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 v1.2.3 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 v1.2.3 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 v1.2.3 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 ++++++---- kernel/cpu.c | 8 +++++--- 2 files changed, 11 insertions(+), 7 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; diff --git a/kernel/cpu.c b/kernel/cpu.c index fe31ff3d3809..9d4e1c28c053 100644 --- a/kernel/cpu.c +++ b/kernel/cpu.c @@ -462,7 +462,6 @@ out: #endif /* CONFIG_SMP */ -#ifndef CONFIG_HAVE_CPUMASK_OF_CPU_MAP /* 64 bits of zeros, for initializers. */ #if BITS_PER_LONG == 32 #define Z64 0, 0 @@ -509,7 +508,11 @@ out: /* We want this statically initialized, just to be safe. We try not * to waste too much space, either. */ -static const cpumask_t cpumask_map[] = { +static const cpumask_t cpumask_map[] +#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP +__initdata +#endif += { CMI0(0), CMI0(1), CMI0(2), CMI0(3), #if NR_CPUS > 4 CMI0(4), CMI0(5), CMI0(6), CMI0(7), @@ -569,4 +572,3 @@ static const cpumask_t cpumask_map[] = { }; const cpumask_t *cpumask_of_cpu_map = cpumask_map; -#endif /* !CONFIG_HAVE_CPUMASK_OF_CPU_MAP */ -- cgit v1.2.3 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 +++-------- drivers/acpi/processor_throttling.c | 11 +++-------- drivers/firmware/dcdbas.c | 3 +-- drivers/misc/sgi-xp/xpc_main.c | 3 +-- kernel/stop_machine.c | 3 +-- kernel/time/tick-common.c | 8 +++----- kernel/trace/trace_sysprof.c | 4 +--- lib/smp_processor_id.c | 5 +---- net/sunrpc/svc.c | 3 +-- 17 files changed, 37 insertions(+), 83 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. diff --git a/drivers/acpi/processor_throttling.c b/drivers/acpi/processor_throttling.c index a2c3f9cfa549..a56fc6c4394b 100644 --- a/drivers/acpi/processor_throttling.c +++ b/drivers/acpi/processor_throttling.c @@ -827,7 +827,6 @@ static int acpi_processor_get_throttling_ptc(struct acpi_processor *pr) static int acpi_processor_get_throttling(struct acpi_processor *pr) { cpumask_t saved_mask; - cpumask_of_cpu_ptr_declare(new_mask); int ret; if (!pr) @@ -839,8 +838,7 @@ static int acpi_processor_get_throttling(struct acpi_processor *pr) * Migrate task to the cpu pointed by pr. */ saved_mask = current->cpus_allowed; - cpumask_of_cpu_ptr_next(new_mask, pr->id); - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(pr->id)); ret = pr->throttling.acpi_processor_get_throttling(pr); /* restore the previous state */ set_cpus_allowed_ptr(current, &saved_mask); @@ -989,7 +987,6 @@ static int acpi_processor_set_throttling_ptc(struct acpi_processor *pr, int acpi_processor_set_throttling(struct acpi_processor *pr, int state) { cpumask_t saved_mask; - cpumask_of_cpu_ptr_declare(new_mask); int ret = 0; unsigned int i; struct acpi_processor *match_pr; @@ -1028,8 +1025,7 @@ int acpi_processor_set_throttling(struct acpi_processor *pr, int state) * it can be called only for the cpu pointed by pr. */ if (p_throttling->shared_type == DOMAIN_COORD_TYPE_SW_ANY) { - cpumask_of_cpu_ptr_next(new_mask, pr->id); - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(pr->id)); ret = p_throttling->acpi_processor_set_throttling(pr, t_state.target_state); } else { @@ -1060,8 +1056,7 @@ int acpi_processor_set_throttling(struct acpi_processor *pr, int state) continue; } t_state.cpu = i; - cpumask_of_cpu_ptr_next(new_mask, i); - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(i)); ret = match_pr->throttling. acpi_processor_set_throttling( match_pr, t_state.target_state); diff --git a/drivers/firmware/dcdbas.c b/drivers/firmware/dcdbas.c index c66817e7717b..50a071f1c945 100644 --- a/drivers/firmware/dcdbas.c +++ b/drivers/firmware/dcdbas.c @@ -245,7 +245,6 @@ static ssize_t host_control_on_shutdown_store(struct device *dev, static int smi_request(struct smi_cmd *smi_cmd) { cpumask_t old_mask; - cpumask_of_cpu_ptr(new_mask, 0); int ret = 0; if (smi_cmd->magic != SMI_CMD_MAGIC) { @@ -256,7 +255,7 @@ static int smi_request(struct smi_cmd *smi_cmd) /* SMI requires CPU 0 */ old_mask = current->cpus_allowed; - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(0)); if (smp_processor_id() != 0) { dev_dbg(&dcdbas_pdev->dev, "%s: failed to get CPU 0\n", __func__); diff --git a/drivers/misc/sgi-xp/xpc_main.c b/drivers/misc/sgi-xp/xpc_main.c index 579b01ff82d4..c3b4227f48a5 100644 --- a/drivers/misc/sgi-xp/xpc_main.c +++ b/drivers/misc/sgi-xp/xpc_main.c @@ -229,11 +229,10 @@ xpc_hb_checker(void *ignore) int last_IRQ_count = 0; int new_IRQ_count; int force_IRQ = 0; - cpumask_of_cpu_ptr(cpumask, XPC_HB_CHECK_CPU); /* this thread was marked active by xpc_hb_init() */ - set_cpus_allowed_ptr(current, cpumask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(XPC_HB_CHECK_CPU)); /* set our heartbeating to other partitions into motion */ xpc_hb_check_timeout = jiffies + (xpc_hb_check_interval * HZ); diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c index 738b411ff2d3..ba9b2054ecbd 100644 --- a/kernel/stop_machine.c +++ b/kernel/stop_machine.c @@ -33,9 +33,8 @@ static int stopmachine(void *cpu) { int irqs_disabled = 0; int prepared = 0; - cpumask_of_cpu_ptr(cpumask, (int)(long)cpu); - set_cpus_allowed_ptr(current, cpumask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu((int)(long)cpu)); /* Ack: we are alive */ smp_mb(); /* Theoretically the ack = 0 might