index initialisation Impact: fix /proc/cpuinfo output on x86/Voyager Ever since | commit 92cb7612aee39642d109b8d935ad265e602c0563 | Author: Mike Travis | Date: Fri Oct 19 20:35:04 2007 +0200 | | x86: convert cpuinfo_x86 array to a per_cpu array We've had an extra field in cpuinfo_x86 which is cpu_index. Unfortunately, voyager has never initialised this, although the only noticeable impact seems to be that /proc/cpuinfo shows all zeros for the processor ids. Anyway, fix this by initialising the boot CPU properly and setting the index when the secondaries update. Signed-off-by: James Bottomley Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/common.c | 2 ++ arch/x86/mach-voyager/voyager_smp.c | 1 + 2 files changed, 3 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 93e9393ea64..da8f15ac7a6 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -549,6 +549,8 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c) this_cpu->c_early_init(c); validate_pat_support(c); + + c->cpu_index = boot_cpu_id; } void __init early_cpu_init(void) diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c index 01285af5782..7f4c6af1435 100644 --- a/arch/x86/mach-voyager/voyager_smp.c +++ b/arch/x86/mach-voyager/voyager_smp.c @@ -420,6 +420,7 @@ void __init smp_store_cpu_info(int id) struct cpuinfo_x86 *c = &cpu_data(id); *c = boot_cpu_data; + c->cpu_index = id; identify_secondary_cpu(c); } -- cgit v1.2.3 From 1c4acdb467f8a6704855a5670ff3d82e3c18eb0b Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Fri, 31 Oct 2008 00:43:03 +0100 Subject: x86: cpu_index build fix fix: arch/x86/kernel/cpu/common.c: In function 'early_identify_cpu': arch/x86/kernel/cpu/common.c:553: error: 'struct cpuinfo_x86' has no member named 'cpu_index' as cpu_index is only available on SMP. Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/common.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index da8f15ac7a6..003a65395bd 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -550,7 +550,9 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c) validate_pat_support(c); +#ifdef CONFIG_SMP c->cpu_index = boot_cpu_id; +#endif } void __init early_cpu_init(void) -- cgit v1.2.3 From ad5173ff8a387191dbacf889becb92c59aba5d59 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Fri, 31 Oct 2008 11:24:27 -0500 Subject: lguest: fix early_ioremap. dmi_scan_machine breaks under lguest: lguest: unhandled trap 14 at 0xc04edeae (0xffa00000) This is because we use current_cr3 for the read_cr3() paravirt function, and it isn't set until the first cr3 change. We got away with it until this happened. Signed-off-by: Rusty Russell --- arch/x86/lguest/boot.c | 29 +++++++++++++++++------------ 1 file changed, 17 insertions(+), 12 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 48ee4f9435f..4e22fa08d62 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -367,10 +367,9 @@ static void lguest_cpuid(unsigned int *ax, unsigned int *bx, * lazily after a task switch, and Linux uses that gratefully, but wouldn't a * name like "FPUTRAP bit" be a little less cryptic? * - * We store cr0 (and cr3) locally, because the Host never changes it. The - * Guest sometimes wants to read it and we'd prefer not to bother the Host - * unnecessarily. */ -static unsigned long current_cr0, current_cr3; + * We store cr0 locally because the Host never changes it. The Guest sometimes + * wants to read it and we'd prefer not to bother the Host unnecessarily. */ +static unsigned long current_cr0; static void lguest_write_cr0(unsigned long val) { lazy_hcall(LHCALL_TS, val & X86_CR0_TS, 0, 0); @@ -399,17 +398,23 @@ static unsigned long lguest_read_cr2(void) return lguest_data.cr2; } +/* See lguest_set_pte() below. */ +static bool cr3_changed = false; + /* cr3 is the current toplevel pagetable page: the principle is the same as - * cr0. Keep a local copy, and tell the Host when it changes. */ + * cr0. Keep a local copy, and tell the Host when it changes. The only + * difference is that our local copy is in lguest_data because the Host needs + * to set it upon our initial hypercall. */ static void lguest_write_cr3(unsigned long cr3) { + lguest_data.pgdir = cr3; lazy_hcall(LHCALL_NEW_PGTABLE, cr3, 0, 0); - current_cr3 = cr3; + cr3_changed = true; } static unsigned long lguest_read_cr3(void) { - return current_cr3; + return lguest_data.pgdir; } /* cr4 is used to enable and disable PGE, but we don't care. */ @@ -498,13 +503,13 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval) * to forget all of them. Fortunately, this is very rare. * * ... except in early boot when the kernel sets up the initial pagetables, - * which makes booting astonishingly slow. So we don't even tell the Host - * anything changed until we've done the first page table switch. */ + * which makes booting astonishingly slow: 1.83 seconds! So we don't even tell + * the Host anything changed until we've done the first page table switch, + * which brings boot back to 0.25 seconds. */ static void lguest_set_pte(pte_t *ptep, pte_t pteval) { *ptep = pteval; - /* Don't bother with hypercall before initial setup. */ - if (current_cr3) + if (cr3_changed) lazy_hcall(LHCALL_FLUSH_TLB, 1, 0, 0); } @@ -521,7 +526,7 @@ static void lguest_set_pte(pte_t *ptep, pte_t pteval) static void lguest_flush_tlb_single(unsigned long addr) { /* Simply set it to zero: if it was not, it will fault back in. */ - lazy_hcall(LHCALL_SET_PTE, current_cr3, addr, 0); + lazy_hcall(LHCALL_SET_PTE, lguest_data.pgdir, addr, 0); } /* This is what happens after the Guest has removed a large number of entries. -- cgit v1.2.3 From 526e5ab200ce483dcdf146806f4936bd58daa800 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Fri, 31 Oct 2008 11:24:27 -0500 Subject: lguest: fix irq vectors. do_IRQ: cannot handle IRQ -1 vector 0x20 cpu 0 ------------[ cut here ]------------ kernel BUG at arch/x86/kernel/irq_32.c:219! We're not ISA: we have a 1:1 mapping from vectors to irqs. Signed-off-by: Rusty Russell --- arch/x86/lguest/boot.