From 323ec001c6bb98eeabb5abbdbb8c8055d9496554 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Sun, 29 Jun 2008 14:19:31 +0400 Subject: x86: use generic per-device dma coherent allocator Signed-off-by: Dmitry Baryshkov Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 1 + arch/x86/kernel/pci-dma.c | 122 +--------------------------------------------- 2 files changed, 3 insertions(+), 120 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e0edaaa6920a..9d4714e0020c 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -25,6 +25,7 @@ config X86 select HAVE_KRETPROBES select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64) select HAVE_ARCH_KGDB if !X86_VOYAGER + select HAVE_GENERIC_DMA_COHERENT if X86_32 config ARCH_DEFCONFIG string diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index dc00a1331ace..b75c81a8d3e6 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -197,124 +197,6 @@ static __init int iommu_setup(char *p) } early_param("iommu", iommu_setup); -#ifdef CONFIG_X86_32 -int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr, - dma_addr_t device_addr, size_t size, int flags) -{ - void __iomem *mem_base = NULL; - int pages = size >> PAGE_SHIFT; - int bitmap_size = BITS_TO_LONGS(pages) * sizeof(long); - - if ((flags & (DMA_MEMORY_MAP | DMA_MEMORY_IO)) == 0) - goto out; - if (!size) - goto out; - if (dev->dma_mem) - goto out; - - /* FIXME: this routine just ignores DMA_MEMORY_INCLUDES_CHILDREN */ - - mem_base = ioremap(bus_addr, size); - if (!mem_base) - goto out; - - dev->dma_mem = kzalloc(sizeof(struct dma_coherent_mem), GFP_KERNEL); - if (!dev->dma_mem) - goto out; - dev->dma_mem->bitmap = kzalloc(bitmap_size, GFP_KERNEL); - if (!dev->dma_mem->bitmap) - goto free1_out; - - dev->dma_mem->virt_base = mem_base; - dev->dma_mem->device_base = device_addr; - dev->dma_mem->size = pages; - dev->dma_mem->flags = flags; - - if (flags & DMA_MEMORY_MAP) - return DMA_MEMORY_MAP; - - return DMA_MEMORY_IO; - - free1_out: - kfree(dev->dma_mem); - out: - if (mem_base) - iounmap(mem_base); - return 0; -} -EXPORT_SYMBOL(dma_declare_coherent_memory); - -void dma_release_declared_memory(struct device *dev) -{ - struct dma_coherent_mem *mem = dev->dma_mem; - - if (!mem) - return; - dev->dma_mem = NULL; - iounmap(mem->virt_base); - kfree(mem->bitmap); - kfree(mem); -} -EXPORT_SYMBOL(dma_release_declared_memory); - -void *dma_mark_declared_memory_occupied(struct device *dev, - dma_addr_t device_addr, size_t size) -{ - struct dma_coherent_mem *mem = dev->dma_mem; - int pos, err; - int pages = (size + (device_addr & ~PAGE_MASK) + PAGE_SIZE - 1); - - pages >>= PAGE_SHIFT; - - if (!mem) - return ERR_PTR(-EINVAL); - - pos = (device_addr - mem->device_base) >> PAGE_SHIFT; - err = bitmap_allocate_region(mem->bitmap, pos, get_order(pages)); - if (err != 0) - return ERR_PTR(err); - return mem->virt_base + (pos << PAGE_SHIFT); -} -EXPORT_SYMBOL(dma_mark_declared_memory_occupied); - -static int dma_alloc_from_coherent_mem(struct device *dev, ssize_t size, - dma_addr_t *dma_handle, void **ret) -{ - struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL; - int order = get_order(size); - - if (mem) { - int page = bitmap_find_free_region(mem->bitmap, mem->size, - order); - if (page >= 0) { - *dma_handle = mem->device_base + (page << PAGE_SHIFT); - *ret = mem->virt_base + (page << PAGE_SHIFT); - memset(*ret, 0, size); - } - if (mem->flags & DMA_MEMORY_EXCLUSIVE) - *ret = NULL; - } - return (mem != NULL); -} - -static int dma_release_coherent(struct device *dev, int order, void *vaddr) -{ - struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL; - - if (mem && vaddr >= mem->virt_base && vaddr < - (mem->virt_base + (mem->size << PAGE_SHIFT))) { - int page = (vaddr - mem->virt_base) >> PAGE_SHIFT; - - bitmap_release_region(mem->bitmap, page, order); - return 1; - } - return 0; -} -#else -#define dma_alloc_from_coherent_mem(dev, size, handle, ret) (0) -#define dma_release_coherent(dev, order, vaddr) (0) -#endif /* CONFIG_X86_32 */ - int dma_supported(struct device *dev, u64 mask) { #ifdef CONFIG_PCI @@ -383,7 +265,7 @@ dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, /* ignore region specifiers */ gfp &= ~(__GFP_DMA | __GFP_HIGHMEM | __GFP_DMA32); - if (dma_alloc_from_coherent_mem(dev, size, dma_handle, &memory)) + if (dma_alloc_from_coherent(dev, size, dma_handle, &memory)) return memory; if (!dev) { @@ -486,7 +368,7 @@ void dma_free_coherent(struct device *dev, size_t size, { int order = get_order(size); WARN_ON(irqs_disabled()); /* for portability */ - if (dma_release_coherent(dev, order, vaddr)) + if (dma_release_from_coherent(dev, order, vaddr)) return; if (dma_ops->unmap_single) dma_ops->unmap_single(dev, bus, size, 0); -- cgit v1.2.2 From 91467bdf6e53058af13fd255375d6634ba0c70e0 Mon Sep 17 00:00:00 2001 From: Bernhard Walle Date: Fri, 18 Jul 2008 19:07:53 +0200 Subject: x86: move dma32_reserve_bootmem() after reserve_crashkernel() On a x86-64 machine (nothing special I could encounter) I had the problem that crashkernel reservation with the usual "64M@16M" failed. While debugging that, I encountered that dma32_reserve_bootmem() reserves a memory region which is in that area. Because dma32_reserve_bootmem() does not rely on a specific offset but crashkernel does, it makes sense to move the dma32_reserve_bootmem() reservation down a bit. I tested that patch and it works without problems. I don't see any negative effects of that move, but maybe I oversaw something ... While we strictly don't need that patch in 2.6.27 because we have the automatic, dynamic offset detection, it makes sense to also include it here because: - it's easier to get it in -stable then, - many people are still used to the 'crashkernel=...@16M' syntax, - not everybody may be using a reloatable kernel. Signed-off-by: Bernhard Walle Cc: kexec@lists.infradead.org Cc: vgoyal@redhat.com Cc: akpm@linux-foundation.org Cc: Bernhard Walle Cc: yhlu.kernel@gmail.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/setup.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 531b55b8e81a..74d110ef2690 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -792,10 +792,6 @@ void __init setup_arch(char **cmdline_p) initmem_init(0, max_pfn); -#ifdef CONFIG_X86_64 - dma32_reserve_bootmem(); -#endif - #ifdef CONFIG_ACPI_SLEEP /* * Reserve low memory region for sleep support. @@ -810,6 +806,15 @@ void __init setup_arch(char **cmdline_p) #endif reserve_crashkernel(); +#ifdef CONFIG_X86_64 + /* + * dma32_reserve_bootmem() allocates bootmem which may conflict + * with the crashkernel command line, so do that after + * reserve_crashkernel() + */ + dma32_reserve_bootmem(); +#endif + reserve_ibft_region(); #ifdef CONFIG_KVM_CLOCK -- cgit v1.2.2 From 1f972768a1df1518f45adb6b8ffbf04fa1c99737 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 26 Jul 2008 13:52:50 +0200 Subject: x86, RDC321x: add to mach-default first step to add RDC321x support to the default PC architecture. Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 26 +++++++++++--------------- arch/x86/Makefile | 5 ----- 2 files changed, 11 insertions(+), 20 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e3cba0b45600..39ae67985950 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -23,7 +23,7 @@ config X86 select HAVE_OPROFILE select HAVE_IOREMAP_PROT select HAVE_KPROBES - select ARCH_WANT_OPTIONAL_GPIOLIB if !X86_RDC321X + select ARCH_WANT_OPTIONAL_GPIOLIB select HAVE_KRETPROBES select HAVE_DYNAMIC_FTRACE select HAVE_FTRACE @@ -332,20 +332,6 @@ config X86_BIGSMP endif -config X86_RDC321X - bool "RDC R-321x SoC" - depends on X86_32 - select M486 - select X86_REBOOTFIXUPS - select GENERIC_GPIO - select LEDS_CLASS - select LEDS_GPIO - select NEW_LEDS - help - This option is needed for RDC R-321x system-on-chip, also known - as R-8610-(G). - If you don't have one of these chips, you should say N here. - config X86_VSMP bool "Support for ScaleMP vSMP" select PARAVIRT @@ -369,6 +355,16 @@ config X86_VISWS A kernel compiled for the Visual Workstation will run on general PCs as well. See for details. +config X86_RDC321X + bool "RDC R-321x SoC" + depends on X86_32 + select M486 + select X86_REBOOTFIXUPS + help + This option is needed for RDC R-321x system-on-chip, also known + as R-8610-(G). + If you don't have one of these chips, you should say N here. + config SCHED_NO_NO_OMIT_FRAME_POINTER def_bool y prompt "Single-depth WCHAN output" diff --git a/arch/x86/Makefile b/arch/x86/Makefile index 919ce21ea654..f5631da585b6 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -118,11 +118,6 @@ mflags-$(CONFIG_X86_GENERICARCH):= -Iinclude/asm-x86/mach-generic fcore-$(CONFIG_X86_GENERICARCH) += arch/x86/mach-generic/ mcore-$(CONFIG_X86_GENERICARCH) := arch/x86/mach-default/ -# RDC R-321x subarch support -mflags-$(CONFIG_X86_RDC321X) := -Iinclude/asm-x86/mach-rdc321x -mcore-$(CONFIG_X86_RDC321X) := arch/x86/mach-default/ -core-$(CONFIG_X86_RDC321X) += arch/x86/mach-rdc321x/ - # default subarch .h files mflags-y += -Iinclude/asm-x86/mach-default -- cgit v1.2.2 From a8132e5fe2c4f3f780b8bd3cce7851640f79f1c7 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Fri, 25 Jul 2008 14:57:59 +0200 Subject: x86, AMD IOMMU: replace to_pages macro with iommu_num_pages This patch removes the to_pages macro from AMD IOMMU code and calls the generic iommu_num_pages function instead. Signed-off-by: Joerg Roedel Cc: iommu@lists.linux-foundation.org Cc: bhavna.sarathy@amd.com Cc: robert.richter@amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/amd_iommu.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index c25210e6ac88..79eff735fde6 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -29,9 +29,6 @@ #define CMD_SET_TYPE(cmd, t) ((cmd)->data[1] |= ((t) << 28)) -#define to_pages(addr, size) \ - (round_up(((addr) & ~PAGE_MASK) + (size), PAGE_SIZE) >> PAGE_SHIFT) - #define EXIT_LOOP_COUNT 10000000 static DEFINE_RWLOCK(amd_iommu_devtable_lock); @@ -185,7 +182,7 @@ static int iommu_flush_pages(struct amd_iommu *iommu, u16 domid, u64 address, size_t size) { int s = 0; - unsigned pages = to_pages(address, size); + unsigned pages = iommu_num_pages(address, size); address &= PAGE_MASK; @@ -557,8 +554,8 @@ static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu, if (iommu->exclusion_start && iommu->exclusion_start < dma_dom->aperture_size) { unsigned long startpage = iommu->exclusion_start >> PAGE_SHIFT; - int pages = to_pages(iommu->exclusion_start, - iommu->exclusion_length); + int pages = iommu_num_pages(iommu->exclusion_start, + iommu->exclusion_length); dma_ops_reserve_addresses(dma_dom, startpage, pages); } @@ -767,7 +764,7 @@ static dma_addr_t __map_single(struct device *dev, unsigned int pages; int i; - pages = to_pages(paddr, size); + pages = iommu_num_pages(paddr, size); paddr &= PAGE_MASK; address = dma_ops_alloc_addresses(dev, dma_dom, pages); @@ -802,7 +799,7 @@ static void __unmap_single(struct amd_iommu *iommu, if ((dma_addr == 0) || (dma_addr + size > dma_dom->aperture_size)) return; - pages = to_pages(dma_addr, size); + pages = iommu_num_pages(dma_addr, size); dma_addr &= PAGE_MASK; start = dma_addr; -- cgit v1.2.2 From 87e39ea5714dd59ba31e36c25833d2b20255a29d Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Fri, 25 Jul 2008 14:58:00 +0200 Subject: x86 gart: replace to_pages macro with iommu_num_pages This patch removes the to_pages macro from x86 GART code and calls the generic iommu_num_pages function instead. Signed-off-by: Joerg Roedel Cc: iommu@lists.linux-foundation.org Cc: bhavna.sarathy@amd.com Cc: robert.richter@amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/pci-gart_64.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index df5f142657d2..c98240431561 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -67,9 +67,6 @@ static u32 gart_unmapped_entry; (((x) & 0xfffff000) | (((x) >> 32) << 4) | GPTE_VALID | GPTE_COHERENT) #define GPTE_DECODE(x) (((x) & 0xfffff000) | (((u64)(x) & 0xff0) << 28)) -#define to_pages(addr, size) \ - (round_up(((addr) & ~PAGE_MASK) + (size), PAGE_SIZE) >> PAGE_SHIFT) - #define EMERGENCY_PAGES 32 /* = 128KB */ #ifdef CONFIG_AGP @@ -241,7 +238,7 @@ nonforced_iommu(struct device *dev, unsigned long addr, size_t size) static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem, size_t size, int dir) { - unsigned long npages = to_pages(phys_mem, size); + unsigned long npages = iommu_num_pages(phys_mem, size); unsigned long iommu_page = alloc_iommu(dev, npages); int i; @@ -304,7 +301,7 @@ static void gart_unmap_single(struct device *dev, dma_addr_t dma_addr, return; iommu_page = (dma_addr - iommu_bus_base)>>PAGE_SHIFT; - npages = to_pages(dma_addr, size); + npages = iommu_num_pages(dma_addr, size); for (i = 0; i < npages; i++) { iommu_gatt_base[iommu_page + i] = gart_unmapped_entry; CLEAR_LEAK(iommu_page + i); @@ -387,7 +384,7 @@ static int __dma_map_cont(struct device *dev, struct scatterlist *start, } addr = phys_addr; - pages = to_pages(s->offset, s->length); + pages = iommu_num_pages(s->offset, s->length); while (pages--) { iommu_gatt_base[iommu_page] = GPTE_ENCODE(addr); SET_LEAK(iommu_page); @@ -470,7 +467,7 @@ gart_map_sg(struct device *dev, struct scatterlist *sg, int nents, int dir) seg_size += s->length; need = nextneed; - pages += to_pages(s->offset, s->length); + pages += iommu_num_pages(s->offset, s->length); ps = s; } if (dma_map_cont(dev, start_sg, i - start, sgmap, pages, need) < 0) -- cgit v1.2.2 From 3a61ec387c9092dfc91a5959145d36835a72fc4c Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Fri, 25 Jul 2008 13:07:50 +0200 Subject: x86, AMD IOMMU: include amd_iommu_last_bdf in device initialization All the values read while searching for amd_iommu_last_bdf are defined as inclusive. Let the code handle this value as such. Found by Wei Wang. Thanks Wei. Signed-off-by: Joerg Roedel Cc: iommu@lists.linux-foundation.org Cc: bhavna.sarathy@amd.com Cc: robert.richter@amd.com Cc: Wei Wang Signed-off-by: Ingo Molnar --- arch/x86/kernel/amd_iommu.c | 4 ++-- arch/x86/kernel/amd_iommu_init.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index c25210e6ac88..74697408576f 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -667,7 +667,7 @@ static int get_device_resources(struct device *dev, _bdf = calc_devid(pcidev->bus->number, pcidev->devfn); /* device not translated by any IOMMU in the system? */ - if (_bdf >= amd_iommu_last_bdf) { + if (_bdf > amd_iommu_last_bdf) { *iommu = NULL; *domain = NULL; *bdf = 0xffff; @@ -1085,7 +1085,7 @@ void prealloc_protection_domains(void) while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) { devid = (dev->bus->number << 8) | dev->devfn; - if (devid >= amd_iommu_last_bdf) + if (devid > amd_iommu_last_bdf) continue; devid = amd_iommu_alias_table[devid]; if (domain_for_device(devid)) diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c index c9d8ff2eb130..d9a9da597e79 100644 --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -732,7 +732,7 @@ static int __init init_exclusion_range(struct ivmd_header *m) set_device_exclusion_range(m->devid, m); break; case ACPI_IVMD_TYPE_ALL: - for (i = 0; i < amd_iommu_last_bdf; ++i) + for (i = 0; i <= amd_iommu_last_bdf; ++i) set_device_exclusion_range(i, m); break; case ACPI_IVMD_TYPE_RANGE: @@ -934,7 +934,7 @@ int __init amd_iommu_init(void) /* * let all alias entries point to itself */ - for (i = 0; i < amd_iommu_last_bdf; ++i) + for (i = 0; i <= amd_iommu_last_bdf; ++i) amd_iommu_alias_table[i] = i; /* -- cgit v1.2.2 From 6524d938b3360504b43a1278b5a8403e85383d1a Mon Sep 17 00:00:00 2001 From: Mike Travis Date: Thu, 24 Jul 2008 18:21:30 -0700 Subject: cpumask: put cpumask_of_cpu_map in the initdata section * Create the cpumask_of_cpu_map statically in the init data section using NR_CPUS but replace it during boot up with one sized by nr_cpu_ids (num possible cpus). Signed-off-by: Mike Travis Cc: Andrew Morton Cc: Jack Steiner Cc: Rusty Russell Signed-off-by: Ingo Molnar --- arch/x86/kernel/setup_percpu.