From 0e83815be719d3391bf5ea24b7fe696c07dbd417 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Mon, 27 Jul 2009 19:43:52 +0200 Subject: x86: fix section mismatch for i386 init code Startup code for i386 in arch/x86/kernel/head_32.S is using the reference variable initial_code that is located in the .cpuinit.data section. If CONFIG_HOTPLUG_CPU is enabled, startup code is not in an init section and can be called later too. In this case the reference initial_code must be kept too. This patch fixes this. See below for the section mismatch warning. WARNING: vmlinux.o(.cpuinit.data+0x0): Section mismatch in reference from the variable initial_code to the function .init.text:i386_start_kernel() The variable __cpuinitdata initial_code references a function __init i386_start_kernel(). If i386_start_kernel is only used by initial_code then annotate i386_start_kernel with a matching annotation. Signed-off-by: Robert Richter LKML-Reference: <1248716632-26844-1-git-send-email-robert.richter@amd.com> Signed-off-by: H. Peter Anvin --- arch/x86/kernel/head_32.S | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index 8663afb5653..0d98a01cbdb 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -602,7 +602,11 @@ ignore_int: #endif iret -.section .cpuinit.data,"wa" +#ifndef CONFIG_HOTPLUG_CPU + __CPUINITDATA +#else + __REFDATA +#endif .align 4 ENTRY(initial_code) .long i386_start_kernel -- cgit v1.2.3-70-g09d2 From bdc6340f4eb68295b1e7c0ade2356b56dca93d93 Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Thu, 30 Jul 2009 14:43:19 -0700 Subject: x86, pat: Fix set_memory_wc related corruption Changeset 3869c4aa18835c8c61b44bd0f3ace36e9d3b5bd0 that went in after 2.6.30-rc1 was a seemingly small change to _set_memory_wc() to make it complaint with SDM requirements. But, introduced a nasty bug, which can result in crash and/or strange corruptions when set_memory_wc is used. One such crash reported here http://lkml.org/lkml/2009/7/30/94 Actually, that changeset introduced two bugs. * change_page_attr_set() takes &addr as first argument and can the addr value might have changed on return, even for single page change_page_attr_set() call. That will make the second change_page_attr_set() in this routine operate on unrelated addr, that can eventually cause strange corruptions and bad page state crash. * The second change_page_attr_set() call, before setting _PAGE_CACHE_WC, should clear the earlier _PAGE_CACHE_UC_MINUS, as otherwise cache attribute will not be WC (will be UC instead). The patch below fixes both these problems. Sending a single patch to fix both the problems, as the change is to the same line of code. The change to have a addr_copy is not very clean. But, it is simpler than making more changes through various routines in pageattr.c. A huge thanks to Jerome for reporting this problem and providing a simple test case that helped us root cause the problem. Reported-by: Jerome Glisse Signed-off-by: Venkatesh Pallipadi Signed-off-by: Suresh Siddha LKML-Reference: <20090730214319.GA1889@linux-os.sc.intel.com> Acked-by: Dave Airlie Signed-off-by: H. Peter Anvin --- arch/x86/mm/pageattr.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index 1b734d7a896..895d90e1a81 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -997,12 +997,15 @@ EXPORT_SYMBOL(set_memory_array_uc); int _set_memory_wc(unsigned long addr, int numpages) { int ret; + unsigned long addr_copy = addr; + ret = change_page_attr_set(&addr, numpages, __pgprot(_PAGE_CACHE_UC_MINUS), 0); - if (!ret) { - ret = change_page_attr_set(&addr, numpages, - __pgprot(_PAGE_CACHE_WC), 0); + ret = change_page_attr_set_clr(&addr_copy, numpages, + __pgprot(_PAGE_CACHE_WC), + __pgprot(_PAGE_CACHE_MASK), + 0, 0, NULL); } return ret; } -- cgit v1.2.3-70-g09d2 From 8523acfe40efc1a8d3da8f473ca67cb195b06f0c Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Mon, 3 Aug 2009 09:25:45 +0200 Subject: x86: Fix CPA memtype reserving in the set_pages_array*() cases The code was incorrectly reserving memtypes using the page virtual address instead of the physical address. Furthermore, the