not be on this CPU yet. */ diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index bf43284d6855..80c4336f4188 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -196,12 +196,10 @@ static int tick_check_new_device(struct clock_event_device *newdev) struct tick_device *td; int cpu, ret = NOTIFY_OK; unsigned long flags; - cpumask_of_cpu_ptr_declare(cpumask); spin_lock_irqsave(&tick_device_lock, flags); cpu = smp_processor_id(); - cpumask_of_cpu_ptr_next(cpumask, cpu); if (!cpu_isset(cpu, newdev->cpumask)) goto out_bc; @@ -209,7 +207,7 @@ static int tick_check_new_device(struct clock_event_device *newdev) curdev = td->evtdev; /* cpu local device ? */ - if (!cpus_equal(newdev->cpumask, *cpumask)) { + if (!cpus_equal(newdev->cpumask, cpumask_of_cpu(cpu))) { /* * If the cpu affinity of the device interrupt can not @@ -222,7 +220,7 @@ static int tick_check_new_device(struct clock_event_device *newdev) * If we have a cpu local device already, do not replace it * by a non cpu local device */ - if (curdev && cpus_equal(curdev->cpumask, *cpumask)) + if (curdev && cpus_equal(curdev->cpumask, cpumask_of_cpu(cpu))) goto out_bc; } @@ -254,7 +252,7 @@ static int tick_check_new_device(struct clock_event_device *newdev) curdev = NULL; } clockevents_exchange_device(curdev, newdev); - tick_setup_device(td, newdev, cpu, cpumask); + tick_setup_device(td, newdev, cpu, &cpumask_of_cpu(cpu)); if (newdev->features & CLOCK_EVT_FEAT_ONESHOT) tick_oneshot_notify(); diff --git a/kernel/trace/trace_sysprof.c b/kernel/trace/trace_sysprof.c index ce2d723c10e1..bb948e52ce20 100644 --- a/kernel/trace/trace_sysprof.c +++ b/kernel/trace/trace_sysprof.c @@ -213,9 +213,7 @@ static void start_stack_timers(void) int cpu; for_each_online_cpu(cpu) { - cpumask_of_cpu_ptr(new_mask, cpu); - - set_cpus_allowed_ptr(current, new_mask); + set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); start_stack_timer(cpu); } set_cpus_allowed_ptr(current, &saved_mask); diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c index c4381d9516f6..0f8fc22ed103 100644 --- a/lib/smp_processor_id.c +++ b/lib/smp_processor_id.c @@ -11,7 +11,6 @@ notrace unsigned int debug_smp_processor_id(void) { unsigned long preempt_count = preempt_count(); int this_cpu = raw_smp_processor_id(); - cpumask_of_cpu_ptr_declare(this_mask); if (likely(preempt_count)) goto out; @@ -23,9 +22,7 @@ notrace unsigned int debug_smp_processor_id(void) * Kernel threads bound to a single CPU can safely use * smp_processor_id(): */ - cpumask_of_cpu_ptr_next(this_mask, this_cpu); - - if (cpus_equal(current->cpus_allowed, *this_mask)) + if (cpus_equal(current->cpus_allowed, cpumask_of_cpu(this_cpu))) goto out; /* diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 835d27413083..5a32cb7c4bb4 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -310,8 +310,7 @@ svc_pool_map_set_cpumask(struct task_struct *task, unsigned int pidx) switch (m->mode) { case SVC_POOL_PERCPU: { - cpumask_of_cpu_ptr(cpumask, node); - set_cpus_allowed_ptr(task, cpumask); + set_cpus_allowed_ptr(task, &cpumask_of_cpu(node)); break; } case SVC_POOL_PERNODE: -- cgit v1.2.3 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 ++- include/asm-x86/kvm_host.h | 1 + 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 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; diff --git a/include/asm-x86/kvm_host.h b/include/asm-x86/kvm_host.h index fdde0bedaa90..bc34dc21f178 100644 --- a/include/asm-x86/kvm_host.h +++ b/include/asm-x86/kvm_host.h @@ -556,6 +556,7 @@ int kvm_fix_hypercall(struct kvm_vcpu *vcpu); int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gva_t gva, u32 error_code); void kvm_enable_tdp(void); +void kvm_disable_tdp(void); int load_pdptrs(struct kvm_vcpu *vcpu, unsigned long cr3); int complete_pio(struct kvm_vcpu *vcpu); -- cgit v1.2.3 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 v1.2.3 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 v1.2.3 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 v1.2.3 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 v1.2.3 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 v1.2.3 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 v1.2.3 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 -------- include/linux/cpumask.h | 26 ++++++++- kernel/cpu.c | 128 +++++++---------------------------------- 3 files changed, 43 insertions(+), 134 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 diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h index 8fa3b6d4a320..96d0509fb8d8 100644 --- a/include/linux/cpumask.h +++ b/include/linux/cpumask.h @@ -265,10 +265,30 @@ static inline void __cpus_shift_left(cpumask_t *dstp, bitmap_shift_left(dstp->bits, srcp->bits, n, nbits); } +/* + * Special-case data structure for "single bit set only" constant CPU masks. + * + * We pre-generate all the 64 (or 32) possible bit positions, with enough + * padding to the left and the right, and return the constant pointer + * appropriately offset. + */ +extern const unsigned long + cpu_bit_bitmap[BITS_PER_LONG+1][BITS_TO_LONGS(NR_CPUS)]; + +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; +} + +/* + * In cases where we take the address of the cpumask immediately, + * gcc optimizes it out (it's a constant) and there's no huge stack + * variable created: + */ +#define cpumask_of_cpu(cpu) ({ *get_cpu_mask(cpu); }) -/* cpumask_of_cpu_map[] is in kernel/cpu.c */ -extern const cpumask_t *cpumask_of_cpu_map; -#define cpumask_of_cpu(cpu) (cpumask_of_cpu_map[cpu]) #define CPU_MASK_LAST_WORD BITMAP_LAST_WORD_MASK(NR_CPUS) diff --git a/kernel/cpu.c b/kernel/cpu.c index a35d8995dc8c..06a8358bb418 100644 --- a/kernel/cpu.c +++ b/kernel/cpu.c @@ -462,115 +462,27 @@ out: #endif /* CONFIG_SMP */ -/* 64 bits of zeros, for initializers. */ -#if BITS_PER_LONG == 32 -#define Z64 0, 0 -#else -#define Z64 0 -#endif +/* + * cpu_bit_bitmap[] is a special, "compressed" data structure that + * represents all NR_CPUS bits binary values of 1< 4 - CMI0(4), CMI0(5), CMI0(6), CMI0(7), -#endif -#if NR_CPUS > 8 - CMI0(8), CMI0(9), CMI0(10), CMI0(11), - CMI0(12), CMI0(13), CMI0(14), CMI0(15), -#endif -#if NR_CPUS > 16 - CMI0(16), CMI0(17), CMI0(18), CMI0(19), - CMI0(20), CMI0(21), CMI0(22), CMI0(23), - CMI0(24), CMI0(25), CMI0(26), CMI0(27), - CMI0(28), CMI0(29), CMI0(30), CMI0(31), -#endif -#if NR_CPUS > 32 -#if BITS_PER_LONG == 32 - CMI(32, 0), CMI(33, 0), CMI(34, 0), CMI(35, 0), - CMI(36, 0), CMI(37, 0), CMI(38, 0), CMI(39, 0), - CMI(40, 0), CMI(41, 0), CMI(42, 0), CMI(43, 0), - CMI(44, 0), CMI(45, 0), CMI(46, 0), CMI(47, 0), - CMI(48, 0), CMI(49, 0), CMI(50, 0), CMI(51, 0), - CMI(52, 0), CMI(53, 0), CMI(54, 0), CMI(55, 0), - CMI(56, 0), CMI(57, 0), CMI(58, 0), CMI(59, 0), - CMI(60, 0), CMI(61, 0), CMI(62, 0), CMI(63, 0), -#else - CMI0(32), CMI0(33), CMI0(34), CMI0(35), - CMI0(36), CMI0(37), CMI0(38), CMI0(39), - CMI0(40), CMI0(41), CMI0(42), CMI0(43), - CMI0(44), CMI0(45), CMI0(46), CMI0(47), - CMI0(48), CMI0(49), CMI0(50), CMI0(51), - CMI0(52), CMI0(53), CMI0(54), CMI0(55), - CMI0(56), CMI0(57), CMI0(58), CMI0(59), - CMI0(60), CMI0(61), CMI0(62), CMI0(63), -#endif /* BITS_PER_LONG == 64 */ -#endif -#if NR_CPUS > 64 - CMI64(64, Z64), -#endif -#if NR_CPUS > 128 - CMI64(128, Z64, Z64), CMI64(192, Z64, Z64, Z64), -#endif -#if NR_CPUS > 256 - CMI256(256, Z256), -#endif -#if NR_CPUS > 512 - CMI256(512, Z256, Z256), CMI256(768, Z256, Z256, Z256), -#endif -#if NR_CPUS > 1024 - CMI1024(1024, Z1024), -#endif -#if NR_CPUS > 2048 - CMI1024(2048, Z1024, Z1024), CMI1024(3072, Z1024, Z1024, Z1024), -#endif -#if NR_CPUS > 4096 -#error NR_CPUS too big. Fix initializers or set CONFIG_HAVE_CPUMASK_OF_CPU_MAP +const unsigned long cpu_bit_bitmap[BITS_PER_LONG+1][BITS_TO_LONGS(NR_CPUS)] = { + + MASK_DECLARE_8(0), MASK_DECLARE_8(8), + MASK_DECLARE_8(16), MASK_DECLARE_8(24), +#if BITS_PER_LONG > 32 + MASK_DECLARE_8(32), MASK_DECLARE_8(40), + MASK_DECLARE_8(48), MASK_DECLARE_8(56), #endif }; - -const cpumask_t *cpumask_of_cpu_map = cpumask_map; - -EXPORT_SYMBOL_GPL(cpumask_of_cpu_map); +EXPORT_SYMBOL_GPL(cpu_bit_bitmap); -- cgit v1.2.3 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 v1.2.3 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 + include/linux/mm_types.h | 4 + include/linux/mmu_notifier.h | 279 +++++++++++++++++++++++++++++++++++++++++++ kernel/fork.c | 3 + mm/Kconfig | 3 + mm/Makefile | 1 + mm/filemap_xip.c | 3 +- mm/fremap.c | 3 + mm/hugetlb.c | 3 + mm/memory.c | 35 +++++- mm/mmap.c | 2 + mm/mmu_notifier.c | 277 ++++++++++++++++++++++++++++++++++++++++++ mm/mprotect.c | 3 + mm/mremap.c | 6 + mm/rmap.c | 13 +- 15 files changed, 623 insertions(+), 13 deletions(-) create mode 100644 include/linux/mmu_notifier.h create mode 100644 mm/mmu_notifier.c (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 diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 746f975b58ef..386edbe2cb4e 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -10,6 +10,7 @@ #include #include #include +#include #include #include @@ -253,6 +254,9 @@ struct mm_struct { struct file *exe_file; unsigned long num_exe_file_vmas; #endif +#ifdef CONFIG_MMU_NOTIFIER + struct mmu_notifier_mm *mmu_notifier_mm; +#endif }; #endif /* _LINUX_MM_TYPES_H */ diff --git a/include/linux/mmu_notifier.h b/include/linux/mmu_notifier.h new file mode 100644 index 000000000000..b77486d152cd --- /dev/null +++ b/include/linux/mmu_notifier.h @@ -0,0 +1,279 @@ +#ifndef _LINUX_MMU_NOTIFIER_H +#define _LINUX_MMU_NOTIFIER_H + +#include +#include +#include + +struct mmu_notifier; +struct mmu_notifier_ops; + +#ifdef CONFIG_MMU_NOTIFIER + +/* + * The mmu notifier_mm structure is allocated and installed in + * mm->mmu_notifier_mm inside the mm_take_all_locks() protected + * critical section and it's released only when mm_count reaches zero + * in mmdrop(). + */ +struct mmu_notifier_mm { + /* all mmu notifiers registerd in this mm are queued in this list */ + struct hlist_head list; + /* to serialize the list modifications and hlist_unhashed */ + spinlock_t lock; +}; + +struct mmu_notifier_ops { + /* + * Called either by mmu_notifier_unregister or when the mm is + * being destroyed by exit_mmap, always before all pages are + * freed. This can run concurrently with other mmu notifier + * methods (the ones invoked outside the mm context) and it + * should tear down all secondary mmu mappings and freeze the + * secondary mmu. If this method isn't implemented you've to + * be sure that nothing could possibly write to the pages + * through the secondary mmu by the time the last thread with + * tsk->mm == mm exits. + * + * As side note: the pages freed after ->release returns could + * be immediately reallocated by the gart at an alias physical + * address with a