diff options
Diffstat (limited to 'arch/x86/kvm/mmu.c')
| -rw-r--r-- | arch/x86/kvm/mmu.c | 1782 |
1 files changed, 1177 insertions, 605 deletions
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 224b02c3cda..931467881da 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -22,6 +22,7 @@ #include "mmu.h" #include "x86.h" #include "kvm_cache_regs.h" +#include "cpuid.h" #include <linux/kvm_host.h> #include <linux/types.h> @@ -90,7 +91,7 @@ module_param(dbg, bool, 0644); #define PTE_PREFETCH_NUM 8 -#define PT_FIRST_AVAIL_BITS_SHIFT 9 +#define PT_FIRST_AVAIL_BITS_SHIFT 10 #define PT64_SECOND_AVAIL_BITS_SHIFT 52 #define PT64_LEVEL_BITS 9 @@ -132,10 +133,8 @@ module_param(dbg, bool, 0644); (PAGE_MASK & ~((1ULL << (PAGE_SHIFT + (((level) - 1) \ * PT32_LEVEL_BITS))) - 1)) -#define PT64_PERM_MASK (PT_PRESENT_MASK | PT_WRITABLE_MASK | PT_USER_MASK \ - | PT64_NX_MASK) - -#define PTE_LIST_EXT 4 +#define PT64_PERM_MASK (PT_PRESENT_MASK | PT_WRITABLE_MASK | shadow_user_mask \ + | shadow_x_mask | shadow_nx_mask) #define ACC_EXEC_MASK 1 #define ACC_WRITE_MASK PT_WRITABLE_MASK @@ -147,10 +146,14 @@ module_param(dbg, bool, 0644); #define CREATE_TRACE_POINTS #include "mmutrace.h" -#define SPTE_HOST_WRITEABLE (1ULL << PT_FIRST_AVAIL_BITS_SHIFT) +#define SPTE_HOST_WRITEABLE (1ULL << PT_FIRST_AVAIL_BITS_SHIFT) +#define SPTE_MMU_WRITEABLE (1ULL << (PT_FIRST_AVAIL_BITS_SHIFT + 1)) #define SHADOW_PT_INDEX(addr, level) PT64_INDEX(addr, level) +/* make pte_list_desc fit well in cache line */ +#define PTE_LIST_EXT 3 + struct pte_list_desc { u64 *sptes[PTE_LIST_EXT]; struct pte_list_desc *more; @@ -187,6 +190,7 @@ static u64 __read_mostly shadow_dirty_mask; static u64 __read_mostly shadow_mmio_mask; static void mmu_spte_set(u64 *sptep, u64 spte); +static void mmu_free_roots(struct kvm_vcpu *vcpu); void kvm_mmu_set_mmio_spte_mask(u64 mmio_mask) { @@ -194,12 +198,63 @@ void kvm_mmu_set_mmio_spte_mask(u64 mmio_mask) } EXPORT_SYMBOL_GPL(kvm_mmu_set_mmio_spte_mask); -static void mark_mmio_spte(u64 *sptep, u64 gfn, unsigned access) +/* + * spte bits of bit 3 ~ bit 11 are used as low 9 bits of generation number, + * the bits of bits 52 ~ bit 61 are used as high 10 bits of generation + * number. + */ +#define MMIO_SPTE_GEN_LOW_SHIFT 3 +#define MMIO_SPTE_GEN_HIGH_SHIFT 52 + +#define MMIO_GEN_SHIFT 19 +#define MMIO_GEN_LOW_SHIFT 9 +#define MMIO_GEN_LOW_MASK ((1 << MMIO_GEN_LOW_SHIFT) - 1) +#define MMIO_GEN_MASK ((1 << MMIO_GEN_SHIFT) - 1) +#define MMIO_MAX_GEN ((1 << MMIO_GEN_SHIFT) - 1) + +static u64 generation_mmio_spte_mask(unsigned int gen) +{ + u64 mask; + + WARN_ON(gen > MMIO_MAX_GEN); + + mask = (gen & MMIO_GEN_LOW_MASK) << MMIO_SPTE_GEN_LOW_SHIFT; + mask |= ((u64)gen >> MMIO_GEN_LOW_SHIFT) << MMIO_SPTE_GEN_HIGH_SHIFT; + return mask; +} + +static unsigned int get_mmio_spte_generation(u64 spte) +{ + unsigned int gen; + + spte &= ~shadow_mmio_mask; + + gen = (spte >> MMIO_SPTE_GEN_LOW_SHIFT) & MMIO_GEN_LOW_MASK; + gen |= (spte >> MMIO_SPTE_GEN_HIGH_SHIFT) << MMIO_GEN_LOW_SHIFT; + return gen; +} + +static unsigned int kvm_current_mmio_generation(struct kvm *kvm) +{ + /* + * Init kvm generation close to MMIO_MAX_GEN to easily test the + * code of handling generation number wrap-around. + */ + return (kvm_memslots(kvm)->generation + + MMIO_MAX_GEN - 150) & MMIO_GEN_MASK; +} + +static void mark_mmio_spte(struct kvm *kvm, u64 *sptep, u64 gfn, + unsigned access) { + unsigned int gen = kvm_current_mmio_generation(kvm); + u64 mask = generation_mmio_spte_mask(gen); + access &= ACC_WRITE_MASK | ACC_USER_MASK; + mask |= shadow_mmio_mask | access | gfn << PAGE_SHIFT; - trace_mark_mmio_spte(sptep, gfn, access); - mmu_spte_set(sptep, shadow_mmio_mask | access | gfn << PAGE_SHIFT); + trace_mark_mmio_spte(sptep, gfn, access, gen); + mmu_spte_set(sptep, mask); } static bool is_mmio_spte(u64 spte) @@ -209,24 +264,38 @@ static bool is_mmio_spte(u64 spte) static gfn_t get_mmio_spte_gfn(u64 spte) { - return (spte & ~shadow_mmio_mask) >> PAGE_SHIFT; + u64 mask = generation_mmio_spte_mask(MMIO_MAX_GEN) | shadow_mmio_mask; + return (spte & ~mask) >> PAGE_SHIFT; } static unsigned get_mmio_spte_access(u64 spte) { - return (spte & ~shadow_mmio_mask) & ~PAGE_MASK; + u64 mask = generation_mmio_spte_mask(MMIO_MAX_GEN) | shadow_mmio_mask; + return (spte & ~mask) & ~PAGE_MASK; } -static bool set_mmio_spte(u64 *sptep, gfn_t gfn, pfn_t pfn, unsigned access) +static bool set_mmio_spte(struct kvm *kvm, u64 *sptep, gfn_t gfn, + pfn_t pfn, unsigned access) { if (unlikely(is_noslot_pfn(pfn))) { - mark_mmio_spte(sptep, gfn, access); + mark_mmio_spte(kvm, sptep, gfn, access); return true; } return false; } +static bool check_mmio_spte(struct kvm *kvm, u64 spte) +{ + unsigned int kvm_gen, spte_gen; + + kvm_gen = kvm_current_mmio_generation(kvm); + spte_gen = get_mmio_spte_generation(spte); + + trace_check_mmio_spte(spte, kvm_gen, spte_gen); + return likely(kvm_gen == spte_gen); +} + static inline u64 rsvd_bits(int s, int e) { return ((1ULL << (e - s + 1)) - 1) << s; @@ -263,11 +332,6 @@ static int is_large_pte(u64 pte) return pte & PT_PAGE_SIZE_MASK; } -static int is_dirty_gpte(unsigned long pte) -{ - return pte & PT_DIRTY_MASK; -} - static int is_rmap_spte(u64 pte) { return is_shadow_present_pte(pte); @@ -398,9 +462,20 @@ static u64 __update_clear_spte_slow(u64 *sptep, u64 spte) /* * The idea using the light way get the spte on x86_32 guest is from * gup_get_pte(arch/x86/mm/gup.c). - * The difference is we can not catch the spte tlb flush if we leave - * guest mode, so we emulate it by increase clear_spte_count when spte - * is cleared. + * + * An spte tlb flush may be pending, because kvm_set_pte_rmapp + * coalesces them and we are running out of the MMU lock. Therefore + * we need to protect against in-progress updates of the spte. + * + * Reading the spte while an update is in progress may get the old value + * for the high part of the spte. The race is fine for a present->non-present + * change (because the high part of the spte is ignored for non-present spte), + * but for a present->present change we must reread the spte. + * + * All such changes are done in two steps (present->non-present and + * non-present->present), hence it is enough to count the number of + * present->non-present updates: if it changed while reading the spte, + * we might have hit the race. This is done using clear_spte_count. */ static u64 __get_spte_lockless(u64 *sptep) { @@ -443,8 +518,23 @@ static bool __check_direct_spte_mmio_pf(u64 spte) } #endif +static bool spte_is_locklessly_modifiable(u64 spte) +{ + return (spte & (SPTE_HOST_WRITEABLE | SPTE_MMU_WRITEABLE)) == + (SPTE_HOST_WRITEABLE | SPTE_MMU_WRITEABLE); +} + static bool spte_has_volatile_bits(u64 spte) { + /* + * Always atomicly update spte if it can be updated + * out of mmu-lock, it can ensure dirty bit is not lost, + * also, it can help us to get a stable is_writable_pte() + * to ensure tlb flush is not missed. + */ + if (spte_is_locklessly_modifiable(spte)) + return true; + if (!shadow_accessed_mask) return false; @@ -477,34 +567,48 @@ static void mmu_spte_set(u64 *sptep, u64 new_spte) /* Rules for using mmu_spte_update: * Update the state bits, it means the mapped pfn is not changged. + * + * Whenever we overwrite a writable spte with a read-only one we + * should flush remote TLBs. Otherwise rmap_write_protect + * will find a read-only spte, even though the writable spte + * might be cached on a CPU's TLB, the return value indicates this + * case. */ -static void mmu_spte_update(u64 *sptep, u64 new_spte) +static bool mmu_spte_update(u64 *sptep, u64 new_spte) { - u64 mask, old_spte = *sptep; + u64 old_spte = *sptep; + bool ret = false; WARN_ON(!is_rmap_spte(new_spte)); - if (!is_shadow_present_pte(old_spte)) - return mmu_spte_set(sptep, new_spte); - - new_spte |= old_spte & shadow_dirty_mask; - - mask = shadow_accessed_mask; - if (is_writable_pte(old_spte)) - mask |= shadow_dirty_mask; + if (!is_shadow_present_pte(old_spte)) { + mmu_spte_set(sptep, new_spte); + return ret; + } - if (!spte_has_volatile_bits(old_spte) || (new_spte & mask) == mask) + if (!spte_has_volatile_bits(old_spte)) __update_clear_spte_fast(sptep, new_spte); else old_spte = __update_clear_spte_slow(sptep, new_spte); + /* + * For the spte updated out of mmu-lock is safe, since + * we always atomicly update it, see the comments in + * spte_has_volatile_bits(). + */ + if (spte_is_locklessly_modifiable(old_spte) && + !is_writable_pte(new_spte)) + ret = true; + if (!shadow_accessed_mask) - return; + return ret; if (spte_is_bit_cleared(old_spte, new_spte, shadow_accessed_mask)) kvm_set_pfn_accessed(spte_to_pfn(old_spte)); if (spte_is_bit_cleared(old_spte, new_spte, shadow_dirty_mask)) kvm_set_pfn_dirty(spte_to_pfn(old_spte)); + + return ret; } /* @@ -526,6 +630,14 @@ static int mmu_spte_clear_track_bits(u64 *sptep) return 0; pfn = spte_to_pfn(old_spte); + + /* + * KVM does not hold the refcount of the page used by + * kvm mmu, before reclaiming the page, we should + * unmap it from mmu first. + */ + WARN_ON(!kvm_is_mmio_pfn(pfn) && !page_count(pfn_to_page(pfn))); + if (!shadow_accessed_mask || old_spte & shadow_accessed_mask) kvm_set_pfn_accessed(pfn); if (!shadow_dirty_mask || (old_spte & shadow_dirty_mask)) @@ -550,19 +662,29 @@ static u64 mmu_spte_get_lockless(u64 *sptep) static void walk_shadow_page_lockless_begin(struct kvm_vcpu *vcpu) { - rcu_read_lock(); - atomic_inc(&vcpu->kvm->arch.reader_counter); - - /* Increase the counter before walking shadow page table */ - smp_mb__after_atomic_inc(); + /* + * Prevent page table teardown by making any free-er wait during + * kvm_flush_remote_tlbs() IPI to all active vcpus. + */ + local_irq_disable(); + vcpu->mode = READING_SHADOW_PAGE_TABLES; + /* + * Make sure a following spte read is not reordered ahead of the write + * to vcpu->mode. + */ + smp_mb(); } static void walk_shadow_page_lockless_end(struct kvm_vcpu *vcpu) { - /* Decrease the counter after walking shadow page table finished */ - smp_mb__before_atomic_dec(); - atomic_dec(&vcpu->kvm->arch.reader_counter); - rcu_read_unlock(); + /* + * Make sure the write to vcpu->mode is not reordered in front of + * reads to sptes. If it does, kvm_commit_zap_page() can see us + * OUTSIDE_GUEST_MODE and proceed to free the shadow page table. + */ + smp_mb(); + vcpu->mode = OUTSIDE_GUEST_MODE; + local_irq_enable(); } static int mmu_topup_memory_cache(struct kvm_mmu_memory_cache *cache, @@ -641,8 +763,7 @@ static void mmu_free_memory_caches(struct kvm_vcpu *vcpu) mmu_page_header_cache); } -static void *mmu_memory_cache_alloc(struct kvm_mmu_memory_cache *mc, - size_t size) +static void *mmu_memory_cache_alloc(struct kvm_mmu_memory_cache *mc) { void *p; @@ -653,8 +774,7 @@ static void *mmu_memory_cache_alloc(struct kvm_mmu_memory_cache *mc, static struct pte_list_desc *mmu_alloc_pte_list_desc(struct kvm_vcpu *vcpu) { - return mmu_memory_cache_alloc(&vcpu->arch.mmu_pte_list_desc_cache, - sizeof(struct pte_list_desc)); + return mmu_memory_cache_alloc(&vcpu->arch.mmu_pte_list_desc_cache); } static void mmu_free_pte_list_desc(struct pte_list_desc *pte_list_desc) @@ -688,9 +808,8 @@ static struct kvm_lpage_info *lpage_info_slot(gfn_t gfn, { unsigned long idx; - idx = (gfn >> KVM_HPAGE_GFN_SHIFT(level)) - - (slot->base_gfn >> KVM_HPAGE_GFN_SHIFT(level)); - return &slot->lpage_info[level - 2][idx]; + idx = gfn_to_index(gfn, slot->base_gfn, level); + return &slot->arch.lpage_info[level - 2][idx]; } static void account_shadowed(struct kvm *kvm, gfn_t gfn) @@ -786,8 +905,7 @@ static int mapping_level(struct kvm_vcpu *vcpu, gfn_t large_gfn) if (host_level == PT_PAGE_TABLE_LEVEL) return host_level; - max_level = kvm_x86_ops->get_lpage_level() < host_level ? - kvm_x86_ops->get_lpage_level() : host_level; + max_level = min(kvm_x86_ops->get_lpage_level(), host_level); for (level = PT_DIRECTORY_LEVEL; level <= max_level; ++level) if (has_wrprotected_page(vcpu->kvm, large_gfn, level)) @@ -842,32 +960,6 @@ static int pte_list_add(struct kvm_vcpu *vcpu, u64 *spte, return count; } -static u64 *pte_list_next(unsigned long *pte_list, u64 *spte) -{ - struct pte_list_desc *desc; - u64 *prev_spte; - int i; - - if (!