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 4e22fa08d62..a5d8e1ace1c 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -586,6 +586,9 @@ static void __init lguest_init_IRQ(void) for (i = 0; i < LGUEST_IRQS; i++) { int vector = FIRST_EXTERNAL_VECTOR + i; + /* Some systems map "vectors" to interrupts weirdly. Lguest has + * a straightforward 1 to 1 mapping, so force that here. */ + __get_cpu_var(vector_irq)[vector] = i; if (vector != SYSCALL_VECTOR) { set_intr_gate(vector, interrupt[vector]); set_irq_chip_and_handler_name(i, &lguest_irq_controller, -- cgit v1.2.3 From b342797c1e5116a130841527b47dfaa462ed0968 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Fri, 31 Oct 2008 09:31:38 +0100 Subject: x86: build fix Impact: build fix on certain UP configs fix: arch/x86/kernel/cpu/common.c: In function 'cpu_init': arch/x86/kernel/cpu/common.c:1141: error: 'boot_cpu_id' undeclared (first use in this function) arch/x86/kernel/cpu/common.c:1141: error: (Each undeclared identifier is reported only once arch/x86/kernel/cpu/common.c:1141: error: for each function it appears in.) Pull in asm/smp.h on UP, so that we get the definition of boot_cpu_id. Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/common.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 003a65395bd..b9c9ea0217a 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -20,6 +20,7 @@ #include #include #include +#include #ifdef CONFIG_X86_LOCAL_APIC #include #include -- cgit v1.2.3 From fd9409343521eac22b6ed51686128a643c7c976b Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Thu, 30 Oct 2008 19:37:09 -0700 Subject: x86: add iomap_atomic*()/iounmap_atomic() on 32-bit using fixmaps Impact: introduce new APIs, separate kmap code from CONFIG_HIGHMEM This takes the code used for CONFIG_HIGHMEM memory mappings except that it's designed for dynamic IO resource mapping. These fixmaps are available even with CONFIG_HIGHMEM turned off. Signed-off-by: Keith Packard Signed-off-by: Eric Anholt Signed-off-by: Ingo Molnar --- arch/x86/include/asm/fixmap.h | 4 +++ arch/x86/include/asm/fixmap_32.h | 4 --- arch/x86/include/asm/highmem.h | 5 +--- arch/x86/mm/Makefile | 2 +- arch/x86/mm/init_32.c | 3 +- arch/x86/mm/iomap_32.c | 59 ++++++++++++++++++++++++++++++++++++++++ 6 files changed, 66 insertions(+), 11 deletions(-) create mode 100644 arch/x86/mm/iomap_32.c (limited to 'arch/x86') diff --git a/arch/x86/include/asm/fixmap.h b/arch/x86/include/asm/fixmap.h index 8668a94f850..23696d44a0a 100644 --- a/arch/x86/include/asm/fixmap.h +++ b/arch/x86/include/asm/fixmap.h @@ -9,6 +9,10 @@ extern int fixmaps_set; +extern pte_t *kmap_pte; +extern pgprot_t kmap_prot; +extern pte_t *pkmap_page_table; + void __native_set_fixmap(enum fixed_addresses idx, pte_t pte); void native_set_fixmap(enum fixed_addresses idx, unsigned long phys, pgprot_t flags); diff --git a/arch/x86/include/asm/fixmap_32.h b/arch/x86/include/asm/fixmap_32.h index 09f29ab5c13..c7115c1d721 100644 --- a/arch/x86/include/asm/fixmap_32.h +++ b/arch/x86/include/asm/fixmap_32.h @@ -28,10 +28,8 @@ extern unsigned long __FIXADDR_TOP; #include #include #include -#ifdef CONFIG_HIGHMEM #include #include -#endif /* * Here we define all the compile-time 'special' virtual @@ -75,10 +73,8 @@ enum fixed_addresses { #ifdef CONFIG_X86_CYCLONE_TIMER FIX_CYCLONE_TIMER, /*cyclone timer register*/ #endif -#ifdef CONFIG_HIGHMEM FIX_KMAP_BEGIN, /* reserved pte's for temporary kernel mappings */ FIX_KMAP_END = FIX_KMAP_BEGIN+(KM_TYPE_NR*NR_CPUS)-1, -#endif #ifdef CONFIG_PCI_MMCONFIG FIX_PCIE_MCFG, #endif diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h index a3b3b7c3027..bf9276bea66 100644 --- a/arch/x86/include/asm/highmem.h +++ b/arch/x86/include/asm/highmem.h @@ -25,14 +25,11 @@ #include #include #include +#include /* declarations for highmem.c */ extern unsigned long highstart_pfn, highend_pfn; -extern pte_t *kmap_pte; -extern pgprot_t kmap_prot; -extern pte_t *pkmap_page_table; - /* * Right now we initialize only a single pte table. It can be extended * easily, subsequent pte tables have to be allocated in one physical diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 59f89b434b4..fea4565ff57 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -1,7 +1,7 @@ obj-y := init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \ pat.o pgtable.o gup.o -obj-$(CONFIG_X86_32) += pgtable_32.o +obj-$(CONFIG_X86_32) += pgtable_32.o iomap_32.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o obj-$(CONFIG_X86_PTDUMP) += dump_pagetables.o diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index 8396868e82c..c483f424207 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -334,7 +334,6 @@ int devmem_is_allowed(unsigned long pagenr) return 0; } -#ifdef CONFIG_HIGHMEM pte_t *kmap_pte; pgprot_t kmap_prot; @@ -357,6 +356,7 @@ static void __init kmap_init(void) kmap_prot = PAGE_KERNEL; } +#ifdef CONFIG_HIGHMEM static void __init permanent_kmaps_init(pgd_t *pgd_base) { unsigned long vaddr; @@ -436,7 +436,6 @@ static void __init set_highmem_pages_init(void) #endif /* !CONFIG_NUMA */ #else -# define kmap_init() do { } while (0) # define permanent_kmaps_init(pgd_base) do { } while (0) # define set_highmem_pages_init() do { } while (0) #endif /* CONFIG_HIGHMEM */ diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c new file mode 100644 index 00000000000..d0151d8ce45 --- /dev/null +++ b/arch/x86/mm/iomap_32.c @@ -0,0 +1,59 @@ +/* + * Copyright © 2008 Ingo Molnar + * + * 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 +#include + +/* Map 'pfn' using fixed map 'type' and protections 'prot' + */ +void * +iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) +{ + enum fixed_addresses idx; + unsigned long vaddr; + + pagefault_disable(); + + idx = type + KM_TYPE_NR*smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + set_pte(kmap_pte-idx, pfn_pte(pfn, prot)); + arch_flush_lazy_mmu_mode(); + + return (void*) vaddr; +} +EXPORT_SYMBOL_GPL(iomap_atomic_prot_pfn); + +void +iounmap_atomic(void *kvaddr, enum km_type type) +{ + unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK; + enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id(); + + /* + * Force