From 0b1faeef5f9243bb5fc5713a34bbf1ceab0de562 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Sun, 15 Jun 2008 12:32:15 +0100 Subject: x86: section/warning fixes WARNING: arch/x86/mm/built-in.o(.text+0x3a1): Section mismatch in reference from the function set_pte_phys() to the function .init.text:spp_getpage() The function set_pte_phys() references the function __init spp_getpage(). This is often because set_pte_phys lacks a __init annotation or the annotation of spp_getpage is wrong. arch/x86/mm/init_64.c: In function 'early_memtest': arch/x86/mm/init_64.c:520: warning: passing argument 2 of 'find_e820_area_size' from incompatible pointer type Signed-off-by: Daniel J Blueman Cc: "Linus Torvalds" Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index 156e6d7b0e3..f6d20be7a8f 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -135,7 +135,7 @@ static __init void *spp_getpage(void) return ptr; } -static void +static __init void set_pte_phys(unsigned long vaddr, unsigned long phys, pgprot_t prot) { pgd_t *pgd; @@ -214,7 +214,7 @@ void __init cleanup_highmap(void) } /* NOTE: this is meant to be run only at boot */ -void __set_fixmap(enum fixed_addresses idx, unsigned long phys, pgprot_t prot) +void __init __set_fixmap(enum fixed_addresses idx, unsigned long phys, pgprot_t prot) { unsigned long address = __fix_to_virt(idx); @@ -506,7 +506,7 @@ early_param("memtest", parse_memtest); static void __init early_memtest(unsigned long start, unsigned long end) { - u64 t_start, t_size; + unsigned long t_start, t_size; unsigned pattern; if (!memtest_pattern) @@ -525,7 +525,7 @@ static void __init early_memtest(unsigned long start, unsigned long end) if (t_start + t_size > end) t_size = end - t_start; - printk(KERN_CONT "\n %016llx - %016llx pattern %d", + printk(KERN_CONT "\n %016lx - %016lx pattern %d", t_start, t_start + t_size, pattern); memtest(t_start, t_size, pattern); -- cgit v1.2.3 From fcb43042ef55d2f46b0efa5d7746967cef38f056 Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Tue, 24 Jun 2008 16:06:23 +0800 Subject: x86: fix cpu hotplug crash Vegard Nossum reported crashes during cpu hotplug tests: http://marc.info/?l=linux-kernel&m=121413950227884&w=4 In function _cpu_up, the panic happens when calling __raw_notifier_call_chain at the second time. Kernel doesn't panic when calling it at the first time. If just say because of nr_cpu_ids, that's not right. By checking the source code, I found that function do_boot_cpu is the culprit. Consider below call chain: _cpu_up=>__cpu_up=>smp_ops.cpu_up=>native_cpu_up=>do_boot_cpu. So do_boot_cpu is called in the end. In do_boot_cpu, if boot_error==true, cpu_clear(cpu, cpu_possible_map) is executed. So later on, when _cpu_up calls __raw_notifier_call_chain at the second time to report CPU_UP_CANCELED, because this cpu is already cleared from cpu_possible_map, get_cpu_sysdev returns NULL. Many resources are related to cpu_possible_map, so it's better not to change it. Below patch against 2.6.26-rc7 fixes it by removing the bit clearing in cpu_possible_map. Signed-off-by: Zhang Yanmin Tested-by: Vegard Nossum Acked-by: Rusty Russell Signed-off-by: Ingo Molnar --- arch/x86/kernel/smpboot.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 56078d61c79..3e1cecedde4 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -996,7 +996,6 @@ do_rest: #endif cpu_clear(cpu, cpu_callout_map); /* was set by do_boot_cpu() */ cpu_clear(cpu, cpu_initialized); /* was set by cpu_init() */ - cpu_clear(cpu, cpu_possible_map); cpu_clear(cpu, cpu_present_map); per_cpu(x86_cpu_to_apicid, cpu) = BAD_APICID; } -- cgit v1.2.3 From 11dbc963a8f6128595d0f6ecf138dc369e144997 Mon Sep 17 00:00:00 2001 From: TAKADA Yoshihito Date: Mon, 30 Jun 2008 13:44:45 +0900 Subject: ptrace GET/SET FPXREGS broken When I update kernel 2.6.25 from 2.6.24, gdb does not work. On 2.6.25, ptrace(PTRACE_GETFPXREGS, ...) returns ENODEV. But 2.6.24 kernel's ptrace() returns EIO. It