*pte_list) - return NULL; - else if (!(*pte_list & 1)) { - if (!spte) - return (u64 *)*pte_list; - return NULL; - } - desc = (struct pte_list_desc *)(*pte_list & ~1ul); - prev_spte = NULL; - while (desc) { - for (i = 0; i < PTE_LIST_EXT && desc->sptes[i]; ++i) { - if (prev_spte == spte) - return desc->sptes[i]; - prev_spte = desc->sptes[i]; - } - desc = desc->more; - } - return NULL; -} - static void pte_list_desc_remove_entry(unsigned long *pte_list, struct pte_list_desc *desc, int i, struct pte_list_desc *prev_desc) @@ -946,16 +1038,13 @@ static void pte_list_walk(unsigned long *pte_list, pte_list_walk_fn fn) } } -static unsigned long *__gfn_to_rmap(struct kvm *kvm, gfn_t gfn, int level, +static unsigned long *__gfn_to_rmap(gfn_t gfn, int level, struct kvm_memory_slot *slot) { - struct kvm_lpage_info *linfo; - - if (likely(level == PT_PAGE_TABLE_LEVEL)) - return &slot->rmap[gfn - slot->base_gfn]; + unsigned long idx; - linfo = lpage_info_slot(gfn, slot, level); - return &linfo->rmap_pde; + idx = gfn_to_index(gfn, slot->base_gfn, level); + return &slot->arch.rmap[level - PT_PAGE_TABLE_LEVEL][idx]; } /* @@ -966,7 +1055,7 @@ static unsigned long *gfn_to_rmap(struct kvm *kvm, gfn_t gfn, int level) struct kvm_memory_slot *slot; slot = gfn_to_memslot(kvm, gfn); - return __gfn_to_rmap(kvm, gfn, level, slot); + return __gfn_to_rmap(gfn, level, slot); } static bool rmap_can_add(struct kvm_vcpu *vcpu) @@ -988,11 +1077,6 @@ static int rmap_add(struct kvm_vcpu *vcpu, u64 *spte, gfn_t gfn) return pte_list_add(vcpu, spte, rmapp); } -static u64 *rmap_next(struct kvm *kvm, unsigned long *rmapp, u64 *spte) -{ - return pte_list_next(rmapp, spte); -} - static void rmap_remove(struct kvm *kvm, u64 *spte) { struct kvm_mmu_page *sp; @@ -1005,149 +1089,305 @@ static void rmap_remove(struct kvm *kvm, u64 *spte) pte_list_remove(spte, rmapp); } +/* + * Used by the following functions to iterate through the sptes linked by a + * rmap. All fields are private and not assumed to be used outside. + */ +struct rmap_iterator { + /* private fields */ + struct pte_list_desc *desc; /* holds the sptep if not NULL */ + int pos; /* index of the sptep */ +}; + +/* + * Iteration must be started by this function. This should also be used after + * removing/dropping sptes from the rmap link because in such cases the + * information in the itererator may not be valid. + * + * Returns sptep if found, NULL otherwise. + */ +static u64 *rmap_get_first(unsigned long rmap, struct rmap_iterator *iter) +{ + if (!rmap) + return NULL; + + if (!(rmap & 1)) { + iter->desc = NULL; + return (u64 *)rmap; + } + + iter->desc = (struct pte_list_desc *)(rmap & ~1ul); + iter->pos = 0; + return iter->desc->sptes[iter->pos]; +} + +/* + * Must be used with a valid iterator: e.g. after rmap_get_first(). + * + * Returns sptep if found, NULL otherwise. + */ +static u64 *rmap_get_next(struct rmap_iterator *iter) +{ + if (iter->desc) { + if (iter->pos < PTE_LIST_EXT - 1) { + u64 *sptep; + + ++iter->pos; + sptep = iter->desc->sptes[iter->pos]; + if (sptep) + return sptep; + } + + iter->desc = iter->desc->more; + + if (iter->desc) { + iter->pos = 0; + /* desc->sptes[0] cannot be NULL */ + return iter->desc->sptes[iter->pos]; + } + } + + return NULL; +} + static void drop_spte(struct kvm *kvm, u64 *sptep) { if (mmu_spte_clear_track_bits(sptep)) rmap_remove(kvm, sptep); } -int kvm_mmu_rmap_write_protect(struct kvm *kvm, u64 gfn, - struct kvm_memory_slot *slot) + +static bool __drop_large_spte(struct kvm *kvm, u64 *sptep) { - unsigned long *rmapp; - u64 *spte; - int i, write_protected = 0; - - rmapp = __gfn_to_rmap(kvm, gfn, PT_PAGE_TABLE_LEVEL, slot); - spte = rmap_next(kvm, rmapp, NULL); - while (spte) { - BUG_ON(!(*spte & PT_PRESENT_MASK)); - rmap_printk("rmap_write_protect: spte %p %llx\n", spte, *spte); - if (is_writable_pte(*spte)) { - mmu_spte_update(spte, *spte & ~PT_WRITABLE_MASK); - write_protected = 1; - } - spte = rmap_next(kvm, rmapp, spte); + if (is_large_pte(*sptep)) { + WARN_ON(page_header(__pa(sptep))->role.level == + PT_PAGE_TABLE_LEVEL); + drop_spte(kvm, sptep); + --kvm->stat.lpages; + return true; } - /* check for huge page mappings */ - for (i = PT_DIRECTORY_LEVEL; - i < PT_PAGE_TABLE_LEVEL + KVM_NR_PAGE_SIZES; ++i) { - rmapp = __gfn_to_rmap(kvm, gfn, i, slot); - spte = rmap_next(kvm, rmapp, NULL); - while (spte) { - BUG_ON(!(*spte & PT_PRESENT_MASK)); - BUG_ON(!is_large_pte(*spte)); - pgprintk("rmap_write_protect(large): spte %p %llx %lld\n", spte, *spte, gfn); - if (is_writable_pte(*spte)) { - drop_spte(kvm, spte); - --kvm->stat.lpages; - spte = NULL; - write_protected = 1; - } - spte = rmap_next(kvm, rmapp, spte); - } + return false; +} + +static void drop_large_spte(struct kvm_vcpu *vcpu, u64 *sptep) +{ + if (__drop_large_spte(vcpu->kvm, sptep)) + kvm_flush_remote_tlbs(vcpu->kvm); +} + +/* + * Write-protect on the specified @sptep, @pt_protect indicates whether + * spte write-protection is caused by protecting shadow page table. + * + * Note: write protection is difference between drity logging and spte + * protection: + * - for dirty logging, the spte can be set to writable at anytime if + * its dirty bitmap is properly set. + * - for spte protection, the spte can be writable only after unsync-ing + * shadow page. + * + * Return true if tlb need be flushed. + */ +static bool spte_write_protect(struct kvm *kvm, u64 *sptep, bool pt_protect) +{ + u64 spte = *sptep; + + if (!is_writable_pte(spte) && + !(pt_protect && spte_is_locklessly_modifiable(spte))) + return false; + + rmap_printk("rmap_write_protect: spte %p %llx\n", sptep, *sptep); + + if (pt_protect) + spte &= ~SPTE_MMU_WRITEABLE; + spte = spte & ~PT_WRITABLE_MASK; + + return mmu_spte_update(sptep, spte); +} + +static bool __rmap_write_protect(struct kvm *kvm, unsigned long *rmapp, + bool pt_protect) +{ + u64 *sptep; + struct rmap_iterator iter; + bool flush = false; + + for (sptep = rmap_get_first(*rmapp, &iter); sptep;) { + BUG_ON(!(*sptep & PT_PRESENT_MASK)); + + flush |= spte_write_protect(kvm, sptep, pt_protect); + sptep = rmap_get_next(&iter); } - return write_protected; + return flush; } -static int rmap_write_protect(struct kvm *kvm, u64 gfn) +/** + * kvm_mmu_write_protect_pt_masked - write protect selected PT level pages + * @kvm: kvm instance + * @slot: slot to protect + * @gfn_offset: start of the BITS_PER_LONG pages we care about + * @mask: indicates which pages we should protect + * + * Used when we do not need to care about huge page mappings: e.g. during dirty + * logging we do not have any such mappings. + */ +void kvm_mmu_write_protect_pt_masked(struct kvm *kvm, + struct kvm_memory_slot *slot, + gfn_t gfn_offset, unsigned long mask) +{ + unsigned long *rmapp; + + while (mask) { + rmapp = __gfn_to_rmap(slot->base_gfn + gfn_offset + __ffs(mask), + PT_PAGE_TABLE_LEVEL, slot); + __rmap_write_protect(kvm, rmapp, false); + + /* clear the first set bit */ + mask &= mask - 1; + } +} + +static bool rmap_write_protect(struct kvm *kvm, u64 gfn) { struct kvm_memory_slot *slot; + unsigned long *rmapp; + int i; + bool write_protected = false; slot = gfn_to_memslot(kvm, gfn); - return kvm_mmu_rmap_write_protect(kvm, gfn, slot); + + for (i = PT_PAGE_TABLE_LEVEL; + i < PT_PAGE_TABLE_LEVEL + KVM_NR_PAGE_SIZES; ++i) { + rmapp = __gfn_to_rmap(gfn, i, slot); + write_protected |= __rmap_write_protect(kvm, rmapp, true); + } + + return write_protected; } static int kvm_unmap_rmapp(struct kvm *kvm, unsigned long *rmapp, - unsigned long data) + struct kvm_memory_slot *slot, unsigned long data) { - u64 *spte; + u64 *sptep; + struct rmap_iterator iter; int need_tlb_flush = 0; - while ((spte = rmap_next(kvm, rmapp, NULL))) { - BUG_ON(!(*spte & PT_PRESENT_MASK)); - rmap_printk("kvm_rmap_unmap_hva: spte %p %llx\n", spte, *spte); - drop_spte(kvm, spte); + while ((sptep = rmap_get_first(*rmapp, &iter))) { + BUG_ON(!(*sptep & PT_PRESENT_MASK)); + rmap_printk("kvm_rmap_unmap_hva: spte %p %llx\n", sptep, *sptep); + + drop_spte(kvm, sptep); need_tlb_flush = 1; } + return need_tlb_flush; } static int kvm_set_pte_rmapp(struct kvm *kvm, unsigned long *rmapp, - unsigned long data) + struct kvm_memory_slot *slot, unsigned long data) { + u64 *sptep; + struct rmap_iterator iter; int need_flush = 0; - u64 *spte, new_spte; + u64 new_spte; pte_t *ptep = (pte_t *)data; pfn_t new_pfn; WARN_ON(pte_huge(*ptep)); new_pfn = pte_pfn(*ptep); - spte = rmap_next(kvm, rmapp, NULL); - while (spte) { - BUG_ON(!is_shadow_present_pte(*spte)); - rmap_printk("kvm_set_pte_rmapp: spte %p %llx\n", spte, *spte); + + for (sptep = rmap_get_first(*rmapp, &iter); sptep;) { + BUG_ON(!is_shadow_present_pte(*sptep)); + rmap_printk("kvm_set_pte_rmapp: spte %p %llx\n", sptep, *sptep); + need_flush = 1; + if (pte_write(*ptep)) { - drop_spte(kvm, spte); - spte = rmap_next(kvm, rmapp, NULL); + drop_spte(kvm, sptep); + sptep = rmap_get_first(*rmapp, &iter); } else { - new_spte = *spte &~ (PT64_BASE_ADDR_MASK); + new_spte = *sptep & ~PT64_BASE_ADDR_MASK; new_spte |= (u64)new_pfn << PAGE_SHIFT; new_spte &= ~PT_WRITABLE_MASK; new_spte &= ~SPTE_HOST_WRITEABLE; new_spte &= ~shadow_accessed_mask; - mmu_spte_clear_track_bits(spte); - mmu_spte_set(spte, new_spte); - spte = rmap_next(kvm, rmapp, spte); + + mmu_spte_clear_track_bits(sptep); + mmu_spte_set(sptep, new_spte); + sptep = rmap_get_next(&iter); } } + if (need_flush) kvm_flush_remote_tlbs(kvm); return 0; } -static int kvm_handle_hva(struct kvm *kvm, unsigned long hva, - unsigned long data, - int (*handler)(struct kvm *kvm, unsigned long *rmapp, - unsigned long data)) +static int kvm_handle_hva_range(struct kvm *kvm, + unsigned long start, + unsigned long end, + unsigned long data, + int (*handler)(struct kvm *kvm, + unsigned long *rmapp, + struct kvm_memory_slot *slot, + unsigned long data)) { int j; - int ret; - int retval = 0; + int ret = 0; struct kvm_memslots *slots; struct kvm_memory_slot *memslot; slots = kvm_memslots(kvm); kvm_for_each_memslot(memslot, slots) { - unsigned long start = memslot->userspace_addr; - unsigned long end; + unsigned long hva_start, hva_end; + gfn_t gfn_start, gfn_end; - end = start + (memslot->npages << PAGE_SHIFT); - if (hva >= start && hva < end) { - gfn_t gfn_offset = (hva - start) >> PAGE_SHIFT; - gfn_t gfn = memslot->base_gfn + gfn_offset; + hva_start = max(start, memslot->userspace_addr); + hva_end = min(end, memslot->userspace_addr + + (memslot->npages << PAGE_SHIFT)); + if (hva_start >= hva_end) + continue; + /* + * {gfn(page) | page intersects with [hva_start, hva_end)} = + * {gfn_start, gfn_start+1, ..., gfn_end-1}. + */ + gfn_start = hva_to_gfn_memslot(hva_start, memslot); + gfn_end = hva_to_gfn_memslot(hva_end + PAGE_SIZE - 1, memslot); - ret = handler(kvm, &memslot->rmap[gfn_offset], data); + for (j = PT_PAGE_TABLE_LEVEL; + j < PT_PAGE_TABLE_LEVEL + KVM_NR_PAGE_SIZES; ++j) { + unsigned long idx, idx_end; + unsigned long *rmapp; - for (j = 0; j < KVM_NR_PAGE_SIZES - 1; ++j) { - struct kvm_lpage_info *linfo; + /* + * {idx(page_j) | page_j intersects with + * [hva_start, hva_end)} = {idx, idx+1, ..., idx_end}. + */ + idx = gfn_to_index(gfn_start, memslot->base_gfn, j); + idx_end = gfn_to_index(gfn_end - 1, memslot->base_gfn, j); - linfo = lpage_info_slot(gfn, memslot, - PT_DIRECTORY_LEVEL + j); - ret |= handler(kvm, &linfo->rmap_pde, data); - } - trace_kvm_age_page(hva, memslot, ret); - retval |= ret; + rmapp = __gfn_to_rmap(gfn_start, j, memslot); + + for (; idx <= idx_end; ++idx) + ret |= handler(kvm, rmapp++, memslot, data); } } - return retval; + return ret; +} + +static int kvm_handle_hva(struct kvm *kvm, unsigned long hva, + unsigned long data, + int (*handler)(struct kvm *kvm, unsigned long *rmapp, + struct kvm_memory_slot *slot, + unsigned long data)) +{ + return kvm_handle_hva_range(kvm, hva, hva + 1, data, handler); } int kvm_unmap_hva(struct kvm *kvm, unsigned long hva) @@ -1155,46 +1395,57 @@ int kvm_unmap_hva(struct kvm *kvm, unsigned long hva) return kvm_handle_hva(kvm, hva, 0, kvm_unmap_rmapp); } +int kvm_unmap_hva_range(struct kvm *kvm, unsigned long start, unsigned long end) +{ + return kvm_handle_hva_range(kvm, start, end, 0, kvm_unmap_rmapp); +} + void kvm_set_spte_hva(struct kvm *kvm, unsigned long hva, pte_t pte) { kvm_handle_hva(kvm, hva, (unsigned long)&pte, kvm_set_pte_rmapp); } static int kvm_age_rmapp(struct kvm *kvm, unsigned long *rmapp, - unsigned long data) + struct kvm_memory_slot *slot, unsigned long data) { - u64 *spte; + u64 *sptep; + struct rmap_iterator uninitialized_var(iter); int young = 0; /* - * Emulate the accessed bit for EPT, by checking if this page has + * In case of absence of EPT Access and Dirty Bits supports, + * emulate the accessed bit for EPT, by checking if this page has * an EPT mapping, and clearing it if it does. On the next access, * a new EPT mapping will be established. * This has some overhead, but not as much as the cost of swapping * out actively used pages or breaking up actively used hugepages. */ - if (!shadow_accessed_mask) - return kvm_unmap_rmapp(kvm, rmapp, data); - - spte = rmap_next(kvm, rmapp, NULL); - while (spte) { - int _young; - u64 _spte = *spte; - BUG_ON(!(_spte & PT_PRESENT_MASK)); - _young = _spte & PT_ACCESSED_MASK; - if (_young) { + if (!shadow_accessed_mask) { + young = kvm_unmap_rmapp(kvm, rmapp, slot, data); + goto out; + } + + for (sptep = rmap_get_first(*rmapp, &iter); sptep; + sptep = rmap_get_next(&iter)) { + BUG_ON(!is_shadow_present_pte(*sptep)); + + if (*sptep & shadow_accessed_mask) { young = 1; - clear_bit(PT_ACCESSED_SHIFT, (unsigned long *)spte); + clear_bit((ffs(shadow_accessed_mask) - 1), + (unsigned long *)sptep); } - spte = rmap_next(kvm, rmapp, spte); } +out: + /* @data has hva passed to kvm_age_hva(). */ + trace_kvm_age_page(data, slot, young); return young; } static int kvm_test_age_rmapp(struct kvm *kvm, unsigned long *rmapp, - unsigned long data) + struct kvm_memory_slot *slot, unsigned long data) { - u64 *spte; + u64 *sptep; + struct rmap_iterator iter; int young = 0; /* @@ -1205,16 +1456,14 @@ static int kvm_test_age_rmapp(struct kvm *kvm, unsigned long *rmapp, if (!shadow_accessed_mask) goto out; - spte = rmap_next(kvm, rmapp, NULL); - while (spte) { - u64 _spte = *spte; - BUG_ON(!(_spte & PT_PRESENT_MASK)); - young = _spte & PT_ACCESSED_MASK; - if (young) { + for (sptep = rmap_get_first(*rmapp, &iter); sptep; + sptep = rmap_get_next(&iter)) { + BUG_ON(!is_shadow_present_pte(*sptep)); + + if (*sptep & shadow_accessed_mask) { young = 1; break; } - spte = rmap_next(kvm, rmapp, spte); } out: return young; @@ -1231,13 +1480,13 @@ static void rmap_recycle(struct kvm_vcpu *vcpu, u64 *spte, gfn_t gfn) rmapp = gfn_to_rmap(vcpu->kvm, gfn, sp->role.level); - kvm_unmap_rmapp(vcpu->kvm, rmapp, 0); + kvm_unmap_rmapp(vcpu->kvm, rmapp, NULL, 0); kvm_flush_remote_tlbs(vcpu->kvm); } int kvm_age_hva(struct kvm *kvm, unsigned long hva) { - return kvm_handle_hva(kvm, hva, 0, kvm_age_rmapp); + return kvm_handle_hva(kvm, hva, hva, kvm_age_rmapp); } int kvm_test_age_hva(struct kvm *kvm, unsigned long hva) @@ -1273,28 +1522,14 @@ static inline void kvm_mod_used_mmu_pages(struct kvm *kvm, int nr) percpu_counter_add(&kvm_total_used_mmu_pages, nr); } -/* - * Remove the sp from shadow page cache, after call it, - * we can not find this sp from the cache, and the shadow - * page table is still valid. - * It should be under the protection of mmu lock. - */ -static void kvm_mmu_isolate_page(struct kvm_mmu_page *sp) +static void kvm_mmu_free_page(struct kvm_mmu_page *sp) { ASSERT(is_empty_shadow_page(sp->spt)); hlist_del(&sp->hash_link); - if (!sp->role.direct) - free_page((unsigned long)sp->gfns); -} - -/* - * Free the shadow page table and the sp, we can do it - * out of the protection of mmu lock. - */ -static void kvm_mmu_free_page(struct kvm_mmu_page *sp) -{ list_del(&sp->link); free_page((unsigned long)sp->spt); + if (!sp->role.direct) + free_page((unsigned long)sp->gfns); kmem_cache_free(mmu_page_header_cache, sp); } @@ -1329,15 +1564,19 @@ static struct kvm_mmu_page *kvm_mmu_alloc_page(struct kvm_vcpu *vcpu, u64 *parent_pte, int direct) { struct kvm_mmu_page *sp; - sp = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_header_cache, - sizeof *sp); - sp->spt = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_cache, PAGE_SIZE); + + sp = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_header_cache); + sp->spt = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_cache); if (!direct) - sp->gfns = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_cache, - PAGE_SIZE); + sp->gfns = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_cache); set_page_private(virt_to_page(sp->spt), (unsigned long)sp); + + /* + * The active_mmu_pages list is the FIFO list, do not move the + * page until it is zapped. kvm_zap_obsolete_pages depends on + * this feature. See the comments in kvm_zap_obsolete_pages(). + */ list_add(&sp->link, &vcpu->kvm->arch.active_mmu_pages); - bitmap_zero(sp->slot_bitmap, KVM_MEM_SLOTS_NUM); sp->parent_ptes = 0; mmu_page_add_parent_pte(vcpu, sp, parent_pte); kvm_mod_used_mmu_pages(vcpu->kvm, +1); @@ -1391,11 +1630,6 @@ struct kvm_mmu_pages { unsigned int nr; }; -#define for_each_unsync_children(bitmap, idx) \ - for (idx = find_first_bit(bitmap, 512); \ - idx < 512; \ - idx = find_next_bit(bitmap, 512, idx+1)) - static int mmu_pages_add(struct kvm_mmu_pages *pvec, struct kvm_mmu_page *sp, int idx) { @@ -1417,7 +1651,7 @@ static int __mmu_unsync_walk(struct kvm_mmu_page *sp, { int i, ret, nr_unsync_leaf = 0; - for_each_unsync_children(sp->unsync_child_bitmap, i) { + for_each_set_bit(i, sp->unsync_child_bitmap, 512) { struct kvm_mmu_page *child; u64 ent = sp->spt[i]; @@ -1479,16 +1713,24 @@ static int kvm_mmu_prepare_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp, static void kvm_mmu_commit_zap_page(struct kvm *kvm, struct list_head *invalid_list); -#define for_each_gfn_sp(kvm, sp, gfn, pos) \ - hlist_for_each_entry(sp, pos, \ - &(kvm)->arch.mmu_page_hash[kvm_page_table_hashfn(gfn)], hash_link) \ - if ((sp)->gfn != (gfn)) {} else +/* + * NOTE: we should pay more attention on the zapped-obsolete page + * (is_obsolete_sp(sp) && sp->role.invalid) when you do hash list walk + * since it has been deleted from active_mmu_pages but still can be found + * at hast list. + * + * for_each_gfn_indirect_valid_sp has skipped that kind of page and + * kvm_mmu_get_page(), the only user of for_each_gfn_sp(), has skipped + * all the obsolete pages. + */ +#define for_each_gfn_sp(_kvm, _sp, _gfn) \ + hlist_for_each_entry(_sp, \ + &(_kvm)->arch.mmu_page_hash[kvm_page_table_hashfn(_gfn)], hash_link) \ + if ((_sp)->gfn != (_gfn)) {} else -#define for_each_gfn_indirect_valid_sp(kvm, sp, gfn, pos) \ - hlist_for_each_entry(sp, pos, \ - &(kvm)->arch.mmu_page_hash[kvm_page_table_hashfn(gfn)], hash_link) \ - if ((sp)->gfn != (gfn) || (sp)->role.direct || \ - (sp)->role.invalid) {} else +#define for_each_gfn_indirect_valid_sp(_kvm, _sp, _gfn) \ + for_each_gfn_sp(_kvm, _sp, _gfn) \ + if ((_sp)->role.direct || (_sp)->role.invalid) {} else /* @sp->gfn should be write-protected at the call site */ static int __kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp, @@ -1541,11 +1783,10 @@ static int kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp, static void kvm_sync_pages(struct kvm_vcpu *vcpu, gfn_t gfn) { struct kvm_mmu_page *s; - struct hlist_node *node; LIST_HEAD(invalid_list); bool flush = false; - for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn, node) { + for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn) { if (!s->unsync) continue; @@ -1634,7 +1875,7 @@ static void mmu_sync_children(struct kvm_vcpu *vcpu, kvm_mmu_pages_init(parent, &parents, &pages); while (mmu_unsync_walk(parent, &pages)) { - int protected = 0; + bool protected = false; for_each_sp(pages, sp, parents, i) protected |= rmap_write_protect(vcpu->kvm, sp->gfn); @@ -1672,6 +1913,11 @@ static void clear_sp_write_flooding_count(u64 *spte) __clear_sp_write_flooding_count(sp); } +static bool is_obsolete_sp(struct kvm *kvm, struct kvm_mmu_page *sp) +{ + return unlikely(sp->mmu_valid_gen != kvm->arch.mmu_valid_gen); +} + static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu, gfn_t gfn, gva_t gaddr, @@ -1683,7 +1929,6 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu, union kvm_mmu_page_role role; unsigned quadrant; struct kvm_mmu_page *sp; - struct hlist_node *node; bool need_sync = false; role = vcpu->arch.mmu.base_role; @@ -1698,7 +1943,10 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu, quadrant &= (1 << ((PT32_PT_BITS - PT64_PT_BITS) * level)) - 1; role.quadrant = quadrant; } - for_each_gfn_sp(vcpu->kvm, sp, gfn, node) { + for_each_gfn_sp(vcpu->kvm, sp, gfn) { + if (is_obsolete_sp(vcpu->kvm, sp)) + continue; + if (!need_sync && sp->unsync) need_sync = true; @@ -1735,6 +1983,7 