code was not ignoring highmem pages as it ought to. ( upstream does not pass in highmem pages yet - but upcoming graphics code will do it and there's no reason to not handle this properly in the CPA APIs.) Fixes: http://bugzilla.kernel.org/show_bug.cgi?id=13884 Signed-off-by: Thomas Hellstrom Acked-by: Suresh Siddha Cc: Cc: dri-devel@lists.sourceforge.net Cc: venkatesh.pallipadi@intel.com LKML-Reference: <1249284345-7654-1-git-send-email-thellstrom@vmware.com> Signed-off-by: Ingo Molnar --- arch/x86/mm/pageattr.c | 30 +++++++++++++++++++++--------- 1 file changed, 21 insertions(+), 9 deletions(-) diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index 895d90e1a81..7e600c1962d 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -591,9 +591,12 @@ static int __change_page_attr(struct cpa_data *cpa, int primary) unsigned int level; pte_t *kpte, old_pte; - if (cpa->flags & CPA_PAGES_ARRAY) - address = (unsigned long)page_address(cpa->pages[cpa->curpage]); - else if (cpa->flags & CPA_ARRAY) + if (cpa->flags & CPA_PAGES_ARRAY) { + struct page *page = cpa->pages[cpa->curpage]; + if (unlikely(PageHighMem(page))) + return 0; + address = (unsigned long)page_address(page); + } else if (cpa->flags & CPA_ARRAY) address = cpa->vaddr[cpa->curpage]; else address = *cpa->vaddr; @@ -697,9 +700,12 @@ static int cpa_process_alias(struct cpa_data *cpa) * No need to redo, when the primary call touched the direct * mapping already: */ - if (cpa->flags & CPA_PAGES_ARRAY) - vaddr = (unsigned long)page_address(cpa->pages[cpa->curpage]); - else if (cpa->flags & CPA_ARRAY) + if (cpa->flags & CPA_PAGES_ARRAY) { + struct page *page = cpa->pages[cpa->curpage]; + if (unlikely(PageHighMem(page))) + return 0; + vaddr = (unsigned long)page_address(page); + } else if (cpa->flags & CPA_ARRAY) vaddr = cpa->vaddr[cpa->curpage]; else vaddr = *cpa->vaddr; @@ -1122,7 +1128,9 @@ int set_pages_array_uc(struct page **pages, int addrinarray) int free_idx; for (i = 0; i < addrinarray; i++) { - start = (unsigned long)page_address(pages[i]); + if (PageHighMem(pages[i])) + continue; + start = page_to_pfn(pages[i]) << PAGE_SHIFT; end = start + PAGE_SIZE; if (reserve_memtype(start, end, _PAGE_CACHE_UC_MINUS, NULL)) goto err_out; @@ -1135,7 +1143,9 @@ int set_pages_array_uc(struct page **pages, int addrinarray) err_out: free_idx = i; for (i = 0; i < free_idx; i++) { - start = (unsigned long)page_address(pages[i]); + if (PageHighMem(pages[i])) + continue; + start = page_to_pfn(pages[i]) << PAGE_SHIFT; end = start + PAGE_SIZE; free_memtype(start, end); } @@ -1164,7 +1174,9 @@ int set_pages_array_wb(struct page **pages, int addrinarray) return retval; for (i = 0; i < addrinarray; i++) { - start = (unsigned long)page_address(pages[i]); + if (PageHighMem(pages[i])) + continue; + start = page_to_pfn(pages[i]) << PAGE_SHIFT; end = start + PAGE_SIZE; free_memtype(start, end); } -- cgit v1.2.3-70-g09d2 From 6c6c51e4cc11a5456fb1172008f7c69d955af9f6 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Mon, 3 Aug 2009 22:47:32 +1000 Subject: x86: Add quirk to make Apple MacBook5,2 use reboot=pci The latest Apple MacBook (MacBook5,2) doesn't reboot successfully under Linux; neither the EFI reboot method nor the default method using the keyboard controller works (the system just hangs and doesn't reset). However, the method using the "PCI reset register" at 0xcf9 does work. This adds a quirk to detect this machine via DMI and force the reboot_type to BOOT_CF9. With this it reboots successfully without requiring a command-line option. Note that the EFI code forces reboot_type to BOOT_EFI when the machine is booted via EFI, but this overrides that since the core_initcall runs after the EFI initialization code. Signed-off-by: Paul Mackerras LKML-Reference: <19062.56420.501516.316181@cargo.ozlabs.ibm.com> Signed-off-by: H. Peter Anvin --- arch/x86/kernel/reboot.c | 34 +++++++++++++++++++++++++++++++++- 1 file changed, 33 insertions(+), 1 deletion(-) diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 508e982dd07..834c9da8bf9 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -3,6 +3,7 @@ #include #include #include +#include #include #include #include @@ -17,7 +18,6 @@ #include #ifdef CONFIG_X86_32 -# include # include # include #else @@ -404,6 +404,38 @@ EXPORT_SYMBOL(machine_real_restart); #endif /* CONFIG_X86_32 */ +/* + * Apple MacBook5,2 (2009 MacBook) needs reboot=p + */ +static int __init set_pci_reboot(const struct dmi_system_id *d) +{ + if (reboot_type != BOOT_CF9) { + reboot_type = BOOT_CF9; + printk(KERN_INFO "%s series board detected. " + "Selecting PCI-method for reboots.