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index f7745f94c006..1cd53dfcd309 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -81,10 +81,12 @@ static void __init setup_per_cpu_maps(void) } #ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP -cpumask_t *cpumask_of_cpu_map __read_mostly; -EXPORT_SYMBOL(cpumask_of_cpu_map); - -/* requires nr_cpu_ids to be initialized */ +/* + * Replace static cpumask_of_cpu_map in the initdata section, + * with one that's allocated sized by the possible number of cpus. + * + * (requires nr_cpu_ids to be initialized) + */ static void __init setup_cpumask_of_cpu(void) { int i; -- cgit v1.2.2 From 0bc3cc03fa6e1c20aecb5a33356bcaae410640b9 Mon Sep 17 00:00:00 2001 From: Mike Travis Date: Thu, 24 Jul 2008 18:21:31 -0700 Subject: cpumask: change cpumask_of_cpu_ptr to use new cpumask_of_cpu * Replace previous instances of the cpumask_of_cpu_ptr* macros with a the new (lvalue capable) generic cpumask_of_cpu(). Signed-off-by: Mike Travis Cc: Andrew Morton Cc: Jack Steiner Cc: Rusty Russell Signed-off-by: Ingo Molnar --- arch/x86/kernel/acpi/cstate.c | 3 +-- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 10 +++------- arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 15 +++++---------- arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c | 12 ++++-------- arch/x86/kernel/cpu/cpufreq/speedstep-ich.c | 3 +-- arch/x86/kernel/cpu/intel_cacheinfo.c | 3 +-- arch/x86/kernel/ldt.c | 6 ++---- arch/x86/kernel/microcode.c | 17 +++++------------ arch/x86/kernel/reboot.c | 11 +++-------- 9 files changed, 25 insertions(+), 55 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/acpi/cstate.c b/arch/x86/kernel/acpi/cstate.c index 9220cf46aa10..c2502eb9aa83 100644 --- a/arch/x86/kernel/acpi/cstate.c +++ b/arch/x86/kernel/acpi/cstate.c @@ -73,7 +73,6 @@ int acpi_processor_ffh_cstate_probe(unsigned int cpu, struct cpuinfo_x86 *c = &cpu_data(cpu); cpumask_t saved_mask; - cpumask_of_cpu_ptr(new_mask, cpu); int retval; unsigned int eax, ebx, ecx, edx; unsigned int edx_part; @@ -92,7 +91,7 @@ int acpi_processor_ffh_cstate_probe(unsigned int cpu, /* Make sure we are running on right CPU */ saved_mask = current->cpus_allowed; - retval = set_cpus_allowed_ptr(current, new_mask); + retval = set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); if (retval) return -1; diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index ff2fff56f0a8..dd097b835839 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -200,12 +200,10 @@ static void drv_read(struct drv_cmd *cmd) static void drv_write(struct drv_cmd *cmd) { cpumask_t saved_mask = current->cpus_allowed; - cpumask_of_cpu_ptr_declare(cpu_mask); unsigned int i; for_each_cpu_mask_nr(i, cmd->mask) { - cpumask_of_cpu_ptr_next(cpu_mask, i); - set_cpus_allowed_ptr(current, cpu_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(i)); do_drv_write(cmd); } @@ -269,12 +267,11 @@ static unsigned int get_measured_perf(unsigned int cpu) } aperf_cur, mperf_cur; cpumask_t saved_mask; - cpumask_of_cpu_ptr(cpu_mask, cpu); unsigned int perf_percent; unsigned int retval; saved_mask = current->cpus_allowed; - set_cpus_allowed_ptr(current, cpu_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); if (get_cpu() != cpu) { /* We were not able to run on requested processor */ put_cpu(); @@ -340,7 +337,6 @@ static unsigned int get_measured_perf(unsigned int cpu) static unsigned int get_cur_freq_on_cpu(unsigned int cpu) { - cpumask_of_cpu_ptr(cpu_mask, cpu); struct acpi_cpufreq_data *data = per_cpu(drv_data, cpu); unsigned int freq; unsigned int cached_freq; @@ -353,7 +349,7 @@ static unsigned int get_cur_freq_on_cpu(unsigned int cpu) } cached_freq = data->freq_table[data->acpi_data->state].frequency; - freq = extract_freq(get_cur_val(cpu_mask), data); + freq = extract_freq(get_cur_val(&cpumask_of_cpu(cpu)), data); if (freq != cached_freq) { /* * The dreaded BIOS frequency change behind our back. diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index 53c7b6936973..c45ca6d4dce1 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -479,12 +479,11 @@ static int core_voltage_post_transition(struct powernow_k8_data *data, u32 reqvi static int check_supported_cpu(unsigned int cpu) { cpumask_t oldmask; - cpumask_of_cpu_ptr(cpu_mask, cpu); u32 eax, ebx, ecx, edx; unsigned int rc = 0; oldmask = current->cpus_allowed; - set_cpus_allowed_ptr(current, cpu_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); if (smp_processor_id() != cpu) { printk(KERN_ERR PFX "limiting to cpu %u failed\n", cpu); @@ -1017,7 +1016,6 @@ static int transition_frequency_pstate(struct powernow_k8_data *data, unsigned i static int powernowk8_target(struct cpufreq_policy *pol, unsigned targfreq, unsigned relation) { cpumask_t oldmask; - cpumask_of_cpu_ptr(cpu_mask, pol->cpu); struct powernow_k8_data *data = per_cpu(powernow_data, pol->cpu); u32 checkfid; u32 checkvid; @@ -1032,7 +1030,7 @@ static int powernowk8_target(struct cpufreq_policy *pol, unsigned targfreq, unsi /* only run on specific CPU from here on */ oldmask = current->cpus_allowed; - set_cpus_allowed_ptr(current, cpu_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(pol->cpu)); if (smp_processor_id() != pol->cpu) { printk(KERN_ERR PFX "limiting to cpu %u failed\n", pol->cpu); @@ -1107,7 +1105,6 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) { struct powernow_k8_data *data; cpumask_t oldmask; - cpumask_of_cpu_ptr_declare(newmask); int rc; if (!cpu_online(pol->cpu)) @@ -1159,8 +1156,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) /* only run on specific CPU from here on */ oldmask = current->cpus_allowed; - cpumask_of_cpu_ptr_next(newmask, pol->cpu); - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(pol->cpu)); if (smp_processor_id() != pol->cpu) { printk(KERN_ERR PFX "limiting to cpu %u failed\n", pol->cpu); @@ -1182,7 +1178,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) set_cpus_allowed_ptr(current, &oldmask); if (cpu_family == CPU_HW_PSTATE) - pol->cpus = *newmask; + pol->cpus = cpumask_of_cpu(pol->cpu); else pol->cpus = per_cpu(cpu_core_map, pol->cpu); data->available_cores = &(pol->cpus); @@ -1248,7 +1244,6 @@ static unsigned int powernowk8_get (unsigned int cpu) { struct powernow_k8_data *data; cpumask_t oldmask = current->cpus_allowed; - cpumask_of_cpu_ptr(newmask, cpu); unsigned int khz = 0; unsigned int first; @@ -1258,7 +1253,7 @@ static unsigned int powernowk8_get (unsigned int cpu) if (!data) return -EINVAL; - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); if (smp_processor_id() != cpu) { printk(KERN_ERR PFX "limiting to CPU %d failed in powernowk8_get\n", cpu); diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c index ca2ac13b7af2..15e13c01cc36 100644 --- a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c +++ b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c @@ -324,10 +324,9 @@ static unsigned int get_cur_freq(unsigned int cpu) unsigned l, h; unsigned clock_freq; cpumask_t saved_mask; - cpumask_of_cpu_ptr(new_mask, cpu); saved_mask = current->cpus_allowed; - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); if (smp_processor_id() != cpu) return 0; @@ -585,15 +584,12 @@ static int centrino_target (struct cpufreq_policy *policy, * Best effort undo.. */ - if (!cpus_empty(*covered_cpus)) { - cpumask_of_cpu_ptr_declare(new_mask); - + if (!cpus_empty(*covered_cpus)) for_each_cpu_mask_nr(j, *covered_cpus) { - cpumask_of_cpu_ptr_next(new_mask, j); - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, + &cpumask_of_cpu(j)); wrmsr(MSR_IA32_PERF_CTL, oldmsr, h); } - } tmp = freqs.new; freqs.new = freqs.old; diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c b/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c index 2f3728dc24f6..191f7263c61d 100644 --- a/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c +++ b/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c @@ -244,8 +244,7 @@ static unsigned int _speedstep_get(const cpumask_t *cpus) static unsigned int speedstep_get(unsigned int cpu) { - cpumask_of_cpu_ptr(newmask, cpu); - return _speedstep_get(newmask); + return _speedstep_get(&cpumask_of_cpu(cpu)); } /** diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 650d40f7912b..6b0a10b002f1 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -516,7 +516,6 @@ static int __cpuinit detect_cache_attributes(unsigned int cpu) unsigned long j; int retval; cpumask_t oldmask; - cpumask_of_cpu_ptr(newmask, cpu); if (num_cache_leaves == 0) return -ENOENT; @@ -527,7 +526,7 @@ static int __cpuinit detect_cache_attributes(unsigned int cpu) return -ENOMEM; oldmask = current->cpus_allowed; - retval = set_cpus_allowed_ptr(current, newmask); + retval = set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); if (retval) goto out; diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c index 3fee2aa50f3f..b68e21f06f4f 100644 --- a/arch/x86/kernel/ldt.c +++ b/arch/x86/kernel/ldt.c @@ -62,12 +62,10 @@ static int alloc_ldt(mm_context_t *pc, int mincount, int reload) if (reload) { #ifdef CONFIG_SMP - cpumask_of_cpu_ptr_declare(mask); - preempt_disable(); load_LDT(pc); - cpumask_of_cpu_ptr_next(mask, smp_processor_id()); - if (!cpus_equal(current->mm->cpu_vm_mask, *mask)) + if (!cpus_equal(current->mm->cpu_vm_mask, + cpumask_of_cpu(smp_processor_id()))) smp_call_function(flush_ldt, current->mm, 1); preempt_enable(); #else diff --git a/arch/x86/kernel/microcode.c b/arch/x86/kernel/microcode.c index 6994c751590e..652fa5c38ebe 100644 --- a/arch/x86/kernel/microcode.c +++ b/arch/x86/kernel/microcode.c @@ -388,7 +388,6 @@ static int do_microcode_update (void) void *new_mc = NULL; int cpu; cpumask_t old; - cpumask_of_cpu_ptr_declare(newmask); old = current->cpus_allowed; @@ -405,8 +404,7 @@ static int do_microcode_update (void) if (!uci->valid) continue; - cpumask_of_cpu_ptr_next(newmask, cpu); - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); error = get_maching_microcode(new_mc, cpu); if (error < 0) goto out; @@ -576,7 +574,6 @@ static int apply_microcode_check_cpu(int cpu) struct cpuinfo_x86 *c = &cpu_data(cpu); struct ucode_cpu_info *uci = ucode_cpu_info + cpu; cpumask_t old; - cpumask_of_cpu_ptr(newmask, cpu); unsigned int val[2]; int err = 0; @@ -585,7 +582,7 @@ static int apply_microcode_check_cpu(int cpu) return 0; old = current->cpus_allowed; - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); /* Check if the microcode we have in memory matches the CPU */ if (c->x86_vendor != X86_VENDOR_INTEL || c->x86 < 6 || @@ -623,12 +620,11 @@ static int apply_microcode_check_cpu(int cpu) static void microcode_init_cpu(int cpu, int resume) { cpumask_t old; - cpumask_of_cpu_ptr(newmask, cpu); struct ucode_cpu_info *uci = ucode_cpu_info + cpu; old = current->cpus_allowed; - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); mutex_lock(µcode_mutex); collect_cpu_info(cpu); if (uci->valid && system_state == SYSTEM_RUNNING && !resume) @@ -661,13 +657,10 @@ static ssize_t reload_store(struct sys_device *dev, if (end == buf) return -EINVAL; if (val == 1) { - cpumask_t old; - cpumask_of_cpu_ptr(newmask, cpu); - - old = current->cpus_allowed; + cpumask_t old = current->cpus_allowed; get_online_cpus(); - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); mutex_lock(µcode_mutex); if (uci->valid) diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 06a9f643817e..724adfc63cb9 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -414,25 +414,20 @@ void native_machine_shutdown(void) /* The boot cpu is always logical cpu 0 */ int reboot_cpu_id = 0; - cpumask_of_cpu_ptr(newmask, reboot_cpu_id); #ifdef CONFIG_X86_32 /* See if there has been given a command line override */ if ((reboot_cpu != -1) && (reboot_cpu < NR_CPUS) && - cpu_online(reboot_cpu)) { + cpu_online(reboot_cpu)) reboot_cpu_id = reboot_cpu; - cpumask_of_cpu_ptr_next(newmask, reboot_cpu_id); - } #endif /* Make certain the cpu I'm about to reboot on is online */ - if (!cpu_online(reboot_cpu_id)) { + if (!cpu_online(reboot_cpu_id)) reboot_cpu_id = smp_processor_id(); - cpumask_of_cpu_ptr_next(newmask, reboot_cpu_id); - } /* Make certain I only run on the appropriate processor */ - set_cpus_allowed_ptr(current, newmask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(reboot_cpu_id)); /* O.K Now that I'm on the appropriate processor, * stop all of the others. -- cgit v1.2.2 From 8d8bb39b9eba32dd70e87fd5ad5c5dd4ba118e06 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Fri, 25 Jul 2008 19:44:49 -0700 Subject: dma-mapping: add the device argument to dma_mapping_error() Add per-device dma_mapping_ops support for CONFIG_X86_64 as POWER architecture does: This enables us to cleanly fix the Calgary IOMMU issue that some devices are not behind the IOMMU (http://lkml.org/lkml/2008/5/8/423). I think that per-device dma_mapping_ops support would be also helpful for KVM people to support PCI passthrough but Andi thinks that this makes it difficult to support the PCI passthrough (see the above thread). So I CC'ed this to KVM camp. Comments are appreciated. A pointer to dma_mapping_ops to struct dev_archdata is added. If the pointer is non NULL, DMA operations in asm/dma-mapping.h use it. If it's NULL, the system-wide dma_ops pointer is used as before. If it's useful for KVM people, I plan to implement a mechanism to register a hook called when a new pci (or dma capable) device is created (it works with hot plugging). It enables IOMMUs to set up an appropriate dma_mapping_ops per device. The major obstacle is that dma_mapping_error doesn't take a pointer to the device unlike other DMA operations. So x86 can't have dma_mapping_ops per device. Note all the POWER IOMMUs use the same dma_mapping_error function so this is not a problem for POWER but x86 IOMMUs use different dma_mapping_error functions. The first patch adds the device argument to dma_mapping_error. The patch is trivial but large since it touches lots of drivers and dma-mapping.h in all the architecture. This patch: dma_mapping_error() doesn't take a pointer to the device unlike other DMA operations. So we can't have dma_mapping_ops per device. Note that POWER already has dma_mapping_ops per device but all the POWER IOMMUs use the same dma_mapping_error function. x86 IOMMUs use device argument. [akpm@linux-foundation.org: fix sge] [akpm@linux-foundation.org: fix svc_rdma] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: fix bnx2x] [akpm@linux-foundation.org: fix s2io] [akpm@linux-foundation.org: fix pasemi_mac] [akpm@linux-foundation.org: fix sdhci] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: fix sparc] [akpm@linux-foundation.org: fix ibmvscsi] Signed-off-by: FUJITA Tomonori Cc: Muli Ben-Yehuda Cc: Andi Kleen Cc: Thomas Gleixner Cc: Ingo Molnar Cc: Avi Kivity Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/kernel/pci-calgary_64.c | 2 +- arch/x86/kernel/pci-dma.c | 27 ++++++++++++++++----------- arch/x86/kernel/pci-gart_64.c | 3 +-- arch/x86/kernel/pci-nommu.c | 14 +------------- arch/x86/kernel/pci-swiotlb_64.c | 2 +- 5 files changed, 20 insertions(+), 28 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 19e7fc7c2c4f..1eb86be93d7a 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -544,7 +544,7 @@ error: return ret; } -static const struct dma_mapping_ops calgary_dma_ops = { +static struct dma_mapping_ops calgary_dma_ops = { .alloc_coherent = calgary_alloc_coherent, .map_single = calgary_map_single, .unmap_single = calgary_unmap_single, diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index cbecb05551bb..37544123896d 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -11,7 +11,7 @@ static int forbid_dac __read_mostly; -const struct dma_mapping_ops *dma_ops; +struct dma_mapping_ops *dma_ops; EXPORT_SYMBOL(dma_ops); static int iommu_sac_force __read_mostly; @@ -312,6 +312,8 @@ static int dma_release_coherent(struct device *dev, int order, void *vaddr) int dma_supported(struct device *dev, u64 mask) { + struct dma_mapping_ops *ops = get_dma_ops(dev); + #ifdef CONFIG_PCI if (mask > 0xffffffff && forbid_dac > 0) { dev_info(dev, "PCI: Disallowing DAC for device\n"); @@ -319,8 +321,8 @@ int dma_supported(struct device *dev, u64 mask) } #endif - if (dma_ops->dma_supported) - return dma_ops->dma_supported(dev, mask); + if (ops->dma_supported) + return ops->dma_supported(dev, mask); /* Copied from i386. Doesn't make much sense, because it will only work for pci_alloc_coherent. @@ -367,6 +369,7 @@ void * dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, gfp_t gfp) { + struct dma_mapping_ops *ops = get_dma_ops(dev); void *memory = NULL; struct page *page; unsigned long dma_mask = 0; @@ -435,8 +438,8 @@ dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, /* Let low level make its own zone decisions */ gfp &= ~(GFP_DMA32|GFP_DMA); - if (dma_ops->alloc_coherent) - return dma_ops->alloc_coherent(dev, size, + if (ops->alloc_coherent) + return ops->alloc_coherent(dev, size, dma_handle, gfp); return NULL; } @@ -448,14 +451,14 @@ dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, } } - if (dma_ops->alloc_coherent) { + if (ops->alloc_coherent) { free_pages((unsigned long)memory, get_order(size)); gfp &= ~(GFP_DMA|GFP_DMA32); - return dma_ops->alloc_coherent(dev, size, dma_handle, gfp); + return ops->alloc_coherent(dev, size, dma_handle, gfp); } - if (dma_ops->map_simple) { - *dma_handle = dma_ops->map_simple(dev, virt_to_phys(memory), + if (ops->map_simple) { + *dma_handle = ops->map_simple(dev, virt_to_phys(memory), size, PCI_DMA_BIDIRECTIONAL); if (*dma_handle != bad_dma_address) @@ -477,12 +480,14 @@ EXPORT_SYMBOL(dma_alloc_coherent); void dma_free_coherent(struct device *dev, size_t size, void *vaddr, dma_addr_t bus) { + struct dma_mapping_ops *ops = get_dma_ops(dev); + int order = get_order(size); WARN_ON(irqs_disabled()); /* for portability */ if (dma_release_coherent(dev, order, vaddr)) return; - if (dma_ops->unmap_single) - dma_ops->unmap_single(dev, bus, size, 0); + if (ops->unmap_single) + ops->unmap_single(dev, bus, size, 0); free_pages((unsigned long)vaddr, order); } EXPORT_SYMBOL(dma_free_coherent); diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index df5f142657d2..744126e64950 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -692,8 +692,7 @@ static __init int init_k8_gatt(struct agp_kern_info *info) extern int agp_amd64_init(void); -static const struct dma_mapping_ops gart_dma_ops = { - .mapping_error = NULL, +static struct dma_mapping_ops gart_dma_ops = { .map_single = gart_map_single, .map_simple = gart_map_simple, .unmap_single = gart_unmap_single, diff --git a/arch/x86/kernel/pci-nommu.c b/arch/x86/kernel/pci-nommu.c index 792b9179eff3..3f91f71cdc3e 100644 --- a/arch/x86/kernel/pci-nommu.c +++ b/arch/x86/kernel/pci-nommu.c @@ -72,21 +72,9 @@ static int nommu_map_sg(struct device *hwdev, struct scatterlist *sg, return nents; } -/* Make sure we keep the same behaviour */ -static int nommu_mapping_error(dma_addr_t dma_addr) -{ -#ifdef CONFIG_X86_32 - return 0; -#else - return (dma_addr == bad_dma_address); -#endif -} - - -const struct dma_mapping_ops nommu_dma_ops = { +struct dma_mapping_ops nommu_dma_ops = { .map_single = nommu_map_single, .map_sg = nommu_map_sg, - .mapping_error = nommu_mapping_error, .is_phys = 1, }; diff --git a/arch/x86/kernel/pci-swiotlb_64.c b/arch/x86/kernel/pci-swiotlb_64.c index 20df839b9c20..c4ce0332759e 100644 --- a/arch/x86/kernel/pci-swiotlb_64.c +++ b/arch/x86/kernel/pci-swiotlb_64.c @@ -18,7 +18,7 @@ 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); } -const struct dma_mapping_ops swiotlb_dma_ops = { +struct dma_mapping_ops swiotlb_dma_ops = { .mapping_error = swiotlb_dma_mapping_error, .alloc_coherent = swiotlb_alloc_coherent, .free_coherent = swiotlb_free_coherent, -- cgit v1.2.2 From 1956a96de488feb05e95c08c9d5e80f63a4be2b1 Mon Sep 17 00:00:00 2001 From: Alexis Bruemmer Date: Fri, 25 Jul 2008 19:44:51 -0700 Subject: x86 calgary: fix handling of devices that aren't behind the Calgary The calgary code can give drivers addresses above 4GB which is very bad for hardware that is only 32bit DMA addressable. With this patch, the calgary code sets the global dma_ops to swiotlb or nommu properly, and the dma_ops of devices behind the Calgary/CalIOC2 to calgary_dma_ops. So the calgary code can handle devices safely that aren't behind the Calgary/CalIOC2. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Alexis Bruemmer Signed-off-by: FUJITA Tomonori Cc: Muli Ben-Yehuda Cc: Ingo Molnar Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/kernel/pci-calgary_64.c | 71 +++++++++++++++------------------------- 1 file changed, 26 insertions(+), 45 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 1eb86be93d7a..b67a4b1d4eae 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -37,6 +37,7 @@ #include #include #include + #include #include #include @@ -413,22 +414,6 @@ static void calgary_unmap_sg(struct device *dev, } } -static int calgary_nontranslate_map_sg(struct device* dev, - struct scatterlist *sg, int nelems, int direction) -{ - struct scatterlist *s; - int i; - - for_each_sg(sg, s, nelems, i) { - struct page *p = sg_page(s); - - BUG_ON(!p); - s->dma_address = virt_to_bus(sg_virt(s)); - s->dma_length = s->length; - } - return nelems; -} - static int calgary_map_sg(struct device *dev, struct scatterlist *sg, int nelems, int direction) { @@ -439,9 +424,6 @@ static int calgary_map_sg(struct device *dev, struct scatterlist *sg, unsigned long entry; int i; - if (!translation_enabled(tbl)) - return calgary_nontranslate_map_sg(dev, sg, nelems, direction); - for_each_sg(sg, s, nelems, i) { BUG_ON(!sg_page(s)); @@ -477,7 +459,6 @@ error: static dma_addr_t calgary_map_single(struct device *dev, phys_addr_t paddr, size_t size, int direction) { - dma_addr_t dma_handle = bad_dma_address; void *vaddr = phys_to_virt(paddr); unsigned long uaddr; unsigned int npages; @@ -486,12 +467,7 @@ static dma_addr_t calgary_map_single(struct device *dev, phys_addr_t paddr, uaddr = (unsigned long)vaddr; npages = num_dma_pages(uaddr, size); - if (translation_enabled(tbl)) - dma_handle = iommu_alloc(dev, tbl, vaddr, npages, direction); - else - dma_handle = virt_to_bus(vaddr); - - return dma_handle; + return iommu_alloc(dev, tbl, vaddr, npages, direction); } static void calgary_unmap_single(struct device *dev, dma_addr_t dma_handle, @@ -500,9 +476,6 @@ static void calgary_unmap_single(struct device *dev, dma_addr_t dma_handle, struct iommu_table *tbl = find_iommu_table(dev); unsigned int npages; - if (!translation_enabled(tbl)) - return; - npages = num_dma_pages(dma_handle, size); iommu_free(tbl, dma_handle, npages); } @@ -525,18 +498,12 @@ static void* calgary_alloc_coherent(struct device *dev, size_t size, goto error; memset(ret, 0, size); - if (translation_enabled(tbl)) { - /* set up tces to cover the allocated range */ - mapping = iommu_alloc(dev, tbl, ret, npages, DMA_BIDIRECTIONAL); - if (mapping == bad_dma_address) - goto free; - - *dma_handle = mapping; - } else /* non translated slot */ - *dma_handle = virt_to_bus(ret); - + /* set up tces to cover the allocated range */ + mapping = iommu_alloc(dev, tbl, ret, npages, DMA_BIDIRECTIONAL); + if (mapping == bad_dma_address) + goto free; + *dma_handle = mapping; return ret; - free: free_pages((unsigned long)ret, get_order(size)); ret = NULL; @@ -1241,6 +1208,16 @@ static int __init calgary_init(void) goto error; } while (1); + dev = NULL; + for_each_pci_dev(dev) { + struct iommu_table *tbl; + + tbl = find_iommu_table(&dev->dev); + + if (translation_enabled(tbl)) + dev->dev.archdata.dma_ops = &calgary_dma_ops; + } + return ret; error: @@ -1262,6 +1239,7 @@ error: calgary_disable_translation(dev); calgary_free_bus(dev); pci_dev_put(dev); /* Undo calgary_init_one()'s pci_dev_get() */ + dev->dev.archdata.dma_ops = NULL; } while (1); return ret; @@ -1503,6 +1481,10 @@ void __init detect_calgary(void) printk(KERN_INFO "PCI-DMA: Calgary TCE table spec is %d, " "CONFIG_IOMMU_DEBUG is %s.\n", specified_table_size, debugging ? "enabled" : "disabled"); + + /* swiotlb for devices that aren't behind the Calgary. */ + if (max_pfn > MAX_DMA32_PFN) + swiotlb = 1; } return; @@ -1519,7 +1501,7 @@ int __init calgary_iommu_init(void) { int ret; - if (no_iommu || swiotlb) + if (no_iommu || (swiotlb && !calgary_detected)) return -ENODEV; if (!calgary_detected) @@ -1532,15 +1514,14 @@ int __init calgary_iommu_init(void) if (ret) { printk(KERN_ERR "PCI-DMA: Calgary init failed %d, " "falling back to no_iommu\n", ret); - if (max_pfn > MAX_DMA32_PFN) - printk(KERN_ERR "WARNING more than 4GB of memory, " - "32bit PCI may malfunction.\n"); return ret; } force_iommu = 1; bad_dma_address = 0x0; - dma_ops = &calgary_dma_ops; + /* dma_ops is set to swiotlb or nommu */ + if (!dma_ops) + dma_ops = &nommu_dma_ops; return 0; } -- cgit v1.2.2 From 3ab83521378268044a448113c6aa9a9e245f4d2f Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Fri, 25 Jul 2008 19:45:07 -0700 Subject: kexec jump This patch provides an enhancement to kexec/kdump. It implements the following features: - Backup/restore memory used by the original kernel before/after kexec. - Save/restore CPU state before/after kexec. The features of this patch can be used as a general method to call program in physical mode (paging turning off). This can be used to call BIOS code under Linux. kexec-tools needs to be patched to support kexec jump. The patches and the precompiled kexec can be download from the following URL: source: http://khibernation.sourceforge.net/download/release_v10/kexec-tools/kexec-tools-src_git_kh10.tar.bz2 patches: http://khibernation.sourceforge.net/download/release_v10/kexec-tools/kexec-tools-patches_git_kh10.tar.bz2 binary: http://khibernation.sourceforge.net/download/release_v10/kexec-tools/kexec_git_kh10 Usage example of calling some physical mode code and return: 1. Compile and install patched kernel with following options selected: CONFIG_X86_32=y CONFIG_KEXEC=y CONFIG_PM=y CONFIG_KEXEC_JUMP=y 2. Build patched kexec-tool or download the pre-built one. 3. Build some physical mode executable named such as "phy_mode" 4. Boot kernel compiled in step 1. 5. Load physical mode executable with /sbin/kexec. The shell command line can be as follow: /sbin/kexec --load-preserve-context --args-none phy_mode 6. Call physical mode executable with following shell command line: /sbin/kexec -e Implementation point: To support jumping without reserving memory. One shadow backup page (source page) is allocated for each page used by kexeced code image (destination page). When do kexec_load, the image of kexeced code is loaded into source pages, and before executing, the destination pages and the source pages are swapped, so the contents of destination pages are backupped. Before jumping to the kexeced code image and after jumping back to the original kernel, the destination pages and the source pages are swapped too. C ABI (calling convention) is used as communication protocol between kernel and called code. A flag named KEXEC_PRESERVE_CONTEXT for sys_kexec_load is added to indicate that the loaded kernel image is used for jumping back. Now, only the i386 architecture is supported. Signed-off-by: Huang Ying Acked-by: Vivek Goyal Cc: "Eric W. Biederman" Cc: Pavel Machek Cc: Nigel Cunningham Cc: "Rafael J. Wysocki" Cc: Ingo Molnar Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/Kconfig | 7 ++ arch/x86/kernel/machine_kexec_32.c | 27 ++++-- arch/x86/kernel/machine_kexec_64.c | 2 +- arch/x86/kernel/relocate_kernel_32.S | 174 ++++++++++++++++++++++++++++++----- 4 files changed, 179 insertions(+), 31 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e3cba0b45600..7ecb679f0130 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1279,6 +1279,13 @@ config CRASH_DUMP (CONFIG_RELOCATABLE=y). For more details see Documentation/kdump/kdump.txt +config KEXEC_JUMP + bool "kexec jump (EXPERIMENTAL)" + depends on EXPERIMENTAL + depends on KEXEC && PM_SLEEP && X86_32 + help + Invoke code in physical address mode via KEXEC + config PHYSICAL_START hex "Physical address where the kernel is loaded" if (EMBEDDED || CRASH_DUMP) default "0x1000000" if X86_NUMAQ diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index 8864230d55af..2b67609d0a1c 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -22,6 +22,7 @@ #include #include #include +#include #define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE))) static u32 kexec_pgd[1024] PAGE_ALIGNED; @@ -85,10 +86,12 @@ static void load_segments(void) * reboot code buffer to allow us to avoid allocations * later. * - * Currently nothing. + * Make control page executable. */ int machine_kexec_prepare(struct kimage *image) { + if (nx_enabled) + set_pages_x(image->control_code_page, 1); return 0; } @@ -98,16 +101,24 @@ int machine_kexec_prepare(struct kimage *image) */ void machine_kexec_cleanup(struct kimage *image) { + if (nx_enabled) + set_pages_nx(image->control_code_page, 1); } /* * Do not allocate memory (or fail in any way) in machine_kexec(). * We are past the point of no return, committed to rebooting now. */ -NORET_TYPE void machine_kexec(struct kimage *image) +void machine_kexec(struct kimage *image) { unsigned long page_list[PAGES_NR]; void *control_page; + asmlinkage unsigned long + (*relocate_kernel_ptr)(unsigned long indirection_page, + unsigned long control_page, + unsigned long start_address, + unsigned int has_pae, + unsigned int preserve_context); tracer_disable(); @@ -115,10 +126,11 @@ NORET_TYPE void machine_kexec(struct kimage *image) local_irq_disable(); control_page = page_address(image->control_code_page); - memcpy(control_page, relocate_kernel, PAGE_SIZE); + memcpy(control_page, relocate_kernel, PAGE_SIZE/2); + relocate_kernel_ptr = control_page; page_list[PA_CONTROL_PAGE] = __pa(control_page); - page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel; + page_list[VA_CONTROL_PAGE] = (unsigned long)control_page; page_list[PA_PGD] = __pa(kexec_pgd); page_list[VA_PGD] = (unsigned long)kexec_pgd; #ifdef CONFIG_X86_PAE @@ -131,6 +143,7 @@ NORET_TYPE 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); /* The segment registers are funny things, they have both a * visible and an invisible part. Whenever the visible part is @@ -149,8 +162,10 @@ NORET_TYPE void machine_kexec(struct kimage *image) set_idt(phys_to_virt(0),0); /* now call it */ - relocate_kernel((unsigned long)image->head, (unsigned long)page_list, - image->start, cpu_has_pae); + image->start = relocate_kernel_ptr((unsigned long)image->head, + (unsigned long)page_list, + image->start, cpu_has_pae, + image->preserve_context); } void arch_crash_save_vmcoreinfo(void) diff --git a/arch/x86/kernel/machine_kexec_64.c b/arch/x86/kernel/machine_kexec_64.c index 9dd9262693a3..c43caa3a91f3 100644 --- a/arch/x86/kernel/machine_kexec_64.c +++ b/arch/x86/kernel/machine_kexec_64.c @@ -181,7 +181,7 @@ void machine_kexec_cleanup(struct kimage *image) * Do not allocate memory (or fail in any way) in machine_kexec(). * We are past the point of no return, committed to rebooting now. */ -NORET_TYPE void machine_kexec(struct kimage *image) +void machine_kexec(struct kimage *image) { unsigned long page_list[PAGES_NR]; void *control_page; diff --git a/arch/x86/kernel/relocate_kernel_32.S b/arch/x86/kernel/relocate_kernel_32.S index c30fe25d470d..703310a99023 100644 --- a/arch/x86/kernel/relocate_kernel_32.S +++ b/arch/x86/kernel/relocate_kernel_32.S @@ -20,11 +20,44 @@ #define PAGE_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) #define PAE_PGD_ATTR (_PAGE_PRESENT) +/* control_page + PAGE_SIZE/2 ~ control_page + PAGE_SIZE * 3/4 are + * used to save some data for jumping back + */ +#define DATA(offset) (PAGE_SIZE/2+(offset)) + +/* Minimal CPU state */ +#define ESP DATA(0x0) +#define CR0 DATA(0x4) +#define CR3 DATA(0x8) +#define CR4 DATA(0xc) + +/* other data */ +#define CP_VA_CONTROL_PAGE DATA(0x10) +#define CP_PA_PGD DATA(0x14) +#define CP_PA_SWAP_PAGE DATA(0x18) +#define CP_PA_BACKUP_PAGES_MAP DATA(0x1c) + .text .align PAGE_SIZE .globl relocate_kernel relocate_kernel: - movl 8(%esp), %ebp /* list of pages */ + /* Save the CPU context, used for jumping back */ + + pushl %ebx + pushl %esi + pushl %edi + pushl %ebp + pushf + + movl 20+8(%esp), %ebp /* list of pages */ + movl PTR(VA_CONTROL_PAGE)(%ebp), %edi + movl %esp, ESP(%edi) + movl %cr0, %eax + movl %eax, CR0(%edi) + movl %cr3, %eax + movl %eax, CR3(%edi) + movl %cr4, %eax + movl %eax, CR4(%edi) #ifdef CONFIG_X86_PAE /* map the control page at its virtual address */ @@ -138,15 +171,25 @@ relocate_kernel: relocate_new_kernel: /* read the arguments and say goodbye to the stack */ - movl 4(%esp), %ebx /* page_list */ - movl 8(%esp), %ebp /* list of pages */ - movl 12(%esp), %edx /* start address */ - movl 16(%esp), %ecx /* cpu_has_pae */ + movl 20+4(%esp), %ebx /* page_list */ + movl 20+8(%esp), %ebp /* list of pages */ + movl 20+12(%esp), %edx /* start address */ + movl 20+16(%esp), %ecx /* cpu_has_pae */ + movl 20+20(%esp), %esi /* preserve_context */ /* zero out flags, and disable interrupts */ pushl $0 popfl + /* save some information for jumping back */ + movl PTR(VA_CONTROL_PAGE)(%ebp), %edi + movl %edi, CP_VA_CONTROL_PAGE(%edi) + movl PTR(PA_PGD)(%ebp), %eax + movl %eax, CP_PA_PGD(%edi) + movl PTR(PA_SWAP_PAGE)(%ebp), %eax + movl %eax, CP_PA_SWAP_PAGE(%edi) + movl %ebx, CP_PA_BACKUP_PAGES_MAP(%edi) + /* get physical address of control page now */ /* this is impossible after page table switch */ movl PTR(PA_CONTROL_PAGE)(%ebp), %edi @@ -197,8 +240,90 @@ identity_mapped: xorl %eax, %eax movl %eax, %cr3 + movl CP_PA_SWAP_PAGE(%edi), %eax + pushl %eax + pushl %ebx + call swap_pages + addl $8, %esp + + /* To be certain of avoiding problems with self-modifying code + * I need to execute a serializing instruction here. + * So I flush the TLB, it's handy, and not processor dependent. + */ + xorl %eax, %eax + movl %eax, %cr3 + + /* set all of the registers to known values */ + /* leave %esp alone */ + + testl %esi, %esi + jnz 1f + xorl %edi, %edi + xorl %eax, %eax + xorl %ebx, %ebx + xorl %ecx, %ecx + xorl %edx, %edx + xorl %esi, %esi + xorl %ebp, %ebp + ret +1: + popl %edx + movl CP_PA_SWAP_PAGE(%edi), %esp + addl $PAGE_SIZE, %esp +2: + call *%edx + + /* get the re-entry point of the peer system */ + movl 0(%esp), %ebp + call 1f +1: + popl %ebx + subl $(1b - relocate_kernel), %ebx + movl CP_VA_CONTROL_PAGE(%ebx), %edi + lea PAGE_SIZE(%ebx), %esp + movl CP_PA_SWAP_PAGE(%ebx), %eax + movl CP_PA_BACKUP_PAGES_MAP(%ebx), %edx + pushl %eax + pushl %edx + call swap_pages + addl $8, %esp + movl CP_PA_PGD(%ebx), %eax + movl %eax, %cr3 + movl %cr0, %eax + orl $(1<<31), %eax + movl %eax, %cr0 + lea PAGE_SIZE(%edi), %esp + movl %edi, %eax + addl $(virtual_mapped - relocate_kernel), %eax + pushl %eax + ret + +virtual_mapped: + movl CR4(%edi), %eax + movl %eax, %cr4 + movl CR3(%edi), %eax + movl %eax, %cr3 + movl CR0(%edi), %eax + movl %eax, %cr0 + movl ESP(%edi), %esp + movl %ebp, %eax + + popf + popl %ebp + popl %edi + popl %esi + popl %ebx + ret + /* Do the copies */ - movl %ebx, %ecx +swap_pages: + movl 8(%esp), %edx + movl 4(%esp), %ecx + pushl %ebp + pushl %ebx + pushl %edi + pushl %esi + movl %ecx, %ebx jmp 1f 0: /* top, read another word from the indirection page */ @@ -226,27 +351,28 @@ identity_mapped: movl %ecx, %esi /* For every source page do a copy */ andl $0xfffff000, %esi + movl %edi, %eax + movl %esi, %ebp + + movl %edx, %edi movl $1024, %ecx rep ; movsl - jmp 0b -3: - - /* To be certain of avoiding problems with self-modifying code - * I need to execute a serializing instruction here. - * So I flush the TLB, it's handy, and not processor dependent. - */ - xorl %eax, %eax - movl %eax, %cr3 + movl %ebp, %edi + movl %eax, %esi + movl $1024, %ecx + rep ; movsl - /* set all of the registers to known values */ - /* leave %esp alone */ + movl %eax, %edi + movl %edx, %esi + movl $1024, %ecx + rep ; movsl - xorl %eax, %eax - xorl %ebx, %ebx - xorl %ecx, %ecx - xorl %edx, %edx - xorl %esi, %esi - xorl %edi, %edi - xorl %ebp, %ebp + lea PAGE_SIZE(%ebp), %esi + jmp 0b +3: + popl %esi + popl %edi + popl %ebx + popl %ebp ret -- cgit v1.2.2 From 89081d17f7bb81d89fa1aa9b70f821c5cf4d39e9 Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Fri, 25 Jul 2008 19:45:10 -0700 Subject: kexec jump: save/restore device state This patch implements devices state save/restore before after kexec. This patch together with features in kexec_jump patch can be used for following: - A simple hibernation implementation without ACPI support. You can kexec a hibernating kernel, save the memory image of original system and shutdown the system. When resuming, you restore the memory image of original system via ordinary kexec load then jump back. - Kernel/system debug through making system snapshot. You can make system snapshot, jump back, do some thing and make another system snapshot. - Cooperative multi-kernel/system. With kexec jump, you can switch between several kernels/systems quickly without boot process except the first time. This appears like swap a whole kernel/system out/in. - A general method to call program in physical mode (paging turning off). This can be used to invoke BIOS code under Linux. The following user-space tools can be used with kexec jump: - kexec-tools needs to be patched to support kexec jump. The patches and the precompiled kexec can be download from the following URL: source: http://khibernation.sourceforge.net/download/release_v10/kexec-tools/kexec-tools-src_git_kh10.tar.bz2 patches: http://khibernation.sourceforge.net/download/release_v10/kexec-tools/kexec-tools-patches_git_kh10.tar.bz2 binary: http://khibernation.sourceforge.net/download/release_v10/kexec-tools/kexec_git_kh10 - makedumpfile with patches are used as memory image saving tool, it can exclude free pages from original kernel memory image file. The patches and the precompiled makedumpfile can be download from the following URL: source: http://khibernation.sourceforge.net/download/release_v10/makedumpfile/makedumpfile-src_cvs_kh10.tar.bz2 patches: http://khibernation.sourceforge.net/download/release_v10/makedumpfile/makedumpfile-patches_cvs_kh10.tar.bz2 binary: http://khibernation.sourceforge.net/download/release_v10/makedumpfile/makedumpfile_cvs_kh10 - An initramfs image can be used as the root file system of kexeced kernel. An initramfs image built with "BuildRoot" can be downloaded from the following URL: initramfs image: http://khibernation.sourceforge.net/download/release_v10/initramfs/rootfs_cvs_kh10.gz All user space tools above are included in the initramfs image. Usage example of simple hibernation: 1. Compile and install patched kernel with following options selected: CONFIG_X86_32=y CONFIG_RELOCATABLE=y CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PM=y CONFIG_HIBERNATION=y CONFIG_KEXEC_JUMP=y 2. Build an initramfs image contains kexec-tool and makedumpfile, or download the pre-built initramfs image, called rootfs.gz in following text. 3. Prepare a partition to save memory image of original kernel, called hibernating partition in following text. 4. Boot kernel compiled in step 1 (kernel A). 