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu, account_shadowed(vcpu->kvm, gfn); } + sp->mmu_valid_gen = vcpu->kvm->arch.mmu_valid_gen; init_shadow_page_table(sp); trace_kvm_mmu_get_page(sp, true); return sp; @@ -1789,22 +2038,20 @@ static void shadow_walk_next(struct kvm_shadow_walk_iterator *iterator) return __shadow_walk_next(iterator, *iterator->sptep); } -static void link_shadow_page(u64 *sptep, struct kvm_mmu_page *sp) +static void link_shadow_page(u64 *sptep, struct kvm_mmu_page *sp, bool accessed) { u64 spte; - spte = __pa(sp->spt) - | PT_PRESENT_MASK | PT_ACCESSED_MASK - | PT_WRITABLE_MASK | PT_USER_MASK; - mmu_spte_set(sptep, spte); -} + BUILD_BUG_ON(VMX_EPT_READABLE_MASK != PT_PRESENT_MASK || + VMX_EPT_WRITABLE_MASK != PT_WRITABLE_MASK); -static void drop_large_spte(struct kvm_vcpu *vcpu, u64 *sptep) -{ - if (is_large_pte(*sptep)) { - drop_spte(vcpu->kvm, sptep); - kvm_flush_remote_tlbs(vcpu->kvm); - } + spte = __pa(sp->spt) | PT_PRESENT_MASK | PT_WRITABLE_MASK | + shadow_user_mask | shadow_x_mask; + + if (accessed) + spte |= shadow_accessed_mask; + + mmu_spte_set(sptep, spte); } static void validate_direct_spte(struct kvm_vcpu *vcpu, u64 *sptep, @@ -1870,10 +2117,11 @@ static void kvm_mmu_put_page(struct kvm_mmu_page *sp, u64 *parent_pte) static void kvm_mmu_unlink_parents(struct kvm *kvm, struct kvm_mmu_page *sp) { - u64 *parent_pte; + u64 *sptep; + struct rmap_iterator iter; - while ((parent_pte = pte_list_next(&sp->parent_ptes, NULL))) - drop_parent_pte(sp, parent_pte); + while ((sptep = rmap_get_first(sp->parent_ptes, &iter))) + drop_parent_pte(sp, sptep); } static int mmu_zap_unsync_children(struct kvm *kvm, @@ -1912,8 +2160,10 @@ static int kvm_mmu_prepare_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp, ret = mmu_zap_unsync_children(kvm, sp, invalid_list); kvm_mmu_page_unlink_children(kvm, sp); kvm_mmu_unlink_parents(kvm, sp); + if (!sp->role.invalid && !sp->role.direct) unaccount_shadowed(kvm, sp->gfn); + if (sp->unsync) kvm_unlink_unsync_page(kvm, sp); if (!sp->root_count) { @@ -1923,64 +2173,58 @@ static int kvm_mmu_prepare_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp, kvm_mod_used_mmu_pages(kvm, -1); } else { list_move(&sp->link, &kvm->arch.active_mmu_pages); - kvm_reload_remote_mmus(kvm); + + /* + * The obsolete pages can not be used on any vcpus. + * See the comments in kvm_mmu_invalidate_zap_all_pages(). + */ + if (!sp->role.invalid && !is_obsolete_sp(kvm, sp)) + kvm_reload_remote_mmus(kvm); } sp->role.invalid = 1; return ret; } -static void kvm_mmu_isolate_pages(struct list_head *invalid_list) +static void kvm_mmu_commit_zap_page(struct kvm *kvm, + struct list_head *invalid_list) { - struct kvm_mmu_page *sp; + struct kvm_mmu_page *sp, *nsp; - list_for_each_entry(sp, invalid_list, link) - kvm_mmu_isolate_page(sp); -} + if (list_empty(invalid_list)) + return; -static void free_pages_rcu(struct rcu_head *head) -{ - struct kvm_mmu_page *next, *sp; + /* + * wmb: make sure everyone sees our modifications to the page tables + * rmb: make sure we see changes to vcpu->mode + */ + smp_mb(); - sp = container_of(head, struct kvm_mmu_page, rcu); - while (sp) { - if (!list_empty(&sp->link)) - next = list_first_entry(&sp->link, - struct kvm_mmu_page, link); - else - next = NULL; + /* + * Wait for all vcpus to exit guest mode and/or lockless shadow + * page table walks. + */ + kvm_flush_remote_tlbs(kvm); + + list_for_each_entry_safe(sp, nsp, invalid_list, link) { + WARN_ON(!sp->role.invalid || sp->root_count); kvm_mmu_free_page(sp); - sp = next; } } -static void kvm_mmu_commit_zap_page(struct kvm *kvm, - struct list_head *invalid_list) +static bool prepare_zap_oldest_mmu_page(struct kvm *kvm, + struct list_head *invalid_list) { struct kvm_mmu_page *sp; - if (list_empty(invalid_list)) - return; - - kvm_flush_remote_tlbs(kvm); - - if (atomic_read(&kvm->arch.reader_counter)) { - kvm_mmu_isolate_pages(invalid_list); - sp = list_first_entry(invalid_list, struct kvm_mmu_page, link); - list_del_init(invalid_list); - - trace_kvm_mmu_delay_free_pages(sp); - call_rcu(&sp->rcu, free_pages_rcu); - return; - } + if (list_empty(&kvm->arch.active_mmu_pages)) + return false; - do { - sp = list_first_entry(invalid_list, struct kvm_mmu_page, link); - WARN_ON(!sp->role.invalid || sp->root_count); - kvm_mmu_isolate_page(sp); - kvm_mmu_free_page(sp); - } while (!list_empty(invalid_list)); + sp = list_entry(kvm->arch.active_mmu_pages.prev, + struct kvm_mmu_page, link); + kvm_mmu_prepare_zap_page(kvm, sp, invalid_list); + return true; } /* @@ -1990,39 +2234,34 @@ static void kvm_mmu_commit_zap_page(struct kvm *kvm, void kvm_mmu_change_mmu_pages(struct kvm *kvm, unsigned int goal_nr_mmu_pages) { LIST_HEAD(invalid_list); - /* - * If we set the number of mmu pages to be smaller be than the - * number of actived pages , we must to free some mmu pages before we - * change the value - */ + + spin_lock(&kvm->mmu_lock); if (kvm->arch.n_used_mmu_pages > goal_nr_mmu_pages) { - while (kvm->arch.n_used_mmu_pages > goal_nr_mmu_pages && - !list_empty(&kvm->arch.active_mmu_pages)) { - struct kvm_mmu_page *page; + /* Need to free some mmu pages to achieve the goal. */ + while (kvm->arch.n_used_mmu_pages > goal_nr_mmu_pages) + if (!prepare_zap_oldest_mmu_page(kvm, &invalid_list)) + break; - page = container_of(kvm->arch.active_mmu_pages.prev, - struct kvm_mmu_page, link); - kvm_mmu_prepare_zap_page(kvm, page, &invalid_list); - } kvm_mmu_commit_zap_page(kvm, &invalid_list); goal_nr_mmu_pages = kvm->arch.n_used_mmu_pages; } kvm->arch.n_max_mmu_pages = goal_nr_mmu_pages; + + spin_unlock(&kvm->mmu_lock); } int kvm_mmu_unprotect_page(struct kvm *kvm, gfn_t gfn) { struct kvm_mmu_page *sp; - struct hlist_node *node; LIST_HEAD(invalid_list); int r; pgprintk("%s: looking for gfn %llx\n", __func__, gfn); r = 0; spin_lock(&kvm->mmu_lock); - for_each_gfn_indirect_valid_sp(kvm, sp, gfn, node) { + for_each_gfn_indirect_valid_sp(kvm, sp, gfn) { pgprintk("%s: gfn %llx role %x\n", __func__, gfn, sp->role.word); r = 1; @@ -2035,14 +2274,6 @@ int kvm_mmu_unprotect_page(struct kvm *kvm, gfn_t gfn) } EXPORT_SYMBOL_GPL(kvm_mmu_unprotect_page); -static void page_header_update_slot(struct kvm *kvm, void *pte, gfn_t gfn) -{ - int slot = memslot_id(kvm, gfn); - struct kvm_mmu_page *sp = page_header(__pa(pte)); - - __set_bit(slot, sp->slot_bitmap); -} - /* * The function is based on mtrr_type_lookup() in * arch/x86/kernel/cpu/mtrr/generic.c @@ -2160,9 +2391,8 @@ static void __kvm_unsync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp) static void kvm_unsync_pages(struct kvm_vcpu *vcpu, gfn_t gfn) { struct kvm_mmu_page *s; - struct hlist_node *node; - for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn, node) { + for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn) { if (s->unsync) continue; WARN_ON(s->role.level != PT_PAGE_TABLE_LEVEL); @@ -2174,19 +2404,17 @@ static int mmu_need_write_protect(struct kvm_vcpu *vcpu, gfn_t gfn, bool can_unsync) { struct kvm_mmu_page *s; - struct hlist_node *node; bool need_unsync = false; - for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn, node) { + for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn) { if (!can_unsync) return 1; if (s->role.level != PT_PAGE_TABLE_LEVEL) return 1; - if (!need_unsync && !s->unsync) { + if (!s->unsync) need_unsync = true; - } } if (need_unsync) kvm_unsync_pages(vcpu, gfn); @@ -2194,15 +2422,14 @@ static int mmu_need_write_protect(struct kvm_vcpu *vcpu, gfn_t gfn, } static int set_spte(struct kvm_vcpu *vcpu, u64 *sptep, - unsigned pte_access, int user_fault, - int write_fault, int level, + unsigned pte_access, int level, gfn_t gfn, pfn_t pfn, bool speculative, bool can_unsync, bool host_writable) { - u64 spte, entry = *sptep; + u64 spte; int ret = 0; - if (set_mmio_spte(sptep, gfn, pfn, pte_access)) + if (set_mmio_spte(vcpu->kvm, sptep, gfn, pfn, pte_access)) return 0; spte = PT_PRESENT_MASK; @@ -2213,8 +2440,10 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *sptep, spte |= shadow_x_mask; else spte |= shadow_nx_mask; + if (pte_access & ACC_USER_MASK) spte |= shadow_user_mask; + if (level > PT_PAGE_TABLE_LEVEL) spte |= PT_PAGE_SIZE_MASK; if (tdp_enabled) @@ -2228,31 +2457,19 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *sptep, spte |= (u64)pfn << PAGE_SHIFT; - if ((pte_access & ACC_WRITE_MASK) - || (!vcpu->arch.mmu.direct_map && write_fault - && !is_write_protection(vcpu) && !user_fault)) { + if (pte_access & ACC_WRITE_MASK) { + /* + * Other vcpu creates new sp in the window between + * mapping_level() and acquiring mmu-lock. We can + * allow guest to retry the access, the mapping can + * be fixed if guest refault. + */ if (level > PT_PAGE_TABLE_LEVEL && - has_wrprotected_page(vcpu->kvm, gfn, level)) { - ret = 1; - drop_spte(vcpu->kvm, sptep); + has_wrprotected_page(vcpu->kvm, gfn, level)) goto done; - } - - spte |= PT_WRITABLE_MASK; - if (!vcpu->arch.mmu.direct_map - && !(pte_access & ACC_WRITE_MASK)) { - spte &= ~PT_USER_MASK; - /* - * If we converted a user page to a kernel page, - * so that the kernel can write to it when cr0.wp=0, - * then we should prevent the kernel from executing it - * if SMEP is enabled. - */ - if (kvm_read_cr4_bits(vcpu, X86_CR4_SMEP)) - spte |= PT64_NX_MASK; - } + spte |= PT_WRITABLE_MASK | SPTE_MMU_WRITEABLE; /* * Optimization: for pte sync, if spte was writable the hash @@ -2268,8 +2485,7 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *sptep, __func__, gfn); ret = 1; pte_access &= ~ACC_WRITE_MASK; - if (is_writable_pte(spte)) - spte &= ~PT_WRITABLE_MASK; + spte &= ~(PT_WRITABLE_MASK | SPTE_MMU_WRITEABLE); } } @@ -2277,33 +2493,22 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *sptep, mark_page_dirty(vcpu->kvm, gfn); set_pte: - mmu_spte_update(sptep, spte); - /* - * If we overwrite a writable spte with a read-only one we - * should flush remote TLBs. Otherwise rmap_write_protect - * will find a read-only spte, even though the writable spte - * might be cached on a CPU's TLB. - */ - if (is_writable_pte(entry) && !is_writable_pte(*sptep)) + if (mmu_spte_update(sptep, spte)) kvm_flush_remote_tlbs(vcpu->kvm); done: return ret; } static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *sptep, - unsigned pt_access, unsigned pte_access, - int user_fault, int write_fault, - int *emulate, int level, gfn_t gfn, - pfn_t pfn, bool speculative, + unsigned pte_access, int write_fault, int *emulate, + int level, gfn_t gfn, pfn_t pfn, bool speculative, bool host_writable) { int was_rmapped = 0; int rmap_count; - pgprintk("%s: spte %llx access %x write_fault %d" - " user_fault %d gfn %llx\n", - __func__, *sptep, pt_access, - write_fault, user_fault, gfn); + pgprintk("%s: spte %llx write_fault %d gfn %llx\n", __func__, + *sptep, write_fault, gfn); if (is_rmap_spte(*sptep)) { /* @@ -2327,9 +2532,8 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *sptep, was_rmapped = 1; } - if (set_spte(vcpu, sptep, pte_access, user_fault, write_fault, - level, gfn, pfn, speculative, true, - host_writable)) { + if (set_spte(vcpu, sptep, pte_access, level, gfn, pfn, speculative, + true, host_writable)) { if (write_fault) *emulate = 1; kvm_mmu_flush_tlb(vcpu); @@ -2347,35 +2551,26 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *sptep, ++vcpu->kvm->stat.lpages; if (is_shadow_present_pte(*sptep)) { - page_header_update_slot(vcpu->kvm, sptep, gfn); if (!was_rmapped) { rmap_count = rmap_add(vcpu, sptep, gfn); if (rmap_count > RMAP_RECYCLE_THRESHOLD) rmap_recycle(vcpu, sptep, gfn); } } - kvm_release_pfn_clean(pfn); -} -static void nonpaging_new_cr3(struct kvm_vcpu *vcpu) -{ + kvm_release_pfn_clean(pfn); } static pfn_t pte_prefetch_gfn_to_pfn(struct kvm_vcpu *vcpu, gfn_t gfn, bool no_dirty_log) { struct kvm_memory_slot *slot; - unsigned long hva; slot = gfn_to_memslot_dirty_bitmap(vcpu, gfn, no_dirty_log); - if (!slot) { - get_page(fault_page); - return page_to_pfn(fault_page); - } + if (!slot) + return KVM_PFN_ERR_FAULT; - hva = gfn_to_hva_memslot(slot, gfn); - - return hva_to_pfn_atomic(vcpu->kvm, hva); + return gfn_to_pfn_memslot_atomic(slot, gfn); } static int direct_pte_prefetch_many(struct kvm_vcpu *vcpu, @@ -2396,10 +2591,9 @@ static int direct_pte_prefetch_many(struct kvm_vcpu *vcpu, return -1; for (i = 0; i < ret; i++, gfn++, start++) - mmu_set_spte(vcpu, start, ACC_ALL, - access, 0, 0, NULL, - sp->role.level, gfn, - page_to_pfn(pages[i]), true, true); + mmu_set_spte(vcpu, start, access, 0, NULL, + sp->role.level, gfn, page_to_pfn(pages[i]), + true, true); return 0; } @@ -2456,18 +2650,20 @@ static int __direct_map(struct kvm_vcpu *vcpu, gpa_t v, int write, int emulate = 0; gfn_t pseudo_gfn; + if (!VALID_PAGE(vcpu->arch.mmu.root_hpa)) + return 0; + for_each_shadow_entry(vcpu, (u64)gfn << PAGE_SHIFT, iterator) { if (iterator.level == level) { - unsigned pte_access = ACC_ALL; - - mmu_set_spte(vcpu, iterator.sptep, ACC_ALL, pte_access, - 0, write, &emulate, - level, gfn, pfn, prefault, map_writable); + mmu_set_spte(vcpu, iterator.sptep, ACC_ALL, + write, &emulate, level, gfn, pfn, + prefault, map_writable); direct_pte_prefetch(vcpu, iterator.sptep); ++vcpu->stat.pf_fixed; break; } + drop_large_spte(vcpu, iterator.sptep); if (!is_shadow_present_pte(*iterator.sptep)) { u64 base_addr = iterator.addr; @@ -2476,17 +2672,8 @@ static int __direct_map(struct kvm_vcpu *vcpu, gpa_t v, int write, sp = kvm_mmu_get_page(vcpu, pseudo_gfn, iterator.addr, iterator.level - 1, 1, ACC_ALL, iterator.sptep); - if (!sp) { - pgprintk("nonpaging_map: ENOMEM\n"); - kvm_release_pfn_clean(pfn); - return -ENOMEM; - } - mmu_spte_set(iterator.sptep, - __pa(sp->spt) - | PT_PRESENT_MASK | PT_WRITABLE_MASK - | shadow_user_mask | shadow_x_mask - | shadow_accessed_mask); + link_shadow_page(iterator.sptep, sp, true); } } return emulate; @@ -2507,8 +2694,16 @@ static void kvm_send_hwpoison_signal(unsigned long address, struct task_struct * static int kvm_handle_bad_page(struct kvm_vcpu *vcpu, gfn_t gfn, pfn_t pfn) { - kvm_release_pfn_clean(pfn); - if (is_hwpoison_pfn(pfn)) { + /* + * Do not cache the mmio info caused by writing the readonly gfn + * into the spte otherwise read access on readonly gfn also can + * caused mmio page fault and treat it as mmio access. + * Return 1 to tell kvm to emulate it. + */ + if (pfn == KVM_PFN_ERR_RO_FAULT) + return 1; + + if (pfn == KVM_PFN_ERR_HWPOISON) { kvm_send_hwpoison_signal(gfn_to_hva(vcpu->kvm, gfn), current); return 0; } @@ -2529,7 +2724,7 @@ static void transparent_hugepage_adjust(struct kvm_vcpu *vcpu, * PT_PAGE_TABLE_LEVEL and there would be no adjustment done * here. */ - if (!is_error_pfn(pfn) && !kvm_is_mmio_pfn(pfn) && + if (!is_error_noslot_pfn(pfn) && !kvm_is_mmio_pfn(pfn) && level == PT_PAGE_TABLE_LEVEL && PageTransCompound(pfn_to_page(pfn)) && !has_wrprotected_page(vcpu->kvm, gfn, PT_DIRECTORY_LEVEL)) { @@ -2551,25 +2746,19 @@ static void transparent_hugepage_adjust(struct kvm_vcpu *vcpu, *gfnp = gfn; kvm_release_pfn_clean(pfn); pfn &= ~mask; - if (!get_page_unless_zero(pfn_to_page(pfn))) - BUG(); + kvm_get_pfn(pfn); *pfnp = pfn; } } } -static bool mmu_invalid_pfn(pfn_t pfn) -{ - return unlikely(is_invalid_pfn(pfn)); -} - static bool handle_abnormal_pfn(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn, pfn_t pfn, unsigned access, int *ret_val) { bool ret = true; /* The pfn is invalid, report the error! */ - if (unlikely(is_invalid_pfn(pfn))) { + if (unlikely(is_error_pfn(pfn))) { *ret_val = kvm_handle_bad_page(vcpu, gfn, pfn); goto exit; } @@ -2582,18 +2771,142 @@ exit: return ret; } +static bool page_fault_can_be_fast(u32 error_code) +{ + /* + * Do not fix the mmio spte with invalid generation number which + * need to be updated by slow page fault path. + */ + if (unlikely(error_code & PFERR_RSVD_MASK)) + return false; + + /* + * #PF can be fast only if the shadow page table is present and it + * is caused by write-protect, that means we just need change the + * W bit of the spte which can be done out of mmu-lock. + */ + if (!(error_code & PFERR_PRESENT_MASK) || + !(error_code & PFERR_WRITE_MASK)) + return false; + + return true; +} + +static bool +fast_pf_fix_direct_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp, + u64 *sptep, u64 spte) +{ + gfn_t gfn; + + WARN_ON(!sp->role.direct); + + /* + * The gfn of direct spte is stable since it is calculated + * by sp->gfn. + */ + gfn = kvm_mmu_page_get_gfn(sp, sptep - sp->spt); + + if (cmpxchg64(sptep, spte, spte | PT_WRITABLE_MASK) == spte) + mark_page_dirty(vcpu->kvm, gfn); + + return true; +} + +/* + * Return value: + * - true: let the vcpu to access on the same address again. + * - false: let the real page fault path to fix it. + */ +static bool fast_page_fault(struct kvm_vcpu *vcpu, gva_t gva, int level, + u32 error_code) +{ + struct kvm_shadow_walk_iterator iterator; + struct kvm_mmu_page *sp; + bool ret = false; + u64 spte = 0ull; + + if (!VALID_PAGE(vcpu->arch.mmu.root_hpa)) + return false; + + if (!page_fault_can_be_fast(error_code)) + return false; + + walk_shadow_page_lockless_begin(vcpu); + for_each_shadow_entry_lockless(vcpu, gva, iterator, spte) + if (!is_shadow_present_pte(spte) || iterator.level < level) + break; + + /* + * If the mapping has been changed, let the vcpu fault on the + * same address again. + */ + if (!is_rmap_spte(spte)) { + ret = true; + goto exit; + } + + sp = page_header(__pa(iterator.sptep)); + if (!is_last_spte(spte, sp->role.level)) + goto exit; + + /* + * Check if it is a spurious fault caused by TLB lazily flushed. + * + * Need not check the access of upper level table entries since + * they are always ACC_ALL. + */ + if (is_writable_pte(spte)) { + ret = true; + goto exit; + } + + /* + * Currently, to simplify the code, only the spte write-protected + * by dirty-log can be fast fixed. + */ + if (!spte_is_locklessly_modifiable(spte)) + goto exit; + + /* + * Do not fix write-permission on the large spte since we only dirty + * the first page into the dirty-bitmap in fast_pf_fix_direct_spte() + * that means other pages are missed if its slot is dirty-logged. + * + * Instead, we let the slow page fault path create a normal spte to + * fix the access. + * + * See the comments in kvm_arch_commit_memory_region(). + */ + if (sp->role.level > PT_PAGE_TABLE_LEVEL) + goto exit; + + /* + * Currently, fast page fault only works for direct mapping since + * the gfn is not stable for indirect shadow page. + * See Documentation/virtual/kvm/locking.txt to get more detail. + */ + ret = fast_pf_fix_direct_spte(vcpu, sp, iterator.sptep, spte); +exit: + trace_fast_page_fault(vcpu, gva, error_code, iterator.sptep, + spte, ret); + walk_shadow_page_lockless_end(vcpu); + + return ret; +} + static bool try_async_pf(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn, gva_t gva, pfn_t *pfn, bool write, bool *writable); +static void make_mmu_pages_available(struct kvm_vcpu *vcpu); -static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn, - bool prefault) +static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, u32 error_code, + gfn_t gfn, bool prefault) { int r; int level; int force_pt_level; pfn_t pfn; unsigned long mmu_seq; - bool map_writable; + bool map_writable, write = error_code & PFERR_WRITE_MASK; force_pt_level = mapping_level_dirty_bitmap(vcpu, gfn); if (likely(!force_pt_level)) { @@ -2610,6 +2923,9 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn, } else level = PT_PAGE_TABLE_LEVEL; + if (fast_page_fault(vcpu, v, level, error_code)) + return 0; + mmu_seq = vcpu->kvm->mmu_notifier_seq; smp_rmb(); @@ -2620,9 +2936,9 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn, return r; spin_lock(&vcpu->kvm->mmu_lock); - if (mmu_notifier_retry(vcpu, mmu_seq)) + if (mmu_notifier_retry(vcpu->kvm, mmu_seq)) goto out_unlock; - kvm_mmu_free_some_pages(vcpu); + make_mmu_pages_available(vcpu); if (likely(!force_pt_level)) transparent_hugepage_adjust(vcpu, &gfn, &pfn, &level); r = __direct_map(vcpu, v, write, map_writable, level, gfn, pfn, @@ -2647,22 +2963,25 @@ static void mmu_free_roots(struct kvm_vcpu *vcpu) if (!VALID_PAGE(vcpu->arch.mmu.root_hpa)) return; - spin_lock(&vcpu->kvm->mmu_lock); + if (vcpu->arch.mmu.shadow_root_level == PT64_ROOT_LEVEL && (vcpu->arch.mmu.root_level == PT64_ROOT_LEVEL || vcpu->arch.mmu.direct_map)) { hpa_t root = vcpu->arch.mmu.root_hpa; + spin_lock(&vcpu->kvm->mmu_lock); sp = page_header(root); --sp->root_count; if (!sp->root_count && sp->role.invalid) { kvm_mmu_prepare_zap_page(vcpu->kvm, sp, &invalid_list); kvm_mmu_commit_zap_page(vcpu->kvm, &invalid_list); } - vcpu->arch.mmu.root_hpa = INVALID_PAGE; spin_unlock(&vcpu->kvm->mmu_lock); + vcpu->arch.mmu.root_hpa = INVALID_PAGE; return; } + + spin_lock(&vcpu->kvm->mmu_lock); for (i = 0; i < 4; ++i) { hpa_t root = vcpu->arch.mmu.pae_root[i]; @@ -2700,7 +3019,7 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu *vcpu) if (vcpu->arch.mmu.shadow_root_level == PT64_ROOT_LEVEL) { spin_lock(&vcpu->kvm->mmu_lock); - kvm_mmu_free_some_pages(vcpu); + make_mmu_pages_available(vcpu); sp = kvm_mmu_get_page(vcpu, 0, 0, PT64_ROOT_LEVEL, 1, ACC_ALL, NULL); ++sp->root_count; @@ -2712,7 +3031,7 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu *vcpu) ASSERT(!VALID_PAGE(root)); spin_lock(&vcpu->kvm->mmu_lock); - kvm_mmu_free_some_pages(vcpu); + make_mmu_pages_available(vcpu); sp = kvm_mmu_get_page(vcpu, i << (30 - PAGE_SHIFT), i << 30, PT32_ROOT_LEVEL, 1, ACC_ALL, @@ -2751,7 +3070,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu *vcpu) ASSERT(!VALID_PAGE(root)); spin_lock(&vcpu->kvm->mmu_lock); - kvm_mmu_free_some_pages(vcpu); + make_mmu_pages_available(vcpu); sp = kvm_mmu_get_page(vcpu, root_gfn, 0, PT64_ROOT_LEVEL, 0, ACC_ALL, NULL); root = __pa(sp->spt); @@ -2785,7 +3104,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu *vcpu) return 1; } spin_lock(&vcpu->kvm->mmu_lock); - kvm_mmu_free_some_pages(vcpu); + make_mmu_pages_available(vcpu); sp = kvm_mmu_get_page(vcpu, root_gfn, i << 30, PT32_ROOT_LEVEL, 0, ACC_ALL, NULL); @@ -2871,6 +3190,7 @@ void kvm_mmu_sync_roots(struct