From 283d0c65e4c277cf81d71418341c74eee199d149 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Sun, 20 Dec 2009 14:25:19 +0200 Subject: KVM: Disentangle mmu notifiers and coalesced_mmio registration They aren't related. Signed-off-by: Avi Kivity --- virt/kvm/kvm_main.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index a944be392d6e..92f5c98b6c4d 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -374,17 +374,16 @@ static struct kvm *kvm_create_vm(void) #endif #if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER) - { - kvm->mmu_notifier.ops = &kvm_mmu_notifier_ops; - r = mmu_notifier_register(&kvm->mmu_notifier, current->mm); - if (r) { + kvm->mmu_notifier.ops = &kvm_mmu_notifier_ops; + r = mmu_notifier_register(&kvm->mmu_notifier, current->mm); +#endif + + if (r) { #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET - put_page(page); + put_page(page); #endif - goto out_err; - } + goto out_err; } -#endif kvm->mm = current->mm; atomic_inc(&kvm->mm->mm_count); @@ -406,11 +405,8 @@ static struct kvm *kvm_create_vm(void) out: return kvm; -#if defined(KVM_COALESCED_MMIO_PAGE_OFFSET) || \ - (defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)) out_err: hardware_disable_all(); -#endif out_err_nodisable: kfree(kvm); return ERR_PTR(r); -- cgit v1.2.3 From 4c07b0a4b6df45fd8c6d07d838b4faf65eed2ba4 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Sun, 20 Dec 2009 14:54:04 +0200 Subject: KVM: Remove ifdefs from mmu notifier initialization Signed-off-by: Avi Kivity --- virt/kvm/kvm_main.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 92f5c98b6c4d..c10d117c4317 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -341,6 +341,20 @@ static const struct mmu_notifier_ops kvm_mmu_notifier_ops = { .change_pte = kvm_mmu_notifier_change_pte, .release = kvm_mmu_notifier_release, }; + +static int kvm_init_mmu_notifier(struct kvm *kvm) +{ + kvm->mmu_notifier.ops = &kvm_mmu_notifier_ops; + return mmu_notifier_register(&kvm->mmu_notifier, current->mm); +} + +#else /* !(CONFIG_MMU_NOTIFIER && KVM_ARCH_WANT_MMU_NOTIFIER) */ + +static int kvm_init_mmu_notifier(struct kvm *kvm) +{ + return 0; +} + #endif /* CONFIG_MMU_NOTIFIER && KVM_ARCH_WANT_MMU_NOTIFIER */ static struct kvm *kvm_create_vm(void) @@ -373,11 +387,7 @@ static struct kvm *kvm_create_vm(void) (struct kvm_coalesced_mmio_ring *)page_address(page); #endif -#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER) - kvm->mmu_notifier.ops = &kvm_mmu_notifier_ops; - r = mmu_notifier_register(&kvm->mmu_notifier, current->mm); -#endif - + r = kvm_init_mmu_notifier(kvm); if (r) { #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET put_page(page); -- cgit v1.2.3 From 980da6ce573b7c40886406674ff8f022a975e65e Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Sun, 20 Dec 2009 15:13:43 +0200 Subject: KVM: Simplify coalesced mmio initialization - add destructor function - move related allocation into constructor - add stubs for !CONFIG_KVM_MMIO Signed-off-by: Avi Kivity --- virt/kvm/kvm_main.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index c10d117c4317..bc23b8e0609b 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -51,9 +51,7 @@ #include #include -#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET #include "coalesced_mmio.h" -#endif #define CREATE_TRACE_POINTS #include @@ -468,10 +466,7 @@ static void kvm_destroy_vm(struct kvm *kvm) kvm_free_irq_routing(kvm); kvm_io_bus_destroy(&kvm->pio_bus); kvm_io_bus_destroy(&kvm->mmio_bus); -#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET - if (kvm->coalesced_mmio_ring != NULL) - free_page((unsigned long)kvm->coalesced_mmio_ring); -#endif + kvm_coalesced_mmio_free(kvm); #if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER) mmu_notifier_unregister(&kvm->mmu_notifier, kvm->mm); #else -- cgit v1.2.3 From 46a26bf55714c1e2f17e34683292a389acb8e601 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:16 -0200 Subject: KVM: modify memslots layout in struct kvm Have a pointer to an allocated region inside struct kvm. [alex: fix ppc book 3s] Signed-off-by: Alexander Graf Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 36 +++++++++++++++++++++++------------- 1 file changed, 23 insertions(+), 13 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index bc23b8e0609b..86dd8f3d29c9 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -375,12 +375,16 @@ static struct kvm *kvm_create_vm(void) INIT_HLIST_HEAD(&kvm->irq_ack_notifier_list); #endif + r = -ENOMEM; + kvm->memslots = kzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); + if (!kvm->memslots) + goto out_err; + #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET page = alloc_page(GFP_KERNEL | __GFP_ZERO); - if (!page) { - r = -ENOMEM; + if (!page) goto out_err; - } + kvm->coalesced_mmio_ring = (struct kvm_coalesced_mmio_ring *)page_address(page); #endif @@ -416,6 +420,7 @@ out: out_err: hardware_disable_all(); out_err_nodisable: + kfree(kvm->memslots); kfree(kvm); return ERR_PTR(r); } @@ -450,9 +455,12 @@ static void kvm_free_physmem_slot(struct kvm_memory_slot *free, void kvm_free_physmem(struct kvm *kvm) { int i; + struct kvm_memslots *slots = kvm->memslots; + + for (i = 0; i < slots->nmemslots; ++i) + kvm_free_physmem_slot(&slots->memslots[i], NULL); - for (i = 0; i < kvm->nmemslots; ++i) - kvm_free_physmem_slot(&kvm->memslots[i], NULL); + kfree(kvm->memslots); } static void kvm_destroy_vm(struct kvm *kvm) @@ -533,7 +541,7 @@ int __kvm_set_memory_region(struct kvm *kvm, if (mem->guest_phys_addr + mem->memory_size < mem->guest_phys_addr) goto out; - memslot = &kvm->memslots[mem->slot]; + memslot = &kvm->memslots->memslots[mem->slot]; base_gfn = mem->guest_phys_addr >> PAGE_SHIFT; npages = mem->memory_size >> PAGE_SHIFT; @@ -554,7 +562,7 @@ int __kvm_set_memory_region(struct kvm *kvm, /* Check for overlaps */ r = -EEXIST; for (i = 0; i < KVM_MEMORY_SLOTS; ++i) { - struct kvm_memory_slot *s = &kvm->memslots[i]; + struct kvm_memory_slot *s = &kvm->memslots->memslots[i]; if (s == memslot || !s->npages) continue; @@ -656,8 +664,8 @@ skip_lpage: kvm_arch_flush_shadow(kvm); spin_lock(&kvm->mmu_lock); - if (mem->slot >= kvm->nmemslots) - kvm->nmemslots = mem->slot + 1; + if (mem->slot >= kvm->memslots->nmemslots) + kvm->memslots->nmemslots = mem->slot + 1; *memslot = new; spin_unlock(&kvm->mmu_lock); @@ -727,7 +735,7 @@ int kvm_get_dirty_log(struct kvm *kvm, if (log->slot >= KVM_MEMORY_SLOTS) goto out; - memslot = &kvm->memslots[log->slot]; + memslot = &kvm->memslots->memslots[log->slot]; r = -ENOENT; if (!memslot->dirty_bitmap) goto out; @@ -781,9 +789,10 @@ EXPORT_SYMBOL_GPL(kvm_is_error_hva); struct kvm_memory_slot *gfn_to_memslot_unaliased(struct kvm *kvm, gfn_t gfn) { int i; + struct kvm_memslots *slots = kvm->memslots; - for (i = 0; i < kvm->nmemslots; ++i) { - struct kvm_memory_slot *memslot = &kvm->memslots[i]; + for (i = 0; i < slots->nmemslots; ++i) { + struct kvm_memory_slot *memslot = &slots->memslots[i]; if (gfn >= memslot->base_gfn && gfn < memslot->base_gfn + memslot->npages) @@ -802,10 +811,11 @@ struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn) int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn) { int i; + struct kvm_memslots *slots = kvm->memslots; gfn = unalias_gfn(kvm, gfn); for (i = 0; i < KVM_MEMORY_SLOTS; ++i) { - struct kvm_memory_slot *memslot = &kvm->memslots[i]; + struct kvm_memory_slot *memslot = &slots->memslots[i]; if (gfn >= memslot->base_gfn && gfn < memslot->base_gfn + memslot->npages) -- cgit v1.2.3 From f7784b8ec9b6a041fa828cfbe9012fe51933f5ac Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:18 -0200 Subject: KVM: split kvm_arch_set_memory_region into prepare and commit Required for SRCU convertion later. Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 86dd8f3d29c9..c9f6cfe83120 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -663,6 +663,10 @@ skip_lpage: if (!npages) kvm_arch_flush_shadow(kvm); + r = kvm_arch_prepare_memory_region(kvm, &new, old, mem, user_alloc); + if (r) + goto out_free; + spin_lock(&kvm->mmu_lock); if (mem->slot >= kvm->memslots->nmemslots) kvm->memslots->nmemslots = mem->slot + 1; @@ -670,13 +674,7 @@ skip_lpage: *memslot = new; spin_unlock(&kvm->mmu_lock); - r = kvm_arch_set_memory_region(kvm, mem, old, user_alloc); - if (r) { - spin_lock(&kvm->mmu_lock); - *memslot = old; - spin_unlock(&kvm->mmu_lock); - goto out_free; - } + kvm_arch_commit_memory_region(kvm, mem, old, user_alloc); kvm_free_physmem_slot(&old, npages ? &new : NULL); /* Slot deletion case: we have to update the current slot */ -- cgit v1.2.3 From 506f0d6f9c40ae7d9634acf3c26358810f42c24a Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:19 -0200 Subject: KVM: introduce gfn_to_pfn_memslot Which takes a memslot pointer instead of using kvm->memslots. To be used by SRCU convertion later. Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 33 +++++++++++++++++++++++++-------- 1 file changed, 25 insertions(+), 8 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index c9f6cfe83120..4e2321c733f7 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -835,21 +835,14 @@ unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn) } EXPORT_SYMBOL_GPL(gfn_to_hva); -pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn) +static pfn_t hva_to_pfn(struct kvm *kvm, unsigned long addr) { struct page *page[1]; - unsigned long addr; int npages; pfn_t pfn; might_sleep(); - addr = gfn_to_hva(kvm, gfn); - if (kvm_is_error_hva(addr)) { - get_page(bad_page); - return page_to_pfn(bad_page); - } - npages = get_user_pages_fast(addr, 1, 1, page); if (unlikely(npages != 1)) { @@ -874,8 +867,32 @@ pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn) return pfn; } +pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn) +{ + unsigned long addr; + + addr = gfn_to_hva(kvm, gfn); + if (kvm_is_error_hva(addr)) { + get_page(bad_page); + return page_to_pfn(bad_page); + } + + return hva_to_pfn(kvm, addr); +} EXPORT_SYMBOL_GPL(gfn_to_pfn); +static unsigned long gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn) +{ + return (slot->userspace_addr + (gfn - slot->base_gfn) * PAGE_SIZE); +} + +pfn_t gfn_to_pfn_memslot(struct kvm *kvm, + struct kvm_memory_slot *slot, gfn_t gfn) +{ + unsigned long addr = gfn_to_hva_memslot(slot, gfn); + return hva_to_pfn(kvm, addr); +} + struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn) { pfn_t pfn; -- cgit v1.2.3 From 3ad26d8139a82b0510b1e0435ee82ae461d33401 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:20 -0200 Subject: KVM: use gfn_to_pfn_memslot in kvm_iommu_map_pages So its possible to iommu map a memslot before making it visible to kvm. Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 4e2321c733f7..87d296d8b270 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -684,7 +684,7 @@ skip_lpage: spin_unlock(&kvm->mmu_lock); #ifdef CONFIG_DMAR /* map the pages in iommu page table */ - r = kvm_iommu_map_pages(kvm, base_gfn, npages); + r = kvm_iommu_map_pages(kvm, memslot); if (r) goto out; #endif -- cgit v1.2.3 From bc6678a33d9b952981a8e44a4f876c3ad64ca4d8 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:21 -0200 Subject: KVM: introduce kvm->srcu and convert kvm_set_memory_region to SRCU update Use two steps for memslot deletion: mark the slot invalid (which stops instantiation of new shadow pages for that slot, but allows destruction), then instantiate the new empty slot. Also simplifies kvm_handle_hva locking. Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 141 ++++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 105 insertions(+), 36 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 87d296d8b270..2bb24a814fdf 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -44,6 +44,7 @@ #include #include #include +#include #include #include @@ -213,7 +214,7 @@ static void kvm_mmu_notifier_invalidate_page(struct mmu_notifier *mn, unsigned long address) { struct kvm *kvm = mmu_notifier_to_kvm(mn); - int need_tlb_flush; + int need_tlb_flush, idx; /* * When ->invalidate_page runs, the linux pte has been zapped @@ -233,10 +234,12 @@ static void kvm_mmu_notifier_invalidate_page(struct mmu_notifier *mn, * pte after kvm_unmap_hva returned, without noticing the page * is going to be freed. */ + idx = srcu_read_lock(&kvm->srcu); spin_lock(&kvm->mmu_lock); kvm->mmu_notifier_seq++; need_tlb_flush = kvm_unmap_hva(kvm, address); spin_unlock(&kvm->mmu_lock); + srcu_read_unlock(&kvm->srcu, idx); /* we've to flush the tlb before the pages can be freed */ if (need_tlb_flush) @@ -250,11 +253,14 @@ static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn, pte_t pte) { struct kvm *kvm = mmu_notifier_to_kvm(mn); + int idx; + idx = srcu_read_lock(&kvm->srcu); spin_lock(&kvm->mmu_lock); kvm->mmu_notifier_seq++; kvm_set_spte_hva(kvm, address, pte); spin_unlock(&kvm->mmu_lock); + srcu_read_unlock(&kvm->srcu, idx); } static void kvm_mmu_notifier_invalidate_range_start(struct mmu_notifier *mn, @@ -263,8 +269,9 @@ static void kvm_mmu_notifier_invalidate_range_start(struct mmu_notifier *mn, unsigned long end) { struct kvm *kvm = mmu_notifier_to_kvm(mn); - int need_tlb_flush = 0; + int need_tlb_flush = 0, idx; + idx = srcu_read_lock(&kvm->srcu); spin_lock(&kvm->mmu_lock); /* * The count increase must become visible at unlock time as no @@ -275,6 +282,7 @@ static void kvm_mmu_notifier_invalidate_range_start(struct mmu_notifier *mn, for (; start < end; start += PAGE_SIZE) need_tlb_flush |= kvm_unmap_hva(kvm, start); spin_unlock(&kvm->mmu_lock); + srcu_read_unlock(&kvm->srcu, idx); /* we've to flush the tlb before the pages can be freed */ if (need_tlb_flush) @@ -312,11 +320,13 @@ static int kvm_mmu_notifier_clear_flush_young(struct mmu_notifier *mn, unsigned long address) { struct kvm *kvm = mmu_notifier_to_kvm(mn); - int young; + int young, idx; + idx = srcu_read_lock(&kvm->srcu); spin_lock(&kvm->mmu_lock); young = kvm_age_hva(kvm, address); spin_unlock(&kvm->mmu_lock); + srcu_read_unlock(&kvm->srcu, idx); if (young) kvm_flush_remote_tlbs(kvm); @@ -379,11 +389,15 @@ static struct kvm *kvm_create_vm(void) kvm->memslots = kzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); if (!kvm->memslots) goto out_err; + if (init_srcu_struct(&kvm->srcu)) + goto out_err; #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET page = alloc_page(GFP_KERNEL | __GFP_ZERO); - if (!page) + if (!page) { + cleanup_srcu_struct(&kvm->srcu); goto out_err; + } kvm->coalesced_mmio_ring = (struct kvm_coalesced_mmio_ring *)page_address(page); @@ -391,6 +405,7 @@ static struct kvm *kvm_create_vm(void) r = kvm_init_mmu_notifier(kvm); if (r) { + cleanup_srcu_struct(&kvm->srcu); #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET put_page(page); #endif @@ -480,6 +495,7 @@ static void kvm_destroy_vm(struct kvm *kvm) #else kvm_arch_flush_shadow(kvm); #endif + cleanup_srcu_struct(&kvm->srcu); kvm_arch_destroy_vm(kvm); hardware_disable_all(); mmdrop(mm); @@ -521,12 +537,13 @@ int __kvm_set_memory_region(struct kvm *kvm, struct kvm_userspace_memory_region *mem, int user_alloc) { - int r; + int r, flush_shadow = 0; gfn_t base_gfn; unsigned long npages; unsigned long i; struct kvm_memory_slot *memslot; struct kvm_memory_slot old, new; + struct kvm_memslots *slots, *old_memslots; r = -EINVAL; /* General sanity checks */ @@ -588,15 +605,7 @@ int __kvm_set_memory_region(struct kvm *kvm, memset(new.rmap, 0, npages * sizeof(*new.rmap)); new.user_alloc = user_alloc; - /* - * hva_to_rmmap() serialzies with the mmu_lock and to be - * safe it has to ignore memslots with !user_alloc && - * !userspace_addr. - */ - if (user_alloc) - new.userspace_addr = mem->userspace_addr; - else - new.userspace_addr = 0; + new.userspace_addr = mem->userspace_addr; } if (!npages) goto skip_lpage; @@ -651,8 +660,9 @@ skip_lpage: if (!new.dirty_bitmap) goto out_free; memset(new.dirty_bitmap, 0, dirty_bytes); + /* destroy any largepage mappings for dirty tracking */ if (old.npages) - kvm_arch_flush_shadow(kvm); + flush_shadow = 1; } #else /* not defined CONFIG_S390 */ new.user_alloc = user_alloc; @@ -660,34 +670,72 @@ skip_lpage: new.userspace_addr = mem->userspace_addr; #endif /* not defined CONFIG_S390 */ - if (!npages) + if (!npages) { + r = -ENOMEM; + slots = kzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); + if (!slots) + goto out_free; + memcpy(slots, kvm->memslots, sizeof(struct kvm_memslots)); + if (mem->slot >= slots->nmemslots) + slots->nmemslots = mem->slot + 1; + slots->memslots[mem->slot].flags |= KVM_MEMSLOT_INVALID; + + old_memslots = kvm->memslots; + rcu_assign_pointer(kvm->memslots, slots); + synchronize_srcu_expedited(&kvm->srcu); + /* From this point no new shadow pages pointing to a deleted + * memslot will be created. + * + * validation of sp->gfn happens in: + * - gfn_to_hva (kvm_read_guest, gfn_to_pfn) + * - kvm_is_visible_gfn (mmu_check_roots) + */ kvm_arch_flush_shadow(kvm); + kfree(old_memslots); + } r = kvm_arch_prepare_memory_region(kvm, &new, old, mem, user_alloc); if (r) goto out_free; - spin_lock(&kvm->mmu_lock); - if (mem->slot >= kvm->memslots->nmemslots) - kvm->memslots->nmemslots = mem->slot + 1; +#ifdef CONFIG_DMAR + /* map the pages in iommu page table */ + if (npages) { + r = kvm_iommu_map_pages(kvm, &new); + if (r) + goto out_free; + } +#endif - *memslot = new; - spin_unlock(&kvm->mmu_lock); + r = -ENOMEM; + slots = kzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); + if (!slots) + goto out_free; + memcpy(slots, kvm->memslots, sizeof(struct kvm_memslots)); + if (mem->slot >= slots->nmemslots) + slots->nmemslots = mem->slot + 1; + + /* actual memory is freed via old in kvm_free_physmem_slot below */ + if (!npages) { + new.rmap = NULL; + new.dirty_bitmap = NULL; + for (i = 0; i < KVM_NR_PAGE_SIZES - 1; ++i) + new.lpage_info[i] = NULL; + } + + slots->memslots[mem->slot] = new; + old_memslots = kvm->memslots; + rcu_assign_pointer(kvm->memslots, slots); + synchronize_srcu_expedited(&kvm->srcu); kvm_arch_commit_memory_region(kvm, mem, old, user_alloc); - kvm_free_physmem_slot(&old, npages ? &new : NULL); - /* Slot deletion case: we have to update the current slot */ - spin_lock(&kvm->mmu_lock); - if (!npages) - *memslot = old; - spin_unlock(&kvm->mmu_lock); -#ifdef CONFIG_DMAR - /* map the pages in iommu page table */ - r = kvm_iommu_map_pages(kvm, memslot); - if (r) - goto out; -#endif + kvm_free_physmem_slot(&old, &new); + kfree(old_memslots); + + if (flush_shadow) + kvm_arch_flush_shadow(kvm); + return 0; out_free: @@ -787,7 +835,7 @@ EXPORT_SYMBOL_GPL(kvm_is_error_hva); struct kvm_memory_slot *gfn_to_memslot_unaliased(struct kvm *kvm, gfn_t gfn) { int i; - struct kvm_memslots *slots = kvm->memslots; + struct kvm_memslots *slots = rcu_dereference(kvm->memslots); for (i = 0; i < slots->nmemslots; ++i) { struct kvm_memory_slot *memslot = &slots->memslots[i]; @@ -809,12 +857,15 @@ struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn) int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn) { int i; - struct kvm_memslots *slots = kvm->memslots; + struct kvm_memslots *slots = rcu_dereference(kvm->memslots); gfn = unalias_gfn(kvm, gfn); for (i = 0; i < KVM_MEMORY_SLOTS; ++i) { struct kvm_memory_slot *memslot = &slots->memslots[i]; + if (memslot->flags & KVM_MEMSLOT_INVALID) + continue; + if (gfn >= memslot->base_gfn && gfn < memslot->base_gfn + memslot->npages) return 1; @@ -823,13 +874,31 @@ int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn) } EXPORT_SYMBOL_GPL(kvm_is_visible_gfn); +int memslot_id(struct kvm *kvm, gfn_t gfn) +{ + int i; + struct kvm_memslots *slots = rcu_dereference(kvm->memslots); + struct kvm_memory_slot *memslot = NULL; + + gfn = unalias_gfn(kvm, gfn); + for (i = 0; i < slots->nmemslots; ++i) { + memslot = &slots->memslots[i]; + + if (gfn >= memslot->base_gfn + && gfn < memslot->base_gfn + memslot->npages) + break; + } + + return memslot - slots->memslots; +} + unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn) { struct kvm_memory_slot *slot; gfn = unalias_gfn(kvm, gfn); slot = gfn_to_memslot_unaliased(kvm, gfn); - if (!slot) + if (!slot || slot->flags & KVM_MEMSLOT_INVALID) return bad_hva(); return (slot->userspace_addr + (gfn - slot->base_gfn) * PAGE_SIZE); } -- cgit v1.2.3 From a983fb238728e1123177e8058d4f644b949a7d05 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:23 -0200 Subject: KVM: x86: switch kvm_set_memory_alias to SRCU update Using a similar two-step procedure as for memslots. Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 2bb24a814fdf..c680f7b64c6f 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -859,7 +859,7 @@ int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn) int i; struct kvm_memslots *slots = rcu_dereference(kvm->memslots); - gfn = unalias_gfn(kvm, gfn); + gfn = unalias_gfn_instantiation(kvm, gfn); for (i = 0; i < KVM_MEMORY_SLOTS; ++i) { struct kvm_memory_slot *memslot = &slots->memslots[i]; @@ -896,7 +896,7 @@ unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn) { struct kvm_memory_slot *slot; - gfn = unalias_gfn(kvm, gfn); + gfn = unalias_gfn_instantiation(kvm, gfn); slot = gfn_to_memslot_unaliased(kvm, gfn); if (!slot || slot->flags & KVM_MEMSLOT_INVALID) return bad_hva(); -- cgit v1.2.3 From e93f8a0f821e290ac5149830110a5f704db7a1fc Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:24 -0200 Subject: KVM: convert io_bus to SRCU Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 106 +++++++++++++++++++++++++++++++--------------------- 1 file changed, 63 insertions(+), 43 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index c680f7b64c6f..659bc12ad16a 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -85,6 +85,8 @@ static long kvm_vcpu_ioctl(struct file *file, unsigned int ioctl, static int hardware_enable_all(void); static void hardware_disable_all(void); +static void kvm_io_bus_destroy(struct