5. In the kernel A, load kernel compiled in step 1 (kernel B) with /sbin/kexec. The shell command line can be as follow: /sbin/kexec --load-preserve-context /boot/bzImage --mem-min=0x100000 --mem-max=0xffffff --initrd=rootfs.gz 6. Boot the kernel B with following shell command line: /sbin/kexec -e 7. The kernel B will boot as normal kexec. In kernel B the memory image of kernel A can be saved into hibernating partition as follow: jump_back_entry=`cat /proc/cmdline | tr ' ' '\n' | grep kexec_jump_back_entry | cut -d '='` echo $jump_back_entry > kexec_jump_back_entry cp /proc/vmcore dump.elf Then you can shutdown the machine as normal. 8. Boot kernel compiled in step 1 (kernel C). Use the rootfs.gz as root file system. 9. In kernel C, load the memory image of kernel A as follow: /sbin/kexec -l --args-none --entry=`cat kexec_jump_back_entry` dump.elf 10. Jump back to the kernel A as follow: /sbin/kexec -e Then, kernel A is resumed. Implementation point: To support jumping between two kernels, before jumping to (executing) the new kernel and jumping back to the original kernel, the devices are put into quiescent state, and the state of devices and CPU is saved. After jumping back from kexeced kernel and jumping to the new kernel, the state of devices and CPU are restored accordingly. The devices/CPU state save/restore code of software suspend is called to implement corresponding function. Known issues: - Because the segment number supported by sys_kexec_load is limited, hibernation image with many segments may not be load. This is planned to be eliminated by adding a new flag to sys_kexec_load to make a image can be loaded with multiple sys_kexec_load invoking. Now, only the i386 architecture is supported. Signed-off-by: Huang Ying Acked-by: Vivek Goyal Cc: "Eric W. Biederman" Cc: Pavel Machek Cc: Nigel Cunningham Cc: "Rafael J. Wysocki" Cc: Ingo Molnar Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/Kconfig | 5 +++-- arch/x86/kernel/machine_kexec_32.c | 12 ++++++++++++ 2 files changed, 15 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 7ecb679f0130..6b2debfabddc 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1282,9 +1282,10 @@ config CRASH_DUMP config KEXEC_JUMP bool "kexec jump (EXPERIMENTAL)" depends on EXPERIMENTAL - depends on KEXEC && PM_SLEEP && X86_32 + depends on KEXEC && HIBERNATION && X86_32 help - Invoke code in physical address mode via KEXEC + Jump between original kernel and kexeced kernel and invoke + code in physical address mode via KEXEC config PHYSICAL_START hex "Physical address where the kernel is loaded" if (EMBEDDED || CRASH_DUMP) diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index 2b67609d0a1c..9fe478d98406 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -125,6 +125,18 @@ void machine_kexec(struct kimage *image) /* Interrupts aren't acceptable while we reboot */ local_irq_disable(); + if (image->preserve_context) { +#ifdef CONFIG_X86_IO_APIC + /* We need to put APICs in legacy mode so that we can + * get timer interrupts in second kernel. kexec/kdump + * paths already have calls to disable_IO_APIC() in + * one form or other. kexec jump path also need + * one. + */ + disable_IO_APIC(); +#endif + } + control_page = page_address(image->control_code_page); memcpy(control_page, relocate_kernel, PAGE_SIZE/2); -- cgit v1.2.2 From 8174c430e445a93016ef18f717fe570214fa38bf Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Fri, 25 Jul 2008 19:45:24 -0700 Subject: x86: lockless get_user_pages_fast() Implement get_user_pages_fast without locking in the fastpath on x86. Do an optimistic lockless pagetable walk, without taking mmap_sem or any page table locks or even mmap_sem. Page table existence is guaranteed by turning interrupts off (combined with the fact that we're always looking up the current mm, means we can do the lockless page table walk within the constraints of the TLB shootdown design). Basically we can do this lockless pagetable walk in a similar manner to the way the CPU's pagetable walker does not have to take any locks to find present ptes. This patch (combined with the subsequent ones to convert direct IO to use it) was found to give about 10% performance improvement on a 2 socket 8 core Intel Xeon system running an OLTP workload on DB2 v9.5 "To test the effects of the patch, an OLTP workload was run on an IBM x3850 M2 server with 2 processors (quad-core Intel Xeon processors at 2.93 GHz) using IBM DB2 v9.5 running Linux 2.6.24rc7 kernel. Comparing runs with and without the patch resulted in an overall performance benefit of ~9.8%. Correspondingly, oprofiles showed that samples from __up_read and __down_read routines that is seen during thread contention for system resources was reduced from 2.8% down to .05%. Monitoring the /proc/vmstat output from the patched run showed that the counter for fast_gup contained a very high number while the fast_gup_slow value was zero." (fast_gup is the old name for get_user_pages_fast, fast_gup_slow is a counter we had for the number of times the slowpath was invoked). The main reason for the improvement is that DB2 has multiple threads each issuing direct-IO. Direct-IO uses get_user_pages, and thus the threads contend the mmap_sem cacheline, and can also contend on page table locks. I would anticipate larger performance gains on larger systems, however I think DB2 uses an adaptive mix of threads and processes, so it could be that thread contention remains pretty constant as machine size increases. In which case, we stuck with "only" a 10% gain. The downside of using get_user_pages_fast is that if there is not a pte with the correct permissions for the access, we end up falling back to get_user_pages and so the get_user_pages_fast is a bit of extra work. However this should not be the common case in most performance critical code. [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: Kconfig fix] [akpm@linux-foundation.org: Makefile fix/cleanup] [akpm@linux-foundation.org: warning fix] Signed-off-by: Nick Piggin Cc: Dave Kleikamp Cc: Andy Whitcroft Cc: Ingo Molnar Cc: Thomas Gleixner Cc: Andi Kleen Cc: Dave Kleikamp Cc: Badari Pulavarty Cc: Zach Brown Cc: Jens Axboe Reviewed-by: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/Kconfig | 1 + arch/x86/mm/Makefile | 1 + arch/x86/mm/gup.c | 258 +++++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 260 insertions(+) create mode 100644 arch/x86/mm/gup.c (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 6b2debfabddc..6bdde845818e 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -22,6 +22,7 @@ config X86 select HAVE_IDE select HAVE_OPROFILE select HAVE_IOREMAP_PROT + select HAVE_GET_USER_PAGES_FAST select HAVE_KPROBES select ARCH_WANT_OPTIONAL_GPIOLIB if !X86_RDC321X select HAVE_KRETPROBES diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 1fbb844c3d7a..2977ea37791f 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -1,6 +1,7 @@ obj-y := init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \ pat.o pgtable.o +obj-$(CONFIG_HAVE_GET_USER_PAGES_FAST) += gup.o obj-$(CONFIG_X86_32) += pgtable_32.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o diff --git a/arch/x86/mm/gup.c b/arch/x86/mm/gup.c new file mode 100644 index 000000000000..6f733121f32e --- /dev/null +++ b/arch/x86/mm/gup.c @@ -0,0 +1,258 @@ +/* + * Lockless get_user_pages_fast for x86 + * + * Copyright (C) 2008 Nick Piggin + * Copyright (C) 2008 Novell Inc. + */ +#include +#include +#include +#include + +#include + +static inline pte_t gup_get_pte(pte_t *ptep) +{ +#ifndef CONFIG_X86_PAE + return *ptep; +#else + /* + * With get_user_pages_fast, we walk down the pagetables without taking + * any locks. For this we would like to load the pointers atoimcally, + * but that is not possible (without expensive cmpxchg8b) on PAE. What + * we do have is the guarantee that a pte will only either go from not + * present to present, or present to not present or both -- it will not + * switch to a completely different present page without a TLB flush in + * between; something that we are blocking by holding interrupts off. + * + * Setting ptes from not present to present goes: + * ptep->pte_high = h; + * smp_wmb(); + * ptep->pte_low = l; + * + * And present to not present goes: + * ptep->pte_low = 0; + * smp_wmb(); + * ptep->pte_high = 0; + * + * We must ensure here that the load of pte_low sees l iff pte_high + * sees h. We load pte_high *after* loading pte_low, which ensures we + * don't see an older value of pte_high. *Then* we recheck pte_low, + * which ensures that we haven't picked up a changed pte high. We might + * have got rubbish values from pte_low and pte_high, but we are + * guaranteed that pte_low will not have the present bit set *unless* + * it is 'l'. And get_user_pages_fast only operates on present ptes, so + * we're safe. + * + * gup_get_pte should not be used or copied outside gup.c without being + * very careful -- it does not atomically load the pte or anything that + * is likely to be useful for you. + */ + pte_t pte; + +retry: + pte.pte_low = ptep->pte_low; + smp_rmb(); + pte.pte_high = ptep->pte_high; + smp_rmb(); + if (unlikely(pte.pte_low != ptep->pte_low)) + goto retry; + + return pte; +#endif +} + +/* + * The performance critical leaf functions are made noinline otherwise gcc + * inlines everything into a single function which results in too much + * register pressure. + */ +static noinline int gup_pte_range(pmd_t pmd, unsigned long addr, + unsigned long end, int write, struct page **pages, int *nr) +{ + unsigned long mask; + pte_t *ptep; + + mask = _PAGE_PRESENT|_PAGE_USER; + if (write) + mask |= _PAGE_RW; + + ptep = pte_offset_map(&pmd, addr); + do { + pte_t pte = gup_get_pte(ptep); + struct page *page; + + if ((pte_val(pte) & (mask | _PAGE_SPECIAL)) != mask) { + pte_unmap(ptep); + return 0; + } + VM_BUG_ON(!pfn_valid(pte_pfn(pte))); + page = pte_page(pte); + get_page(page); + pages[*nr] = page; + (*nr)++; + + } while (ptep++, addr += PAGE_SIZE, addr != end); + pte_unmap(ptep - 1); + + return 1; +} + +static inline void get_head_page_multiple(struct page *page, int nr) +{ + VM_BUG_ON(page != compound_head(page)); + VM_BUG_ON(page_count(page) == 0); + atomic_add(nr, &page->_count); +} + +static noinline int gup_huge_pmd(pmd_t pmd, unsigned long addr, + unsigned long end, int write, struct page **pages, int *nr) +{ + unsigned long mask; + pte_t pte = *(pte_t *)&pmd; + struct page *head, *page; + int refs; + + mask = _PAGE_PRESENT|_PAGE_USER; + if (write) + mask |= _PAGE_RW; + if ((pte_val(pte) & mask) != mask) + return 0; + /* hugepages are never "special" */ + VM_BUG_ON(pte_val(pte) & _PAGE_SPECIAL); + VM_BUG_ON(!pfn_valid(pte_pfn(pte))); + + refs = 0; + head = pte_page(pte); + page = head + ((addr & ~HPAGE_MASK) >> PAGE_SHIFT); + do { + VM_BUG_ON(compound_head(page) != head); + pages[*nr] = page; + (*nr)++; + page++; + refs++; + } while (addr += PAGE_SIZE, addr != end); + get_head_page_multiple(head, refs); + + return 1; +} + +static int gup_pmd_range(pud_t pud, unsigned long addr, unsigned long end, + int write, struct page **pages, int *nr) +{ + unsigned long next; + pmd_t *pmdp; + + pmdp = pmd_offset(&pud, addr); + do { + pmd_t pmd = *pmdp; + + next = pmd_addr_end(addr, end); + if (pmd_none(pmd)) + return 0; + if (unlikely(pmd_large(pmd))) { + if (!gup_huge_pmd(pmd, addr, next, write, pages, nr)) + return 0; + } else { + if (!gup_pte_range(pmd, addr, next, write, pages, nr)) + return 0; + } + } while (pmdp++, addr = next, addr != end); + + return 1; +} + +static int gup_pud_range(pgd_t pgd, unsigned long addr, unsigned long end, + int write, struct page **pages, int *nr) +{ + unsigned long next; + pud_t *pudp; + + pudp = pud_offset(&pgd, addr); + do { + pud_t pud = *pudp; + + next = pud_addr_end(addr, end); + if (pud_none(pud)) + return 0; + if (!gup_pmd_range(pud, addr, next, write, pages, nr)) + return 0; + } while (pudp++, addr = next, addr != end); + + return 1; +} + +int get_user_pages_fast(unsigned long start, int nr_pages, int write, + struct page **pages) +{ + struct mm_struct *mm = current->mm; + unsigned long end = start + (nr_pages << PAGE_SHIFT); + unsigned long addr = start; + unsigned long next; + pgd_t *pgdp; + int nr = 0; + + if (unlikely(!access_ok(write ? VERIFY_WRITE : VERIFY_READ, + start, nr_pages*PAGE_SIZE))) + goto slow_irqon; + + /* + * XXX: batch / limit 'nr', to avoid large irq off latency + * needs some instrumenting to determine the common sizes used by + * important workloads (eg. DB2), and whether limiting the batch size + * will decrease performance. + * + * It seems like we're in the clear for the moment. Direct-IO is + * the main guy that batches up lots of get_user_pages, and even + * they are limited to 64-at-a-time which is not so many. + */ + /* + * This doesn't prevent pagetable teardown, but does prevent + * the pagetables and pages from being freed on x86. + * + * So long as we atomically load page table pointers versus teardown + * (which we do on x86, with the above PAE exception), we can follow the + * address down to the the page and take a ref on it. + */ + local_irq_disable(); + pgdp = pgd_offset(mm, addr); + do { + pgd_t pgd = *pgdp; + + next = pgd_addr_end(addr, end); + if (pgd_none(pgd)) + goto slow; + if (!gup_pud_range(pgd, addr, next, write, pages, &nr)) + goto slow; + } while (pgdp++, addr = next, addr != end); + local_irq_enable(); + + VM_BUG_ON(nr != (end - start) >> PAGE_SHIFT); + return nr; + + { + int ret; + +slow: + local_irq_enable(); +slow_irqon: + /* Try to get the remaining pages with get_user_pages */ + start += nr << PAGE_SHIFT; + pages += nr; + + down_read(&mm->mmap_sem); + ret = get_user_pages(current, mm, start, + (end - start) >> PAGE_SHIFT, write, 0, pages, NULL); + up_read(&mm->mmap_sem); + + /* Have to be a bit careful with return values */ + if (nr > 0) { + if (ret < 0) + ret = nr; + else + ret += nr; + } + + return ret; + } +} -- cgit v1.2.2 From 652ea695364142b2464744746beac206d050ef19 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Fri, 25 Jul 2008 19:45:27 -0700 Subject: x86: support 1GB hugepages with get_user_pages_lockless() Signed-off-by: Nick Piggin Cc: Ingo Molnar Cc: Thomas Gleixner Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/mm/gup.c | 43 ++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 40 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/gup.c b/arch/x86/mm/gup.c index 6f733121f32e..3085f25b4355 100644 --- a/arch/x86/mm/gup.c +++ b/arch/x86/mm/gup.c @@ -124,7 +124,7 @@ static noinline int gup_huge_pmd(pmd_t pmd, unsigned long addr, refs = 0; head = pte_page(pte); - page = head + ((addr & ~HPAGE_MASK) >> PAGE_SHIFT); + page = head + ((addr & ~PMD_MASK) >> PAGE_SHIFT); do { VM_BUG_ON(compound_head(page) != head); pages[*nr] = page; @@ -162,6 +162,38 @@ static int gup_pmd_range(pud_t pud, unsigned long addr, unsigned long end, return 1; } +static noinline int gup_huge_pud(pud_t pud, unsigned long addr, + unsigned long end, int write, struct page **pages, int *nr) +{ + unsigned long mask; + pte_t pte = *(pte_t *)&pud; + struct page *head, *page; + int refs; + + mask = _PAGE_PRESENT|_PAGE_USER; + if (write) + mask |= _PAGE_RW; + if ((pte_val(pte) & mask) != mask) + return 0; + /* hugepages are never "special" */ + VM_BUG_ON(pte_val(pte) & _PAGE_SPECIAL); + VM_BUG_ON(!pfn_valid(pte_pfn(pte))); + + refs = 0; + head = pte_page(pte); + page = head + ((addr & ~PUD_MASK) >> PAGE_SHIFT); + do { + VM_BUG_ON(compound_head(page) != head); + pages[*nr] = page; + (*nr)++; + page++; + refs++; + } while (addr += PAGE_SIZE, addr != end); + get_head_page_multiple(head, refs); + + return 1; +} + static int gup_pud_range(pgd_t pgd, unsigned long addr, unsigned long end, int write, struct page **pages, int *nr) { @@ -175,8 +207,13 @@ static int gup_pud_range(pgd_t pgd, unsigned long addr, unsigned long end, next = pud_addr_end(addr, end); if (pud_none(pud)) return 0; - if (!gup_pmd_range(pud, addr, next, write, pages, nr)) - return 0; + if (unlikely(pud_large(pud))) { + if (!gup_huge_pud(pud, addr, next, write, pages, nr)) + return 0; + } else { + if (!gup_pmd_range(pud, addr, next, write, pages, nr)) + return 0; + } } while (pudp++, addr = next, addr != end); return 1; -- cgit v1.2.2 From 6341c393fcc37d58727865f1ee2f65e632e9d4f0 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Fri, 25 Jul 2008 19:45:44 -0700 Subject: tracehook: exec This moves all the ptrace hooks related to exec into tracehook.h inlines. This also lifts the calls for tracing out of the binfmt load_binary hooks into search_binary_handler() after it calls into the binfmt module. This change has no effect, since all the binfmt modules' load_binary functions did the call at the end on success, and now search_binary_handler() does it immediately after return if successful. We consolidate the repeated code, and binfmt modules no longer need to import ptrace_notify(). Signed-off-by: Roland McGrath Cc: Oleg Nesterov Reviewed-by: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/ia32/ia32_aout.