kvm_vcpu *vcpu) mmu_sync_roots(vcpu); spin_unlock(&vcpu->kvm->mmu_lock); } +EXPORT_SYMBOL_GPL(kvm_mmu_sync_roots); static gpa_t nonpaging_gva_to_gpa(struct kvm_vcpu *vcpu, gva_t vaddr, u32 access, struct x86_exception *exception) @@ -2917,6 +3237,9 @@ static u64 walk_shadow_page_get_mmio_spte(struct kvm_vcpu *vcpu, u64 addr) struct kvm_shadow_walk_iterator iterator; u64 spte = 0ull; + if (!VALID_PAGE(vcpu->arch.mmu.root_hpa)) + return spte; + walk_shadow_page_lockless_begin(vcpu); for_each_shadow_entry_lockless(vcpu, addr, iterator, spte) if (!is_shadow_present_pte(spte)) @@ -2926,17 +3249,12 @@ static u64 walk_shadow_page_get_mmio_spte(struct kvm_vcpu *vcpu, u64 addr) return spte; } -/* - * If it is a real mmio page fault, return 1 and emulat the instruction - * directly, return 0 to let CPU fault again on the address, -1 is - * returned if bug is detected. - */ int handle_mmio_page_fault_common(struct kvm_vcpu *vcpu, u64 addr, bool direct) { u64 spte; if (quickly_check_mmio_pf(vcpu, addr, direct)) - return 1; + return RET_MMIO_PF_EMULATE; spte = walk_shadow_page_get_mmio_spte(vcpu, addr); @@ -2944,12 +3262,15 @@ int handle_mmio_page_fault_common(struct kvm_vcpu *vcpu, u64 addr, bool direct) gfn_t gfn = get_mmio_spte_gfn(spte); unsigned access = get_mmio_spte_access(spte); + if (!check_mmio_spte(vcpu->kvm, spte)) + return RET_MMIO_PF_INVALID; + if (direct) addr = 0; trace_handle_mmio_page_fault(addr, gfn, access); vcpu_cache_mmio_info(vcpu, addr, gfn, access); - return 1; + return RET_MMIO_PF_EMULATE; } /* @@ -2957,13 +3278,13 @@ int handle_mmio_page_fault_common(struct kvm_vcpu *vcpu, u64 addr, bool direct) * it's a BUG if the gfn is not a mmio page. */ if (direct && !check_direct_spte_mmio_pf(spte)) - return -1; + return RET_MMIO_PF_BUG; /* * If the page table is zapped by other cpus, let CPU fault again on * the address. */ - return 0; + return RET_MMIO_PF_RETRY; } EXPORT_SYMBOL_GPL(handle_mmio_page_fault_common); @@ -2973,7 +3294,7 @@ static int handle_mmio_page_fault(struct kvm_vcpu *vcpu, u64 addr, int ret; ret = handle_mmio_page_fault_common(vcpu, addr, direct); - WARN_ON(ret < 0); + WARN_ON(ret == RET_MMIO_PF_BUG); return ret; } @@ -2985,8 +3306,12 @@ static int nonpaging_page_fault(struct kvm_vcpu *vcpu, gva_t gva, pgprintk("%s: gva %lx error %x\n", __func__, gva, error_code); - if (unlikely(error_code & PFERR_RSVD_MASK)) - return handle_mmio_page_fault(vcpu, gva, error_code, true); + if (unlikely(error_code & PFERR_RSVD_MASK)) { + r = handle_mmio_page_fault(vcpu, gva, error_code, true); + + if (likely(r != RET_MMIO_PF_INVALID)) + return r; + } r = mmu_topup_memory_caches(vcpu); if (r) @@ -2998,7 +3323,7 @@ static int nonpaging_page_fault(struct kvm_vcpu *vcpu, gva_t gva, gfn = gva >> PAGE_SHIFT; return nonpaging_map(vcpu, gva & PAGE_MASK, - error_code & PFERR_WRITE_MASK, gfn, prefault); + error_code, gfn, prefault); } static int kvm_arch_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn) @@ -3010,7 +3335,7 @@ static int kvm_arch_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn) arch.direct_map = vcpu->arch.mmu.direct_map; arch.cr3 = vcpu->arch.mmu.get_cr3(vcpu); - return kvm_setup_async_pf(vcpu, gva, gfn, &arch); + return kvm_setup_async_pf(vcpu, gva, gfn_to_hva(vcpu->kvm, gfn), &arch); } static bool can_do_async_pf(struct kvm_vcpu *vcpu) @@ -3032,8 +3357,6 @@ static bool try_async_pf(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn, if (!async) return false; /* *pfn has correct page already */ - put_page(pfn_to_page(*pfn)); - if (!prefault && can_do_async_pf(vcpu)) { trace_kvm_try_async_get_page(gva, gfn); if (kvm_find_async_pf_gfn(vcpu, gfn)) { @@ -3064,8 +3387,12 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code, ASSERT(vcpu); ASSERT(VALID_PAGE(vcpu->arch.mmu.root_hpa)); - if (unlikely(error_code & PFERR_RSVD_MASK)) - return handle_mmio_page_fault(vcpu, gpa, error_code, true); + if (unlikely(error_code & PFERR_RSVD_MASK)) { + r = handle_mmio_page_fault(vcpu, gpa, error_code, true); + + if (likely(r != RET_MMIO_PF_INVALID)) + return r; + } r = mmu_topup_memory_caches(vcpu); if (r) @@ -3078,6 +3405,9 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code, } else level = PT_PAGE_TABLE_LEVEL; + if (fast_page_fault(vcpu, gpa, level, error_code)) + return 0; + mmu_seq = vcpu->kvm->mmu_notifier_seq; smp_rmb(); @@ -3088,9 +3418,9 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code, return r; spin_lock(&vcpu->kvm->mmu_lock); - if (mmu_notifier_retry(vcpu, mmu_seq)) + if (mmu_notifier_retry(vcpu->kvm, mmu_seq)) goto out_unlock; - kvm_mmu_free_some_pages(vcpu); + make_mmu_pages_available(vcpu); if (likely(!force_pt_level)) transparent_hugepage_adjust(vcpu, &gfn, &pfn, &level); r = __direct_map(vcpu, gpa, write, map_writable, @@ -3105,18 +3435,11 @@ out_unlock: return 0; } -static void nonpaging_free(struct kvm_vcpu *vcpu) +static void nonpaging_init_context(struct kvm_vcpu *vcpu, + struct kvm_mmu *context) { - mmu_free_roots(vcpu); -} - -static int nonpaging_init_context(struct kvm_vcpu *vcpu, - struct kvm_mmu *context) -{ - context->new_cr3 = nonpaging_new_cr3; context->page_fault = nonpaging_page_fault; context->gva_to_gpa = nonpaging_gva_to_gpa; - context->free = nonpaging_free; context->sync_page = nonpaging_sync_page; context->invlpg = nonpaging_invlpg; context->update_pte = nonpaging_update_pte; @@ -3125,7 +3448,6 @@ static int nonpaging_init_context(struct kvm_vcpu *vcpu, context->root_hpa = INVALID_PAGE; context->direct_map = true; context->nx = false; - return 0; } void kvm_mmu_flush_tlb(struct kvm_vcpu *vcpu) @@ -3133,10 +3455,10 @@ void kvm_mmu_flush_tlb(struct kvm_vcpu *vcpu) ++vcpu->stat.tlb_flush; kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu); } +EXPORT_SYMBOL_GPL(kvm_mmu_flush_tlb); -static void paging_new_cr3(struct kvm_vcpu *vcpu) +void kvm_mmu_new_cr3(struct kvm_vcpu *vcpu) { - pgprintk("%s: cr3 %lx\n", __func__, kvm_read_cr3(vcpu)); mmu_free_roots(vcpu); } @@ -3151,21 +3473,8 @@ static void inject_page_fault(struct kvm_vcpu *vcpu, vcpu->arch.mmu.inject_page_fault(vcpu, fault); } -static void paging_free(struct kvm_vcpu *vcpu) -{ - nonpaging_free(vcpu); -} - -static bool is_rsvd_bits_set(struct kvm_mmu *mmu, u64 gpte, int level) -{ - int bit7; - - bit7 = (gpte >> 7) & 1; - return (gpte & mmu->rsvd_bits_mask[bit7][level-1]) != 0; -} - -static bool sync_mmio_spte(u64 *sptep, gfn_t gfn, unsigned access, - int *nr_present) +static bool sync_mmio_spte(struct kvm *kvm, u64 *sptep, gfn_t gfn, + unsigned access, int *nr_present) { if (unlikely(is_mmio_spte(*sptep))) { if (gfn != get_mmio_spte_gfn(*sptep)) { @@ -3174,13 +3483,27 @@ static bool sync_mmio_spte(u64 *sptep, gfn_t gfn, unsigned access, } (*nr_present)++; - mark_mmio_spte(sptep, gfn, access); + mark_mmio_spte(kvm, sptep, gfn, access); return true; } return false; } +static inline bool is_last_gpte(struct kvm_mmu *mmu, unsigned level, unsigned gpte) +{ + unsigned index; + + index = level - 1; + index |= (gpte & PT_PAGE_SIZE_MASK) >> (PT_PAGE_SIZE_SHIFT - 2); + return mmu->last_pte_bitmap & (1 << index); +} + +#define PTTYPE_EPT 18 /* arbitrary */ +#define PTTYPE PTTYPE_EPT +#include "paging_tmpl.h" +#undef PTTYPE + #define PTTYPE 64 #include "paging_tmpl.h" #undef PTTYPE @@ -3190,15 +3513,19 @@ static bool sync_mmio_spte(u64 *sptep, gfn_t gfn, unsigned access, #undef PTTYPE static void reset_rsvds_bits_mask(struct kvm_vcpu *vcpu, - struct kvm_mmu *context, - int level) + struct kvm_mmu *context) { int maxphyaddr = cpuid_maxphyaddr(vcpu); u64 exb_bit_rsvd = 0; + u64 gbpages_bit_rsvd = 0; + + context->bad_mt_xwr = 0; if (!context->nx) exb_bit_rsvd = rsvd_bits(63, 63); - switch (level) { + if (!guest_cpuid_has_gbpages(vcpu)) + gbpages_bit_rsvd = rsvd_bits(7, 7); + switch (context->root_level) { case PT32_ROOT_LEVEL: /* no rsvd bits for 2 level 4K page table entries */ context->rsvd_bits_mask[0][1] = 0; @@ -3220,7 +3547,7 @@ static void reset_rsvds_bits_mask(struct kvm_vcpu *vcpu, case PT32E_ROOT_LEVEL: context->rsvd_bits_mask[0][2] = rsvd_bits(maxphyaddr, 63) | - rsvd_bits(7, 8) | rsvd_bits(1, 2); /* PDPTE */ + rsvd_bits(5, 8) | rsvd_bits(1, 2); /* PDPTE */ context->rsvd_bits_mask[0][1] = exb_bit_rsvd | rsvd_bits(maxphyaddr, 62); /* PDE */ context->rsvd_bits_mask[0][0] = exb_bit_rsvd | @@ -3232,16 +3559,16 @@ static void reset_rsvds_bits_mask(struct kvm_vcpu *vcpu, break; case PT64_ROOT_LEVEL: context->rsvd_bits_mask[0][3] = exb_bit_rsvd | - rsvd_bits(maxphyaddr, 51) | rsvd_bits(7, 8); + rsvd_bits(maxphyaddr, 51) | rsvd_bits(7, 7); context->rsvd_bits_mask[0][2] = exb_bit_rsvd | - rsvd_bits(maxphyaddr, 51) | rsvd_bits(7, 8); + gbpages_bit_rsvd | rsvd_bits(maxphyaddr, 51); context->rsvd_bits_mask[0][1] = exb_bit_rsvd | rsvd_bits(maxphyaddr, 51); context->rsvd_bits_mask[0][0] = exb_bit_rsvd | rsvd_bits(maxphyaddr, 51); context->rsvd_bits_mask[1][3] = context->rsvd_bits_mask[0][3]; context->rsvd_bits_mask[1][2] = exb_bit_rsvd | - rsvd_bits(maxphyaddr, 51) | + gbpages_bit_rsvd | rsvd_bits(maxphyaddr, 51) | rsvd_bits(13, 29); context->rsvd_bits_mask[1][1] = exb_bit_rsvd | rsvd_bits(maxphyaddr, 51) | @@ -3251,70 +3578,181 @@ static void reset_rsvds_bits_mask(struct kvm_vcpu *vcpu, } } -static int paging64_init_context_common(struct kvm_vcpu *vcpu, - struct kvm_mmu *context, - int level) +static void reset_rsvds_bits_mask_ept(struct kvm_vcpu *vcpu, + struct kvm_mmu *context, bool execonly) +{ + int maxphyaddr = cpuid_maxphyaddr(vcpu); + int pte; + + context->rsvd_bits_mask[0][3] = + rsvd_bits(maxphyaddr, 51) | rsvd_bits(3, 7); + context->rsvd_bits_mask[0][2] = + rsvd_bits(maxphyaddr, 51) | rsvd_bits(3, 6); + context->rsvd_bits_mask[0][1] = + rsvd_bits(maxphyaddr, 51) | rsvd_bits(3, 6); + context->rsvd_bits_mask[0][0] = rsvd_bits(maxphyaddr, 51); + + /* large page */ + context->rsvd_bits_mask[1][3] = context->rsvd_bits_mask[0][3]; + context->rsvd_bits_mask[1][2] = + rsvd_bits(maxphyaddr, 51) | rsvd_bits(12, 29); + context->rsvd_bits_mask[1][1] = + rsvd_bits(maxphyaddr, 51) | rsvd_bits(12, 20); + context->rsvd_bits_mask[1][0] = context->rsvd_bits_mask[0][0]; + + for (pte = 0; pte < 64; pte++) { + int rwx_bits = pte & 7; + int mt = pte >> 3; + if (mt == 0x2 || mt == 0x3 || mt == 0x7 || + rwx_bits == 0x2 || rwx_bits == 0x6 || + (rwx_bits == 0x4 && !execonly)) + context->bad_mt_xwr |= (1ull << pte); + } +} + +void update_permission_bitmask(struct kvm_vcpu *vcpu, + struct kvm_mmu *mmu, bool ept) +{ + unsigned bit, byte, pfec; + u8 map; + bool fault, x, w, u, wf, uf, ff, smapf, cr4_smap, cr4_smep, smap = 0; + + cr4_smep = kvm_read_cr4_bits(vcpu, X86_CR4_SMEP); + cr4_smap = kvm_read_cr4_bits(vcpu, X86_CR4_SMAP); + for (byte = 0; byte < ARRAY_SIZE(mmu->permissions); ++byte) { + pfec = byte << 1; + map = 0; + wf = pfec & PFERR_WRITE_MASK; + uf = pfec & PFERR_USER_MASK; + ff = pfec & PFERR_FETCH_MASK; + /* + * PFERR_RSVD_MASK bit is set in PFEC if the access is not + * subject to SMAP restrictions, and cleared otherwise. The + * bit is only meaningful if the SMAP bit is set in CR4. + */ + smapf = !