kvm_io_bus *bus); + static bool kvm_rebooting; static bool largepages_enabled = true; @@ -367,7 +369,7 @@ static int kvm_init_mmu_notifier(struct kvm *kvm) static struct kvm *kvm_create_vm(void) { - int r = 0; + int r = 0, i; struct kvm *kvm = kvm_arch_create_vm(); #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET struct page *page; @@ -391,6 +393,14 @@ static struct kvm *kvm_create_vm(void) goto out_err; if (init_srcu_struct(&kvm->srcu)) goto out_err; + for (i = 0; i < KVM_NR_BUSES; i++) { + kvm->buses[i] = kzalloc(sizeof(struct kvm_io_bus), + GFP_KERNEL); + if (!kvm->buses[i]) { + cleanup_srcu_struct(&kvm->srcu); + goto out_err; + } + } #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET page = alloc_page(GFP_KERNEL | __GFP_ZERO); @@ -416,11 +426,9 @@ static struct kvm *kvm_create_vm(void) atomic_inc(&kvm->mm->mm_count); spin_lock_init(&kvm->mmu_lock); spin_lock_init(&kvm->requests_lock); - kvm_io_bus_init(&kvm->pio_bus); kvm_eventfd_init(kvm); mutex_init(&kvm->lock); mutex_init(&kvm->irq_lock); - kvm_io_bus_init(&kvm->mmio_bus); init_rwsem(&kvm->slots_lock); atomic_set(&kvm->users_count, 1); spin_lock(&kvm_lock); @@ -435,6 +443,8 @@ out: out_err: hardware_disable_all(); out_err_nodisable: + for (i = 0; i < KVM_NR_BUSES; i++) + kfree(kvm->buses[i]); kfree(kvm->memslots); kfree(kvm); return ERR_PTR(r); @@ -480,6 +490,7 @@ void kvm_free_physmem(struct kvm *kvm) static void kvm_destroy_vm(struct kvm *kvm) { + int i; struct mm_struct *mm = kvm->mm; kvm_arch_sync_events(kvm); @@ -487,8 +498,8 @@ static void kvm_destroy_vm(struct kvm *kvm) list_del(&kvm->vm_list); spin_unlock(&kvm_lock); kvm_free_irq_routing(kvm); - kvm_io_bus_destroy(&kvm->pio_bus); - kvm_io_bus_destroy(&kvm->mmio_bus); + for (i = 0; i < KVM_NR_BUSES; i++) + kvm_io_bus_destroy(kvm->buses[i]); kvm_coalesced_mmio_free(kvm); #if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER) mmu_notifier_unregister(&kvm->mmu_notifier, kvm->mm); @@ -1949,12 +1960,7 @@ static struct notifier_block kvm_reboot_notifier = { .priority = 0, }; -void kvm_io_bus_init(struct kvm_io_bus *bus) -{ - memset(bus, 0, sizeof(*bus)); -} - -void kvm_io_bus_destroy(struct kvm_io_bus *bus) +static void kvm_io_bus_destroy(struct kvm_io_bus *bus) { int i; @@ -1963,13 +1969,15 @@ void kvm_io_bus_destroy(struct kvm_io_bus *bus) kvm_iodevice_destructor(pos); } + kfree(bus); } /* kvm_io_bus_write - called under kvm->slots_lock */ -int kvm_io_bus_write(struct kvm_io_bus *bus, gpa_t addr, +int kvm_io_bus_write(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr, int len, const void *val) { int i; + struct kvm_io_bus *bus = rcu_dereference(kvm->buses[bus_idx]); for (i = 0; i < bus->dev_count; i++) if (!kvm_iodevice_write(bus->devs[i], addr, len, val)) return 0; @@ -1977,59 +1985,71 @@ int kvm_io_bus_write(struct kvm_io_bus *bus, gpa_t addr, } /* kvm_io_bus_read - called under kvm->slots_lock */ -int kvm_io_bus_read(struct kvm_io_bus *bus, gpa_t addr, int len, void *val) +int kvm_io_bus_read(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr, + int len, void *val) { int i; + struct kvm_io_bus *bus = rcu_dereference(kvm->buses[bus_idx]); + for (i = 0; i < bus->dev_count; i++) if (!kvm_iodevice_read(bus->devs[i], addr, len, val)) return 0; return -EOPNOTSUPP; } -int kvm_io_bus_register_dev(struct kvm *kvm, struct kvm_io_bus *bus, - struct kvm_io_device *dev) +/* Caller must have write lock on slots_lock. */ +int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, + struct kvm_io_device *dev) { - int ret; - - down_write(&kvm->slots_lock); - ret = __kvm_io_bus_register_dev(bus, dev); - up_write(&kvm->slots_lock); + struct