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c index 58cccb6483b0..a0e1dbe67dc1 100644 --- a/arch/x86/ia32/ia32_aout.c +++ b/arch/x86/ia32/ia32_aout.c @@ -441,12 +441,6 @@ beyond_if: regs->r8 = regs->r9 = regs->r10 = regs->r11 = regs->r12 = regs->r13 = regs->r14 = regs->r15 = 0; set_fs(USER_DS); - if (unlikely(current->ptrace & PT_PTRACED)) { - if (current->ptrace & PT_TRACE_EXEC) - ptrace_notify((PTRACE_EVENT_EXEC << 8) | SIGTRAP); - else - send_sig(SIGTRAP, current, 0); - } return 0; } -- cgit v1.2.2 From 8dad322f5449010c14990dd6934878f576b2ee60 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Fri, 25 Jul 2008 19:46:11 -0700 Subject: x86: use generic show_mem() Remove arch-specific show_mem() in favor of the generic version. This also removes the following redundant information display: - pages in swapcache, printed by show_swap_cache_info() - dirty pages, writeback pages, mapped pages, slab pages, pagetable pages, printed by show_free_areas() where show_mem() calls show_free_areas(), which calls show_swap_cache_info(). Signed-off-by: Johannes Weiner Acked-by: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/mm/init_64.c | 37 ------------------------------------- arch/x86/mm/pgtable_32.c | 47 ----------------------------------------------- 2 files changed, 84 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index ec37121f6709..129618ca0ea2 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -86,43 +86,6 @@ early_param("gbpages", parse_direct_gbpages_on); * around without checking the pgd every time. */ -void show_mem(void) -{ - long i, total = 0, reserved = 0; - long shared = 0, cached = 0; - struct page *page; - pg_data_t *pgdat; - - printk(KERN_INFO "Mem-info:\n"); - show_free_areas(); - for_each_online_pgdat(pgdat) { - for (i = 0; i < pgdat->node_spanned_pages; ++i) { - /* - * This loop can take a while with 256 GB and - * 4k pages so defer the NMI watchdog: - */ - if (unlikely(i % MAX_ORDER_NR_PAGES == 0)) - touch_nmi_watchdog(); - - if (!pfn_valid(pgdat->node_start_pfn + i)) - continue; - - page = pfn_to_page(pgdat->node_start_pfn + i); - total++; - if (PageReserved(page)) - reserved++; - else if (PageSwapCache(page)) - cached++; - else if (page_count(page)) - shared += page_count(page) - 1; - } - } - printk(KERN_INFO "%lu pages of RAM\n", total); - printk(KERN_INFO "%lu reserved pages\n", reserved); - printk(KERN_INFO "%lu pages shared\n", shared); - printk(KERN_INFO "%lu pages swap cached\n", cached); -} - int after_bootmem; static __init void *spp_getpage(void) diff --git a/arch/x86/mm/pgtable_32.c b/arch/x86/mm/pgtable_32.c index b4becbf8c570..cab0abbd1ebe 100644 --- a/arch/x86/mm/pgtable_32.c +++ b/arch/x86/mm/pgtable_32.c @@ -20,53 +20,6 @@ #include #include -void show_mem(void) -{ - int total = 0, reserved = 0; - int shared = 0, cached = 0; - int highmem = 0; - struct page *page; - pg_data_t *pgdat; - unsigned long i; - unsigned long flags; - - printk(KERN_INFO "Mem-info:\n"); - show_free_areas(); - for_each_online_pgdat(pgdat) { - pgdat_resize_lock(pgdat, &flags); - for (i = 0; i < pgdat->node_spanned_pages; ++i) { - if (unlikely(i % MAX_ORDER_NR_PAGES == 0)) - touch_nmi_watchdog(); - page = pgdat_page_nr(pgdat, i); - total++; - if (PageHighMem(page)) - highmem++; - if (PageReserved(page)) - reserved++; - else if (PageSwapCache(page)) - cached++; - else if (page_count(page)) - shared += page_count(page) - 1; - } - pgdat_resize_unlock(pgdat, &flags); - } - printk(KERN_INFO "%d pages of RAM\n", total); - printk(KERN_INFO "%d pages of HIGHMEM\n", highmem); - printk(KERN_INFO "%d reserved pages\n", reserved); - printk(KERN_INFO "%d pages shared\n", shared); - printk(KERN_INFO "%d pages swap cached\n", cached); - - printk(KERN_INFO "%lu pages dirty\n", global_page_state(NR_FILE_DIRTY)); - printk(KERN_INFO "%lu pages writeback\n", - global_page_state(NR_WRITEBACK)); - printk(KERN_INFO "%lu pages mapped\n", global_page_state(NR_FILE_MAPPED)); - printk(KERN_INFO "%lu pages slab\n", - global_page_state(NR_SLAB_RECLAIMABLE) + - global_page_state(NR_SLAB_UNRECLAIMABLE)); - printk(KERN_INFO "%lu pages pagetables\n", - global_page_state(NR_PAGETABLE)); -} - /* * Associate a virtual page frame with a given physical page frame * and protection flags for that frame. -- cgit v1.2.2 From 5f4cb662a0a2533b45656607471571460310a5ca Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 14 Jul 2008 20:36:36 +0200 Subject: KVM: SVM: allow enabling/disabling NPT by reloading only the architecture module If NPT is enabled after loading both KVM modules on AMD and it should be disabled, both KVM modules must be reloaded. If only the architecture module is reloaded the behavior is undefined. With this patch it is possible to disable NPT only by reloading the kvm_amd module. Signed-off-by: Joerg Roedel Signed-off-by: Avi Kivity --- arch/x86/kvm/mmu.c | 6 ++++++ arch/x86/kvm/svm.c | 3 ++- 2 files changed, 8 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index b0e4ddca6c18..d087d9c4f2d9 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -1870,6 +1870,12 @@ void kvm_enable_tdp(void) } EXPORT_SYMBOL_GPL(kvm_enable_tdp); +void kvm_disable_tdp(void) +{ + tdp_enabled = false; +} +EXPORT_SYMBOL_GPL(kvm_disable_tdp); + static void free_mmu_pages(struct kvm_vcpu *vcpu) { struct kvm_mmu_page *sp; diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index b756e876dce3..951b789cc913 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -453,7 +453,8 @@ static __init int svm_hardware_setup(void) if (npt_enabled) { printk(KERN_INFO "kvm: Nested Paging enabled\n"); kvm_enable_tdp(); - } + } else + kvm_disable_tdp(); return 0; -- cgit v1.2.2 From 98899aa0e0bf5de05850082be0eb837058c09ea5 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 16 Jul 2008 19:07:10 -0300 Subject: KVM: task switch: segment base is linear address The segment base is always a linear address, so translate before accessing guest memory. Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 9f1cdb011cff..cd687395e4e7 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3223,6 +3223,7 @@ static void get_segment_descritptor_dtable(struct kvm_vcpu *vcpu, static int load_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector, struct desc_struct *seg_desc) { + gpa_t gpa; struct descriptor_table dtable; u16 index = selector >> 3; @@ -3232,13 +3233,16 @@ static int load_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector, kvm_queue_exception_e(vcpu, GP_VECTOR, selector & 0xfffc); return 1; } - return kvm_read_guest(vcpu->kvm, dtable.base + index * 8, seg_desc, 8); + gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, dtable.base); + gpa += index * 8; + return kvm_read_guest(vcpu->kvm, gpa, seg_desc, 8); } /* allowed just for 8 bytes segments */ static int save_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector, struct desc_struct *seg_desc) { + gpa_t gpa; struct descriptor_table dtable; u16 index = selector >> 3; @@ -3246,7 +3250,9 @@ static int save_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector, if (dtable.limit < index * 8 + 7) return 1; - return kvm_write_guest(vcpu->kvm, dtable.base + index * 8, seg_desc, 8); + gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, dtable.base); + gpa += index * 8; + return kvm_write_guest(vcpu->kvm, gpa, seg_desc, 8); } static u32 get_tss_base_addr(struct kvm_vcpu *vcpu, @@ -3258,7 +3264,7 @@ static u32 get_tss_base_addr(struct kvm_vcpu *vcpu, base_addr |= (seg_desc->base1 << 16); base_addr |= (seg_desc->base2 << 24); - return base_addr; + return vcpu->arch.mmu.gva_to_gpa(vcpu, base_addr); } static int load_tss_segment32(struct kvm_vcpu *vcpu, -- cgit v1.2.2 From 34198bf8426276a2ce1e97056a0f02d43637e5ae Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 16 Jul 2008 19:07:11 -0300 Subject: KVM: task switch: use seg regs provided by subarch instead of reading from GDT There is no guarantee that the old TSS descriptor in the GDT contains the proper base address. This is the case for Windows installation's reboot-via-triplefault. Use guest registers instead. Also translate the address properly. Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 93 ++++++++++++++++++------------------------------------ 1 file changed, 30 insertions(+), 63 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index cd687395e4e7..27c6ece91da6 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3267,54 +3267,6 @@ static u32 get_tss_base_addr(struct kvm_vcpu *vcpu, return vcpu->arch.mmu.gva_to_gpa(vcpu, base_addr); } -static int load_tss_segment32(struct kvm_vcpu *vcpu, - struct desc_struct *seg_desc, - struct tss_segment_32 *tss) -{ - u32 base_addr; - - base_addr = get_tss_base_addr(vcpu, seg_desc); - - return kvm_read_guest(vcpu->kvm, base_addr, tss, - sizeof(struct tss_segment_32)); -} - -static int save_tss_segment32(struct kvm_vcpu *vcpu, - struct desc_struct *seg_desc, - struct tss_segment_32 *tss) -{ - u32 base_addr; - - base_addr = get_tss_base_addr(vcpu, seg_desc); - - return kvm_write_guest(vcpu->kvm, base_addr, tss, - sizeof(struct tss_segment_32)); -} - -static int load_tss_segment16(struct kvm_vcpu *vcpu, - struct desc_struct *seg_desc, - struct tss_segment_16 *tss) -{ - u32 base_addr; - - base_addr = get_tss_base_addr(vcpu, seg_desc); - - return kvm_read_guest(vcpu->kvm, base_addr, tss, - sizeof(struct tss_segment_16)); -} - -static int save_tss_segment16(struct kvm_vcpu *vcpu, - struct desc_struct *seg_desc, - struct tss_segment_16 *tss) -{ - u32 base_addr; - - base_addr = get_tss_base_addr(vcpu, seg_desc); - - return kvm_write_guest(vcpu->kvm, base_addr, tss, - sizeof(struct tss_segment_16)); -} - static u16 get_segment_selector(struct kvm_vcpu *vcpu, int seg) { struct kvm_segment kvm_seg; @@ -3472,20 +3424,26 @@ static int load_state_from_tss16(struct kvm_vcpu *vcpu, } static int kvm_task_switch_16(struct kvm_vcpu *vcpu, u16 tss_selector, - struct desc_struct *cseg_desc, + u32 old_tss_base, struct desc_struct *nseg_desc) { struct tss_segment_16 tss_segment_16; int ret = 0; - if (load_tss_segment16(vcpu, cseg_desc, &tss_segment_16)) + if (kvm_read_guest(vcpu->kvm, old_tss_base, &tss_segment_16, + sizeof tss_segment_16)) goto out; save_state_to_tss16(vcpu, &tss_segment_16); - save_tss_segment16(vcpu, cseg_desc, &tss_segment_16); - if (load_tss_segment16(vcpu, nseg_desc, &tss_segment_16)) + if (kvm_write_guest(vcpu->kvm, old_tss_base, &tss_segment_16, + sizeof tss_segment_16)) goto out; + + if (kvm_read_guest(vcpu->kvm, get_tss_base_addr(vcpu, nseg_desc), + &tss_segment_16, sizeof tss_segment_16)) + goto out; + if (load_state_from_tss16(vcpu, &tss_segment_16)) goto out; @@ -3495,20 +3453,26 @@ out: } static int kvm_task_switch_32(struct kvm_vcpu *vcpu, u16 tss_selector, - struct desc_struct *cseg_desc, + u32 old_tss_base, struct desc_struct *nseg_desc) { struct tss_segment_32 tss_segment_32; int ret = 0; - if (load_tss_segment32(vcpu, cseg_desc, &tss_segment_32)) + if (kvm_read_guest(vcpu->kvm, old_tss_base, &tss_segment_32, + sizeof tss_segment_32)) goto out; save_state_to_tss32(vcpu, &tss_segment_32); - save_tss_segment32(vcpu, cseg_desc, &tss_segment_32); - if (load_tss_segment32(vcpu, nseg_desc, &tss_segment_32)) + if (kvm_write_guest(vcpu->kvm, old_tss_base, &tss_segment_32, + sizeof tss_segment_32)) + goto out; + + if (kvm_read_guest(vcpu->kvm, get_tss_base_addr(vcpu, nseg_desc), + &tss_segment_32, sizeof tss_segment_32)) goto out; + if (load_state_from_tss32(vcpu, &tss_segment_32)) goto out; @@ -3523,16 +3487,20 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason) struct desc_struct cseg_desc; struct desc_struct nseg_desc; int ret = 0; + u32 old_tss_base = get_segment_base(vcpu, VCPU_SREG_TR); + u16 old_tss_sel = get_segment_selector(vcpu, VCPU_SREG_TR); - kvm_get_segment(vcpu, &tr_seg, VCPU_SREG_TR); + old_tss_base = vcpu->arch.mmu.gva_to_gpa(vcpu, old_tss_base); + /* FIXME: Handle errors. Failure to read either TSS or their + * descriptors should generate a pagefault. + */ if (load_guest_segment_descriptor(vcpu, tss_selector, &nseg_desc)) goto out; - if (load_guest_segment_descriptor(vcpu, tr_seg.selector, &cseg_desc)) + if (load_guest_segment_descriptor(vcpu, old_tss_sel, &cseg_desc)) goto out; - if (reason != TASK_SWITCH_IRET) { int cpl; @@ -3550,8 +3518,7 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason) if (reason == TASK_SWITCH_IRET || reason == TASK_SWITCH_JMP) { cseg_desc.type &= ~(1 << 1); //clear the B flag - save_guest_segment_descriptor(vcpu, tr_seg.selector, - &cseg_desc); + save_guest_segment_descriptor(vcpu, old_tss_sel, &cseg_desc); } if (reason == TASK_SWITCH_IRET) { @@ -3563,10 +3530,10 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason) kvm_x86_ops->cache_regs(vcpu); if (nseg_desc.type & 8) - ret = kvm_task_switch_32(vcpu, tss_selector, &cseg_desc, + ret = kvm_task_switch_32(vcpu, tss_selector, old_tss_base, &nseg_desc); else - ret = kvm_task_switch_16(vcpu, tss_selector, &cseg_desc, + ret = kvm_task_switch_16(vcpu, tss_selector, old_tss_base, &nseg_desc); if (reason == TASK_SWITCH_CALL || reason == TASK_SWITCH_GATE) { -- cgit v1.2.2 From 577bdc496614ced56d999bbb425e85adf2386490 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Sat, 19 Jul 2008 08:57:05 +0300 Subject: KVM: Avoid instruction emulation when event delivery is pending When an event (such as an interrupt) is injected, and the stack is shadowed (and therefore write protected), the guest will exit. The current code will see that the stack is shadowed and emulate a few instructions, each time postponing the injection. Eventually the injection may succeed, but at that time the guest may be unwilling to accept the interrupt (for example, the TPR may have changed). This occurs every once in a while during a Windows 2008 boot. Fix by unshadowing the fault address if the fault was due to an event injection. Signed-off-by: Avi Kivity --- arch/x86/kvm/mmu.c | 1 + arch/x86/kvm/svm.c | 7 ++++++- arch/x86/kvm/vmx.c | 2 ++ 3 files changed, 9 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index d087d9c4f2d9..2fa231923cf7 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -1814,6 +1814,7 @@ int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva) spin_unlock(&vcpu->kvm->mmu_lock); return r; } +EXPORT_SYMBOL_GPL(kvm_mmu_unprotect_page_virt); void __kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu) { diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 951b789cc913..e2ee264740c7 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -1008,10 +1008,13 @@ static int pf_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run) struct kvm *kvm = svm->vcpu.kvm; u64 fault_address; u32 error_code; + bool event_injection = false; if (!irqchip_in_kernel(kvm) && - is_external_interrupt(exit_int_info)) + is_external_interrupt(exit_int_info)) { + event_injection = true; push_irq(&svm->vcpu, exit_int_info & SVM_EVTINJ_VEC_MASK); + } fault_address = svm->vmcb->control.exit_info_2; error_code = svm->vmcb->control.exit_info_1; @@ -1025,6 +1028,8 @@ static int pf_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run) (u32)fault_address, (u32)(fault_address >> 32), handler); + if (event_injection) + kvm_mmu_unprotect_page_virt(&svm->vcpu, fault_address); return kvm_mmu_page_fault(&svm->vcpu, fault_address, error_code); } diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index 0cac63701719..b918fc83435c 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -2298,6 +2298,8 @@ static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) cr2 = vmcs_readl(EXIT_QUALIFICATION); KVMTRACE_3D(PAGE_FAULT, vcpu, error_code, (u32)cr2, (u32)((u64)cr2 >> 32), handler); + if (vect_info & VECTORING_INFO_VALID_MASK) + kvm_mmu_unprotect_page_virt(vcpu, cr2); return kvm_mmu_page_fault(vcpu, cr2, error_code); } -- cgit v1.2.2 From c93cd3a58845012df2d658fecd0ac99f7008d753 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Sat, 19 Jul 2008 19:08:07 -0300 Subject: KVM: task switch: translate guest segment limit to virt-extension byte granular field If 'g' is one then limit is 4kb granular. Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 27c6ece91da6..5916191420c7 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3184,6 +3184,10 @@ static void seg_desct_to_kvm_desct(struct desc_struct *seg_desc, u16 selector, kvm_desct->base |= seg_desc->base2 << 24; kvm_desct->limit = seg_desc->limit0; kvm_desct->limit |= seg_desc->limit << 16; + if (seg_desc->g) { + kvm_desct->limit <<= 12; + kvm_desct->limit |= 0xfff; + } kvm_desct->selector = selector; kvm_desct->type = seg_desc->type; kvm_desct->present = seg_desc->p; -- cgit v1.2.2 From 5ec5726a16245138f5d5305b00a752acb5730076 Mon Sep 17 00:00:00 2001 From: Sheng Yang Date: Wed, 16 Jul 2008 09:21:22 +0800 Subject: KVM: VMX: Fix bypass_guest_pf enabling when disable EPT in module parameter Signed-off-by: Sheng Yang Signed-off-by: Avi Kivity --- arch/x86/kvm/vmx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index b918fc83435c..f71151d999e4 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -3305,7 +3305,7 @@ static int __init vmx_init(void) vmx_disable_intercept_for_msr(vmx_msr_bitmap, MSR_IA32_SYSENTER_ESP); vmx_disable_intercept_for_msr(vmx_msr_bitmap, MSR_IA32_SYSENTER_EIP); - if (cpu_has_vmx_ept()) + if (vm_need_ept()) bypass_guest_pf = 0; if (bypass_guest_pf) -- cgit v1.2.2 From 5fdbcb9dd16f1e89ead127d3ee1a38e3a00cf1ea Mon Sep 17 00:00:00 2001 From: Sheng Yang Date: Wed, 16 Jul 2008 09:25:40 +0800 Subject: KVM: VMX: Fix undefined beaviour of EPT after reload kvm-intel.ko As well as move set base/mask ptes to vmx_init(). Signed-off-by: Sheng Yang Signed-off-by: Avi Kivity --- arch/x86/kvm/vmx.