(pfec & PFERR_RSVD_MASK); + for (bit = 0; bit < 8; ++bit) { + x = bit & ACC_EXEC_MASK; + w = bit & ACC_WRITE_MASK; + u = bit & ACC_USER_MASK; + + if (!ept) { + /* Not really needed: !nx will cause pte.nx to fault */ + x |= !mmu->nx; + /* Allow supervisor writes if !cr0.wp */ + w |= !is_write_protection(vcpu) && !uf; + /* Disallow supervisor fetches of user code if cr4.smep */ + x &= !(cr4_smep && u && !uf); + + /* + * SMAP:kernel-mode data accesses from user-mode + * mappings should fault. A fault is considered + * as a SMAP violation if all of the following + * conditions are ture: + * - X86_CR4_SMAP is set in CR4 + * - An user page is accessed + * - Page fault in kernel mode + * - if CPL = 3 or X86_EFLAGS_AC is clear + * + * Here, we cover the first three conditions. + * The fourth is computed dynamically in + * permission_fault() and is in smapf. + * + * Also, SMAP does not affect instruction + * fetches, add the !ff check here to make it + * clearer. + */ + smap = cr4_smap && u && !uf && !ff; + } else + /* Not really needed: no U/S accesses on ept */ + u = 1; + + fault = (ff && !x) || (uf && !u) || (wf && !w) || + (smapf && smap); + map |= fault << bit; + } + mmu->permissions[byte] = map; + } +} + +static void update_last_pte_bitmap(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu) +{ + u8 map; + unsigned level, root_level = mmu->root_level; + const unsigned ps_set_index = 1 << 2; /* bit 2 of index: ps */ + + if (root_level == PT32E_ROOT_LEVEL) + --root_level; + /* PT_PAGE_TABLE_LEVEL always terminates */ + map = 1 | (1 << ps_set_index); + for (level = PT_DIRECTORY_LEVEL; level <= root_level; ++level) { + if (level <= PT_PDPE_LEVEL + && (mmu->root_level >= PT32E_ROOT_LEVEL || is_pse(vcpu))) + map |= 1 << (ps_set_index | (level - 1)); + } + mmu->last_pte_bitmap = map; +} + +static void paging64_init_context_common(struct kvm_vcpu *vcpu, + struct kvm_mmu *context, + int level) { context->nx = is_nx(vcpu); + context->root_level = level; - reset_rsvds_bits_mask(vcpu, context, level); + reset_rsvds_bits_mask(vcpu, context); + update_permission_bitmask(vcpu, context, false); + update_last_pte_bitmap(vcpu, context); ASSERT(is_pae(vcpu)); - context->new_cr3 = paging_new_cr3; context->page_fault = paging64_page_fault; context->gva_to_gpa = paging64_gva_to_gpa; context->sync_page = paging64_sync_page; context->invlpg = paging64_invlpg; context->update_pte = paging64_update_pte; - context->free = paging_free; - context->root_level = level; context->shadow_root_level = level; context->root_hpa = INVALID_PAGE; context->direct_map = false; - return 0; } -static int paging64_init_context(struct kvm_vcpu *vcpu, - struct kvm_mmu *context) +static void paging64_init_context(struct kvm_vcpu *vcpu, + struct kvm_mmu *context) { - return paging64_init_context_common(vcpu, context, PT64_ROOT_LEVEL); + paging64_init_context_common(vcpu, context, PT64_ROOT_LEVEL); } -static int paging32_init_context(struct kvm_vcpu *vcpu, - struct kvm_mmu *context) +static void paging32_init_context(struct kvm_vcpu *vcpu, + struct kvm_mmu *context) { context->nx = false; + context->root_level = PT32_ROOT_LEVEL; - reset_rsvds_bits_mask(vcpu, context, PT32_ROOT_LEVEL); + reset_rsvds_bits_mask(vcpu, context); + update_permission_bitmask(vcpu, context, false); + update_last_pte_bitmap(vcpu, context); - context->new_cr3 = paging_new_cr3; context->page_fault = paging32_page_fault; context->gva_to_gpa = paging32_gva_to_gpa; - context->free = paging_free; context->sync_page = paging32_sync_page; context->invlpg = paging32_invlpg; context->update_pte = paging32_update_pte; - context->root_level = PT32_ROOT_LEVEL; context->shadow_root_level = PT32E_ROOT_LEVEL; context->root_hpa = INVALID_PAGE; context->direct_map = false; - return 0; } -static int paging32E_init_context(struct kvm_vcpu *vcpu, - struct kvm_mmu *context) +static void paging32E_init_context(struct kvm_vcpu *vcpu, + struct kvm_mmu *context) { - return paging64_init_context_common(vcpu, context, PT32E_ROOT_LEVEL); + paging64_init_context_common(vcpu, context, PT32E_ROOT_LEVEL); } -static int init_kvm_tdp_mmu(struct kvm_vcpu *vcpu) +static void init_kvm_tdp_mmu(struct kvm_vcpu *vcpu) { struct kvm_mmu *context = vcpu->arch.walk_mmu; context->base_role.word = 0; - context->new_cr3 = nonpaging_new_cr3; context->page_fault = tdp_page_fault; - context->free = nonpaging_free; context->sync_page = nonpaging_sync_page; context->invlpg = nonpaging_invlpg; context->update_pte = nonpaging_update_pte; @@ -3325,7 +3763,6 @@ static int init_kvm_tdp_mmu(struct kvm_vcpu *vcpu) context->get_cr3 = get_cr3; context->get_pdptr = kvm_pdptr_read; context->inject_page_fault = kvm_inject_page_fault; - context->nx = is_nx(vcpu); if (!is_paging(vcpu)) { context->nx = false; @@ -3333,62 +3770,81 @@ static int init_kvm_tdp_mmu(struct kvm_vcpu *vcpu) context->root_level = 0; } else if (is_long_mode(vcpu)) { context->nx = is_nx(vcpu); - reset_rsvds_bits_mask(vcpu, context, PT64_ROOT_LEVEL); - context->gva_to_gpa = paging64_gva_to_gpa; context->root_level = PT64_ROOT_LEVEL; + reset_rsvds_bits_mask(vcpu, context); + context->gva_to_gpa = paging64_gva_to_gpa; } else if (is_pae(vcpu)) { context->nx = is_nx(vcpu); - reset_rsvds_bits_mask(vcpu, context, PT32E_ROOT_LEVEL); - context->gva_to_gpa = paging64_gva_to_gpa; context->root_level = PT32E_ROOT_LEVEL; + reset_rsvds_bits_mask(vcpu, context); + context->gva_to_gpa = paging64_gva_to_gpa; } else { context->nx = false; - reset_rsvds_bits_mask(vcpu, context, PT32_ROOT_LEVEL); - context->gva_to_gpa = paging32_gva_to_gpa; context->root_level = PT32_ROOT_LEVEL; + reset_rsvds_bits_mask(vcpu, context); + context->gva_to_gpa = paging32_gva_to_gpa; } - return 0; + update_permission_bitmask(vcpu, context, false); + update_last_pte_bitmap(vcpu, context); } -int kvm_init_shadow_mmu(struct kvm_vcpu *vcpu, struct kvm_mmu *context) +void kvm_init_shadow_mmu(struct kvm_vcpu *vcpu, struct kvm_mmu *context) { - int r; bool smep = kvm_read_cr4_bits(vcpu, X86_CR4_SMEP); ASSERT(vcpu); ASSERT(!VALID_PAGE(vcpu->arch.mmu.root_hpa)); if (!is_paging(vcpu)) - r = nonpaging_init_context(vcpu, context); + nonpaging_init_context(vcpu, context); else if (is_long_mode(vcpu)) - r = paging64_init_context(vcpu, context); + paging64_init_context(vcpu, context); else if (is_pae(vcpu)) - r = paging32E_init_context(vcpu, context); + paging32E_init_context(vcpu, context); else - r = paging32_init_context(vcpu, context); + paging32_init_context(vcpu, context); + vcpu->arch.mmu.base_role.nxe = is_nx(vcpu); vcpu->arch.mmu.base_role.cr4_pae = !!is_pae(vcpu); vcpu->arch.mmu.base_role.cr0_wp = is_write_protection(vcpu); vcpu->arch.mmu.base_role.smep_andnot_wp = smep && !is_write_protection(vcpu); - - return r; } EXPORT_SYMBOL_GPL(kvm_init_shadow_mmu); -static int init_kvm_softmmu(struct kvm_vcpu *vcpu) +void kvm_init_shadow_ept_mmu(struct kvm_vcpu *vcpu, struct kvm_mmu *context, + bool execonly) { - int r = kvm_init_shadow_mmu(vcpu, vcpu->arch.walk_mmu); + ASSERT(vcpu); + ASSERT(!VALID_PAGE(vcpu->arch.mmu.root_hpa)); + + context->shadow_root_level = kvm_x86_ops->get_tdp_level(); + + context->nx = true; + context->page_fault = ept_page_fault; + context->gva_to_gpa = ept_gva_to_gpa; + context->sync_page = ept_sync_page; + context->invlpg = ept_invlpg; + context->update_pte = ept_update_pte; + context->root_level = context->shadow_root_level; + context->root_hpa = INVALID_PAGE; + context->direct_map = false; + + update_permission_bitmask(vcpu, context, true); + reset_rsvds_bits_mask_ept(vcpu, context, execonly); +} +EXPORT_SYMBOL_GPL(kvm_init_shadow_ept_mmu); +static void init_kvm_softmmu(struct kvm_vcpu *vcpu) +{ + kvm_init_shadow_mmu(vcpu, vcpu->arch.walk_mmu); vcpu->arch.walk_mmu->set_cr3 = kvm_x86_ops->set_cr3; vcpu->arch.walk_mmu->get_cr3 = get_cr3; vcpu->arch.walk_mmu->get_pdptr = kvm_pdptr_read; vcpu->arch.walk_mmu->inject_page_fault = kvm_inject_page_fault; - - return r; } -static int init_kvm_nested_mmu(struct kvm_vcpu *vcpu) +static void init_kvm_nested_mmu(struct kvm_vcpu *vcpu) { struct kvm_mmu *g_context = &vcpu->arch.nested_mmu; @@ -3408,25 +3864,26 @@ static int init_kvm_nested_mmu(struct kvm_vcpu *vcpu) g_context->gva_to_gpa = nonpaging_gva_to_gpa_nested; } else if (is_long_mode(vcpu)) { g_context->nx = is_nx(vcpu); - reset_rsvds_bits_mask(vcpu, g_context, PT64_ROOT_LEVEL); g_context->root_level = PT64_ROOT_LEVEL; + reset_rsvds_bits_mask(vcpu, g_context); g_context->gva_to_gpa = paging64_gva_to_gpa_nested; } else if (is_pae(vcpu)) { g_context->nx = is_nx(vcpu); - reset_rsvds_bits_mask(vcpu, g_context, PT32E_ROOT_LEVEL); g_context->root_level = PT32E_ROOT_LEVEL; + reset_rsvds_bits_mask(vcpu, g_context); g_context->gva_to_gpa = paging64_gva_to_gpa_nested; } else { g_context->nx = false; - reset_rsvds_bits_mask(vcpu, g_context, PT32_ROOT_LEVEL); g_context->root_level = PT32_ROOT_LEVEL; + reset_rsvds_bits_mask(vcpu, g_context); g_context->gva_to_gpa = paging32_gva_to_gpa_nested; } - return 0; + update_permission_bitmask(vcpu, g_context, false); + update_last_pte_bitmap(vcpu, g_context); } -static int init_kvm_mmu(struct kvm_vcpu *vcpu) +static void init_kvm_mmu(struct kvm_vcpu *vcpu) { if (mmu_is_nested(vcpu)) return init_kvm_nested_mmu(vcpu); @@ -3436,18 +3893,12 @@ static int init_kvm_mmu(struct kvm_vcpu *vcpu) return init_kvm_softmmu(vcpu); } -static void destroy_kvm_mmu(struct kvm_vcpu *vcpu) +void kvm_mmu_reset_context(struct kvm_vcpu *vcpu) { ASSERT(vcpu); - if (VALID_PAGE(vcpu->arch.mmu.root_hpa)) - /* mmu.free() should set root_hpa = INVALID_PAGE */ - vcpu->arch.mmu.free(vcpu); -} -int kvm_mmu_reset_context(struct kvm_vcpu *vcpu) -{ - destroy_kvm_mmu(vcpu); - return init_kvm_mmu(vcpu); + kvm_mmu_unload(vcpu); + init_kvm_mmu(vcpu); } EXPORT_SYMBOL_GPL(kvm_mmu_reset_context); @@ -3459,9 +3910,7 @@ int kvm_mmu_load(struct kvm_vcpu *vcpu) if (r) goto out; r = mmu_alloc_roots(vcpu); - spin_lock(&vcpu->kvm->mmu_lock); - mmu_sync_roots(vcpu); - spin_unlock(&vcpu->kvm->mmu_lock); + kvm_mmu_sync_roots(vcpu); if (r) goto out; /* set_cr3() should ensure TLB has been flushed */ @@ -3474,6 +3923,7 @@ EXPORT_SYMBOL_GPL(kvm_mmu_load); void kvm_mmu_unload(struct kvm_vcpu *vcpu) { mmu_free_roots(vcpu); + WARN_ON(VALID_PAGE(vcpu->arch.mmu.root_hpa)); } EXPORT_SYMBOL_GPL(kvm_mmu_unload); @@ -3498,8 +3948,8 @@ static bool need_remote_flush(u64 old, u64 new) return true; if ((old ^ new) & PT64_BASE_ADDR_MASK) return true; - old ^= PT64_NX_MASK; - new ^= PT64_NX_MASK; + old ^= shadow_nx_mask; + new ^= shadow_nx_mask; return (old & ~new & PT64_PERM_MASK) != 0; } @@ -3530,7 +3980,7 @@ static u64 mmu_pte_write_fetch_gpte(struct kvm_vcpu *vcpu, gpa_t *gpa, /* Handle a 32-bit guest writing two halves of a 64-bit gpte */ *gpa &= ~(gpa_t)7; *bytes = 8; - r = kvm_read_guest(vcpu->kvm, *gpa, &gentry, min(*bytes, 8)); + r = kvm_read_guest(vcpu->kvm, *gpa, &gentry, 8); if (r) gentry = 0; new = (const u8 *)&gentry; @@ -3555,13 +4005,13 @@ static u64 mmu_pte_write_fetch_gpte(struct kvm_vcpu *vcpu, gpa_t *gpa, * If we're seeing too many writes to a page, it may no longer be a page table, * or we may be forking, in which case it is better to unmap the page. */ -static bool detect_write_flooding(struct kvm_mmu_page *sp, u64 *spte) +static bool detect_write_flooding(struct kvm_mmu_page *sp) { /* * Skip write-flooding detected for the sp whose level is 1, because * it can become unsync, then the guest page is not write-protected. */ - if (sp->role.level == 1) + if (sp->role.level == PT_PAGE_TABLE_LEVEL) return false; return ++sp->write_flooding_count >= 3; @@ -3632,7 +4082,6 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa, gfn_t gfn = gpa >> PAGE_SHIFT; union kvm_mmu_page_role mask = { .word = 0 }; struct kvm_mmu_page *sp; - struct hlist_node *node; LIST_HEAD(invalid_list); u64 entry, gentry, *spte; int npte; @@ -3663,11 +4112,9 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa, kvm_mmu_audit(vcpu, AUDIT_PRE_PTE_WRITE); mask.cr0_wp = mask.cr4_pae = mask.nxe = 1; - for_each_gfn_indirect_valid_sp(vcpu->kvm, sp, gfn, node) { - spte = get_written_sptes(sp, gpa, &npte); - + for_each_gfn_indirect_valid_sp(vcpu->kvm, sp, gfn) { if (detect_write_misaligned(sp, gpa, bytes) || - detect_write_flooding(sp, spte)) { + detect_write_flooding(sp)) { zap_page |= !!kvm_mmu_prepare_zap_page(vcpu->kvm, sp, &invalid_list); ++vcpu->kvm->stat.mmu_flooded; @@ -3686,7 +4133,7 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa, !