other mappings to Oops if they'll try to access this pte + * without first remap it. Keeping stale mappings around is a bad idea + * also, in case the page changes cacheability attributes or becomes + * a protected page in a hypervisor. + */ + if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) + kpte_clear_flush(kmap_pte-idx, vaddr); + + arch_flush_lazy_mmu_mode(); + pagefault_enable(); +} +EXPORT_SYMBOL_GPL(iounmap_atomic); -- cgit v1.2.3 From 2576c9991758e431b73e374f6019d6e1e12a8d36 Mon Sep 17 00:00:00 2001 From: Venki Pallipadi Date: Tue, 7 Oct 2008 13:33:12 -0700 Subject: x86: fix AMDC1E and XTOPOLOGY conflict in cpufeature Impact: fix xsave slowdown regression Fix two features from conflicting in feature bits. Fixes this performance regression: Subject: cpu2000(both float and int) 13% regression with 2.6.28-rc1 http://lkml.org/lkml/2008/10/28/36 Reported-by: "Zhang, Yanmin" Bisected-by: "Zhang, Yanmin" Signed-off-by: Venkatesh Pallipadi Signed-off-by: Ingo Molnar --- arch/x86/include/asm/cpufeature.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index f73e95d75b4..cfdf8c2c5c3 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -91,7 +91,7 @@ #define X86_FEATURE_11AP (3*32+19) /* "" Bad local APIC aka 11AP */ #define X86_FEATURE_NOPL (3*32+20) /* The NOPL (0F 1F) instructions */ #define X86_FEATURE_AMDC1E (3*32+21) /* AMD C1E detected */ -#define X86_FEATURE_XTOPOLOGY (3*32+21) /* cpu topology enum extensions */ +#define X86_FEATURE_XTOPOLOGY (3*32+22) /* cpu topology enum extensions */ /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */ #define X86_FEATURE_XMM3 (4*32+ 0) /* "pni" SSE-3 */ -- cgit v1.2.3 From 1f98757776eafe31065be9118db6051afcf8643c Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 1 Nov 2008 10:17:22 -0700 Subject: x86: Clean up late e820 resource allocation This makes the late e820 resources use 'insert_resource_expand_to_fit()' instead of doing a 'reserve_region_with_split()', and also avoids marking them as IORESOURCE_BUSY. This results in us being perfectly happy to use pre-existing PCI resources even if they were marked as being in a reserved region, while still avoiding any _new_ allocations in the reserved regions. It also makes for a simpler and more accurate resource tree. Example resource allocation from Jonathan Corbet, who has firmware that has an e820 reserved entry that covered a big range (e0000000-fed003ff), and that had various PCI resources in it set up by firmware. With old kernels, the reserved range would force us to re-allocate all pre-existing PCI resources, and his reserved range would end up looking like this: e0000000-fed003ff : reserved fec00000-fec00fff : IOAPIC 0 fed00000-fed003ff : HPET 0 where only the pre-allocated special regions (IOAPIC and HPET) were kept around. With 2.6.28-rc2, which uses 'reserve_region_with_split()', Jonathan's resource tree looked like this: e0000000-fe7fffff : reserved fe800000-fe8fffff : PCI Bus 0000:01 fe800000-fe8fffff : reserved fe900000-fe9d9aff : reserved fe9d9b00-fe9d9bff : 0000:00:1f.3 fe9d9b00-fe9d9bff : reserved fe9d9c00-fe9d9fff : 0000:00:1a.7 fe9d9c00-fe9d9fff : reserved fe9da000-fe9dafff : 0000:00:03.3 fe9da000-fe9dafff : reserved fe9db000-fe9dbfff : 0000:00:19.0 fe9db000-fe9dbfff : reserved fe9dc000-fe9dffff : 0000:00:1b.0 fe9dc000-fe9dffff : reserved fe9e0000-fe9fffff : 0000:00:19.0 fe9e0000-fe9fffff : reserved fea00000-fea7ffff : 0000:00:02.0 fea00000-fea7ffff : reserved fea80000-feafffff : 0000:00:02.1 fea80000-feafffff : reserved feb00000-febfffff : 0000:00:02.0 feb00000-febfffff : reserved fec00000-fed003ff : reserved fec00000-fec00fff : IOAPIC 0 fed00000-fed003ff : HPET 0 and because the reserved entry had been split and moved into the individual resources, and because it used the IORESOURCE_BUSY flag, the drivers that actually wanted to _use_ those resources couldn't actually attach to them: e1000e 0000:00:19.0: BAR 0: can't reserve mem region [0xfe9e0000-0xfe9fffff] HDA Intel 0000:00:1b.0: BAR 0: can't reserve mem region [0xfe9dc000-0xfe9dffff] with this patch, the resource tree instead becomes e0000000-fed003ff : reserved fe800000-fe8fffff : PCI Bus 0000:01 fe9d9b00-fe9d9bff : 0000:00:1f.3 fe9d9c00-fe9d9fff : 0000:00:1a.7 fe9d9c00-fe9d9fff : ehci_hcd fe9da000-fe9dafff : 0000:00:03.3 fe9db000-fe9dbfff : 0000:00:19.0 fe9db000-fe9dbfff : e1000e fe9dc000-fe9dffff : 0000:00:1b.0 fe9dc000-fe9dffff : ICH HD audio fe9e0000-fe9fffff : 0000:00:19.0 fe9e0000-fe9fffff : e1000e fea00000-fea7ffff : 0000:00:02.0 fea80000-feafffff : 0000:00:02.1 feb00000-febfffff : 0000:00:02.0 fec00000-fec00fff : IOAPIC 0 fed00000-fed003ff : HPET 0 ie the one reserved region now ends up surrounding all the PCI resources that were allocated inside of it by firmware, and because it is not marked BUSY, drivers have no problem attaching to the pre-allocated resources. Reported-and-tested-by: Jonathan Corbet Cc: Yinghai Lu Cc: Ingo Molnar Cc: Robert Hancock Signed-off-by: Linus Torvalds --- arch/x86/kernel/e820.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index ce97bf3bed1..7aafeb5263e 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -1290,15 +1290,17 @@ void __init e820_reserve_resources(void) res->start = e820.map[i].addr; res->end = end; - res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; + res->flags = IORESOURCE_MEM; /* * don't register the region that could be conflicted with * pci device BAR resource and insert them later in * pcibios_resource_survey() */ - if (e820.map[i].type != E820_RESERVED || res->start < (1ULL<<20)) + if (e820.map[i].type != E820_RESERVED || res->start < (1ULL<<20)) { + res->flags |= IORESOURCE_BUSY; insert_resource(&iomem_resource, res); + } res++; } @@ -1318,7 +1320,7 @@ void __init e820_reserve_resources_late(void) res = e820_res; for (i = 0; i < e820.nr_map; i++) { if (!res->parent && res->end) - reserve_region_with_split(&iomem_resource, res->start, res->end, res->name); + insert_resource_expand_to_fit(&iomem_resource, res); res++; } } -- cgit v1.2.3 From 