different cache model, so if ->release isn't + * implemented because all _software_ driven memory accesses + * through the secondary mmu are terminated by the time the + * last thread of this mm quits, you've also to be sure that + * speculative _hardware_ operations can't allocate dirty + * cachelines in the cpu that could not be snooped and made + * coherent with the other read and write operations happening + * through the gart alias address, so leading to memory + * corruption. + */ + void (*release)(struct mmu_notifier *mn, + struct mm_struct *mm); + + /* + * clear_flush_young is called after the VM is + * test-and-clearing the young/accessed bitflag in the + * pte. This way the VM will provide proper aging to the + * accesses to the page through the secondary MMUs and not + * only to the ones through the Linux pte. + */ + int (*clear_flush_young)(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long address); + + /* + * Before this is invoked any secondary MMU is still ok to + * read/write to the page previously pointed to by the Linux + * pte because the page hasn't been freed yet and it won't be + * freed until this returns. If required set_page_dirty has to + * be called internally to this method. + */ + void (*invalidate_page)(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long address); + + /* + * invalidate_range_start() and invalidate_range_end() must be + * paired and are called only when the mmap_sem and/or the + * locks protecting the reverse maps are held. The subsystem + * must guarantee that no additional references are taken to + * the pages in the range established between the call to + * invalidate_range_start() and the matching call to + * invalidate_range_end(). + * + * Invalidation of multiple concurrent ranges may be + * optionally permitted by the driver. Either way the + * establishment of sptes is forbidden in the range passed to + * invalidate_range_begin/end for the whole duration of the + * invalidate_range_begin/end critical section. + * + * invalidate_range_start() is called when all pages in the + * range are still mapped and have at least a refcount of one. + * + * invalidate_range_end() is called when all pages in the + * range have been unmapped and the pages have been freed by + * the VM. + * + * The VM will remove the page table entries and potentially + * the page between invalidate_range_start() and + * invalidate_range_end(). If the page must not be freed + * because of pending I/O or other circumstances then the + * invalidate_range_start() callback (or the initial mapping + * by the driver) must make sure that the refcount is kept + * elevated. + * + * If the driver increases the refcount when the pages are + * initially mapped into an address space then either + * invalidate_range_start() or invalidate_range_end() may + * decrease the refcount. If the refcount is decreased on + * invalidate_range_start() then the VM can free pages as page + * table entries are removed. If the refcount is only + * droppped on invalidate_range_end() then the driver itself + * will drop the last refcount but it must take care to flush + * any secondary tlb before doing the final free on the + * page. Pages will no longer be referenced by the linux + * address space but may still be referenced by sptes until + * the last refcount is dropped. + */ + void (*invalidate_range_start)(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long start, unsigned long end); + void (*invalidate_range_end)(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long start, unsigned long end); +}; + +/* + * The notifier chains are protected by mmap_sem and/or the reverse map + * semaphores. Notifier chains are only changed when all reverse maps and + * the mmap_sem locks are taken. + * + * Therefore notifier chains can only be traversed when either + * + * 1. mmap_sem is held. + * 2. One of the reverse map locks is held (i_mmap_lock or anon_vma->lock). + * 3. No other concurrent thread can access the list (release) + */ +struct mmu_notifier { + struct hlist_node hlist; + const struct mmu_notifier_ops *ops; +}; + +static inline int mm_has_notifiers(struct mm_struct *mm) +{ + return unlikely(mm->mmu_notifier_mm); +} + +extern int mmu_notifier_register(struct mmu_notifier *mn, + struct mm_struct *mm); +extern int __mmu_notifier_register(struct mmu_notifier *mn, + struct mm_struct *mm); +extern void mmu_notifier_unregister(struct mmu_notifier *mn, + struct mm_struct *mm); +extern void __mmu_notifier_mm_destroy(struct mm_struct *mm); +extern void __mmu_notifier_release(struct mm_struct *mm); +extern int __mmu_notifier_clear_flush_young(struct mm_struct *mm, + unsigned long address); +extern void __mmu_notifier_invalidate_page(struct mm_struct *mm, + unsigned long address); +extern void __mmu_notifier_invalidate_range_start(struct mm_struct *mm, + unsigned long start, unsigned long end); +extern void __mmu_notifier_invalidate_range_end(struct mm_struct *mm, + unsigned long start, unsigned long end); + +static inline void mmu_notifier_release(struct mm_struct *mm) +{ + if (mm_has_notifiers(mm)) + __mmu_notifier_release(mm); +} + +static inline int mmu_notifier_clear_flush_young(struct mm_struct *mm, + unsigned long address) +{ + if (mm_has_notifiers(mm)) + return __mmu_notifier_clear_flush_young(mm, address); + return 0; +} + +static inline void