((sp->role.word ^ vcpu->arch.mmu.base_role.word) & mask.word) && rmap_can_add(vcpu)) mmu_pte_write_new_pte(vcpu, sp, spte, &gentry); - if (!remote_flush && need_remote_flush(entry, *spte)) + if (need_remote_flush(entry, *spte)) remote_flush = true; ++spte; } @@ -3713,17 +4160,17 @@ int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva) } EXPORT_SYMBOL_GPL(kvm_mmu_unprotect_page_virt); -void __kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu) +static void make_mmu_pages_available(struct kvm_vcpu *vcpu) { LIST_HEAD(invalid_list); - while (kvm_mmu_available_pages(vcpu->kvm) < KVM_REFILL_PAGES && - !list_empty(&vcpu->kvm->arch.active_mmu_pages)) { - struct kvm_mmu_page *sp; + if (likely(kvm_mmu_available_pages(vcpu->kvm) >= KVM_MIN_FREE_MMU_PAGES)) + return; + + while (kvm_mmu_available_pages(vcpu->kvm) < KVM_REFILL_PAGES) { + if (!prepare_zap_oldest_mmu_page(vcpu->kvm, &invalid_list)) + break; - sp = container_of(vcpu->kvm->arch.active_mmu_pages.prev, - struct kvm_mmu_page, link); - kvm_mmu_prepare_zap_page(vcpu->kvm, sp, &invalid_list); ++vcpu->kvm->stat.mmu_recycled; } kvm_mmu_commit_zap_page(vcpu->kvm, &invalid_list); @@ -3760,7 +4207,7 @@ int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gva_t cr2, u32 error_code, switch (er) { case EMULATE_DONE: return 1; - case EMULATE_DO_MMIO: + case EMULATE_USER_EXIT: ++vcpu->stat.mmio_exits; /* fall through */ case EMULATE_FAIL: @@ -3835,111 +4282,233 @@ int kvm_mmu_create(struct kvm_vcpu *vcpu) return alloc_mmu_pages(vcpu); } -int kvm_mmu_setup(struct kvm_vcpu *vcpu) +void kvm_mmu_setup(struct kvm_vcpu *vcpu) { ASSERT(vcpu); ASSERT(!VALID_PAGE(vcpu->arch.mmu.root_hpa)); - return init_kvm_mmu(vcpu); + init_kvm_mmu(vcpu); } void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot) { - struct kvm_mmu_page *sp; + struct kvm_memory_slot *memslot; + gfn_t last_gfn; + int i; - list_for_each_entry(sp, &kvm->arch.active_mmu_pages, link) { - int i; - u64 *pt; + memslot = id_to_memslot(kvm->memslots, slot); + last_gfn = memslot->base_gfn + memslot->npages - 1; - if (!test_bit(slot, sp->slot_bitmap)) - continue; + spin_lock(&kvm->mmu_lock); - pt = sp->spt; - for (i = 0; i < PT64_ENT_PER_PAGE; ++i) { - if (!is_shadow_present_pte(pt[i]) || - !is_last_spte(pt[i], sp->role.level)) - continue; + for (i = PT_PAGE_TABLE_LEVEL; + i < PT_PAGE_TABLE_LEVEL + KVM_NR_PAGE_SIZES; ++i) { + unsigned long *rmapp; + unsigned long last_index, index; - if (is_large_pte(pt[i])) { - drop_spte(kvm, &pt[i]); - --kvm->stat.lpages; - continue; - } + rmapp = memslot->arch.rmap[i - PT_PAGE_TABLE_LEVEL]; + last_index = gfn_to_index(last_gfn, memslot->base_gfn, i); - /* avoid RMW */ - if (is_writable_pte(pt[i])) - mmu_spte_update(&pt[i], - pt[i] & ~PT_WRITABLE_MASK); + for (index = 0; index <= last_index; ++index, ++rmapp) { + if (*rmapp) + __rmap_write_protect(kvm, rmapp, false); + + if (need_resched() || spin_needbreak(&kvm->mmu_lock)) + cond_resched_lock(&kvm->mmu_lock); } } + + spin_unlock(&kvm->mmu_lock); + + /* + * kvm_mmu_slot_remove_write_access() and kvm_vm_ioctl_get_dirty_log() + * which do tlb flush out of mmu-lock should be serialized by + * kvm->slots_lock otherwise tlb flush would be missed. + */ + lockdep_assert_held(&kvm->slots_lock); + + /* + * We can flush all the TLBs out of the mmu lock without TLB + * corruption since we just change the spte from writable to + * readonly so that we only need to care the case of changing + * spte from present to present (changing the spte from present + * to nonpresent will flush all the TLBs immediately), in other + * words, the only case we care is mmu_spte_update() where we + * haved checked SPTE_HOST_WRITEABLE | SPTE_MMU_WRITEABLE + * instead of PT_WRITABLE_MASK, that means it does not depend + * on PT_WRITABLE_MASK anymore. + */ kvm_flush_remote_tlbs(kvm); } -void kvm_mmu_zap_all(struct kvm *kvm) +#define BATCH_ZAP_PAGES 10 +static void kvm_zap_obsolete_pages(struct kvm *kvm) { struct kvm_mmu_page *sp, *node; - LIST_HEAD(invalid_list); + int batch = 0; - spin_lock(&kvm->mmu_lock); restart: - list_for_each_entry_safe(sp, node, &kvm->arch.active_mmu_pages, link) - if (kvm_mmu_prepare_zap_page(kvm, sp, &invalid_list)) + list_for_each_entry_safe_reverse(sp, node, + &kvm->arch.active_mmu_pages, link) { + int ret; + + /* + * No obsolete page exists before new created page since + * active_mmu_pages is the FIFO list. + */ + if (!is_obsolete_sp(kvm, sp)) + break; + + /* + * Since we are reversely walking the list and the invalid + * list will be moved to the head, skip the invalid page + * can help us to avoid the infinity list walking. + */ + if (sp->role.invalid) + continue; + + /* + * Need not flush tlb since we only zap the sp with invalid + * generation number. + */ + if (batch >= BATCH_ZAP_PAGES && + cond_resched_lock(&kvm->mmu_lock)) { + batch = 0; goto restart; + } - kvm_mmu_commit_zap_page(kvm, &invalid_list); + ret = kvm_mmu_prepare_zap_page(kvm, sp, + &kvm->arch.zapped_obsolete_pages); + batch += ret; + + if (ret) + goto restart; + } + + /* + * Should flush tlb before free page tables since lockless-walking + * may use the pages. + */ + kvm_mmu_commit_zap_page(kvm, &kvm->arch.zapped_obsolete_pages); +} + +/* + * Fast invalidate all shadow pages and use lock-break technique + * to zap obsolete pages. + * + * It's required when memslot is being deleted or VM is being + * destroyed, in these cases, we should ensure that KVM MMU does + * not use any resource of the being-deleted slot or all slots + * after calling the function. + */ +void kvm_mmu_invalidate_zap_all_pages(struct kvm *kvm) +{ + spin_lock(&kvm->mmu_lock); + trace_kvm_mmu_invalidate_zap_all_pages(kvm); + kvm->arch.mmu_valid_gen++; + + /* + * Notify all vcpus to reload its shadow page table + * and flush TLB. Then all vcpus will switch to new + * shadow page table with the new mmu_valid_gen. + * + * Note: we should do this under the protection of + * mmu-lock, otherwise, vcpu would purge shadow page + * but miss tlb flush. + */ + kvm_reload_remote_mmus(kvm); + + kvm_zap_obsolete_pages(kvm); spin_unlock(&kvm->mmu_lock); } -static void kvm_mmu_remove_some_alloc_mmu_pages(struct kvm *kvm, - struct list_head *invalid_list) +static bool kvm_has_zapped_obsolete_pages(struct kvm *kvm) { - struct kvm_mmu_page *page; + return unlikely(!list_empty_careful(&kvm->arch.zapped_obsolete_pages)); +} - page = container_of(kvm->arch.active_mmu_pages.prev, - struct kvm_mmu_page, link); - kvm_mmu_prepare_zap_page(kvm, page, invalid_list); +void kvm_mmu_invalidate_mmio_sptes(struct kvm *kvm) +{ + /* + * The very rare case: if the generation-number is round, + * zap all shadow pages. + */ + if (unlikely(kvm_current_mmio_generation(kvm) >= MMIO_MAX_GEN)) { + printk_ratelimited(KERN_INFO "kvm: zapping shadow pages for mmio generation wraparound\n"); + kvm_mmu_invalidate_zap_all_pages(kvm); + } } -static int mmu_shrink(struct shrinker *shrink, struct shrink_control *sc) +static unsigned long +mmu_shrink_scan(struct shrinker *shrink, struct shrink_control *sc) { struct kvm *kvm; - struct kvm *kvm_freed = NULL; int nr_to_scan = sc->nr_to_scan; + unsigned long freed = 0; - if (nr_to_scan == 0) - goto out; - - raw_spin_lock(&kvm_lock); + spin_lock(&kvm_lock); list_for_each_entry(kvm, &vm_list, vm_list) { int idx; LIST_HEAD(invalid_list); + /* + * Never scan more than sc->nr_to_scan VM instances. + * Will not hit this condition practically since we do not try + * to shrink more than one VM and it is very unlikely to see + * !n_used_mmu_pages so many times. + */ + if (!nr_to_scan--) + break; + /* + * n_used_mmu_pages is accessed without holding kvm->mmu_lock + * here. We may skip a VM instance errorneosly, but we do not + * want to shrink a VM that only started to populate its MMU + * anyway. + */ + if (!kvm->arch.n_used_mmu_pages && + !kvm_has_zapped_obsolete_pages(kvm)) + continue; + idx = srcu_read_lock(&kvm->srcu); spin_lock(&kvm->mmu_lock); - if (!kvm_freed && nr_to_scan > 0 && - kvm->arch.n_used_mmu_pages > 0) { - kvm_mmu_remove_some_alloc_mmu_pages(kvm, - &invalid_list); - kvm_freed = kvm; + + if (kvm_has_zapped_obsolete_pages(kvm)) { + kvm_mmu_commit_zap_page(kvm, + &kvm->arch.zapped_obsolete_pages); + goto unlock; } - nr_to_scan--; + if (prepare_zap_oldest_mmu_page(kvm, &invalid_list)) + freed++; kvm_mmu_commit_zap_page(kvm, &invalid_list); + +unlock: spin_unlock(&kvm->mmu_lock); srcu_read_unlock(&kvm->srcu, idx); + + /* + * unfair on small ones + * per-vm shrinkers cry out + * sadness comes quickly + */ + list_move_tail(&kvm->vm_list, &vm_list); + break; } - if (kvm_freed) - list_move_tail(&kvm_freed->vm_list, &vm_list); - raw_spin_unlock(&kvm_lock); + spin_unlock(&kvm_lock); + return freed; +} -out: +static unsigned long +mmu_shrink_count(struct shrinker *shrink, struct shrink_control *sc) +{ return percpu_counter_read_positive(&kvm_total_used_mmu_pages); } static struct shrinker mmu_shrinker = { - .shrink = mmu_shrink, + .count_objects = mmu_shrink_count, + .scan_objects = mmu_shrink_scan, .seeks = DEFAULT_SEEKS * 10, }; @@ -4005,6 +4574,9 @@ int kvm_mmu_get_spte_hierarchy(struct kvm_vcpu *vcpu, u64 addr, u64 sptes[4]) u64 spte; int nr_sptes = 0; + if (!VALID_PAGE(vcpu->arch.mmu.root_hpa)) + return nr_sptes; + walk_shadow_page_lockless_begin(vcpu); for_each_shadow_entry_lockless(vcpu, addr, iterator, spte) { sptes[iterator.level-1] = spte; @@ -4022,7 +4594,7 @@ void kvm_mmu_destroy(struct kvm_vcpu *vcpu) { ASSERT(vcpu); - destroy_kvm_mmu(vcpu); + kvm_mmu_unload(vcpu); free_mmu_pages(vcpu); mmu_free_memory_caches(vcpu); } |