is issue of compatibility. I attached test program as pt.c and patch for fix it. #include #include #include #include #include #include #include struct user_fxsr_struct { unsigned short cwd; unsigned short swd; unsigned short twd; unsigned short fop; long fip; long fcs; long foo; long fos; long mxcsr; long reserved; long st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */ long xmm_space[32]; /* 8*16 bytes for each XMM-reg = 128 bytes */ long padding[56]; }; int main(void) { pid_t pid; pid = fork(); switch(pid){ case -1:/* error */ break; case 0:/* child */ child(); break; default: parent(pid); break; } return 0; } int child(void) { ptrace(PTRACE_TRACEME); kill(getpid(), SIGSTOP); sleep(10); return 0; } int parent(pid_t pid) { int ret; struct user_fxsr_struct fpxregs; ret = ptrace(PTRACE_GETFPXREGS, pid, 0, &fpxregs); if(ret < 0){ printf("%d: %s.\n", errno, strerror(errno)); } kill(pid, SIGCONT); wait(pid); return 0; } /* in the kerel, at kernel/i387.c get_fpxregs() */ Signed-off-by: Ingo Molnar --- arch/x86/kernel/i387.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index eb9ddd8efb8..95e80e5033c 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -162,7 +162,7 @@ int xfpregs_get(struct task_struct *target, const struct user_regset *regset, int ret; if (!cpu_has_fxsr) - return -ENODEV; + return -EIO; ret = init_fpu(target); if (ret) @@ -179,7 +179,7 @@ int xfpregs_set(struct task_struct *target, const struct user_regset *regset, int ret; if (!cpu_has_fxsr) - return -ENODEV; + return -EIO; ret = init_fpu(target); if (ret) -- cgit v1.2.3 From efac41894df57d32b483ac622d03541b5b2692c0 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 1 Jul 2008 08:56:32 +0200 Subject: x86: fix NODES_SHIFT Kconfig range commit 4323838215184f5a2f081e0d17b8d60731b03164 x86: change size of node ids from u8 to s16 set the range for NODES_SHIFT to 1..15. The possible range is 1..9 Fixes Bugzilla #10726 Reported-by: Dave Jones Signed-off-by: Thomas Gleixner --- arch/x86/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 52e18e6d2ba..8a07f417f5f 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -961,8 +961,8 @@ config NUMA_EMU number of nodes. This is only useful for debugging. config NODES_SHIFT - int "Max num nodes shift(1-15)" - range 1 15 if X86_64 + int "Max num nodes shift(1-9)" + range 1 9 if X86_64 default "6" if X86_64 default "4" if X86_NUMAQ default "3" -- cgit v1.2.3 From 216705d2720dedf630b55d641737f430ead0c228 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Wed, 2 Jul 2008 22:48:03 +0100 Subject: x86: fix Intel Mac booting with EFI Fedora reports that mem_init()'s zap_low_mappings(), extended to SMP in 61165d7a035f6571c7576e7f51e7230157724c8d x86: fix app crashes after SMP resume causes 32-bit Intel Mac machines to reboot very early when booting with EFI. The EFI code appears to manage low mappings for itself when needed; but like many before it, confuses PSE with PAE. So it has only been mapping half the space it needed when PSE but not PAE. This remained unnoticed until we moved the SMP zap_low_mappings() before efi_enter_virtual_mode(). Presumably could have been noticed years ago if anyone ran a UP kernel on such machines? Reported-by: Peter Jones Signed-off-by: Hugh Dickins Cc: Peter Jones Cc: Glauber Costa Cc: Andrew Morton Cc: Linus Torvalds Signed-off-by: Ingo Molnar Tested-by: Peter Jones --- arch/x86/kernel/efi_32.