\n", d->ident); + } + return 0; +} + +static struct dmi_system_id __initdata pci_reboot_dmi_table[] = { + { /* Handle problems with rebooting on Apple MacBook5,2 */ + .callback = set_pci_reboot, + .ident = "Apple MacBook", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5,2"), + }, + }, + { } +}; + +static int __init pci_reboot_init(void) +{ + dmi_check_system(pci_reboot_dmi_table); + return 0; +} +core_initcall(pci_reboot_init); + static inline void kb_wait(void) { int i; -- cgit v1.2.3-70-g09d2 From 6a7bbd57ed50bb62c9a81ae5f2e202ca689e5964 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Mon, 3 Aug 2009 22:38:10 +1000 Subject: x86: Make 64-bit efi_ioremap use ioremap on MMIO regions Booting current 64-bit x86 kernels on the latest Apple MacBook (MacBook5,2) via EFI gives the following warning: [ 0.182209] ------------[ cut here ]------------ [ 0.182222] WARNING: at arch/x86/mm/pageattr.c:581 __cpa_process_fault+0x44/0xa0() [ 0.182227] Hardware name: MacBook5,2 [ 0.182231] CPA: called for zero pte. vaddr = ffff8800ffe00000 cpa->vaddr = ffff8800ffe00000 [ 0.182236] Modules linked in: [ 0.182242] Pid: 0, comm: swapper Not tainted 2.6.31-rc4 #6 [ 0.182246] Call Trace: [ 0.182254] [] ? __cpa_process_fault+0x44/0xa0 [ 0.182261] [] warn_slowpath_common+0x78/0xd0 [ 0.182266] [] warn_slowpath_fmt+0x64/0x70 [ 0.182272] [] ? update_page_count+0x3c/0x50 [ 0.182280] [] ? phys_pmd_init+0x140/0x22e [ 0.182286] [] __cpa_process_fault+0x44/0xa0 [ 0.182292] [] __change_page_attr_set_clr+0x5f0/0xb40 [ 0.182301] [] ? vm_unmap_aliases+0x175/0x190 [ 0.182307] [] change_page_attr_set_clr+0xfe/0x3d0 [ 0.182314] [] _set_memory_uc+0x2a/0x30 [ 0.182319] [] set_memory_uc+0x7b/0xb0 [ 0.182327] [] efi_enter_virtual_mode+0x2ad/0x2c9 [ 0.182334] [] start_kernel+0x2db/0x3f4 [ 0.182340] [] x86_64_start_reservations+0x99/0xb9 [ 0.182345] [] x86_64_start_kernel+0xe0/0xf2 [ 0.182357] ---[ end trace 4eaa2a86a8e2da22 ]--- [ 0.182982] init_memory_mapping: 00000000ffffc000-0000000100000000 [ 0.182993] 00ffffc000 - 0100000000 page 4k This happens because the 64-bit version of efi_ioremap calls init_memory_mapping for all addresses, regardless of whether they are RAM or MMIO. The EFI tables on this machine ask for runtime access to some MMIO regions: [ 0.000000] EFI: mem195: type=11, attr=0x8000000000000000, range=[0x0000000093400000-0x0000000093401000) (0MB) [ 0.000000] EFI: mem196: type=11, attr=0x8000000000000000, range=[0x00000000ffc00000-0x00000000ffc40000) (0MB) [ 0.000000] EFI: mem197: type=11, attr=0x8000000000000000, range=[0x00000000ffc40000-0x00000000ffc80000) (0MB) [ 0.000000] EFI: mem198: type=11, attr=0x8000000000000000, range=[0x00000000ffc80000-0x00000000ffca4000) (0MB) [ 0.000000] EFI: mem199: type=11, attr=0x8000000000000000, range=[0x00000000ffca4000-0x00000000ffcb4000) (0MB) [ 0.000000] EFI: mem200: type=11, attr=0x8000000000000000, range=[0x00000000ffcb4000-0x00000000ffffc000) (3MB) [ 0.000000] EFI: mem201: type=11, attr=0x8000000000000000, range=[0x00000000ffffc000-0x0000000100000000) (0MB) This arranges to pass the EFI memory type through to efi_ioremap, and makes efi_ioremap use ioremap rather than init_memory_mapping if the type is EFI_MEMORY_MAPPED_IO. With this, the above warning goes away. Signed-off-by: Paul Mackerras LKML-Reference: <19062.55858.533494.471153@cargo.ozlabs.ibm.com> Cc: Huang Ying Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/efi.h | 5 +++-- arch/x86/kernel/efi.c | 2 +- arch/x86/kernel/efi_64.c | 6 +++++- 3 files changed, 9 insertions(+), 4 deletions(-) diff --git