mmu_notifier_invalidate_page(struct mm_struct *mm, + unsigned long address) +{ + if (mm_has_notifiers(mm)) + __mmu_notifier_invalidate_page(mm, address); +} + +static inline void mmu_notifier_invalidate_range_start(struct mm_struct *mm, + unsigned long start, unsigned long end) +{ + if (mm_has_notifiers(mm)) + __mmu_notifier_invalidate_range_start(mm, start, end); +} + +static inline void mmu_notifier_invalidate_range_end(struct mm_struct *mm, + unsigned long start, unsigned long end) +{ + if (mm_has_notifiers(mm)) + __mmu_notifier_invalidate_range_end(mm, start, end); +} + +static inline void mmu_notifier_mm_init(struct mm_struct *mm) +{ + mm->mmu_notifier_mm = NULL; +} + +static inline void mmu_notifier_mm_destroy(struct mm_struct *mm) +{ + if (mm_has_notifiers(mm)) + __mmu_notifier_mm_destroy(mm); +} + +/* + * These two macros will sometime replace ptep_clear_flush. + * ptep_clear_flush is impleemnted as macro itself, so this also is + * implemented as a macro until ptep_clear_flush will converted to an + * inline function, to diminish the risk of compilation failure. The + * invalidate_page method over time can be moved outside the PT lock + * and these two macros can be later removed. + */ +#define ptep_clear_flush_notify(__vma, __address, __ptep) \ +({ \ + pte_t __pte; \ + struct vm_area_struct *___vma = __vma; \ + unsigned long ___address = __address; \ + __pte = ptep_clear_flush(___vma, ___address, __ptep); \ + mmu_notifier_invalidate_page(___vma->vm_mm, ___address); \ + __pte; \ +}) + +#define ptep_clear_flush_young_notify(__vma, __address, __ptep) \ +({ \ + int __young; \ + struct vm_area_struct *___vma = __vma; \ + unsigned long ___address = __address; \ + __young = ptep_clear_flush_young(___vma, ___address, __ptep); \ + __young |= mmu_notifier_clear_flush_young(___vma->vm_mm, \ + ___address); \ + __young; \ +}) + +#else /* CONFIG_MMU_NOTIFIER */ + +static inline void mmu_notifier_release(struct mm_struct *mm) +{ +} + +static inline int mmu_notifier_clear_flush_young(struct mm_struct *mm, + unsigned long address) +{ + return 0; +} + +static inline void mmu_notifier_invalidate_page(struct mm_struct *mm, + unsigned long address) +{ +} + +static inline void mmu_notifier_invalidate_range_start(struct mm_struct *mm, + unsigned long start, unsigned long end) +{ +} + +static inline void mmu_notifier_invalidate_range_end(struct mm_struct *mm, + unsigned long start, unsigned long end) +{ +} + +static inline void mmu_notifier_mm_init(struct mm_struct *mm) +{ +} + +static inline void mmu_notifier_mm_destroy(struct mm_struct *mm) +{ +} + +#define ptep_clear_flush_young_notify ptep_clear_flush_young +#define ptep_clear_flush_notify ptep_clear_flush + +#endif /* CONFIG_MMU_NOTIFIER */ + +#endif /* _LINUX_MMU_NOTIFIER_H */ diff --git a/kernel/fork.c b/kernel/fork.c index 8214ba7c8bb1..7ce2ebe84796 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include #include @@ -414,6 +415,7 @@ static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p) if (likely(!mm_alloc_pgd(mm))) { mm->def_flags = 0; + mmu_notifier_mm_init(mm); return mm; } @@ -446,6 +448,7 @@ void __mmdrop(struct mm_struct *mm) BUG_ON(mm == &init_mm); mm_free_pgd(mm); destroy_context(mm); + mmu_notifier_mm_destroy(mm); free_mm(mm); } EXPORT_SYMBOL_GPL(__mmdrop); diff --git a/mm/Kconfig b/mm/Kconfig index efee5d379df4..446c6588c753 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -208,3 +208,6 @@ config NR_QUICK config VIRT_TO_BUS def_bool y depends on !ARCH_NO_VIRT_TO_BUS + +config MMU_NOTIFIER + bool diff --git a/mm/Makefile b/mm/Makefile index 06ca2381fef1..da4ccf015aea 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -25,6 +25,7 @@ obj-$(CONFIG_SHMEM) += shmem.o obj-$(CONFIG_TMPFS_POSIX_ACL) += shmem_acl.o obj-$(CONFIG_TINY_SHMEM) += tiny-shmem.o obj-$(CONFIG_SLOB) += slob.o +obj-$(CONFIG_MMU_NOTIFIER) += mmu_notifier.o obj-$(CONFIG_SLAB) += slab.o obj-$(CONFIG_SLUB) += slub.o obj-$(CONFIG_MEMORY_HOTPLUG) += memory_hotplug.o diff --git a/mm/filemap_xip.c b/mm/filemap_xip.c index 98a3f31ccd6a..380ab402d711 100644 --- a/mm/filemap_xip.c +++ b/mm/filemap_xip.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -188,7 +189,7 @@ __xip_unmap (struct address_space * mapping, if (pte) { /* Nuke the page table entry. */ flush_cache_page(vma, address, pte_pfn(*pte)); - pteval = ptep_clear_flush(vma, address, pte); + pteval = ptep_clear_flush_notify(vma, address, pte); page_remove_rmap(page, vma); dec_mm_counter(mm, file_rss); BUG_ON(pte_dirty(pteval)); diff --git a/mm/fremap.c b/mm/fremap.c index 07a9c82ce1a3..7881638e4a12 100644 --- a/mm/fremap.c +++ b/mm/fremap.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -214,7 +215,9 @@ asmlinkage long sys_remap_file_pages(unsigned long start, unsigned long size, spin_unlock(&mapping->i_mmap_lock); } + mmu_notifier_invalidate_range_start(mm, start, start + size); err = populate_range(mm, vma, start, size, pgoff); + mmu_notifier_invalidate_range_end(mm, start, start + size); if (!err && !