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/efi_32.c b/arch/x86/kernel/efi_32.c index 5d23d85624d..4b63c8e1f13 100644 --- a/arch/x86/kernel/efi_32.c +++ b/arch/x86/kernel/efi_32.c @@ -49,13 +49,13 @@ void efi_call_phys_prelog(void) local_irq_save(efi_rt_eflags); /* - * If I don't have PSE, I should just duplicate two entries in page - * directory. If I have PSE, I just need to duplicate one entry in + * If I don't have PAE, I should just duplicate two entries in page + * directory. If I have PAE, I just need to duplicate one entry in * page directory. */ cr4 = read_cr4(); - if (cr4 & X86_CR4_PSE) { + if (cr4 & X86_CR4_PAE) { efi_bak_pg_dir_pointer[0].pgd = swapper_pg_dir[pgd_index(0)].pgd; swapper_pg_dir[0].pgd = @@ -93,7 +93,7 @@ void efi_call_phys_epilog(void) cr4 = read_cr4(); - if (cr4 & X86_CR4_PSE) { + if (cr4 & X86_CR4_PAE) { swapper_pg_dir[pgd_index(0)].pgd = efi_bak_pg_dir_pointer[0].pgd; } else { -- cgit v1.2.3 From 27df66a406a171308b138bd84938cb735392e15c Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 3 Jul 2008 10:14:10 +0200 Subject: arch/x86/mm/init_64.c: early_memtest(): fix types fix this warning: arch/x86/mm/init_64.c: In function 'early_memtest': arch/x86/mm/init_64.c:524: warning: passing argument 2 of 'find_e820_area_size' from incompatible pointer type Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index f6d20be7a8f..819dad973b1 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -506,7 +506,7 @@ early_param("memtest", parse_memtest); static void __init early_memtest(unsigned long start, unsigned long end) { - unsigned long t_start, t_size; + u64 t_start, t_size; unsigned pattern; if (!memtest_pattern) @@ -525,8 +525,9 @@ static void __init early_memtest(unsigned long start, unsigned long end) if (t_start + t_size > end) t_size = end - t_start; - printk(KERN_CONT "\n %016lx - %016lx pattern %d", - t_start, t_start + t_size, pattern); + printk(KERN_CONT "\n %016llx - %016llx pattern %d", + (unsigned long long)t_start, + (unsigned long long)t_start + t_size, pattern); memtest(t_start, t_size, pattern); -- cgit v1.2.3 From d8355aca23863be659ec5b7e0393cfbfa91ec221 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Thu, 3 Jul 2008 22:10:18 -0700 Subject: xen: fix address truncation in pte mfn<->pfn conversion When converting the page number in a pte/pmd/pud/pgd between machine and pseudo-physical addresses, the converted result was being truncated at 32-bits. This caused failures on machines with more than 4G of physical memory. Signed-off-by: Jeremy Fitzhardinge Cc: "Christopher S. Aker" Cc: Ian Campbell 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 df40bf74ea7..4e527e7893a 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -185,7 +185,7 @@ static pteval_t pte_mfn_to_pfn(pteval_t val) if (val & _PAGE_PRESENT) { unsigned long mfn = (val & PTE_MASK) >> PAGE_SHIFT; pteval_t flags = val & ~PTE_MASK; - val = (mfn_to_pfn(mfn) << PAGE_SHIFT) | flags; + val = ((pteval_t)mfn_to_pfn(mfn) << PAGE_SHIFT) | flags; } return val; @@ -196,7 +196,7 @@ static pteval_t pte_pfn_to_mfn(pteval_t val) if (val & _PAGE_PRESENT) { unsigned long pfn = (val & PTE_MASK) >> PAGE_SHIFT; pteval_t flags = val & ~PTE_MASK; - val = (pfn_to_mfn(pfn) << PAGE_SHIFT) | flags; + val = ((pteval_t)pfn_to_mfn(pfn) << PAGE_SHIFT) | flags; } return val; -- cgit v1.2.3 From 4b4f7280d7fd1feeff134c2cf2db32fd583b6c29 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Tue, 24 Jun 2008 23:03:48 +0200 Subject: x86 ACPI: normalize segment descriptor register on resume Some Dell laptops enter resume with apparent garbage in the segment descriptor registers (almost certainly the result of a botched transition from protected to real mode.) The only way to clean that up is to enter protected mode ourselves and clean out the descriptor registers. This fixes resume on Dell XPS M1210 and Dell D620. Reference: http://bugzilla.kernel.org/show_bug.cgi?id=10927 Signed-off-by: H. Peter Anvin Cc: Andrew Morton Cc: Pavel Machek Cc: pm list Cc: Len Brown Signed-off-by: Ingo Molnar Tested-by: Kirill A. Shutemov Signed-off-by: Rafael J. Wysocki Signed-off-by: Ingo Molnar --- arch/x86/kernel/acpi/realmode/wakeup.S | 38 +++++++++++++++++++++++++++++++++- arch/x86/kernel/acpi/realmode/wakeup.h | 5 +++++ arch/x86/kernel/acpi/sleep.c | 16 +++++++++++++- 3 files changed, 57 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/acpi/realmode/wakeup.S b/arch/x86/kernel/acpi/realmode/wakeup.S index f9b77fb37e5..3355973b12a 100644 --- a/arch/x86/kernel/acpi/realmode/wakeup.S +++ b/arch/x86/kernel/acpi/realmode/wakeup.S @@ -5,6 +5,7 @@ #include #include #include +#include .code16 .section ".header", "a" @@ -24,6 +25,11 @@ pmode_gdt: .quad 0 realmode_flags: .long 0 real_magic: .long 0 trampoline_segment: .word 0 +_pad1: .byte 0 +wakeup_jmp: .byte 0xea /* ljmpw */ +wakeup_jmp_off: .word 3f +wakeup_jmp_seg: .word 0 +wakeup_gdt: .quad 0, 0, 0 signature: .long 0x51ee1111 .text @@ -34,11 +40,34 @@ _start: cli cld + /* Apparently some dimwit BIOS programmers don't know how to + program a PM to RM transition, and we might end up here with + junk in the data segment descriptor registers. The only way + to repair that is to go into PM and fix it ourselves... */ + movw $16, %cx + lgdtl %cs:wakeup_gdt + movl %cr0, %eax + orb $X86_CR0_PE, %al + movl %eax, %cr0 + jmp 1f +1: ljmpw $8, $2f +2: + movw %cx, %ds + movw %cx, %es + movw %cx, %ss + movw %cx, %fs + movw %cx, %gs + + andb $~X86_CR0_PE, %al + movl %eax, %cr0 + jmp wakeup_jmp +3: /* Set up segments */ movw %cs, %ax movw %ax, %ds movw %ax, %es movw %ax, %ss + lidtl wakeup_idt movl $wakeup_stack_end, %esp @@ -98,7 +127,14 @@ bogus_real_magic: jmp 1b .data - .balign 4 + .balign 8 + + /* This is the standard real-mode IDT */ +wakeup_idt: + .word 0xffff /* limit */ + .long 0 /* address */ + .word 0 + .globl HEAP, heap_end HEAP: .long wakeup_heap diff --git a/arch/x86/kernel/acpi/realmode/wakeup.h b/arch/x86/kernel/acpi/realmode/wakeup.h index ef8166fe802..69d38d0b2b6 100644 --- a/arch/x86/kernel/acpi/realmode/wakeup.h +++ b/arch/x86/kernel/acpi/realmode/wakeup.h @@ -24,6 +24,11 @@ struct wakeup_header { u32 realmode_flags; u32 real_magic; u16 trampoline_segment; /* segment with trampoline code, 64-bit only */ + u8 _pad1; + u8 wakeup_jmp; + u16 wakeup_jmp_off; + u16 wakeup_jmp_seg; + u64 wakeup_gdt[3]; u32 signature; /* To check we have correct structure */ } __attribute__((__packed__)); diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index afc25ee9964..36af01f029e 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -50,6 +50,20 @@ int acpi_save_state_mem(void) header->video_mode = saved_video_mode; + header->wakeup_jmp_seg = acpi_wakeup_address >> 4; + /* GDT[0]: GDT self-pointer */ + header->wakeup_gdt[0] = + (u64)(sizeof(header->wakeup_gdt) - 1) + + ((u64)(acpi_wakeup_address + + ((char *)&header->wakeup_gdt - (char *)acpi_realmode)) + << 16); + /* GDT[1]: real-mode-like code segment */ + header->wakeup_gdt[1] = (0x009bULL << 40) + + ((u64)acpi_wakeup_address << 16) + 0xffff; + /* GDT[2]: real-mode-like data segment */ + header->wakeup_gdt[2] = (0x0093ULL << 40) + + ((u64)acpi_wakeup_address << 16) + 0xffff; + #ifndef CONFIG_64BIT store_gdt((struct desc_ptr *)&header->pmode_gdt); @@ -111,7 +125,7 @@ void __init acpi_reserve_bootmem(void) return; } - acpi_wakeup_address = acpi_realmode; + acpi_wakeup_address = virt_to_phys((void *)acpi_realmode); } -- cgit v1.2.3 From 64e83b5a919a65eb35b63dd7e07c188379ff8ce6 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sat, 5 Jul 2008 00:05:30 +0200 Subject: x86 ACPI: fix resume from suspend to RAM on uniprocessor x86-64 Since the trampoline code is now used for ACPI resume from suspend to RAM, the trampoline page tables have to be fixed up during boot not only on SMP systems, but also on UP systems that use the trampoline. Reference: http://bugzilla.kernel.org/show_bug.cgi?id=10923 Reported-by: Dionisus Torimens Signed-off-by: Rafael J. Wysocki Cc: Andi Kleen Cc: Andrew Morton Cc: pm list Signed-off-by: Ingo Molnar --- arch/x86/kernel/head_64.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index 10a1955bb1d..b817974ef94 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -128,7 +128,7 @@ ident_complete: /* Fixup phys_base */ addq %rbp, phys_base(%rip) -#ifdef CONFIG_SMP +#ifdef CONFIG_X86_TRAMPOLINE addq %rbp, trampoline_level4_pgt + 0(%rip) addq %rbp, trampoline_level4_pgt + (511*8)(%rip) #endif -- cgit v1.2.3