a/arch/x86/include/asm/efi.h b/arch/x86/include/asm/efi.h index edc90f23e70..8406ed7f992 100644 --- a/arch/x86/include/asm/efi.h +++ b/arch/x86/include/asm/efi.h @@ -33,7 +33,7 @@ extern unsigned long asmlinkage efi_call_phys(void *, ...); #define efi_call_virt6(f, a1, a2, a3, a4, a5, a6) \ efi_call_virt(f, a1, a2, a3, a4, a5, a6) -#define efi_ioremap(addr, size) ioremap_cache(addr, size) +#define efi_ioremap(addr, size, type) ioremap_cache(addr, size) #else /* !CONFIG_X86_32 */ @@ -84,7 +84,8 @@ extern u64 efi_call6(void *fp, u64 arg1, u64 arg2, u64 arg3, efi_call6((void *)(efi.systab->runtime->f), (u64)(a1), (u64)(a2), \ (u64)(a3), (u64)(a4), (u64)(a5), (u64)(a6)) -extern void __iomem *efi_ioremap(unsigned long addr, unsigned long size); +extern void __iomem *efi_ioremap(unsigned long addr, unsigned long size, + u32 type); #endif /* CONFIG_X86_32 */ diff --git a/arch/x86/kernel/efi.c b/arch/x86/kernel/efi.c index 96f7ac0bbf0..19ccf6d0dcc 100644 --- a/arch/x86/kernel/efi.c +++ b/arch/x86/kernel/efi.c @@ -512,7 +512,7 @@ void __init efi_enter_virtual_mode(void) && end_pfn <= max_pfn_mapped)) va = __va(md->phys_addr); else - va = efi_ioremap(md->phys_addr, size); + va = efi_ioremap(md->phys_addr, size, md->type); md->virt_addr = (u64) (unsigned long) va; diff --git a/arch/x86/kernel/efi_64.c b/arch/x86/kernel/efi_64.c index 22c3b7828c5..ac0621a7ac3 100644 --- a/arch/x86/kernel/efi_64.c +++ b/arch/x86/kernel/efi_64.c @@ -98,10 +98,14 @@ void __init efi_call_phys_epilog(void) early_runtime_code_mapping_set_exec(0); } -void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size) +void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size, + u32 type) { unsigned long last_map_pfn; + if (type == EFI_MEMORY_MAPPED_IO) + return ioremap(phys_addr, size); + last_map_pfn = init_memory_mapping(phys_addr, phys_addr + size); if ((last_map_pfn << PAGE_SHIFT) < phys_addr + size) return NULL; -- cgit v1.2.3-70-g09d2 From d2ba8b211bb8abc29aa627dbd4dce08cfbc8082b Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 3 Aug 2009 14:44:54 -0700 Subject: x86: Fix assert syntax in vmlinux.lds.S Older versions of binutils did not accept the naked "ASSERT" syntax; it is considered an expression whose value needs to be assigned to something. Reported-tested-and-fixed-by: Jean Delvare Signed-off-by: H. Peter Anvin --- arch/x86/kernel/vmlinux.lds.S | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 59f31d2dd43..78d185d797d 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -393,8 +393,8 @@ SECTIONS #ifdef CONFIG_X86_32 -ASSERT((_end - LOAD_OFFSET <= KERNEL_IMAGE_SIZE), - "kernel image bigger than KERNEL_IMAGE_SIZE") +. = ASSERT((_end - LOAD_OFFSET <= KERNEL_IMAGE_SIZE), + "kernel image bigger than KERNEL_IMAGE_SIZE"); #else /* * Per-cpu symbols which need to be offset from __per_cpu_load @@ -407,12 +407,12 @@ INIT_PER_CPU(irq_stack_union); /* * Build-time check on the image size: */ -ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), - "kernel image bigger than KERNEL_IMAGE_SIZE") +. = ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), + "kernel image bigger than KERNEL_IMAGE_SIZE"); #ifdef CONFIG_SMP -ASSERT((per_cpu__irq_stack_union == 0), - "irq_stack_union is not at start of per-cpu area"); +. = ASSERT((per_cpu__irq_stack_union == 0), + "irq_stack_union is not at start of per-cpu area"); #endif #endif /* CONFIG_X86_32 */ @@ -420,7 +420,7 @@ ASSERT((per_cpu__irq_stack_union == 0), #ifdef CONFIG_KEXEC #include -ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, - "kexec control code size is too big") +. = ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, + "kexec control code size is too big"); #endif -- cgit v1.2.3-70-g09d2 From bab9a3da93bfe09c609407dedae5708b07a7ac56 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Thu, 30 Jul 2009 11:10:01 +0200 Subject: x86, msr: execute on the correct CPU subset Make rdmsr_on_cpus/wrmsr_on_cpus execute on the current CPU only if it is in the supplied bitmask. Signed-off-by: Borislav Petkov Signed-off-by: H. Peter Anvin --- arch/x86/lib/msr.c | 26 ++++++++++---------------- 