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index f71151d999e4..2a69773e3b26 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -3118,15 +3118,6 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id) return ERR_PTR(-ENOMEM); allocate_vpid(vmx); - if (id == 0 && vm_need_ept()) { - kvm_mmu_set_base_ptes(VMX_EPT_READABLE_MASK | - VMX_EPT_WRITABLE_MASK | - VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT); - kvm_mmu_set_mask_ptes(0ull, VMX_EPT_FAKE_ACCESSED_MASK, - VMX_EPT_FAKE_DIRTY_MASK, 0ull, - VMX_EPT_EXECUTABLE_MASK); - kvm_enable_tdp(); - } err = kvm_vcpu_init(&vmx->vcpu, kvm, id); if (err) @@ -3305,8 +3296,17 @@ static int __init vmx_init(void) vmx_disable_intercept_for_msr(vmx_msr_bitmap, MSR_IA32_SYSENTER_ESP); vmx_disable_intercept_for_msr(vmx_msr_bitmap, MSR_IA32_SYSENTER_EIP); - if (vm_need_ept()) + if (vm_need_ept()) { bypass_guest_pf = 0; + kvm_mmu_set_base_ptes(VMX_EPT_READABLE_MASK | + VMX_EPT_WRITABLE_MASK | + VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT); + kvm_mmu_set_mask_ptes(0ull, VMX_EPT_FAKE_ACCESSED_MASK, + VMX_EPT_FAKE_DIRTY_MASK, 0ull, + VMX_EPT_EXECUTABLE_MASK); + kvm_enable_tdp(); + } else + kvm_disable_tdp(); if (bypass_guest_pf) kvm_mmu_set_nonpresent_ptes(~0xffeull, 0ull); -- cgit v1.2.2 From 583323b9d2f624884a8c9563fb5a4d795f39ab07 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 27 Jul 2008 21:43:11 +0200 Subject: x86: fix cpu hotplug on 32bit commit 3e9704739daf46a8ba6593d749c67b5f7cd633d2 ("x86: boot secondary cpus through initial_code") causes the kernel to crash when a CPU is brought online after the read only sections have been write protected. The write to initial_code in do_boot_cpu() fails. Move inital_code to .cpuinit.data section. Signed-off-by: Thomas Gleixner Acked-by: H. Peter Anvin --- arch/x86/kernel/head_32.S | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index f67e93441caf..a7010c3a377a 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -456,9 +456,6 @@ is386: movl $2,%ecx # set MP 1: #endif /* CONFIG_SMP */ jmp *(initial_code) -.align 4 -ENTRY(initial_code) - .long i386_start_kernel /* * We depend on ET to be correct. This checks for 287/387. @@ -601,6 +598,11 @@ ignore_int: #endif iret +.section .cpuinit.data,"wa" +.align 4 +ENTRY(initial_code) + .long i386_start_kernel + .section .text /* * Real beginning of normal "text" segment -- cgit v1.2.2 From e56b3bc7942982ac2589c942fb345e38bc7a341a Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 28 Jul 2008 11:32:33 -0700 Subject: cpu masks: optimize and clean up cpumask_of_cpu() Clean up and optimize cpumask_of_cpu(), by sharing all the zero words. Instead of stupidly generating all possible i=0...NR_CPUS 2^i patterns creating a huge array of constant bitmasks, realize that the zero words can be shared. In other words, on a 64-bit architecture, we only ever need 64 of these arrays - with a different bit set in one single world (with enough zero words around it so that we can create any bitmask by just offsetting in that big array). And then we just put enough zeroes around it that we can point every single cpumask to be one of those things. So when we have 4k CPU's, instead of having 4k arrays (of 4k bits each, with one bit set in each array - 2MB memory total), we have exactly 64 arrays instead, each 8k bits in size (64kB total). And then we just point cpumask(n) to the right position (which we can calculate dynamically). Once we have the right arrays, getting "cpumask(n)" ends up being: static inline const cpumask_t *get_cpu_mask(unsigned int cpu) { const unsigned long *p = cpu_bit_bitmap[1 + cpu % BITS_PER_LONG]; p -= cpu / BITS_PER_LONG; return (const cpumask_t *)p; } This brings other advantages and simplifications as well: - we are not wasting memory that is just filled with a single bit in various different places - we don't need all those games to re-create the arrays in some dense format, because they're already going to be dense enough. if we compile a kernel for up to 4k CPU's, "wasting" that 64kB of memory is a non-issue (especially since by doing this "overlapping" trick we probably get better cache behaviour anyway). [ mingo@elte.hu: Converted Linus's mails into a commit. See: http://lkml.org/lkml/2008/7/27/156 http://lkml.org/lkml/2008/7/28/320 Also applied a family filter - which also has the side-effect of leaving out the bits where Linus calls me an idio... Oh, never mind ;-) ] Signed-off-by: Ingo Molnar Cc: Rusty Russell Cc: Andrew Morton Cc: Al Viro Cc: Mike Travis Signed-off-by: Ingo Molnar --- arch/x86/kernel/setup_percpu.c | 23 ----------------------- 1 file changed, 23 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 1cd53dfcd309..76e305e064f9 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -80,26 +80,6 @@ static void __init setup_per_cpu_maps(void) #endif } -#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP -/* - * Replace static cpumask_of_cpu_map in the initdata section, - * with one that's allocated sized by the possible number of cpus. - * - * (requires nr_cpu_ids to be initialized) - */ -static void __init setup_cpumask_of_cpu(void) -{ - int i; - - /* alloc_bootmem zeroes memory */ - cpumask_of_cpu_map = alloc_bootmem_low(sizeof(cpumask_t) * nr_cpu_ids); - for (i = 0; i < nr_cpu_ids; i++) - cpu_set(i, cpumask_of_cpu_map[i]); -} -#else -static inline void setup_cpumask_of_cpu(void) { } -#endif - #ifdef CONFIG_X86_32 /* * Great future not-so-futuristic plan: make i386 and x86_64 do it @@ -199,9 +179,6 @@ void __init setup_per_cpu_areas(void) /* Setup node to cpumask map */ setup_node_to_cpumask_map(); - - /* Setup cpumask_of_cpu map */ - setup_cpumask_of_cpu(); } #endif -- cgit v1.2.2 From 12c0b20fa4afb5c8a377d6987fb2dcf353e1dce1 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Wed, 23 Jul 2008 17:00:13 -0600 Subject: x86/PCI: use dev_printk when possible Convert printks to use dev_printk(). I converted DBG() to dev_dbg(). This DBG() is from arch/x86/pci/pci.h and requires source-code modification to enable, so dev_dbg() seems roughly equivalent. Signed-off-by: Bjorn Helgaas Signed-off-by: Jesse Barnes --- arch/x86/pci/fixup.c | 3 +- arch/x86/pci/i386.c | 26 +++++------- arch/x86/pci/irq.c | 106 ++++++++++++++++++++++-------------------------- arch/x86/pci/numaq_32.c | 5 ++- 4 files changed, 64 insertions(+), 76 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/pci/fixup.c b/arch/x86/pci/fixup.c index ff3a6a336342..4bdaa590375d 100644 --- a/arch/x86/pci/fixup.c +++ b/arch/x86/pci/fixup.c @@ -23,7 +23,8 @@ static void __devinit pci_fixup_i450nx(struct pci_dev *d) pci_read_config_byte(d, reg++, &busno); pci_read_config_byte(d, reg++, &suba); pci_read_config_byte(d, reg++, &subb); - DBG("i450NX PXB %d: %02x/%02x/%02x\n", pxb, busno, suba, subb); + dev_dbg(&d->dev, "i450NX PXB %d: %02x/%02x/%02x\n", pxb, busno, + suba, subb); if (busno) pci_scan_bus_with_sysdata(busno); /* Bus A */ if (suba < subb) diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c index a09505806b82..5807d1bc73f7 100644 --- a/arch/x86/pci/i386.c +++ b/arch/x86/pci/i386.c @@ -128,10 +128,8 @@ static void __init pcibios_allocate_bus_resources(struct list_head *bus_list) pr = pci_find_parent_resource(dev, r); if (!r->start || !pr || request_resource(pr, r) < 0) { - printk(KERN_ERR "PCI: Cannot allocate " - "resource region %d " - "of bridge %s\n", - idx, pci_name(dev)); + dev_err(&dev->dev, "BAR %d: can't " + "allocate resource\n", idx); /* * Something is wrong with the region. * Invalidate the resource to prevent @@ -166,15 +164,15 @@ static void __init pcibios_allocate_resources(int pass) else disabled = !(command & PCI_COMMAND_MEMORY); if (pass == disabled) { - DBG("PCI: Resource %08lx-%08lx " - "(f=%lx, d=%d, p=%d)\n", - r->start, r->end, r->flags, disabled, pass); + dev_dbg(&dev->dev, "resource %#08llx-%#08llx " + "(f=%lx, d=%d, p=%d)\n", + (unsigned long long) r->start, + (unsigned long long) r->end, + r->flags, disabled, pass); pr = pci_find_parent_resource(dev, r); if (!pr || request_resource(pr, r) < 0) { - printk(KERN_ERR "PCI: Cannot allocate " - "resource region %d " - "of device %s\n", - idx, pci_name(dev)); + dev_err(&dev->dev, "BAR %d: can't " + "allocate resource\n", idx); /* We'll assign a new address later */ r->end -= r->start; r->start = 0; @@ -187,8 +185,7 @@ static void __init pcibios_allocate_resources(int pass) /* Turn the ROM off, leave the resource region, * but keep it unregistered. */ u32 reg; - DBG("PCI: Switching off ROM of %s\n", - pci_name(dev)); + dev_dbg(&dev->dev, "disabling ROM\n"); r->flags &= ~IORESOURCE_ROM_ENABLE; pci_read_config_dword(dev, dev->rom_base_reg, ®); @@ -257,8 +254,7 @@ void pcibios_set_master(struct pci_dev *dev) lat = pcibios_max_latency; else return; - printk(KERN_DEBUG "PCI: Setting latency timer of device %s to %d\n", - pci_name(dev), lat); + dev_printk(KERN_DEBUG, &dev->dev, "setting latency timer to %d\n", lat); pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat); } diff --git a/arch/x86/pci/irq.c b/arch/x86/pci/irq.c index 6a06a2eb0597..fec0123b33a9 100644 --- a/arch/x86/pci/irq.c +++ b/arch/x86/pci/irq.c @@ -436,7 +436,7 @@ static int pirq_vlsi_get(struct pci_dev *router, struct pci_dev *dev, int pirq) { WARN_ON_ONCE(pirq >= 9); if (pirq > 8) { - printk(KERN_INFO "VLSI router pirq escape (%d)\n", pirq); + dev_info(&dev->dev, "VLSI router PIRQ escape (%d)\n", pirq); return 0; } return read_config_nybble(router, 0x74, pirq-1); @@ -446,7 +446,7 @@ static int pirq_vlsi_set(struct pci_dev *router, struct pci_dev *dev, int pirq, { WARN_ON_ONCE(pirq >= 9); if (pirq > 8) { - printk(KERN_INFO "VLSI router pirq escape (%d)\n", pirq); + dev_info(&dev->dev, "VLSI router PIRQ escape (%d)\n", pirq); return 0; } write_config_nybble(router, 0x74, pirq-1, irq); @@ -492,15 +492,17 @@ static int pirq_amd756_get(struct pci_dev *router, struct pci_dev *dev, int pirq irq = 0; if (pirq <= 4) irq = read_config_nybble(router, 0x56, pirq - 1); - printk(KERN_INFO "AMD756: dev %04x:%04x, router pirq : %d get irq : %2d\n", - dev->vendor, dev->device, pirq, irq); + dev_info(&dev->dev, + "AMD756: dev [%04x/%04x], router PIRQ %d get IRQ %d\n", + dev->vendor, dev->device, pirq, irq); return irq; } static int pirq_amd756_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq) { - printk(KERN_INFO "AMD756: dev %04x:%04x, router pirq : %d SET irq : %2d\n", - dev->vendor, dev->device, pirq, irq); + dev_info(&dev->dev, + "AMD756: dev [%04x/%04x], router PIRQ %d set IRQ %d\n", + dev->vendor, dev->device, pirq, irq); if (pirq <= 4) write_config_nybble(router, 0x56, pirq - 1, irq); return 1; @@ -730,7 +732,6 @@ static __init int ali_router_probe(struct irq_router *r, struct pci_dev *router, switch (device) { case PCI_DEVICE_ID_AL_M1533: case PCI_DEVICE_ID_AL_M1563: - printk(KERN_DEBUG "PCI: Using ALI IRQ Router\n"); r->name = "ALI"; r->get = pirq_ali_get; r->set = pirq_ali_set; @@ -840,11 +841,9 @@ static void __init pirq_find_router(struct irq_router *r) h->probe(r, pirq_router_dev, pirq_router_dev->device)) break; } - printk(KERN_INFO "PCI: Using IRQ router %s [%04x/%04x] at %s\n", - pirq_router.name, - pirq_router_dev->vendor, - pirq_router_dev->device, - pci_name(pirq_router_dev)); + dev_info(&pirq_router_dev->dev, "%s IRQ router [%04x/%04x]\n", + pirq_router.name, + pirq_router_dev->vendor, pirq_router_dev->device); /* The device remains referenced for the kernel lifetime */ } @@ -877,7 +876,7 @@ static int pcibios_lookup_irq(struct pci_dev *dev, int assign) /* Find IRQ pin */ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin); if (!pin) { - DBG(KERN_DEBUG " -> no interrupt pin\n"); + dev_dbg(&dev->dev, "no interrupt pin\n"); return 0; } pin = pin - 1; @@ -887,20 +886,20 @@ static int pcibios_lookup_irq(struct pci_dev *dev, int assign) if (!pirq_table) return 0; - DBG(KERN_DEBUG "IRQ for %s[%c]", pci_name(dev), 'A' + pin); info = pirq_get_info(dev); if (!info) { - DBG(" -> not found in routing table\n" KERN_DEBUG); + dev_dbg(&dev->dev, "PCI INT %c not found in routing table\n", + 'A' + pin); return 0; } pirq = info->irq[pin].link; mask = info->irq[pin].bitmap; if (!pirq) { - DBG(" -> not routed\n" KERN_DEBUG); + dev_dbg(&dev->dev, "PCI INT %c not routed\n", 'A' + pin); return 0; } - DBG(" -> PIRQ %02x, mask %04x, excl %04x", pirq, mask, - pirq_table->exclusive_irqs); + dev_dbg(&dev->dev, "PCI INT %c -> PIRQ %02x, mask %04x, excl %04x", + 'A' + pin, pirq, mask, pirq_table->exclusive_irqs); mask &= pcibios_irq_mask; /* Work around broken HP Pavilion Notebooks which assign USB to @@ -930,10 +929,8 @@ static int pcibios_lookup_irq(struct pci_dev *dev, int assign) if (pci_probe & PCI_USE_PIRQ_MASK) newirq = 0; else - printk("\n" KERN_WARNING - "PCI: IRQ %i for device %s doesn't match PIRQ mask - try pci=usepirqmask\n" - KERN_DEBUG, newirq, - pci_name(dev)); + dev_warn(&dev->dev, "IRQ %d doesn't match PIRQ mask " + "%#x; try pci=usepirqmask\n", newirq, mask); } if (!newirq && assign) { for (i = 0; i < 16; i++) { @@ -944,39 +941,35 @@ static int pcibios_lookup_irq(struct pci_dev *dev, int assign) newirq = i; } } - DBG(" -> newirq=%d", newirq); + dev_dbg(&dev->dev, "PCI INT %c -> newirq %d", 'A' + pin, newirq); /* Check if it is hardcoded */ if ((pirq & 0xf0) == 0xf0) { irq = pirq & 0xf; - DBG(" -> hardcoded IRQ %d\n", irq); - msg = "Hardcoded"; + msg = "hardcoded"; } else if (r->get && (irq = r->get(pirq_router_dev, dev, pirq)) && \ ((!(pci_probe & PCI_USE_PIRQ_MASK)) || ((1 << irq) & mask))) { - DBG(" -> got IRQ %d\n", irq); - msg = "Found"; + msg = "found"; eisa_set_level_irq(irq); } else if (newirq && r->set && (dev->class >> 8) != PCI_CLASS_DISPLAY_VGA) { - DBG(" -> assigning IRQ %d", newirq); if (r->set(pirq_router_dev, dev, pirq, newirq)) { eisa_set_level_irq(newirq); - DBG(" ... OK\n"); - msg = "Assigned"; + msg = "assigned"; irq = newirq; } } if (!irq) { - DBG(" ... failed\n"); if (newirq && mask == (1 << newirq)) { - msg = "Guessed"; + msg = "guessed"; irq = newirq; - } else + } else { + dev_dbg(&dev->dev, "can't route interrupt\n"); return 0; + } } - printk(KERN_INFO "PCI: %s IRQ %d for device %s\n", msg, irq, - pci_name(dev)); + dev_info(&dev->dev, "%s PCI INT %c -> IRQ %d\n", msg, 'A' + pin, irq); /* Update IRQ for all devices with the same pirq value */ while ((dev2 = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev2)) != NULL) { @@ -996,17 +989,17 @@ static int pcibios_lookup_irq(struct pci_dev *dev, int assign) (!(pci_probe & PCI_USE_PIRQ_MASK) || \ ((1 << dev2->irq) & mask))) { #ifndef CONFIG_PCI_MSI - printk(KERN_INFO "IRQ routing conflict for %s, have irq %d, want irq %d\n", - pci_name(dev2), dev2->irq, irq); + dev_info(&dev2->dev, "IRQ routing conflict: " + "have IRQ %d, want IRQ %d\n", + dev2->irq, irq); #endif continue; } dev2->irq = irq; pirq_penalty[irq]++; if (dev != dev2) - printk(KERN_INFO - "PCI: Sharing IRQ %d with %s\n", - irq, pci_name(dev2)); + dev_info(&dev->dev, "sharing IRQ %d with %s\n", + irq, pci_name(dev2)); } } return 1; @@ -1025,8 +1018,7 @@ static void __init pcibios_fixup_irqs(void) * already in use. */ if (dev->irq >= 16) { - DBG(KERN_DEBUG "%s: ignoring bogus IRQ %d\n", - pci_name(dev), dev->irq); + dev_dbg(&dev->dev, "ignoring bogus IRQ %d\n", dev->irq); dev->irq = 0; } /* @@ -1070,12 +1062,12 @@ static void __init pcibios_fixup_irqs(void) irq = IO_APIC_get_PCI_irq_vector(bridge->bus->number, PCI_SLOT(bridge->devfn), pin); if (irq >= 0) - printk(KERN_WARNING "PCI: using PPB %s[%c] to get irq %d\n", - pci_name(bridge), 'A' + pin, irq); + dev_warn(&dev->dev, "using bridge %s INT %c to get IRQ %d\n", + pci_name(bridge), + 'A' + pin, irq); } if (irq >= 0) { - printk(KERN_INFO "PCI->APIC IRQ transform: %s[%c] -> IRQ %d\n", - pci_name(dev), 'A' + pin, irq); + dev_info(&dev->dev, "PCI->APIC IRQ transform: INT %c -> IRQ %d\n", 'A' + pin, irq); dev->irq = irq; } } @@ -1231,25 +1223,24 @@ static int pirq_enable_irq(struct pci_dev *dev) irq = IO_APIC_get_PCI_irq_vector(bridge->bus->number, PCI_SLOT(bridge->devfn), pin); if (irq >= 0) - printk(KERN_WARNING - "PCI: using PPB %s[%c] to get irq %d\n", - pci_name(bridge), - 'A' + pin, irq); + dev_warn(&dev->dev, "using bridge %s " + "INT %c to get IRQ %d\n", + pci_name(bridge), 'A' + pin, + irq); dev = bridge; } dev = temp_dev; if (irq >= 0) { - printk(KERN_INFO - "PCI->APIC IRQ transform: %s[%c] -> IRQ %d\n", - pci_name(dev), 'A' + pin, irq); + dev_info(&dev->dev, "PCI->APIC IRQ transform: " + "INT %c -> IRQ %d\n", 'A' + pin, irq); dev->irq = irq; return 0; } else - msg = " Probably buggy MP table."; + msg = "; probably buggy MP table"; } else if (pci_probe & PCI_BIOS_IRQ_SCAN) msg = ""; else - msg = " Please try using pci=biosirq."; + msg = "; please try using pci=biosirq"; /* * With IDE legacy devices the IRQ lookup failure is not @@ -1259,9 +1250,8 @@ static int pirq_enable_irq(struct pci_dev *dev) !(dev->class & 0x5)) return 0; - printk(KERN_WARNING - "PCI: No IRQ known for interrupt pin %c of device %s.