73557af5bf32c3db973050de1fb73423e8fc873e Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Fri, 31 Oct 2008 13:59:49 -0400 Subject: x86, voyager: fix smp_intr_init() compile breakage Impact: fix x86/Voyager build Looks like this became static on the rest of x86. Fix it up by adding an external definition to mach-voyager/setup.c Signed-off-by: Ingo Molnar --- arch/x86/include/asm/voyager.h | 1 + arch/x86/mach-voyager/setup.c | 2 +- arch/x86/mach-voyager/voyager_smp.c | 2 +- 3 files changed, 3 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/voyager.h b/arch/x86/include/asm/voyager.h index 9c811d2e6f9..b3e64730762 100644 --- a/arch/x86/include/asm/voyager.h +++ b/arch/x86/include/asm/voyager.h @@ -520,6 +520,7 @@ extern void voyager_restart(void); extern void voyager_cat_power_off(void); extern void voyager_cat_do_common_interrupt(void); extern void voyager_handle_nmi(void); +extern void voyager_smp_intr_init(void); /* Commands for the following are */ #define VOYAGER_PSI_READ 0 #define VOYAGER_PSI_WRITE 1 diff --git a/arch/x86/mach-voyager/setup.c b/arch/x86/mach-voyager/setup.c index 6bbdd633864..a580b9562e7 100644 --- a/arch/x86/mach-voyager/setup.c +++ b/arch/x86/mach-voyager/setup.c @@ -27,7 +27,7 @@ static struct irqaction irq2 = { void __init intr_init_hook(void) { #ifdef CONFIG_SMP - smp_intr_init(); + voyager_smp_intr_init(); #endif setup_irq(2, &irq2); diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c index 7f4c6af1435..0e331652681 100644 --- a/arch/x86/mach-voyager/voyager_smp.c +++ b/arch/x86/mach-voyager/voyager_smp.c @@ -1258,7 +1258,7 @@ static void handle_vic_irq(unsigned int irq, struct irq_desc *desc) #define QIC_SET_GATE(cpi, vector) \ set_intr_gate((cpi) + QIC_DEFAULT_CPI_BASE, (vector)) -void __init smp_intr_init(void) +void __init voyager_smp_intr_init(void) { int i; -- cgit v1.2.3 From e5beae16901795223d677f15aa2fe192976278ee Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Mon, 3 Nov 2008 18:21:45 +0100 Subject: io mapping: clean up #ifdefs Impact: cleanup clean up ifdefs: change #ifdef CONFIG_X86_32/64 to CONFIG_HAVE_ATOMIC_IOMAP. flip around the #ifdef sections to clean up the structure. Signed-off-by: Keith Packard Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 6f20718d315..e60c59b81bd 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1894,6 +1894,10 @@ config SYSVIPC_COMPAT endmenu +config HAVE_ATOMIC_IOMAP + def_bool y + depends on X86_32 + source "net/Kconfig" source "drivers/Kconfig" -- cgit v1.2.3 From 70de9a97049e0ba79dc040868564408d5ce697f9 Mon Sep 17 00:00:00 2001 From: Alok Kataria Date: Mon, 3 Nov 2008 11:18:47 -0800 Subject: x86: don't use tsc_khz to calculate lpj if notsc is passed Impact: fix udelay when "notsc" boot parameter is passed With notsc passed on commandline, tsc may not be used for udelays, make sure that we do not use tsc_khz to calculate the lpj value in such cases. Reported-by: Bartlomiej Zolnierkiewicz Signed-off-by: Alok N Kataria Cc: Signed-off-by: Ingo Molnar --- arch/x86/kernel/tsc.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 62348e4fd8d..2ef80e30192 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -813,10 +813,6 @@ void __init tsc_init(void) cpu_khz = calibrate_cpu(); #endif - lpj = ((u64)tsc_khz * 1000); - do_div(lpj, HZ); - lpj_fine = lpj; - printk("Detected %lu.%03lu MHz processor.\n", (unsigned long)cpu_khz / 1000, (unsigned long)cpu_khz % 1000); @@ -836,6 +832,10 @@ void __init tsc_init(void) /* now allow native_sched_clock() to use rdtsc */ tsc_disabled = 0; + lpj = ((u64)tsc_khz * 1000); + do_div(lpj, HZ); + lpj_fine = lpj; + use_tsc_delay(); /* Check and install the TSC clocksource */ dmi_check_system(bad_tsc_dmi_table); -- cgit v1.2.3 From 9fcd18c9e63e325dbd2b4c726623f760788d5aa8 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Wed, 5 Nov 2008 16:52:08 +0100 Subject: sched: re-tune balancing Impact: improve wakeup affinity on NUMA systems, tweak SMP systems Given the fixes+tweaks to the wakeup-buddy code, re-tweak the domain balancing defaults on NUMA and SMP systems. Turn on SD_WAKE_AFFINE which was off on x86 NUMA - there's no reason why we would not want to have wakeup affinity across nodes as well. (we already do this in the standard NUMA template.) lat_ctx on a NUMA box is particularly happy about this change: before: | phoenix:~/l> ./lat_ctx -s 0 2 | "size=0k ovr=2.60 | 2 5.70 after: | phoenix:~/l> ./lat_ctx -s 0 2 | "size=0k ovr=2.65 | 2 2.07 a 2.75x speedup. pipe-test is similarly happy about it too: | phoenix:~/sched-tests> ./pipe-test | 18.26 usecs/loop. | 14.70 usecs/loop. | 14.38 usecs/loop. | 10.55 usecs/loop. # +WAKE_AFFINE on domain0+domain1 | 8.63 usecs/loop. | 8.59 usecs/loop. | 9.03 usecs/loop. | 8.94 usecs/loop. | 8.96 usecs/loop. | 8.63 usecs/loop. Also: - disable SD_BALANCE_NEWIDLE on NUMA and SMP domains (keep it for siblings) - enable SD_WAKE_BALANCE on SMP domains Sysbench+postgresql improves all around the board, quite significantly: .28-rc3-11474e2c .28-rc3-11474e2c-tune ------------------------------------------------- 1: 571 688 +17.08% 2: 1236 1206 -2.55% 4: 2381 2642 +9.89% 8: 4958 5164 +3.99% 16: 9580 9574 -0.07% 32: 7128 8118 +12.20% 64: 7342 8266 +11.18% 128: 7342 8064 +8.95% 256: 7519 7884 +4.62% 512: 7350 7731 +4.93% ------------------------------------------------- SUM: 55412 59341 +6.62% So it's a win both for the runup portion, the peak area and the tail. Signed-off-by: Ingo Molnar --- arch/x86/include/asm/topology.h | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index 90ac7718469..4850e4b02b6 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -154,7 +154,7 @@ extern unsigned long node_remap_size[]; #endif -/* sched_domains SD_NODE_INIT for NUMAQ machines */ +/* sched_domains SD_NODE_INIT for NUMA machines */ #define SD_NODE_INIT (struct sched_domain) { \ .min_interval = 8, \ .max_interval = 32, \ @@ -169,8 +169,9 @@ extern unsigned long node_remap_size[]; .flags = SD_LOAD_BALANCE \ | SD_BALANCE_EXEC \ | SD_BALANCE_FORK \ - | SD_SERIALIZE \ - | SD_WAKE_BALANCE, \ + | SD_WAKE_AFFINE \ + | SD_WAKE_BALANCE \ + | SD_SERIALIZE, \ .last_balance = jiffies, \ .balance_interval = 1, \ } -- cgit v1.2.3 From c78d0cf2925bffae8a6f00e7d9b8e971b0392edd Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Wed, 5 Nov 2008 12:04:46 +0000 Subject: x86: don't allow nr_irqs > NR_IRQS Impact: fix boot hang on 32-bit systems with more than 224 IO-APIC pins On some 32-bit systems with a lot of IO-APICs probe_nr_irqs() can return a value larger than NR_IRQS. This will lead to probe_irq_on() overrunning the irq_desc array. I hit this when running net-next-2.6 (close to 2.6.28-rc3) on a Supermicro dual Xeon system. NR_IRQS is 224 but probe_nr_irqs() detects 5 IOAPICs and returns 240. Here are the log messages: Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x01] address[0xfec00000] gsi_base[0]) Tue Nov 4 16:53:47 2008 IOAPIC[0]: apic_id 1, version 32, address 0xfec00000, GSI 0-23 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x02] address[0xfec81000] gsi_base[24]) Tue Nov 4 16:53:47 2008 IOAPIC[1]: apic_id 2, version 32, address 0xfec81000, GSI 24-47 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x03] address[0xfec81400] gsi_base[48]) Tue Nov 4 16:53:47 2008 IOAPIC[2]: apic_id 3, version 32, address 0xfec81400, GSI 48-71 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x04] address[0xfec82000] gsi_base[72]) Tue Nov 4 16:53:47 2008 IOAPIC[3]: apic_id 4, version 32, address 0xfec82000, GSI 72-95 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x05] address[0xfec82400] gsi_base[96]) Tue Nov 4 16:53:47 2008 IOAPIC[4]: apic_id 5, version 32, address 0xfec82400, GSI 96-119 Tue Nov 4 16:53:47 2008 ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 high edge) Tue Nov 4 16:53:47 2008 ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 high level) Tue Nov 4 16:53:47 2008 Enabling APIC mode: Flat. Using 5 I/O APICs Signed-off-by: Ben Hutchings Acked-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/kernel/io_apic.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c index b764d7429c6..7a3f2028e2e 100644 --- a/arch/x86/kernel/io_apic.c +++ b/arch/x86/kernel/io_apic.c @@ -3611,6 +3611,8 @@ int __init probe_nr_irqs(void) /* something wrong ? */ if (nr < nr_min) nr = nr_min; + if (WARN_ON(nr > NR_IRQS)) + nr = NR_IRQS; return nr; } -- cgit v1.2.3 From 1b4897688011cd05e07f00dcfe6af3331eb36a3c Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 4 Nov 2008 14:10:13 -0800 Subject: x86: size NR_IRQS on 32-bit systems the same way as 64-bit Impact: make NR_IRQS big enough for system with lots of apic/pins If lots of IO_APIC's are there (or can be there), size the same way as 64-bit, depending on MAX_IO_APICS and NR_CPUS. This fixes the boot problem reported by Ben Hutchings on a 32-bit server with 5 IO-APICs and 240 IO-APIC pins. Signed-off-by: Yinghai Tested-by: Ben Hutchings Signed-off-by: Ingo Molnar --- arch/x86/include/asm/irq_vectors.h | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h index d843ed0e9b2..503aadc4ad3 100644 --- a/arch/x86/include/asm/irq_vectors.h +++ b/arch/x86/include/asm/irq_vectors.h @@ -101,30 +101,22 @@ #define LAST_VM86_IRQ 15 #define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15) -#ifdef CONFIG_X86_64 +#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_PARAVIRT) && !defined(CONFIG_X86_VISWS) && !defined(CONFIG_X86_VOYAGER) # if NR_CPUS < MAX_IO_APICS # define NR_IRQS (NR_VECTORS + (32 * NR_CPUS)) # else # define NR_IRQS (NR_VECTORS + (32 * MAX_IO_APICS)) # endif -#elif !defined(CONFIG_X86_VOYAGER) +#elif defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) || defined(CONFIG_X86_VOYAGER) -# if defined(CONFIG_X86_IO_APIC) || defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) - -# define NR_IRQS 224 - -# else /* IO_APIC || PARAVIRT */ - -# define NR_IRQS 16 - -# endif +# define NR_IRQS 224 -#else /* !VISWS && !VOYAGER */ +#else /* IO_APIC || PARAVIRT */ -# define NR_IRQS 224 +# define NR_IRQS 16 -#endif /* VISWS */ +#endif /* Voyager specific defines */ /* These define the CPIs we use in linux */ -- cgit v1.2.3 From da85f865b1dcec0853c48b763ed312441ce0c7df Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Wed, 5 Nov 2008 13:37:27 -0600 Subject: x86: mention ACPI in top-level Kconfig menu Impact: clarify menuconfig text Mention ACPI in the top-level menu to give a clue as to where it lives. This matches what ia64 does. Signed-off-by: Bjorn Helgaas Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 6f20718d315..5d6aa4013dc 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1494,7 +1494,7 @@ config HAVE_ARCH_EARLY_PFN_TO_NID def_bool X86_64 depends on NUMA -menu "Power management options" +menu "Power management and ACPI options" depends on !X86_VOYAGER config ARCH_HIBERNATION_HEADER -- cgit v1.2.3 From 7db282fa67b58daff8a57f9e1c93d4474b5908ff Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 5 Nov 2008 23:36:48 -0800 Subject: x86: remove VISWS and PARAVIRT around NR_IRQS puzzle Impact: fix warning message when PARAVIRT is set in config Remove stale #ifdef components from our IRQ sizing logic. x86/Voyager is the only holdout. Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/include/asm/irq_vectors.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h index 503aadc4ad3..0005adb0f94 100644 --- a/arch/x86/include/asm/irq_vectors.h +++ b/arch/x86/include/asm/irq_vectors.h @@ -101,18 +101,18 @@ #define LAST_VM86_IRQ 15 #define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15) -#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_PARAVIRT) && !defined(CONFIG_X86_VISWS) && !defined(CONFIG_X86_VOYAGER) +#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_X86_VOYAGER) # if NR_CPUS < MAX_IO_APICS # define NR_IRQS (NR_VECTORS + (32 * NR_CPUS)) # else # define NR_IRQS (NR_VECTORS + (32 * MAX_IO_APICS)) # endif -#elif defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) || defined(CONFIG_X86_VOYAGER) +#elif defined(CONFIG_X86_VOYAGER) # define NR_IRQS 224 -#else /* IO_APIC || PARAVIRT */ +#else /* IO_APIC || VOYAGER */ # define NR_IRQS 16 -- cgit v1.2.3 From d6f0f39b7d05e62b347c4352d070e4afb3ade4b5 Mon