1 file changed, 10 insertions(+), 16 deletions(-) diff --git a/arch/x86/lib/msr.c b/arch/x86/lib/msr.c index 1440b9c0547..caa24aca811 100644 --- a/arch/x86/lib/msr.c +++ b/arch/x86/lib/msr.c @@ -89,16 +89,13 @@ void rdmsr_on_cpus(const cpumask_t *mask, u32 msr_no, struct msr *msrs) rv.msrs = msrs; rv.msr_no = msr_no; - preempt_disable(); - /* - * FIXME: handle the CPU we're executing on separately for now until - * smp_call_function_many has been fixed to not skip it. - */ - this_cpu = raw_smp_processor_id(); - smp_call_function_single(this_cpu, __rdmsr_on_cpu, &rv, 1); + this_cpu = get_cpu(); + + if (cpumask_test_cpu(this_cpu, mask)) + __rdmsr_on_cpu(&rv); smp_call_function_many(mask, __rdmsr_on_cpu, &rv, 1); - preempt_enable(); + put_cpu(); } EXPORT_SYMBOL(rdmsr_on_cpus); @@ -121,16 +118,13 @@ void wrmsr_on_cpus(const cpumask_t *mask, u32 msr_no, struct msr *msrs) rv.msrs = msrs; rv.msr_no = msr_no; - preempt_disable(); - /* - * FIXME: handle the CPU we're executing on separately for now until - * smp_call_function_many has been fixed to not skip it. - */ - this_cpu = raw_smp_processor_id(); - smp_call_function_single(this_cpu, __wrmsr_on_cpu, &rv, 1); + this_cpu = get_cpu(); + + if (cpumask_test_cpu(this_cpu, mask)) + __wrmsr_on_cpu(&rv); smp_call_function_many(mask, __wrmsr_on_cpu, &rv, 1); - preempt_enable(); + put_cpu(); } EXPORT_SYMBOL(wrmsr_on_cpus); -- cgit v1.2.3-70-g09d2 From f1f029c7bfbf4ee1918b90a431ab823bed812504 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 3 Aug 2009 16:33:40 -0700 Subject: x86: fix assembly constraints in native_save_fl() From Gabe Black in bugzilla 13888: native_save_fl is implemented as follows: 11static inline unsigned long native_save_fl(void) 12{ 13 unsigned long flags; 14 15 asm volatile("# __raw_save_flags\n\t" 16 "pushf ; pop %0" 17 : "=g" (flags) 18 : /* no input */ 19 : "memory"); 20 21 return flags; 22} If gcc chooses to put flags on the stack, for instance because this is inlined into a larger function with more register pressure, the offset of the flags variable from the stack pointer will change when the pushf is performed. gcc doesn't attempt to understand that fact, and address used for pop will still be the same. It will write to somewhere near flags on the stack but not actually into it and overwrite some other value. I saw this happen in the ide_device_add_all function when running in a simulator I work on. I'm assuming that some quirk of how the simulated hardware is set up caused the code path this is on to be executed when it normally wouldn't. A simple fix might be to change "=g" to "=r". Reported-by: Gabe Black Signed-off-by: H. Peter Anvin Cc: Stable Team --- arch/x86/include/asm/irqflags.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/arch/x86/include/asm/irqflags.h b/arch/x86/include/asm/irqflags.h index 2bdab21f089..c6ccbe7e81a 100644 --- a/arch/x86/include/asm/irqflags.h +++ b/arch/x86/include/asm/irqflags.h @@ -12,9 +12,15 @@ static inline unsigned long native_save_fl(void) { unsigned long flags; + /* + * Note: this needs to be "=r" not "=rm", because we have the + * stack offset from what gcc expects at the time the "pop" is + * executed, and so a memory reference with respect to the stack + * would end up using the wrong address. + */ asm volatile("# __raw_save_flags\n\t" "pushf ; pop %0" - : "=g" (flags) + : "=r" (flags) : /* no input */ : "memory"); -- cgit v1.2.3-70-g09d2 From 6c7184b77464261b7d55583a48accbd1350923a3 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Mon, 27 Jul 2009 09:35:07 -0500 Subject: x86, UV: Handle missing blade-local memory correctly UV blades may not have any blade-local memory. Add a field (nid) to the UV blade structure to indicates whether the node has local memory. This is needed by the GRU driver (pushed separately). Signed-off-by: Jack Steiner Cc: linux-mm@kvack.org LKML-Reference: <20090727143507.GA7006@sgi.com> Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_hub.h | 7 +++++++ arch/x86/kernel/apic/x2apic_uv_x.c | 5 +++++ 2 files changed, 12 insertions(+) diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h