%s\n", - 'A' + pin, pci_name(dev), msg); + dev_warn(&dev->dev, "can't find IRQ for PCI INT %c%s\n", + 'A' + pin, msg); } return 0; } diff --git a/arch/x86/pci/numaq_32.c b/arch/x86/pci/numaq_32.c index f4b16dc11dad..1177845d3186 100644 --- a/arch/x86/pci/numaq_32.c +++ b/arch/x86/pci/numaq_32.c @@ -131,13 +131,14 @@ static void __devinit pci_fixup_i450nx(struct pci_dev *d) u8 busno, suba, subb; int quad = BUS2QUAD(d->bus->number); - printk("PCI: Searching for i450NX host bridges on %s\n", pci_name(d)); + dev_info(&d->dev, "searching for i450NX host bridges\n"); reg = 0xd0; for(pxb=0; pxb<2; pxb++) { pci_read_config_byte(d, reg++, &busno); pci_read_config_byte(d, reg++, &suba); pci_read_config_byte(d, reg++, &subb); - DBG("i450NX PXB %d: %02x/%02x/%02x\n", pxb, busno, suba, subb); + dev_dbg(&d->dev, "i450NX PXB %d: %02x/%02x/%02x\n", + pxb, busno, suba, subb); if (busno) { /* Bus A */ pci_scan_bus_with_sysdata(QUADLOCAL2BUS(quad, busno)); -- cgit v1.2.2 From cddb8a5c14aa89810b40495d94d3d2a0faee6619 Mon Sep 17 00:00:00 2001 From: Andrea Arcangeli Date: Mon, 28 Jul 2008 15:46:29 -0700 Subject: mmu-notifiers: core With KVM/GFP/XPMEM there isn't just the primary CPU MMU pointing to pages. There are secondary MMUs (with secondary sptes and secondary tlbs) too. sptes in the kvm case are shadow pagetables, but when I say spte in mmu-notifier context, I mean "secondary pte". In GRU case there's no actual secondary pte and there's only a secondary tlb because the GRU secondary MMU has no knowledge about sptes and every secondary tlb miss event in the MMU always generates a page fault that has to be resolved by the CPU (this is not the case of KVM where the a secondary tlb miss will walk sptes in hardware and it will refill the secondary tlb transparently to software if the corresponding spte is present). The same way zap_page_range has to invalidate the pte before freeing the page, the spte (and secondary tlb) must also be invalidated before any page is freed and reused. Currently we take a page_count pin on every page mapped by sptes, but that means the pages can't be swapped whenever they're mapped by any spte because they're part of the guest working set. Furthermore a spte unmap event can immediately lead to a page to be freed when the pin is released (so requiring the same complex and relatively slow tlb_gather smp safe logic we have in zap_page_range and that can be avoided completely if the spte unmap event doesn't require an unpin of the page previously mapped in the secondary MMU). The mmu notifiers allow kvm/GRU/XPMEM to attach to the tsk->mm and know when the VM is swapping or freeing or doing anything on the primary MMU so that the secondary MMU code can drop sptes before the pages are freed, avoiding all page pinning and allowing 100% reliable swapping of guest physical address space. Furthermore it avoids the code that teardown the mappings of the secondary MMU, to implement a logic like tlb_gather in zap_page_range that would require many IPI to flush other cpu tlbs, for each fixed number of spte unmapped. To make an example: if what happens on the primary MMU is a protection downgrade (from writeable to wrprotect) the secondary MMU mappings will be invalidated, and the next secondary-mmu-page-fault will call get_user_pages and trigger a do_wp_page through get_user_pages if it called get_user_pages with write=1, and it'll re-establishing an updated spte or secondary-tlb-mapping on the copied page. Or it will setup a readonly spte or readonly tlb mapping if it's a guest-read, if it calls get_user_pages with write=0. This is just an example. This allows to map any page pointed by any pte (and in turn visible in the primary CPU MMU), into a secondary MMU (be it a pure tlb like GRU, or an full MMU with both sptes and secondary-tlb like the shadow-pagetable layer with kvm), or a remote DMA in software like XPMEM (hence needing of schedule in XPMEM code to send the invalidate to the remote node, while no need to schedule in kvm/gru as it's an immediate event like invalidating primary-mmu pte). At least for KVM without this patch it's impossible to swap guests reliably. And having this feature and removing the page pin allows several other optimizations that simplify life considerably. Dependencies: 1) mm_take_all_locks() to register the mmu notifier when the whole VM isn't doing anything with "mm". This allows mmu notifier users to keep track if the VM is in the middle of the invalidate_range_begin/end critical section with an atomic counter incraese in range_begin and decreased in range_end. No secondary MMU page fault is allowed to map any spte or secondary tlb reference, while the VM is in the middle of range_begin/end as any page returned by get_user_pages in that critical section could later immediately be freed without any further ->invalidate_page notification (invalidate_range_begin/end works on ranges and ->invalidate_page isn't called immediately before freeing the page). To stop all page freeing and pagetable overwrites the mmap_sem must be taken in write mode and all other anon_vma/i_mmap locks must be taken too. 2) It'd be a waste to add branches in the VM if nobody could possibly run KVM/GRU/XPMEM on the kernel, so mmu notifiers will only enabled if CONFIG_KVM=m/y. In the current kernel kvm won't yet take advantage of mmu notifiers, but this already allows to compile a KVM external module against a kernel with mmu notifiers enabled and from the next pull from kvm.git we'll start using them. And GRU/XPMEM will also be able to continue the development by enabling KVM=m in their config, until they submit all GRU/XPMEM GPLv2 code to the mainline kernel. Then they can also enable MMU_NOTIFIERS in the same way KVM does it (even if KVM=n). This guarantees nobody selects MMU_NOTIFIER=y if KVM and GRU and XPMEM are all =n. The mmu_notifier_register call can fail because mm_take_all_locks may be interrupted by a signal and return -EINTR. Because mmu_notifier_reigster is used when a driver startup, a failure can be gracefully handled. Here an example of the change applied to kvm to register the mmu notifiers. Usually when a driver startups other allocations are required anyway and -ENOMEM failure paths exists already. struct kvm *kvm_arch_create_vm(void) { struct kvm *kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL); + int err; if (!kvm) return ERR_PTR(-ENOMEM); INIT_LIST_HEAD(&kvm->arch.active_mmu_pages); + kvm->arch.mmu_notifier.ops = &kvm_mmu_notifier_ops; + err = mmu_notifier_register(&kvm->arch.mmu_notifier, current->mm); + if (err) { + kfree(kvm); + return ERR_PTR(err); + } + return kvm; } mmu_notifier_unregister returns void and it's reliable. The patch also adds a few needed but missing includes that would prevent kernel to compile after these changes on non-x86 archs (x86 didn't need them by luck). [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: fix mm/filemap_xip.c build] [akpm@linux-foundation.org: fix mm/mmu_notifier.c build] Signed-off-by: Andrea Arcangeli Signed-off-by: Nick Piggin Signed-off-by: Christoph Lameter Cc: Jack Steiner Cc: Robin Holt Cc: Nick Piggin Cc: Peter Zijlstra Cc: Kanoj Sarcar Cc: Roland Dreier Cc: Steve Wise Cc: Avi Kivity Cc: Hugh Dickins Cc: Rusty Russell Cc: Anthony Liguori Cc: Chris Wright Cc: Marcelo Tosatti Cc: Eric Dumazet Cc: "Paul E. McKenney" Cc: Izik Eidus Cc: Anthony Liguori Cc: Rik van Riel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/kvm/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig index 8d45fabc5f3b..ce3251ce5504 100644 --- a/arch/x86/kvm/Kconfig +++ b/arch/x86/kvm/Kconfig @@ -21,6 +21,7 @@ config KVM tristate "Kernel-based Virtual Machine (KVM) support" depends on HAVE_KVM select PREEMPT_NOTIFIERS + select MMU_NOTIFIER select ANON_INODES ---help--- Support hosting fully virtualized guest machines using hardware -- cgit v1.2.2 From 5d006d8d09e82f086ca0baf79a2907f2c1e25af7 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Tue, 29 Jul 2008 09:58:29 -0500 Subject: lguest: set max_pfn_mapped, growl loudly at Yinghai Lu 6af61a7614a306fe882a0c2b4ddc63b65aa66efc 'x86: clean up max_pfn_mapped usage - 32-bit' makes the following comment: XEN PV and lguest may need to assign max_pfn_mapped too. But no CC. Yinghai, wasting fellow developers' time is a VERY bad habit. If you do it again, I will hunt you down and try to extract the three hours of my life I just lost :) Signed-off-by: Rusty Russell Cc: Yinghai Lu --- 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 0313a5eec412..d9249a882aa5 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1014,6 +1014,9 @@ __init void lguest_init(void) init_pg_tables_start = __pa(pg0); init_pg_tables_end = __pa(pg0); + /* As described in head_32.S, we map the first 128M of memory. */ + max_pfn_mapped = (128*1024*1024) >> PAGE_SHIFT; + /* Load the %fs segment register (the per-cpu segment register) with * the normal data segment to get through booting. */ asm volatile ("mov %0, %%fs" : : "r" (__KERNEL_DS) : "memory"); -- cgit v1.2.2 From 9b79022ca909b66e2cd0cfd9248f832fc165f77f Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 28 Jul 2008 17:54:21 -0700 Subject: Fix 'get_user_pages_fast()' with non-page-aligned start address Alexey Dobriyan reported trouble with LTP with the new fast-gup code, and Johannes Weiner debugged it to non-page-aligned addresses, where the new get_user_pages_fast() code would do all the wrong things, including just traversing past the end of the requested area due to 'addr' never matching 'end' exactly. This is not a pretty fix, and we may actually want to move the alignment into generic code, leaving just the core code per-arch, but Alexey verified that the vmsplice01 LTP test doesn't crash with this. Reported-and-tested-by: Alexey Dobriyan Debugged-by: Johannes Weiner Cc: Nick Piggin Cc: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/mm/gup.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/gup.c b/arch/x86/mm/gup.c index 3085f25b4355..007bb06c7504 100644 --- a/arch/x86/mm/gup.c +++ b/arch/x86/mm/gup.c @@ -223,14 +223,17 @@ int get_user_pages_fast(unsigned long start, int nr_pages, int write, struct page **pages) { struct mm_struct *mm = current->mm; - unsigned long end = start + (nr_pages << PAGE_SHIFT); - unsigned long addr = start; + unsigned long addr, len, end; unsigned long next; pgd_t *pgdp; int nr = 0; + start &= PAGE_MASK; + addr = start; + len = (unsigned long) nr_pages << PAGE_SHIFT; + end = start + len; if (unlikely(!access_ok(write ? VERIFY_WRITE : VERIFY_READ, - start, nr_pages*PAGE_SIZE))) + start, len))) goto slow_irqon; /* -- cgit v1.2.2 From a1708ce8a362c4999f1201237ae7b77c4d13af82 Mon Sep 17 00:00:00 2001 From: Andrea Arcangeli Date: Fri, 25 Jul 2008 16:26:39 +0200 Subject: KVM: Allow reading aliases with mmu_lock This allows the mmu notifier code to run unalias_gfn with only the mmu_lock held. Only alias writes need the mmu_lock held. Readers will either take the slots_lock in read mode or the mmu_lock. Signed-off-by: Andrea Arcangeli Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 5916191420c7..9870ce422920 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1495,6 +1495,7 @@ static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm, goto out; down_write(&kvm->slots_lock); + spin_lock(&kvm->mmu_lock); p = &kvm->arch.aliases[alias->slot]; p->base_gfn = alias->guest_phys_addr >> PAGE_SHIFT; @@ -1506,6 +1507,7 @@ static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm, break; kvm->arch.naliases = n; + spin_unlock(&kvm->mmu_lock); kvm_mmu_zap_all(kvm); up_write(&kvm->slots_lock); -- cgit v1.2.2 From 604b38ac0369bd50fcbb33344aa5553c071009f7 Mon Sep 17 00:00:00 2001 From: Andrea Arcangeli Date: Fri, 25 Jul 2008 16:32:03 +0200 Subject: KVM: Allow browsing memslots with mmu_lock This allows reading memslots with only the mmu_lock hold for mmu notifiers that runs in atomic context and with mmu_lock held. Signed-off-by: Andrea Arcangeli Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 9870ce422920..c7b01efe0646 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3974,16 +3974,23 @@ int kvm_arch_set_memory_region(struct kvm *kvm, */ if (!user_alloc) { if (npages && !old.rmap) { + unsigned long userspace_addr; + down_write(¤t->mm->mmap_sem); - memslot->userspace_addr = do_mmap(NULL, 0, - npages * PAGE_SIZE, - PROT_READ | PROT_WRITE, - MAP_SHARED | MAP_ANONYMOUS, - 0); + userspace_addr = do_mmap(NULL, 0, + npages * PAGE_SIZE, + PROT_READ | PROT_WRITE, + MAP_SHARED | MAP_ANONYMOUS, + 0); up_write(¤t->mm->mmap_sem); - if (IS_ERR((void *)memslot->userspace_addr)) - return PTR_ERR((void *)memslot->userspace_addr); + if (IS_ERR((void *)userspace_addr)) + return PTR_ERR((void *)userspace_addr); + + /* set userspace_addr atomically for kvm_hva_to_rmapp */ + spin_lock(&kvm->mmu_lock); + memslot->userspace_addr = userspace_addr; + spin_unlock(&kvm->mmu_lock); } else { if (!old.user_alloc && old.rmap) { int ret; -- cgit v1.2.2 From e930bffe95e1e886a1ede80726ea38df5838d067 Mon Sep 17 00:00:00 2001 From: Andrea Arcangeli Date: Fri, 25 Jul 2008 16:24:52 +0200 Subject: KVM: Synchronize guest physical memory map to host virtual memory map Synchronize changes to host virtual addresses which are part of a KVM memory slot to the KVM shadow mmu. This allows pte operations like swapping, page migration, and madvise() to transparently work with KVM. Signed-off-by: Andrea Arcangeli Signed-off-by: Avi Kivity --- arch/x86/kvm/mmu.c | 100 +++++++++++++++++++++++++++++++++++++++++++++ arch/x86/kvm/paging_tmpl.h | 12 ++++++ 2 files changed, 112 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 2fa231923cf7..0bfe2bd305eb 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -653,6 +653,84 @@ static void rmap_write_protect(struct kvm *kvm, u64 gfn) account_shadowed(kvm, gfn); } +static int kvm_unmap_rmapp(struct kvm *kvm, unsigned long *rmapp) +{ + u64 *spte; + int need_tlb_flush = 0; + + while ((spte = rmap_next(kvm, rmapp, NULL))) { + BUG_ON(!(*spte & PT_PRESENT_MASK)); + rmap_printk("kvm_rmap_unmap_hva: spte %p %llx\n", spte, *spte); + rmap_remove(kvm, spte); + set_shadow_pte(spte, shadow_trap_nonpresent_pte); + need_tlb_flush = 1; + } + return need_tlb_flush; +} + +static int kvm_handle_hva(struct kvm *kvm, unsigned long hva, + int (*handler)(struct kvm *kvm, unsigned long *rmapp)) +{ + int i; + int retval = 0; + + /* + * If mmap_sem isn't taken, we can look the memslots with only + * the mmu_lock by skipping over the slots with userspace_addr == 0. + */ + for (i = 0; i < kvm->nmemslots; i++) { + struct kvm_memory_slot *memslot = &kvm->memslots[i]; + unsigned long start = memslot->userspace_addr; + unsigned long end; + + /* mmu_lock protects userspace_addr */ + if (!start) + continue; + + end = start + (memslot->npages << PAGE_SHIFT); + if (hva >= start && hva < end) { + gfn_t gfn_offset = (hva - start) >> PAGE_SHIFT; + retval |= handler(kvm, &memslot->rmap[gfn_offset]); + retval |= handler(kvm, + &memslot->lpage_info[ + gfn_offset / + KVM_PAGES_PER_HPAGE].rmap_pde); + } + } + + return retval; +} + +int kvm_unmap_hva(struct kvm *kvm, unsigned long hva) +{ + return kvm_handle_hva(kvm, hva, kvm_unmap_rmapp); +} + +static int kvm_age_rmapp(struct kvm *kvm, unsigned long *rmapp) +{ + u64 *spte; + int young = 0; + + spte = rmap_next(kvm, rmapp, NULL); + while (spte) { + int _young; + u64 _spte = *spte; + BUG_ON(!(_spte & PT_PRESENT_MASK)); + _young = _spte & PT_ACCESSED_MASK; + if (_young) { + young = 1; + clear_bit(PT_ACCESSED_SHIFT, (unsigned long *)spte); + } + spte = rmap_next(kvm, rmapp, spte); + } + return young; +} + +int kvm_age_hva(struct kvm *kvm, unsigned long hva) +{ + return kvm_handle_hva(kvm, hva, kvm_age_rmapp); +} + #ifdef MMU_DEBUG static int is_empty_shadow_page(u64 *spt) { @@ -1203,6 +1281,7 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn) int r; int largepage = 0; pfn_t pfn; + unsigned long mmu_seq; down_read(¤t->mm->mmap_sem); if (is_largepage_backed(vcpu, gfn & ~(KVM_PAGES_PER_HPAGE-1))) { @@ -1210,6 +1289,8 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn) largepage = 1; } + mmu_seq = vcpu->kvm->mmu_notifier_seq; + /* implicit mb(), we'll read before PT lock is unlocked */ pfn = gfn_to_pfn(vcpu->kvm, gfn); up_read(¤t->mm->mmap_sem); @@ -1220,6 +1301,8 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn) } spin_lock(&vcpu->kvm->mmu_lock); + if (mmu_notifier_retry(vcpu, mmu_seq)) + goto out_unlock; kvm_mmu_free_some_pages(vcpu); r = __direct_map(vcpu, v, write, largepage, gfn, pfn, PT32E_ROOT_LEVEL); @@ -1227,6 +1310,11 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn) return r; + +out_unlock: + spin_unlock(&vcpu->kvm->mmu_lock); + kvm_release_pfn_clean(pfn); + return 0; } @@ -1345,6 +1433,7 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, int r; int largepage = 0; gfn_t gfn = gpa >> PAGE_SHIFT; + unsigned long mmu_seq; ASSERT(vcpu); ASSERT(VALID_PAGE(vcpu->arch.mmu.root_hpa)); @@ -1358,6 +1447,8 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, gfn &= ~(KVM_PAGES_PER_HPAGE-1); largepage = 1; } + mmu_seq = vcpu->kvm->mmu_notifier_seq; + /* implicit mb(), we'll read before PT lock is unlocked */ pfn = gfn_to_pfn(vcpu->kvm, gfn); up_read(¤t->mm->mmap_sem); if (is_error_pfn(pfn)) { @@ -1365,12 +1456,19 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, return 1; } spin_lock(&vcpu->kvm->mmu_lock); + if (mmu_notifier_retry(vcpu, mmu_seq)) + goto out_unlock; kvm_mmu_free_some_pages(vcpu); r = __direct_map(vcpu, gpa, error_code & PFERR_WRITE_MASK, largepage, gfn, pfn, kvm_x86_ops->get_tdp_level()); spin_unlock(&vcpu->kvm->mmu_lock); return r; + +out_unlock: + spin_unlock(&vcpu->kvm->mmu_lock); + kvm_release_pfn_clean(pfn); + return 0; } static void nonpaging_free(struct kvm_vcpu *vcpu) @@ -1670,6 +1768,8 @@ static void mmu_guess_page_from_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa, gfn &= ~(KVM_PAGES_PER_HPAGE-1); vcpu->arch.update_pte.largepage = 1; } + vcpu->arch.update_pte.mmu_seq = vcpu->kvm->mmu_notifier_seq; + /* implicit mb(), we'll read before PT lock is unlocked */ pfn = gfn_to_pfn(vcpu->kvm, gfn); up_read(¤t->mm->mmap_sem); diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h index 4d918220baeb..f72ac1fa35f0 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -263,6 +263,8 @@ static void FNAME(update_pte)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *page, pfn = vcpu->arch.update_pte.pfn; if (is_error_pfn(pfn)) return; + if (mmu_notifier_retry(vcpu, vcpu->arch.update_pte.mmu_seq)) + return; kvm_get_pfn(pfn); mmu_set_spte(vcpu, spte, page->role.access, pte_access, 0, 0, gpte & PT_DIRTY_MASK, NULL, largepage, gpte_to_gfn(gpte), @@ -380,6 +382,7 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, int r; pfn_t pfn; int largepage = 0; + unsigned long mmu_seq; pgprintk("%s: addr %lx err %x\n", __func__, addr, error_code); kvm_mmu_audit(vcpu, "pre page fault"); @@ -413,6 +416,8 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, largepage = 1; } } + mmu_seq = vcpu->kvm->mmu_notifier_seq; + /* implicit mb(), we'll read before PT lock is unlocked */ pfn = gfn_to_pfn(vcpu->kvm, walker.gfn); up_read(¤t->mm->mmap_sem); @@ -424,6 +429,8 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, } spin_lock(&vcpu->kvm->mmu_lock); + if (mmu_notifier_retry(vcpu, mmu_seq)) + goto out_unlock; kvm_mmu_free_some_pages(vcpu); shadow_pte = FNAME(fetch)(vcpu, addr, &walker, user_fault, write_fault, largepage, &write_pt, pfn); @@ -439,6 +446,11 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, spin_unlock(&vcpu->kvm->mmu_lock); return write_pt; + +out_unlock: + spin_unlock(&vcpu->kvm->mmu_lock); + kvm_release_pfn_clean(pfn); + return 0; } static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t vaddr) -- cgit v1.2.2 From ed8486243379ef3e6c61363df915882945c0eaec Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Tue, 29 Jul 2008 11:30:57 +0300 Subject: KVM: Advertise synchronized mmu support to userspace Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index c7b01efe0646..0d682fc6aeb3 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -883,6 +883,7 @@ int kvm_dev_ioctl_check_extension(long ext) case KVM_CAP_PIT: case KVM_CAP_NOP_IO_DELAY: case KVM_CAP_MP_STATE: + case KVM_CAP_SYNC_MMU: r = 1; break; case KVM_CAP_COALESCED_MMIO: -- cgit v1.2.2 From 8978b74253280d59e97cf49a3ec2c0cbccd5b801 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Tue, 29 Jul 2008 13:38:53 +0900 Subject: generic, x86: fix add iommu_num_pages helper function This IOMMU helper function doesn't work for some architectures: http://marc.info/?l=linux-kernel&m=121699304403202&w=2 It also breaks POWER and SPARC builds: http://marc.info/?l=linux-kernel&m=121730388001890&w=2 Currently, only x86 IOMMUs use this so let's move it to x86 for now. Reported-by: Stephen Rothwell Signed-off-by: FUJITA Tomonori Signed-off-by: Ingo Molnar --- arch/x86/kernel/pci-dma.