Sep 17 00:00:00 2001 From: Suresh Siddha Date: Tue, 4 Nov 2008 13:53:04 -0800 Subject: x86: add smp_mb() before sending INVALIDATE_TLB_VECTOR Impact: fix rare x2apic hang On x86, x2apic mode accesses for sending IPI's don't have serializing semantics. If the IPI receivner refers(in lock-free fashion) to some memory setup by the sender, the need for smp_mb() before sending the IPI becomes critical in x2apic mode. Add the smp_mb() in native_flush_tlb_others() before sending the IPI. Signed-off-by: Suresh Siddha Signed-off-by: Ingo Molnar --- arch/x86/kernel/tlb_32.c | 6 ++++++ arch/x86/kernel/tlb_64.c | 5 +++++ 2 files changed, 11 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/tlb_32.c b/arch/x86/kernel/tlb_32.c index e00534b3353..f4049f3513b 100644 --- a/arch/x86/kernel/tlb_32.c +++ b/arch/x86/kernel/tlb_32.c @@ -154,6 +154,12 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, flush_mm = mm; flush_va = va; cpus_or(flush_cpumask, cpumask, flush_cpumask); + + /* + * Make the above memory operations globally visible before + * sending the IPI. + */ + smp_mb(); /* * We have to send the IPI only to * CPUs affected. diff --git a/arch/x86/kernel/tlb_64.c b/arch/x86/kernel/tlb_64.c index dcbf7a1159e..8f919ca6949 100644 --- a/arch/x86/kernel/tlb_64.c +++ b/arch/x86/kernel/tlb_64.c @@ -182,6 +182,11 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, f->flush_va = va; cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask); + /* + * Make the above memory operations globally visible before + * sending the IPI. + */ + smp_mb(); /* * We have to send the IPI only to * CPUs affected. -- cgit v1.2.3 From 80be308dfa3798c7bad0fc81760b2faf83870e91 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 6 Nov 2008 14:59:05 +0100 Subject: AMD IOMMU: fix lazy IO/TLB flushing in unmap path Lazy flushing needs to take care of the unmap path too which is not yet implemented and leads to stale IO/TLB entries. This is fixed by this patch. Signed-off-by: Joerg Roedel --- arch/x86/kernel/amd_iommu.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 38e88d40ab1..4755bbc7ae5 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -526,6 +526,9 @@ static void dma_ops_free_addresses(struct dma_ops_domain *dom, { address >>= PAGE_SHIFT; iommu_area_free(dom->bitmap, address, pages); + + if (address + pages >= dom->next_bit) + dom->need_flush = true; } /**************************************************************************** @@ -981,8 +984,10 @@ static void __unmap_single(struct amd_iommu *iommu, dma_ops_free_addresses(dma_dom, dma_addr, pages); - if (amd_iommu_unmap_flush) + if (amd_iommu_unmap_flush || dma_dom->need_flush) { iommu_flush_pages(iommu, dma_dom->domain.id, dma_addr, size); + dma_dom->need_flush = false; + } } /* -- cgit v1.2.3 From b9c3bfc24e1088d260de4091b2b41808c7398355 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 6 Nov 2008 12:05:40 +0000 Subject: x86: align DirectMap in /proc/meminfo Impact: right-align /proc/meminfo consistent with other fields When the split-LRU patches added Inactive(anon) and Inactive(file) lines to /proc/meminfo, all counts were moved two columns rightwards to fit in. Now move x86's DirectMap lines two columns rightwards to line up. Signed-off-by: Hugh Dickins Signed-off-by: Ingo Molnar --- arch/x86/mm/pageattr.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index f1dc1b75d16..e89d24815f2 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -67,18 +67,18 @@ static void split_page_count(int level) void arch_report_meminfo(struct seq_file *m) { - seq_printf(m, "DirectMap4k: %8lu kB\n", + seq_printf(m, "DirectMap4k: %8lu kB\n", direct_pages_count[PG_LEVEL_4K] << 2); #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) - seq_printf(m, "DirectMap2M: %8lu kB\n", + seq_printf(m, "DirectMap2M: %8lu kB\n", direct_pages_count[PG_LEVEL_2M] << 11); #else - seq_printf(m, "DirectMap4M: %8lu kB\n", + seq_printf(m, "DirectMap4M: %8lu kB\n", direct_pages_count[PG_LEVEL_2M] << 12); #endif #ifdef CONFIG_X86_64 if (direct_gbpages) - seq_printf(m, "DirectMap1G: %8lu kB\n", + seq_printf(m, "DirectMap1G: %8lu kB\n", direct_pages_count[PG_LEVEL_1G] << 20); #endif } -- cgit v1.2.3 From 8d00450d296dedec9ada38d43b83e79cca6fd5a3 Mon Sep 17 00:00:00 2001 From: Eduardo Habkost Date: Tue, 4 Nov 2008 12:52:44 -0200 Subject: Revert "x86: default to reboot via ACPI" This reverts commit c7ffa6c26277b403920e2255d10df849bd613380. the assumptio of this change was that this would not break any existing machine. Andrey Borzenkov reported troubles with the ACPI reboot method: the system would hang on reboot, necessiating a power cycle. Probably more systems are affected as well. Also, there are patches queued up for v2.6.29 to disable virtualization on emergency_restart() - which was the original motivation of this change. Reported-by: Andrey Borzenkov Bisected-by: Andrey Borzenkov Signed-off-by: Eduardo Habkost Acked-by: Avi Kivity Signed-off-by: Ingo Molnar --- arch/x86/kernel/reboot.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index f4c93f1cfc1..724adfc63cb 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -29,11 +29,7 @@ EXPORT_SYMBOL(pm_power_off); static const struct desc_ptr no_idt = {}; static int reboot_mode; -/* - * Keyboard reset and triple fault may result in INIT, not RESET, which - * doesn't work when we're in vmx root mode. Try ACPI first. - */ -enum reboot_type reboot_type = BOOT_ACPI; +enum reboot_type reboot_type = BOOT_KBD; int reboot_force; #if defined(CONFIG_X86_32) && defined(CONFIG_SMP) -- cgit v1.2.3 From 47cb2ed9df2789fc4a3fe1201e475078f93c4839 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Thu, 6 Nov 2008 13:48:24 -0800 Subject: x86, xen: fix use of pgd_page now that it really does return a page Impact: fix 32-bit Xen guest boot crash On 32-bit PAE, pud_page, for no good reason, didn't really return a struct page *. Since Jan Beulich's fix "i386/PAE: fix pud_page()", pud_page does return a struct page *. Because PAE has 3 pagetable levels, the pud level is folded into the pgd level, so pgd_page() is the same as pud_page(), and now returns a struct page *. Update the xen/mmu.c code which uses pgd_page() accordingly. Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Ingo Molnar --- arch/x86/xen/mmu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index aba77b2b7d1..49697d86c6a 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -874,7 +874,7 @@ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) #else /* CONFIG_X86_32 */ #ifdef CONFIG_X86_PAE /* Need to make sure unshared kernel PMD is pinnable */ - xen_pin_page(mm, virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), + xen_pin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]), PT_PMD); #endif xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd))); @@ -991,7 +991,7 @@ static void __xen_pgd_unpin(struct mm_struct *mm, pgd_t *pgd) #ifdef CONFIG_X86_PAE /* Need to make sure unshared kernel PMD is unpinned */ - xen_unpin_page(mm, virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), + xen_unpin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]), PT_PMD); #endif -- cgit v1.2.3 From d05fdf316067cd311d5e7add08da26ded8a58080 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Tue, 28 Oct 2008 19:23:06 +1100 Subject: xen: make sure stray alias mappings are gone before pinning Xen requires that all mappings of pagetable pages are read-only, so that they can't be updated illegally. As a result, if a page is being turned into a pagetable page, we need to make sure all its mappings are RO. If the page had been used for ioremap or vmalloc, it may still have left over mappings as a result of not having been lazily unmapped. This change makes sure we explicitly mop them all up before pinning the page. Unlike aliases created by kmap, the there can be vmalloc aliases even for non-high pages, so we must do the flush unconditionally. Signed-off-by: Jeremy Fitzhardinge Cc: Linux Memory Management List Cc: Nick Piggin Signed-off-by: Ingo Molnar --- arch/x86/xen/enlighten.c | 5 +++-- arch/x86/xen/mmu.c | 9 ++++++--- 2 files changed, 9 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index b61534c7a4c..5e4686d70f6 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -863,15 +863,16 @@ static void xen_alloc_ptpage(struct mm_struct *mm, unsigned long pfn, unsigned l if (PagePinned(virt_to_page(mm->pgd))) { SetPagePinned(page); + vm_unmap_aliases(); if (!PageHighMem(page)) { make_lowmem_page_readonly(__va(PFN_PHYS((unsigned long)pfn))); if (level == PT_PTE && USE_SPLIT_PTLOCKS) pin_pagetable_pfn(MMUEXT_PIN_L1_TABLE, pfn); - } else + } else { /* make sure there are no stray mappings of this page */ kmap_flush_unused(); - vm_unmap_aliases(); + } } } diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index aba77b2b7d1..89f3b6edc65 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -850,13 +850,16 @@ static int xen_pin_page(struct mm_struct *mm, struct page *page, read-only, and can be pinned. */ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) { + vm_unmap_aliases(); + xen_mc_batch(); - if (xen_pgd_walk(mm, xen_pin_page, USER_LIMIT)) { - /* re-enable interrupts for kmap_flush_unused */ + if (xen_pgd_walk(mm, xen_pin_page, USER_LIMIT)) { + /* re-enable interrupts for flushing */ xen_mc_issue(0); + kmap_flush_unused(); - vm_unmap_aliases(); + xen_mc_batch(); } -- cgit v1.2.3 From 7c64ade53a6f977d73f16243865c42ceae999aea Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Fri, 7 Nov 2008 14:02:49 +0100 Subject: oprofile: Fix p6 counter overflow check Fix the counter overflow check for CPUs with counter width > 32 I had a similar change in a different patch that I didn't submit and I didn't notice the problem earlier because it was always tested together. Signed-off-by: Andi Kleen Signed-off-by: Robert Richter --- arch/x86/oprofile/op_model_ppro.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index 0620d6d45f7..3f1b81a83e2 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -27,8 +27,7 @@ static int num_counters = 2; static int counter_width = 32; #define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0) -#define CTR_READ(l, h, msrs, c) do {rdmsr(msrs->counters[(c)].addr, (l), (h)); } while (0) -#define CTR_OVERFLOWED(n) (!((n) & (1U<<(counter_width-1)))) +#define CTR_OVERFLOWED(n) (!((n) & (1ULL<<(counter_width-1)))) #define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0) #define CTRL_READ(l, h, msrs, c) do {rdmsr((msrs->controls[(c)].addr), (l), (h)); } while (0) @@ -124,14 +123,14 @@ static void ppro_setup_ctrs(struct op_msrs const * const msrs) static int ppro_check_ctrs(struct pt_regs * const regs, struct op_msrs const * const msrs) { - unsigned int low, high; + u64 val; int i; for (i = 0 ; i < num_counters; ++i) { if (!reset_value[i]) continue; - CTR_READ(low, high, msrs, i); - if (CTR_OVERFLOWED(low)) { + rdmsrl(msrs->counters[i].addr, val); + if (CTR_OVERFLOWED(val)) { oprofile_add_sample(regs, i); wrmsrl(msrs->counters[i].addr, -reset_value[i]); } -- cgit v1.2.3 From 0d12cdd5f883f508d33b85c1bae98fa28987c8c7 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 8 Nov 2008 16:19:55 +0100 Subject: sched: improve sched_clock() performance in scheduler-intense workloads native_read_tsc() overhead accounts for 20% of the system overhead: 659567 system_call 41222.9375 686796 schedule 435.7843 718382 __switch_to 665.1685 823875 switch_mm 4526.7857 1883122 native_read_tsc 55385.9412 9761990 total 2.8468 this is large part due to the rdtsc_barrier() that is done before and after reading the TSC. But sched_clock() is not a precise clock in the GTOD sense, using such barriers is completely pointless. So remove the barriers and only use them in vget_cycles(). This improves lat_ctx performance by about 5%. Signed-off-by: Ingo Molnar --- arch/x86/include/asm/msr.h | 2 -- arch/x86/include/asm/tsc.h | 8 +++++++- 2 files changed, 7 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/msr.h b/arch/x86/include/asm/msr.h index 46be2fa7ac2..c2a812ebde8 100644 --- a/arch/x86/include/asm/msr.h +++ b/arch/x86/include/asm/msr.h @@ -108,9 +108,7 @@ static __always_inline unsigned long