index 341070f7ad5..a6dde059e02 100644 --- a/arch/x86/include/asm/uv/uv_hub.h +++ b/arch/x86/include/asm/uv/uv_hub.h @@ -327,6 +327,7 @@ struct uv_blade_info { unsigned short nr_possible_cpus; unsigned short nr_online_cpus; unsigned short pnode; + short memory_nid; }; extern struct uv_blade_info *uv_blade_info; extern short *uv_node_to_blade; @@ -363,6 +364,12 @@ static inline int uv_blade_to_pnode(int bid) return uv_blade_info[bid].pnode; } +/* Nid of memory node on blade. -1 if no blade-local memory */ +static inline int uv_blade_to_memory_nid(int bid) +{ + return uv_blade_info[bid].memory_nid; +} + /* Determine the number of possible cpus on a blade */ static inline int uv_blade_nr_possible_cpus(int bid) { diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 096d19aea2f..ad3f0a984ca 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -591,6 +591,8 @@ void __init uv_system_init(void) bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades(); uv_blade_info = kmalloc(bytes, GFP_KERNEL); BUG_ON(!uv_blade_info); + for (blade = 0; blade < uv_num_possible_blades(); blade++) + uv_blade_info[blade].memory_nid = -1; get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); @@ -629,6 +631,9 @@ void __init uv_system_init(void) lcpu = uv_blade_info[blade].nr_possible_cpus; uv_blade_info[blade].nr_possible_cpus++; + /* Any node on the blade, else will contain -1. */ + uv_blade_info[blade].memory_nid = nid; + uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base; uv_cpu_hub_info(cpu)->lowmem_remap_top = lowmem_redir_size; uv_cpu_hub_info(cpu)->m_val = m_val; -- cgit v1.2.3-70-g09d2 From cc5e4fa1bd4d2f56da07f9092281afdcd2374ab9 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Mon, 27 Jul 2009 09:36:56 -0500 Subject: x86, UV: Delete mapping of MMR rangs mapped by BIOS The UV BIOS has added additional MMR ranges that are mapped via EFI virtual mode mappings. These ranges should be deleted from ranges mapped by uv_system_init(). Signed-off-by: Jack Steiner Cc: linux-mm@kvack.org LKML-Reference: <20090727143656.GA7698@sgi.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/x2apic_uv_x.c | 31 +------------------------------ 1 file changed, 1 insertion(+), 30 deletions(-) diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index ad3f0a984ca..445c210daf8 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -362,12 +362,6 @@ static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size) BUG(); } -static __init void map_low_mmrs(void) -{ - init_extra_mapping_uc(UV_GLOBAL_MMR32_BASE, UV_GLOBAL_MMR32_SIZE); - init_extra_mapping_uc(UV_LOCAL_MMR_BASE, UV_LOCAL_MMR_SIZE); -} - enum map_type {map_wb, map_uc}; static __init void map_high(char *id, unsigned long base, int shift, @@ -395,26 +389,6 @@ static __init void map_gru_high(int max_pnode) map_high("GRU", gru.s.base, shift, max_pnode, map_wb); } -static __init void map_config_high(int max_pnode) -{ - union uvh_rh_gam_cfg_overlay_config_mmr_u cfg; - int shift = UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_BASE_SHFT; - - cfg.v = uv_read_local_mmr(UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR); - if (cfg.s.enable) - map_high("CONFIG", cfg.s.base, shift, max_pnode, map_uc); -} - -static __init void map_mmr_high(int max_pnode) -{ - union uvh_rh_gam_mmr_overlay_config_mmr_u mmr; - int shift = UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_SHFT; - - mmr.v = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR); - if (mmr.s.enable) - map_high("MMR", mmr.s.base, shift, max_pnode, map_uc); -} - static __init void map_mmioh_high(int max_pnode) { union uvh_rh_gam_mmioh_overlay_config_mmr_u mmioh; @@ -566,8 +540,6 @@ void __init uv_system_init(void) unsigned long mmr_base, present, paddr; unsigned short pnode_mask; - map_low_mmrs(); - m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG); m_val = m_n_config.s.m_skt; n_val = m_n_config.s.n_skt; @@ -667,11 +639,10 @@ void __init uv_system_init(void) pnode = (paddr >> m_val) & pnode_mask; blade = boot_pnode_to_blade(pnode); uv_node_to_blade[nid] = blade; + max_pnode = max(pnode, max_pnode); } map_gru_high(max_pnode); - map_mmr_high(max_pnode); - map_config_high(max_pnode); map_mmioh_high(max_pnode); uv_cpu_init(); -- cgit v1.2.3-70-g09d2 From 67e83f309ed0baaf01a2c956b5174905bcdc1242 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Mon, 27 Jul 2009 09:38:08 -0500 Subject: x86, UV: Fix macros for accessing large node numbers The UV chipset automatically supplies the upper bits on nodes being referenced by MMR accesses. These bit can be deleted from the hub addressing macros. Signed-off-by: Jack Steiner LKML-Reference: <20090727143808.GA8076@sgi.com> Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_hub.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h index a6dde059e02..77a68505419 100644 --- a/arch/x86/include/asm/uv/uv_hub.h +++ b/arch/x86/include/asm/uv/uv_hub.h @@ -175,7 +175,7 @@ DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); #define UV_GLOBAL_MMR32_PNODE_BITS(p) ((p) << (UV_GLOBAL_MMR32_PNODE_SHIFT)) #define UV_GLOBAL_MMR64_PNODE_BITS(p) \ - ((unsigned long)(UV_PNODE_TO_GNODE(p)) << UV_GLOBAL_MMR64_PNODE_SHIFT) + (((unsigned long)(p)) << UV_GLOBAL_MMR64_PNODE_SHIFT) #define UV_APIC_PNODE_SHIFT 6 -- cgit v1.2.3-70-g09d2 From c5997fa8d7aca3c9876a6ff71bacf27c41095ce9 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Mon, 27 Jul 2009 09:38:56 -0500 Subject: x86, UV: Fix UV apic mode Change SGI UV default apicid mode to "physical". This is required to match settings in the UV hub chip. Signed-off-by: Jack Steiner LKML-Reference: <20090727143856.GA8905@sgi.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/x2apic_uv_x.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 445c210daf8..832e908adcb 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -261,7 +261,7 @@ struct apic apic_x2apic_uv_x = { .apic_id_registered = uv_apic_id_registered, .irq_delivery_mode = dest_Fixed, - .irq_dest_mode = 1, /* logical */ + .irq_dest_mode = 0, /* physical */ .target_cpus = uv_target_cpus, .disable_esr = 0, -- cgit v1.2.3-70-g09d2 From d8c7eb34c2db6268909ae8c3958be63bde254292 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sat, 25 Jul 2009 03:23:09 -0700 Subject: x86: Don't use current_cpu_data in x2apic phys_pkg_id One system has socket 1 come up as BSP. kexeced kernel reports BSP as: [ 1.524550] Initializing cgroup subsys cpuacct [ 1.536064] initial_apicid:20 [ 1.537135] ht_mask_width:1 [ 1.538128] core_select_mask:f [ 1.539126] core_plus_mask_width:5 [ 1.558479] CPU: Physical Processor ID: 0 [ 1.559501] CPU: Processor Core ID: 0 [ 1.560539] CPU: L1 I cache: 32K, L1 D cache: 32K [ 1.579098] CPU: L2 cache: 256K [ 1.580085] CPU: L3 cache: 24576K [ 1.581108] CPU 0/0x20 -> Node 0 [ 1.596193] CPU 0 microcode level: 0xffff0008 It doesn't have correct physical processor id and will get an error: [ 38.840859] CPU0 attaching sched-domain: [ 38.848287] domain 0: span 0,8,72 level SIBLING [ 38.851151] groups: 0 8 72 [ 38.858137] domain 1: span 0,8-15,72-79 level MC [ 38.868944] groups: 0,8,72 9,73 10,74 11,75 12,76 13,77 14,78 15,79 [ 38.881383] ERROR: parent span is not a superset of domain->span [ 38.890724] domain 2: span 0-7,64-71 level CPU [ 38.899237] ERROR: domain->groups does not contain CPU0 [ 38.909229] groups: 8-15,72-79 [ 38.912547] ERROR: groups don't span domain->span [ 38.919665] domain 3: span 0-127 level NODE [ 38.930739] groups: 0-7,64-71 8-15,72-79 16-23,80-87 24-31,88-95 32-39,96-103 40-47,104-111 48-55,112-119 56-63,120-127 it turns out: we can not use current_cpu_data in phys_pgd_id for x2apic. identify_boot_cpu() is called by check_bugs() before smp_prepare_cpus() and till smp_prepare_cpus() current_cpu_data for bsp is assigned with boot_cpu_data. Just make phys_pkg_id for x2apic is aligned to xapic. Signed-off-by: Yinghai Lu Acked-by: Suresh Siddha Cc: Andrew Morton LKML-Reference: <4A6ADD0D.10002@kernel.org> Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/x2apic_cluster.c | 2 +- arch/x86/kernel/apic/x2apic_phys.