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index 8dbffb846de9..87d4d6964ec2 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -123,6 +123,14 @@ void __init pci_iommu_alloc(void) pci_swiotlb_init(); } + +unsigned long iommu_num_pages(unsigned long addr, unsigned long len) +{ + unsigned long size = roundup((addr & ~PAGE_MASK) + len, PAGE_SIZE); + + return size >> PAGE_SHIFT; +} +EXPORT_SYMBOL(iommu_num_pages); #endif /* -- cgit v1.2.2 From 0d39741a27d86d305cc75ba626392be410dcbab9 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Tue, 29 Jul 2008 22:34:01 -0700 Subject: GRU Driver: export is_uv_system(), zap_page_range() & follow_page() Exports needed by the GRU driver. Signed-off-by: Jack Steiner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/kernel/genapic_64.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/genapic_64.c b/arch/x86/kernel/genapic_64.c index 1fa8be5bd217..eaff0bbb1444 100644 --- a/arch/x86/kernel/genapic_64.c +++ b/arch/x86/kernel/genapic_64.c @@ -99,3 +99,4 @@ int is_uv_system(void) { return uv_system_type != UV_NONE; } +EXPORT_SYMBOL_GPL(is_uv_system); -- cgit v1.2.2 From afd962a9e8708c571c5c0c4a6d098f931742c229 Mon Sep 17 00:00:00 2001 From: Vitaly Mayatskikh Date: Wed, 30 Jul 2008 13:30:14 +0200 Subject: x86: wrong register was used in align macro New ALIGN_DESTINATION macro has sad typo: r8d register was used instead of ecx in fixup section. This can be considered as a regression. Register ecx was also wrongly loaded with value in r8d in copy_user_nocache routine. Signed-off-by: Vitaly Mayatskikh Signed-off-by: Linus Torvalds --- arch/x86/lib/copy_user_64.S | 2 +- arch/x86/lib/copy_user_nocache_64.S | 3 +-- 2 files changed, 2 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/lib/copy_user_64.S b/arch/x86/lib/copy_user_64.S index dfdf428975c0..f118c110af32 100644 --- a/arch/x86/lib/copy_user_64.S +++ b/arch/x86/lib/copy_user_64.S @@ -52,7 +52,7 @@ jnz 100b 102: .section .fixup,"ax" -103: addl %r8d,%edx /* ecx is zerorest also */ +103: addl %ecx,%edx /* ecx is zerorest also */ jmp copy_user_handle_tail .previous diff --git a/arch/x86/lib/copy_user_nocache_64.S b/arch/x86/lib/copy_user_nocache_64.S index 40e0e309d27e..cb0c112386fb 100644 --- a/arch/x86/lib/copy_user_nocache_64.S +++ b/arch/x86/lib/copy_user_nocache_64.S @@ -32,7 +32,7 @@ jnz 100b 102: .section .fixup,"ax" -103: addl %r8d,%edx /* ecx is zerorest also */ +103: addl %ecx,%edx /* ecx is zerorest also */ jmp copy_user_handle_tail .previous @@ -108,7 +108,6 @@ ENTRY(__copy_user_nocache) jmp 60f 50: movl %ecx,%edx 60: sfence - movl %r8d,%ecx jmp copy_user_handle_tail .previous -- cgit v1.2.2 From ec983f7060cd73e14cdd3edd910339127a8a4e96 Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Wed, 30 Jul 2008 12:05:03 -0400 Subject: [CPUFREQ] Remove EXPERIMENTAL annotation from VIA C7 powersaver kconfig. This has been pretty solid, and doesn't see much change at all. Noticed by Harald Welte. Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/Kconfig b/arch/x86/kernel/cpu/cpufreq/Kconfig index cb7a5715596d..efae3b22a0ff 100644 --- a/arch/x86/kernel/cpu/cpufreq/Kconfig +++ b/arch/x86/kernel/cpu/cpufreq/Kconfig @@ -235,9 +235,9 @@ config X86_LONGHAUL If in doubt, say N. config X86_E_POWERSAVER - tristate "VIA C7 Enhanced PowerSaver (EXPERIMENTAL)" + tristate "VIA C7 Enhanced PowerSaver" select CPU_FREQ_TABLE - depends on X86_32 && EXPERIMENTAL + depends on X86_32 help This adds the CPUFreq driver for VIA C7 processors. -- cgit v1.2.2 From 460f5ef2835dbc33825f611f408eb09c29be4b85 Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Wed, 30 Jul 2008 13:01:42 -0400 Subject: [CPUFREQ] Fix warning in elanfreq arch/x86/kernel/cpu/cpufreq/elanfreq.c:47:26: warning: symbol 'elan_multiplier' was not declared. Should it be static? Yes, yes it should. Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/elanfreq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/elanfreq.c b/arch/x86/kernel/cpu/cpufreq/elanfreq.c index 94619c22f563..e4a4bf870e94 100644 --- a/arch/x86/kernel/cpu/cpufreq/elanfreq.c +++ b/arch/x86/kernel/cpu/cpufreq/elanfreq.c @@ -44,7 +44,7 @@ struct s_elan_multiplier { * It is important that the frequencies * are listed in ascending order here! */ -struct s_elan_multiplier elan_multiplier[] = { +static struct s_elan_multiplier elan_multiplier[] = { {1000, 0x02, 0x18}, {2000, 0x02, 0x10}, {4000, 0x02, 0x08}, -- cgit v1.2.2 From 23431b495fcbbde6d09aba30019b6a224d543cf9 Mon Sep 17 00:00:00 2001 From: Mark Langsdorf Date: Thu, 31 Jul 2008 12:39:05 -0500 Subject: [CPUFREQ][1/2] whitespace fix for powernow-k8 Trivial whitespace fix for powernow-k8. Signed-off-by: Mark Langsdorf Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index c45ca6d4dce1..84bb395038d8 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -66,7 +66,6 @@ static u32 find_freq_from_fid(u32 fid) return 800 + (fid * 100); } - /* Return a frequency in KHz, given an input fid */ static u32 find_khz_freq_from_fid(u32 fid) { @@ -78,7 +77,6 @@ static u32 find_khz_freq_from_pstate(struct cpufreq_frequency_table *data, u32 p return data[pstate].frequency; } - /* Return the vco fid for an input fid * * Each "low" fid has corresponding "high" fid, and you can get to "low" fids @@ -166,7 +164,6 @@ static void fidvid_msr_init(void) wrmsr(MSR_FIDVID_CTL, lo, hi); } - /* write the new fid value along with the other control fields to the msr */ static int write_new_fid(struct powernow_k8_data *data, u32 fid) { -- cgit v1.2.2 From 34ae7f35a21694aa5cb8829dc5142c39d73d6ba0 Mon Sep 17 00:00:00 2001 From: Mark Langsdorf Date: Thu, 31 Jul 2008 12:39:12 -0500 Subject: [CPUFREQ][2/2] preregister support for powernow-k8 This patch provides support for the _PSD ACPI object in the Powernow-k8 driver. Although it looks like an invasive patch, most of it is simply the consequence of turning the static acpi_performance_data structure into a pointer. AMD has tested it on several machines over the past few days without issue. [trivial checkpatch warnings fixed up by davej] [X86_POWERNOW_K8_ACPI=n buildfix from Randy Dunlap] Signed-off-by: Mark Langsdorf Tested-by: Frank Arnold Signed-off-by: Randy Dunlap Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 109 ++++++++++++++++++++---------- arch/x86/kernel/cpu/cpufreq/powernow-k8.h | 3 +- 2 files changed, 75 insertions(+), 37 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index 84bb395038d8..4e7271999a74 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -737,44 +737,63 @@ static int find_psb_table(struct powernow_k8_data *data) #ifdef CONFIG_X86_POWERNOW_K8_ACPI static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index) { - if (!data->acpi_data.state_count || (cpu_family == CPU_HW_PSTATE)) + if (!data->acpi_data->state_count || (cpu_family == CPU_HW_PSTATE)) return; - data->irt = (data->acpi_data.states[index].control >> IRT_SHIFT) & IRT_MASK; - data->rvo = (data->acpi_data.states[index].control >> RVO_SHIFT) & RVO_MASK; - data->exttype = (data->acpi_data.states[index].control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK; - data->plllock = (data->acpi_data.states[index].control >> PLL_L_SHIFT) & PLL_L_MASK; - data->vidmvs = 1 << ((data->acpi_data.states[index].control >> MVS_SHIFT) & MVS_MASK); - data->vstable = (data->acpi_data.states[index].control >> VST_SHIFT) & VST_MASK; + data->irt = (data->acpi_data->states[index].control >> IRT_SHIFT) & IRT_MASK; + data->rvo = (data->acpi_data->states[index].control >> RVO_SHIFT) & RVO_MASK; + data->exttype = (data->acpi_data->states[index].control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK; + data->plllock = (data->acpi_data->states[index].control >> PLL_L_SHIFT) & PLL_L_MASK; + data->vidmvs = 1 << ((data->acpi_data->states[index].control >> MVS_SHIFT) & MVS_MASK); + data->vstable = (data->acpi_data->states[index].control >> VST_SHIFT) & VST_MASK; +} + + +static struct acpi_processor_performance *acpi_perf_data; +static int preregister_valid; + +static int powernow_k8_cpu_preinit_acpi(void) +{ + acpi_perf_data = alloc_percpu(struct acpi_processor_performance); + if (!acpi_perf_data) + return -ENODEV; + + if (acpi_processor_preregister_performance(acpi_perf_data)) + return -ENODEV; + else + preregister_valid = 1; + return 0; } static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) { struct cpufreq_frequency_table *powernow_table; int ret_val; + int cpu = 0; - if (acpi_processor_register_performance(&data->acpi_data, data->cpu)) { + data->acpi_data = percpu_ptr(acpi_perf_data, cpu); + if (acpi_processor_register_performance(data->acpi_data, data->cpu)) { dprintk("register performance failed: bad ACPI data\n"); return -EIO; } /* verify the data contained in the ACPI structures */ - if (data->acpi_data.state_count <= 1) { + if (data->acpi_data->state_count <= 1) { dprintk("No ACPI P-States\n"); goto err_out; } - if ((data->acpi_data.control_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) || - (data->acpi_data.status_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) { + if ((data->acpi_data->control_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) || + (data->acpi_data->status_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) { dprintk("Invalid control/status registers (%x - %x)\n", - data->acpi_data.control_register.space_id, - data->acpi_data.status_register.space_id); + data->acpi_data->control_register.space_id, + data->acpi_data->status_register.space_id); goto err_out; } /* fill in data->powernow_table */ powernow_table = kmalloc((sizeof(struct cpufreq_frequency_table) - * (data->acpi_data.state_count + 1)), GFP_KERNEL); + * (data->acpi_data->state_count + 1)), GFP_KERNEL); if (!powernow_table) { dprintk("powernow_table memory alloc failure\n"); goto err_out; @@ -787,12 +806,12 @@ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) if (ret_val) goto err_out_mem; - powernow_table[data->acpi_data.state_count].frequency = CPUFREQ_TABLE_END; - powernow_table[data->acpi_data.state_count].index = 0; + powernow_table[data->acpi_data->state_count].frequency = CPUFREQ_TABLE_END; + powernow_table[data->acpi_data->state_count].index = 0; data->powernow_table = powernow_table; /* fill in data */ - data->numps = data->acpi_data.state_count; + data->numps = data->acpi_data->state_count; if (first_cpu(per_cpu(cpu_core_map, data->cpu)) == data->cpu) print_basics(data); powernow_k8_acpi_pst_values(data, 0); @@ -800,16 +819,31 @@ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) /* notify BIOS that we exist */ acpi_processor_notify_smm(THIS_MODULE); + /* determine affinity, from ACPI if available */ + if (preregister_valid) { + if ((data->acpi_data->shared_type == CPUFREQ_SHARED_TYPE_ALL) || + (data->acpi_data->shared_type == CPUFREQ_SHARED_TYPE_ANY)) + data->starting_core_affinity = data->acpi_data->shared_cpu_map; + else + data->starting_core_affinity = cpumask_of_cpu(data->cpu); + } else { + /* best guess from family if not */ + if (cpu_family == CPU_HW_PSTATE) + data->starting_core_affinity = cpumask_of_cpu(data->cpu); + else + data->starting_core_affinity = per_cpu(cpu_core_map, data->cpu); + } + return 0; err_out_mem: kfree(powernow_table); err_out: - acpi_processor_unregister_performance(&data->acpi_data, data->cpu); + acpi_processor_unregister_performance(data->acpi_data, data->cpu); /* data->acpi_data.state_count informs us at ->exit() whether ACPI was used */ - data->acpi_data.state_count = 0; + data->acpi_data->state_count = 0; return -ENODEV; } @@ -821,10 +855,10 @@ static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpuf rdmsr(MSR_PSTATE_CUR_LIMIT, hi, lo); data->max_hw_pstate = (hi & HW_PSTATE_MAX_MASK) >> HW_PSTATE_MAX_SHIFT; - for (i = 0; i < data->acpi_data.state_count; i++) { + for (i = 0; i < data->acpi_data->state_count; i++) { u32 index; - index = data->acpi_data.states[i].control & HW_PSTATE_MASK; + index = data->acpi_data->states[i].control & HW_PSTATE_MASK; if (index > data->max_hw_pstate) { printk(KERN_ERR PFX "invalid pstate %d - bad value %d.\n", i, index); printk(KERN_ERR PFX "Please report to BIOS manufacturer\n"); @@ -840,7 +874,7 @@ static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpuf powernow_table[i].index = index; - powernow_table[i].frequency = data->acpi_data.states[i].core_frequency * 1000; + powernow_table[i].frequency = data->acpi_data->states[i].core_frequency * 1000; } return 0; } @@ -849,16 +883,16 @@ static int fill_powernow_table_fidvid(struct powernow_k8_data *data, struct cpuf { int i; int cntlofreq = 0; - for (i = 0; i < data->acpi_data.state_count; i++) { + for (i = 0; i < data->acpi_data->state_count; i++) { u32 fid; u32 vid; if (data->exttype) { - fid = data->acpi_data.states[i].status & EXT_FID_MASK; - vid = (data->acpi_data.states[i].status >> VID_SHIFT) & EXT_VID_MASK; + fid = data->acpi_data->states[i].status & EXT_FID_MASK; + vid = (data->acpi_data->states[i].status >> VID_SHIFT) & EXT_VID_MASK; } else { - fid = data->acpi_data.states[i].control & FID_MASK; - vid = (data->acpi_data.states[i].control >> VID_SHIFT) & VID_MASK; + fid = data->acpi_data->states[i].control & FID_MASK; + vid = (data->acpi_data->states[i].control >> VID_SHIFT) & VID_MASK; } dprintk(" %d : fid 0x%x, vid 0x%x\n", i, fid, vid); @@ -899,10 +933,10 @@ static int fill_powernow_table_fidvid(struct powernow_k8_data *data, struct cpuf cntlofreq = i; } - if (powernow_table[i].frequency != (data->acpi_data.states[i].core_frequency * 1000)) { + if (powernow_table[i].frequency != (data->acpi_data->states[i].core_frequency * 1000)) { printk(KERN_INFO PFX "invalid freq entries %u kHz vs. %u kHz\n", powernow_table[i].frequency, - (unsigned int) (data->acpi_data.states[i].core_frequency * 1000)); + (unsigned int) (data->acpi_data->states[i].core_frequency * 1000)); powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID; continue; } @@ -912,11 +946,12 @@ static int fill_powernow_table_fidvid(struct powernow_k8_data *data, struct cpuf static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data) { - if (data->acpi_data.state_count) - acpi_processor_unregister_performance(&data->acpi_data, data->cpu); + if (data->acpi_data->state_count) + acpi_processor_unregister_performance(data->acpi_data, data->cpu); } #else +static int powernow_k8_cpu_preinit_acpi(void) { return -ENODEV; } static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) { return -ENODEV; } static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data) { return; } static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index) { return; } @@ -1101,7 +1136,7 @@ static int powernowk8_verify(struct cpufreq_policy *pol) static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) { struct powernow_k8_data *data; - cpumask_t oldmask; + cpumask_t oldmask = CPU_MASK_ALL; int rc; if (!cpu_online(pol->cpu)) @@ -1174,10 +1209,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) /* run on any CPU again */ set_cpus_allowed_ptr(current, &oldmask); - if (cpu_family == CPU_HW_PSTATE) - pol->cpus = cpumask_of_cpu(pol->cpu); - else - pol->cpus = per_cpu(cpu_core_map, pol->cpu); + pol->cpus = data->starting_core_affinity; data->available_cores = &(pol->cpus); /* Take a crude guess here. @@ -1300,6 +1332,7 @@ static int __cpuinit powernowk8_init(void) } if (supported_cpus == num_online_cpus()) { + powernow_k8_cpu_preinit_acpi(); printk(KERN_INFO PFX "Found %d %s " "processors (%d cpu cores) (" VERSION ")\n", num_online_nodes(), @@ -1316,6 +1349,10 @@ static void __exit powernowk8_exit(void) dprintk("exit\n"); cpufreq_unregister_driver(&cpufreq_amd64_driver); + +#ifdef CONFIG_X86_POWERNOW_K8_ACPI + free_percpu(acpi_perf_data); +#endif } MODULE_AUTHOR("Paul Devriendt and Mark Langsdorf "); diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h index ab48cfed4d96..a62612cd4be8 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h @@ -33,12 +33,13 @@ struct powernow_k8_data { #ifdef CONFIG_X86_POWERNOW_K8_ACPI /* the acpi table needs to be kept. it's only available if ACPI was * used to determine valid frequency/vid/fid states */ - struct acpi_processor_performance acpi_data; + struct acpi_processor_performance *acpi_data; #endif /* we need to keep track of associated cores, but let cpufreq * handle hotplug events - so just point at cpufreq pol->cpus * structure */ cpumask_t *available_cores; + cpumask_t starting_core_affinity; }; -- cgit v1.2.2