(flags & MAP_NONBLOCK)) { if (unlikely(has_write_lock)) { downgrade_write(&mm->mmap_sem); diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 3be79dc18c5c..80eb0d31d0d3 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include #include @@ -1672,6 +1673,7 @@ void __unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start, BUG_ON(start & ~huge_page_mask(h)); BUG_ON(end & ~huge_page_mask(h)); + mmu_notifier_invalidate_range_start(mm, start, end); spin_lock(&mm->page_table_lock); for (address = start; address < end; address += sz) { ptep = huge_pte_offset(mm, address); @@ -1713,6 +1715,7 @@ void __unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start, } spin_unlock(&mm->page_table_lock); flush_tlb_range(vma, start, end); + mmu_notifier_invalidate_range_end(mm, start, end); list_for_each_entry_safe(page, tmp, &page_list, lru) { list_del(&page->lru); put_page(page); diff --git a/mm/memory.c b/mm/memory.c index a8ca04faaea6..67f0ab9077d9 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -51,6 +51,7 @@ #include #include #include +#include #include #include @@ -652,6 +653,7 @@ int copy_page_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, unsigned long next; unsigned long addr = vma->vm_start; unsigned long end = vma->vm_end; + int ret; /* * Don't copy ptes where a page fault will fill them correctly. @@ -667,17 +669,33 @@ int copy_page_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, if (is_vm_hugetlb_page(vma)) return copy_hugetlb_page_range(dst_mm, src_mm, vma); + /* + * We need to invalidate the secondary MMU mappings only when + * there could be a permission downgrade on the ptes of the + * parent mm. And a permission downgrade will only happen if + * is_cow_mapping() returns true. + */ + if (is_cow_mapping(vma->vm_flags)) + mmu_notifier_invalidate_range_start(src_mm, addr, end); + + ret = 0; dst_pgd = pgd_offset(dst_mm, addr); src_pgd = pgd_offset(src_mm, addr); do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(src_pgd)) continue; - if (copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd, - vma, addr, next)) - return -ENOMEM; + if (unlikely(copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd, + vma, addr, next))) { + ret = -ENOMEM; + break; + } } while (dst_pgd++, src_pgd++, addr = next, addr != end); - return 0; + + if (is_cow_mapping(vma->vm_flags)) + mmu_notifier_invalidate_range_end(src_mm, + vma->vm_start, end); + return ret; } static unsigned long zap_pte_range(struct mmu_gather *tlb, @@ -881,7 +899,9 @@ unsigned long unmap_vmas(struct mmu_gather **tlbp, unsigned long start = start_addr; spinlock_t *i_mmap_lock = details? details->i_mmap_lock: NULL; int fullmm = (*tlbp)->fullmm; + struct mm_struct *mm = vma->vm_mm; + mmu_notifier_invalidate_range_start(mm, start_addr, end_addr); for ( ; vma && vma->vm_start < end_addr; vma = vma->vm_next) { unsigned long end; @@ -946,6 +966,7 @@ unsigned long unmap_vmas(struct mmu_gather **tlbp, } } out: + mmu_notifier_invalidate_range_end(mm, start_addr, end_addr); return start; /* which is now the end (or restart) address */ } @@ -1616,10 +1637,11 @@ int apply_to_page_range(struct mm_struct *mm, unsigned long addr, { pgd_t *pgd; unsigned long next; - unsigned long end = addr + size; + unsigned long start = addr, end = addr + size; int err; BUG_ON(addr >= end); + mmu_notifier_invalidate_range_start(mm, start, end); pgd = pgd_offset(mm, addr); do { next = pgd_addr_end(addr, end); @@ -1627,6 +1649,7 @@ int apply_to_page_range(struct mm_struct *mm, unsigned long addr, if (err) break; } while (pgd++, addr = next, addr != end); + mmu_notifier_invalidate_range_end(mm, start, end); return err; } EXPORT_SYMBOL_GPL(apply_to_page_range); @@ -1839,7 +1862,7 @@ gotten: * seen in the presence of one thread doing SMC and another * thread doing COW. */ - ptep_clear_flush(vma, address, page_table); + ptep_clear_flush_notify(vma, address, page_table); set_pte_at(mm, address, page_table, entry); update_mmu_cache(vma, address, entry); lru_cache_add_active(new_page); diff --git a/mm/mmap.c b/mm/mmap.c index e5f9cb83d6d4..245c3d69067b 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include @@ -2061,6 +2062,7 @@ void exit_mmap(struct mm_struct *mm) /* mm's last user has gone, and its about to be pulled down */ arch_exit_mmap(mm); + mmu_notifier_release(mm); lru_add_drain(); flush_cache_mm(mm); diff --git a/mm/mmu_notifier.c b/mm/mmu_notifier.c new file mode 100644 index 000000000000..5f4ef0250bee --- /dev/null +++ b/mm/mmu_notifier.c @@ -0,0 +1,277 @@ +/* + * linux/mm/mmu_notifier.c + * + * Copyright (C) 2008 Qumranet, Inc. + * Copyright (C) 2008 SGI + * Christoph Lameter + * + * This work is licensed under the terms of the GNU GPL, version 2. See + * the COPYING file in the top-level directory. + */ + +#include +#include +#include +#include +#include +#include +#include + +/* + * This function can't run concurrently against mmu_notifier_register + * because mm->mm_users > 0 during mmu_notifier_register and exit_mmap + * runs with mm_users == 0. Other tasks may still invoke mmu notifiers + * in parallel despite there being no task using this mm any more, + * through the vmas outside of the exit_mmap context, such as with + * vmtruncate. This serializes against mmu_notifier_unregister with + * the mmu_notifier_mm->lock in addition to RCU and it serializes + * against the other mmu notifiers with RCU. struct mmu_notifier_mm + * can't go away from under us as exit_mmap holds an mm_count pin + * itself. + */ +void __mmu_notifier_release(struct mm_struct *mm) +{ + struct mmu_notifier *mn; + + spin_lock(&mm->mmu_notifier_mm->lock); + while (unlikely(!hlist_empty(&mm->mmu_notifier_mm->list))) { + mn = hlist_entry(mm->mmu_notifier_mm->list.first, + struct