long __native_read_tsc(void) { DECLARE_ARGS(val, low, high); - rdtsc_barrier(); asm volatile("rdtsc" : EAX_EDX_RET(val, low, high)); - rdtsc_barrier(); return EAX_EDX_VAL(val, low, high); } diff --git a/arch/x86/include/asm/tsc.h b/arch/x86/include/asm/tsc.h index 38ae163cc91..9cd83a8e40d 100644 --- a/arch/x86/include/asm/tsc.h +++ b/arch/x86/include/asm/tsc.h @@ -34,6 +34,8 @@ static inline cycles_t get_cycles(void) static __always_inline cycles_t vget_cycles(void) { + cycles_t cycles; + /* * We only do VDSOs on TSC capable CPUs, so this shouldnt * access boot_cpu_data (which is not VDSO-safe): @@ -42,7 +44,11 @@ static __always_inline cycles_t vget_cycles(void) if (!cpu_has_tsc) return 0; #endif - return (cycles_t)__native_read_tsc(); + rdtsc_barrier(); + cycles = (cycles_t)__native_read_tsc(); + rdtsc_barrier(); + + return cycles; } extern void tsc_init(void); -- cgit v1.2.3 From 7cbaef9c83e58bbd4bdd534b09052b6c5ec457d5 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 8 Nov 2008 17:05:38 +0100 Subject: sched: optimize sched_clock() a bit sched_clock() uses cycles_2_ns() needlessly - which is an irq-disabling variant of __cycles_2_ns(). Most of the time sched_clock() is called with irqs disabled already. The few places that call it with irqs enabled need to be updated. Signed-off-by: Ingo Molnar --- arch/x86/kernel/tsc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 2ef80e30192..424093b157d 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -55,7 +55,7 @@ u64 native_sched_clock(void) rdtscll(this_offset); /* return the value in ns */ - return cycles_2_ns(this_offset); + return __cycles_2_ns(this_offset); } /* We need to define a real function for sched_clock, to override the -- cgit v1.2.3 From 1de5b0854623d30d01d72cd4ea323eb5f39d1f16 Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Sun, 2 Nov 2008 16:04:18 +0000 Subject: x86: HPET: convert WARN_ON to WARN_ON_ONCE It is possible to flood the console with call traces if the WARN_ON condition is true because of the frequency with which this function is called. Signed-off-by: Matt Fleming Cc: mingo@elte.hu Cc: venkatesh.pallipadi@intel.com Signed-off-by: Thomas Gleixner --- arch/x86/kernel/hpet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 77017e834cf..f10f9461a43 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -322,7 +322,7 @@ static int hpet_next_event(unsigned long delta, * what we wrote hit the chip before we compare it to the * counter. */ - WARN_ON((u32)hpet_readl(HPET_T0_CMP) != cnt); + WARN_ON_ONCE((u32)hpet_readl(HPET_T0_CMP) != cnt); return (s32)((u32)hpet_readl(HPET_COUNTER) - cnt) >= 0 ? -ETIME : 0; } -- cgit v1.2.3 From 89d77a1eb60be916d85d9394bedbfa2037af89c5 Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Sun, 2 Nov 2008 16:04:20 +0000 Subject: x86: HPET: read from HPET_Tn_CMP() not HPET_T0_CMP In hpet_next_event() we check that the value we just wrote to HPET_Tn_CMP(timer) has reached the chip. Currently, we're checking that the value we wrote to HPET_Tn_CMP(timer) is in HPET_T0_CMP, which, if timer is anything other than timer 0, is likely to fail. Signed-off-by: Matt Fleming Cc: mingo@elte.hu Cc: venkatesh.pallipadi@intel.com Signed-off-by: Thomas Gleixner --- arch/x86/kernel/hpet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index f10f9461a43..cfe6aa56f71 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -322,7 +322,7 @@ static int hpet_next_event(unsigned long delta, * what we wrote hit the chip before we compare it to the * counter. */ - WARN_ON_ONCE((u32)hpet_readl(HPET_T0_CMP) != cnt); + WARN_ON_ONCE((u32)hpet_readl(HPET_Tn_CMP(timer)) != cnt); return (s32)((u32)hpet_readl(HPET_COUNTER) - cnt) >= 0 ? -ETIME : 0; } -- cgit v1.2.3 From 5ceb1a04187553e08c6ab60d30cee7c454ee139a Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Sun, 2 Nov 2008 22:23:13 +0000 Subject: x86: HPET: enter hpet_interrupt_handler with interrupts disabled Some functions that may be called from this handler require that interrupts are disabled. Also, combining IRQF_DISABLED and IRQF_SHARED does not reliably disable interrupts in a handler, so remove IRQF_SHARED from the irq flags (this irq is not shared anyway). Signed-off-by: Matt Fleming Cc: mingo@elte.hu Cc: venkatesh.pallipadi@intel.com Cc: "Will Newton" Signed-off-by: Thomas Gleixner --- arch/x86/kernel/hpet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index cfe6aa56f71..067d8de913f 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -445,7 +445,7 @@ static int hpet_setup_irq(struct hpet_dev *dev) { if (request_irq(dev->irq, hpet_interrupt_handler, - IRQF_SHARED|IRQF_NOBALANCING, dev->name, dev)) + IRQF_DISABLED|IRQF_NOBALANCING, dev->name, dev)) return -1; disable_irq(dev->irq); -- cgit v1.2.3 From 4694516d1987303dd83bfd0efdd36fa5b65d701b Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 10 Nov 2008 21:52:47 +0100 Subject: x86: Make NUMA on 32-bit depend on BROKEN While investigating the failure of hibernation on 32-bit x86 with CONFIG_NUMA set, as described in this message http://marc.info/?l=linux-kernel&m=122634118116226&w=4 I asked some people for help and I was told that it wasn't really worth the effort, because CONFIG_NUMA was generally broken on 32-bit x86 systems and it shouldn't be used in such configs. For this reason, make CONFIG_NUMA depend on BROKEN instead of EXPERIMENTAL on x86-32. Signed-off-by: Rafael J. Wysocki Cc: Andi Kleen Cc: Pavel Machek Cc: Peter Zijlstra Signed-off-by: Linus Torvalds --- arch/x86/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 4cf0ab13d18..93224b56918 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -957,7 +957,7 @@ config ARCH_PHYS_ADDR_T_64BIT config NUMA bool "Numa Memory Allocation and Scheduler Support (EXPERIMENTAL)" depends on SMP - depends on X86_64 || (X86_32 && HIGHMEM64G && (X86_NUMAQ || X86_BIGSMP || X86_SUMMIT && ACPI) && EXPERIMENTAL) + depends on X86_64 || (X86_32 && HIGHMEM64G && (X86_NUMAQ || X86_BIGSMP || X86_SUMMIT && ACPI) && BROKEN) default n if X86_PC default y if (X86_NUMAQ || X86_SUMMIT || X86_BIGSMP) help -- cgit v1.2.3