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index 8e4cbb255c3..2ed4e2bb3b3 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -170,7 +170,7 @@ static unsigned long set_apic_id(unsigned int id) static int x2apic_cluster_phys_pkg_id(int initial_apicid, int index_msb) { - return current_cpu_data.initial_apicid >> index_msb; + return initial_apicid >> index_msb; } static void x2apic_send_IPI_self(int vector) diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index a284359627e..0b631c6a2e0 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -162,7 +162,7 @@ static unsigned long set_apic_id(unsigned int id) static int x2apic_phys_pkg_id(int initial_apicid, int index_msb) { - return current_cpu_data.initial_apicid >> index_msb; + return initial_apicid >> index_msb; } static void x2apic_send_IPI_self(int vector) -- cgit v1.2.3-70-g09d2 From 6abf65510944d33b47575d151c6b318993c8d2b5 Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Thu, 30 Jul 2009 16:49:30 +0100 Subject: x86, 32-bit: Fix double accounting in reserve_top_address() With VMALLOC_END included in the calculation of MAXMEM (as of 2.6.28) it is no longer correct to also bump __VMALLOC_RESERVE in reserve_top_address(). Doing so results in needlessly small lowmem. Signed-off-by: Jan Beulich LKML-Reference: <4A71DD2A020000780000D482@vpn.id2.novell.com> Signed-off-by: Ingo Molnar --- arch/x86/mm/pgtable.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c index af8f9650058..ed34f5e3599 100644 --- a/arch/x86/mm/pgtable.c +++ b/arch/x86/mm/pgtable.c @@ -329,7 +329,6 @@ void __init reserve_top_address(unsigned long reserve) printk(KERN_INFO "Reserving virtual address space above 0x%08x\n", (int)-reserve); __FIXADDR_TOP = -reserve - PAGE_SIZE; - __VMALLOC_RESERVE += reserve; #endif } -- cgit v1.2.3-70-g09d2 From 2a5ef41661b56cf4eee042a6967c4e14b63e8eac Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Mon, 20 Jul 2009 09:28:41 -0500 Subject: x86, UV: Complete IRQ interrupt migration in arch_enable_uv_irq() In uv_setup_irq(), the call to create_irq() initially assigns IRQ vectors to cpu 0. The subsequent call to assign_irq_vector() in arch_enable_uv_irq() migrates the IRQ to another cpu and frees the cpu 0 vector - at least it will be freed as soon as the "IRQ move" completes. arch_enable_uv_irq() needs to send a cleanup IPI to complete the IRQ move. Otherwise, assignment of GRU interrupts on large systems (>200 cpus) will exhaust the cpu 0 interrupt vectors and initialization of the GRU driver will fail. Signed-off-by: Jack Steiner LKML-Reference: <20090720142840.GA8885@sgi.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/io_apic.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 2284a4812b6..d2ed6c5ddc8 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -3793,6 +3793,9 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade, mmr_pnode = uv_blade_to_pnode(mmr_blade); uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value); + if (cfg->move_in_progress) + send_cleanup_vector(cfg); + return irq; } -- cgit v1.2.3-70-g09d2 From dc731fbbadf5d65c98fcd6c86472aa286c16458a Mon Sep 17 00:00:00 2001 From: Subrata Modak Date: Tue, 21 Jul 2009 08:02:27 +0530 Subject: x86: Work around compilation warning in arch/x86/kernel/apm_32.c The following fix was initially inspired by David Howells fix few days back: http://lkml.org/lkml/2009/7/9/109 However, Ingo disapproves such fixes as it's dangerous (it can hide future, relevant warnings) - in something as performance-uncritical. So, initialize 'err' to '0' to work around a GCC false positive warning: http://lkml.org/lkml/2009/7/18/89 Signed-off-by: Subrata Modak Cc: Sachin P Sant Cc: David Howells Cc: Balbir Singh Cc: Stephen Rothwell LKML-Reference: <20090721023226.31855.67236.sendpatchset@subratamodak.linux.ibm.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/apm_32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index 79302e9a33a..442b5508893 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -811,7 +811,7 @@ static int apm_do_idle(void) u8 ret = 0; int idled = 0; int polling; - int err; + int err = 0; polling = !!(current_thread_info()->status & TS_POLLING); if (polling) { -- cgit v1.2.3-70-g09d2