mmu_notifier, + hlist); + /* + * We arrived before mmu_notifier_unregister so + * mmu_notifier_unregister will do nothing other than + * to wait ->release to finish and + * mmu_notifier_unregister to return. + */ + hlist_del_init_rcu(&mn->hlist); + /* + * RCU here will block mmu_notifier_unregister until + * ->release returns. + */ + rcu_read_lock(); + spin_unlock(&mm->mmu_notifier_mm->lock); + /* + * if ->release runs before mmu_notifier_unregister it + * must be handled as it's the only way for the driver + * to flush all existing sptes and stop the driver + * from establishing any more sptes before all the + * pages in the mm are freed. + */ + if (mn->ops->release) + mn->ops->release(mn, mm); + rcu_read_unlock(); + spin_lock(&mm->mmu_notifier_mm->lock); + } + spin_unlock(&mm->mmu_notifier_mm->lock); + + /* + * synchronize_rcu here prevents mmu_notifier_release to + * return to exit_mmap (which would proceed freeing all pages + * in the mm) until the ->release method returns, if it was + * invoked by mmu_notifier_unregister. + * + * The mmu_notifier_mm can't go away from under us because one + * mm_count is hold by exit_mmap. + */ + synchronize_rcu(); +} + +/* + * If no young bitflag is supported by the hardware, ->clear_flush_young can + * unmap the address and return 1 or 0 depending if the mapping previously + * existed or not. + */ +int __mmu_notifier_clear_flush_young(struct mm_struct *mm, + unsigned long address) +{ + struct mmu_notifier *mn; + struct hlist_node *n; + int young = 0; + + rcu_read_lock(); + hlist_for_each_entry_rcu(mn, n, &mm->mmu_notifier_mm->list, hlist) { + if (mn->ops->clear_flush_young) + young |= mn->ops->clear_flush_young(mn, mm, address); + } + rcu_read_unlock(); + + return young; +} + +void __mmu_notifier_invalidate_page(struct mm_struct *mm, + unsigned long address) +{ + struct mmu_notifier *mn; + struct hlist_node *n; + + rcu_read_lock(); + hlist_for_each_entry_rcu(mn, n, &mm->mmu_notifier_mm->list, hlist) { + if (mn->ops->invalidate_page) + mn->ops->invalidate_page(mn, mm, address); + } + rcu_read_unlock(); +} + +void __mmu_notifier_invalidate_range_start(struct mm_struct *mm, + unsigned long start, unsigned long end) +{ + struct mmu_notifier *mn; + struct hlist_node *n; + + rcu_read_lock(); + hlist_for_each_entry_rcu(mn, n, &mm->mmu_notifier_mm->list, hlist) { + if (mn->ops->invalidate_range_start) + mn->ops->invalidate_range_start(mn, mm, start, end); + } + rcu_read_unlock(); +} + +void __mmu_notifier_invalidate_range_end(struct mm_struct *mm, + unsigned long start, unsigned long end) +{ + struct mmu_notifier *mn; + struct hlist_node *n; + + rcu_read_lock(); + hlist_for_each_entry_rcu(mn, n, &mm->mmu_notifier_mm->list, hlist) { + if (mn->ops->invalidate_range_end) + mn->ops->invalidate_range_end(mn, mm, start, end); + } + rcu_read_unlock(); +} + +static int do_mmu_notifier_register(struct mmu_notifier *mn, + struct mm_struct *mm, + int take_mmap_sem) +{ + struct mmu_notifier_mm *mmu_notifier_mm; + int ret; + + BUG_ON(atomic_read(&mm->mm_users) <= 0); + + ret = -ENOMEM; + mmu_notifier_mm = kmalloc(sizeof(struct mmu_notifier_mm), GFP_KERNEL); + if (unlikely(!mmu_notifier_mm)) + goto out; + + if (take_mmap_sem) + down_write(&mm->mmap_sem); + ret = mm_take_all_locks(mm); + if (unlikely(ret)) + goto out_cleanup; + + if (!mm_has_notifiers(mm)) { + INIT_HLIST_HEAD(&mmu_notifier_mm->list); + spin_lock_init(&mmu_notifier_mm->lock); + mm->mmu_notifier_mm = mmu_notifier_mm; + mmu_notifier_mm = NULL; + } + atomic_inc(&mm->mm_count); + + /* + * Serialize the update against mmu_notifier_unregister. A + * side note: mmu_notifier_release can't run concurrently with + * us because we hold the mm_users pin (either implicitly as + * current->mm or explicitly with get_task_mm() or similar). + * We can't race against any other mmu notifier method either + * thanks to mm_take_all_locks(). + */ + spin_lock(&mm->mmu_notifier_mm->lock); + hlist_add_head(&mn->hlist, &mm->mmu_notifier_mm->list); + spin_unlock(&mm->mmu_notifier_mm->lock); + + mm_drop_all_locks(mm); +out_cleanup: + if (take_mmap_sem) + up_write(&mm->mmap_sem); + /* kfree() does nothing if mmu_notifier_mm is NULL */ + kfree(mmu_notifier_mm); +out: + BUG_ON(atomic_read(&mm->mm_users) <= 0); + return ret; +} + +/* + * Must not hold mmap_sem nor any other VM related lock when calling + * this registration function. Must also ensure mm_users can't go down + * to zero while this runs to avoid races with mmu_notifier_release, + * so mm has to be current->mm or the mm should be pinned safely such + * as with get_task_mm(). If the mm is not current->mm, the mm_users + * pin should be released by calling mmput after mmu_notifier_register + * returns. mmu_notifier_unregister must be always called to + * unregister the notifier. mm_count is automatically pinned to allow + * mmu_notifier_unregister to safely run at any time later, before or + * after exit_mmap. ->release will always be called before exit_mmap + * frees the pages. + */ +int mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm) +{ + return do_mmu_notifier_register(mn, mm, 1); +} +EXPORT_SYMBOL_GPL(mmu_notifier_register); + +/* + * Same as mmu_notifier_register but here the caller must hold the + * mmap_sem in write mode. + */ +int __mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm) +{ + return do_mmu_notifier_register(mn, mm, 0); +} +EXPORT_SYMBOL_GPL(__mmu_notifier_register); + +/* this is called after the last mmu_notifier_unregister() returned */ +void __mmu_notifier_mm_destroy(struct mm_struct *mm) +{ + BUG_ON(!hlist_empty(&mm->mmu_notifier_mm->list)); + kfree(mm->mmu_notifier_mm); + mm->mmu_notifier_mm = LIST_POISON1; /* debug */ +} + +/* + * This releases the mm_count pin automatically and frees the mm + * structure if it was the last user of it. It serializes against + * running mmu notifiers with RCU and against mmu_notifier_unregister + * with the unregister lock + RCU. All sptes must be dropped before + * calling mmu_notifier_unregister. ->release or any other notifier + * method may be invoked concurrently with mmu_notifier_unregister, + * and only after mmu_notifier_unregister returned we're guaranteed + * that ->release or any other method can't run anymore. + */ +void mmu_notifier_unregister(struct mmu_notifier *mn, struct mm_struct *mm) +{ + BUG_ON(atomic_read(&mm->mm_count) <= 0); + + spin_lock(&mm->mmu_notifier_mm->lock); + if (!hlist_unhashed(&mn->hlist)) { + hlist_del_rcu(&mn->hlist); + + /* + * RCU here will force exit_mmap to wait ->release to finish + * before freeing the pages. + */ + rcu_read_lock(); + spin_unlock(&mm->mmu_notifier_mm->lock); + /* + * exit_mmap will block in mmu_notifier_release to + * guarantee ->release is called before freeing the + * pages. + */ + if (mn->ops->release) + mn->ops->release(mn, mm); + rcu_read_unlock(); + } else + spin_unlock(&mm->mmu_notifier_mm->lock); + + /* + * Wait any running method to finish, of course including + * ->release if it was run by mmu_notifier_relase instead of us. + */ + synchronize_rcu(); + + BUG_ON(atomic_read(&mm->mm_count) <= 0); + + mmdrop(mm); +} +EXPORT_SYMBOL_GPL(mmu_notifier_unregister); diff --git a/mm/mprotect.c b/mm/mprotect.c index abd645a3b0a0..fded06f923f4 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include @@ -203,10 +204,12 @@ success: dirty_accountable = 1; } + mmu_notifier_invalidate_range_start(mm, start, end); if (is_vm_hugetlb_page(vma)) hugetlb_change_protection(vma, start, end, vma->vm_page_prot); else change_protection(vma, start, end, vma->vm_page_prot, dirty_accountable); + mmu_notifier_invalidate_range_end(mm, start, end); vm_stat_account(mm, oldflags, vma->vm_file, -nrpages); vm_stat_account(mm, newflags, vma->vm_file, nrpages); return 0; diff --git a/mm/mremap.c b/mm/mremap.c index 08e3c7f2bd15..1a7743923c8c 100644 --- a/mm/mremap.c +++ b/mm/mremap.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -74,7 +75,11 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, struct mm_struct *mm = vma->vm_mm; pte_t *old_pte, *new_pte, pte; spinlock_t *old_ptl, *new_ptl; + unsigned long old_start; + old_start = old_addr; + mmu_notifier_invalidate_range_start(vma->vm_mm, + old_start, old_end); if (vma->vm_file) { /* * Subtle point from Rajesh Venkatasubramanian: before @@ -116,6 +121,7 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, pte_unmap_unlock(old_pte - 1, old_ptl); if (mapping) spin_unlock(&mapping->i_mmap_lock); + mmu_notifier_invalidate_range_end(vma->vm_mm, old_start, old_end); } #define LATENCY_LIMIT (64 * PAGE_SIZE) diff --git a/mm/rmap.c b/mm/rmap.c index 39ae5a9bf382..99bc3f9cd796 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -49,6 +49,7 @@ #include #include #include +#include #include @@ -287,7 +288,7 @@ static int page_referenced_one(struct page *page, if (vma->vm_flags & VM_LOCKED) { referenced++; *mapcount = 1; /* break early from loop */ - } else if (ptep_clear_flush_young(vma, address, pte)) + } else if (ptep_clear_flush_young_notify(vma, address, pte)) referenced++; /* Pretend the page is referenced if the task has the @@ -457,7 +458,7 @@ static int page_mkclean_one(struct page *page, struct vm_area_struct *vma) pte_t entry; flush_cache_page(vma, address, pte_pfn(*pte)); - entry = ptep_clear_flush(vma, address, pte); + entry = ptep_clear_flush_notify(vma, address, pte); entry = pte_wrprotect(entry); entry = pte_mkclean(entry); set_pte_at(mm, address, pte, entry); @@ -705,14 +706,14 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma, * skipped over this mm) then we should reactivate it. */ if (!migration && ((vma->vm_flags & VM_LOCKED) || - (ptep_clear_flush_young(vma, address, pte)))) { + (ptep_clear_flush_young_notify(vma, address, pte)))) { ret = SWAP_FAIL; goto out_unmap; } /* Nuke the page table entry. */ flush_cache_page(vma, address, page_to_pfn(page)); - pteval = ptep_clear_flush(vma, address, pte); + pteval = ptep_clear_flush_notify(vma, address, pte); /* Move the dirty bit to the physical page now the pte is gone. */ if (pte_dirty(pteval)) @@ -837,12 +838,12 @@ static void try_to_unmap_cluster(unsigned long cursor, page = vm_normal_page(vma, address, *pte); BUG_ON(!page || PageAnon(page)); - if (ptep_clear_flush_young(vma, address, pte)) + if (ptep_clear_flush_young_notify(vma, address, pte)) continue; /* Nuke the page table entry. */ flush_cache_page(vma, address, pte_pfn(*pte)); - pteval = ptep_clear_flush(vma, address, pte); + pteval = ptep_clear_flush_notify(vma, address, pte); /* If nonlinear, store the file page offset in the pte. */ if (page->index != linear_page_index(vma, address)) -- cgit v1.2.3 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 v1.2.3 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 v1.2.3