kvm_io_bus *new_bus, *bus; - return ret; -} - -/* An unlocked version. Caller must have write lock on slots_lock. */ -int __kvm_io_bus_register_dev(struct kvm_io_bus *bus, - struct kvm_io_device *dev) -{ + bus = kvm->buses[bus_idx]; if (bus->dev_count > NR_IOBUS_DEVS-1) return -ENOSPC; - bus->devs[bus->dev_count++] = dev; + new_bus = kzalloc(sizeof(struct kvm_io_bus), GFP_KERNEL); + if (!new_bus) + return -ENOMEM; + memcpy(new_bus, bus, sizeof(struct kvm_io_bus)); + new_bus->devs[new_bus->dev_count++] = dev; + rcu_assign_pointer(kvm->buses[bus_idx], new_bus); + synchronize_srcu_expedited(&kvm->srcu); + kfree(bus); return 0; } -void kvm_io_bus_unregister_dev(struct kvm *kvm, - struct kvm_io_bus *bus, - struct kvm_io_device *dev) +/* Caller must have write lock on slots_lock. */ +int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx, + struct kvm_io_device *dev) { - down_write(&kvm->slots_lock); - __kvm_io_bus_unregister_dev(bus, dev); - up_write(&kvm->slots_lock); -} + int i, r; + struct kvm_io_bus *new_bus, *bus; -/* An unlocked version. Caller must have write lock on slots_lock. */ -void __kvm_io_bus_unregister_dev(struct kvm_io_bus *bus, - struct kvm_io_device *dev) -{ - int i; + new_bus = kzalloc(sizeof(struct kvm_io_bus), GFP_KERNEL); + if (!new_bus) + return -ENOMEM; - for (i = 0; i < bus->dev_count; i++) - if (bus->devs[i] == dev) { - bus->devs[i] = bus->devs[--bus->dev_count]; + bus = kvm->buses[bus_idx]; + memcpy(new_bus, bus, sizeof(struct kvm_io_bus)); + + r = -ENOENT; + for (i = 0; i < new_bus->dev_count; i++) + if (new_bus->devs[i] == dev) { + r = 0; + new_bus->devs[i] = new_bus->devs[--new_bus->dev_count]; break; } + + if (r) { + kfree(new_bus); + return r; + } + + rcu_assign_pointer(kvm->buses[bus_idx], new_bus); + synchronize_srcu_expedited(&kvm->srcu); + kfree(bus); + return r; } static struct notifier_block kvm_cpu_notifier = { -- cgit v1.2.3 From 79fac95ecfa3969aab8119d37ccd7226165f933a Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 23 Dec 2009 14:35:26 -0200 Subject: KVM: convert slots_lock to a mutex Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 659bc12ad16a..2b7cd6c0d9ca 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -429,7 +429,7 @@ static struct kvm *kvm_create_vm(void) kvm_eventfd_init(kvm); mutex_init(&kvm->lock); mutex_init(&kvm->irq_lock); - init_rwsem(&kvm->slots_lock); + mutex_init(&kvm->slots_lock); atomic_set(&kvm->users_count, 1); spin_lock(&kvm_lock); list_add(&kvm->vm_list, &vm_list); @@ -763,9 +763,9 @@ int kvm_set_memory_region(struct kvm *kvm, { int r; - down_write(&kvm->slots_lock); + mutex_lock(&kvm->slots_lock); r = __kvm_set_memory_region(kvm, mem, user_alloc); - up_write(&kvm->slots_lock); + mutex_unlock(&kvm->slots_lock); return r; } EXPORT_SYMBOL_GPL(kvm_set_memory_region); @@ -1997,7 +1997,7 @@ int kvm_io_bus_read(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr, return -EOPNOTSUPP; } -/* Caller must have write lock on slots_lock. */ +/* Caller must hold slots_lock. */ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, struct kvm_io_device *dev) { @@ -2019,7 +2019,7 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, return 0; } -/* Caller must have write lock on slots_lock. */ +/* Caller must hold slots_lock. */ int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx, struct kvm_io_device *dev) { -- cgit v1.2.3 From 647492047763c3ee8fe51ecf9a04f39040aa495b Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Tue, 19 Jan 2010 12:45:23 -0200 Subject: KVM: fix cleanup_srcu_struct on vm destruction cleanup_srcu_struct on VM destruction remains broken: BUG: unable to handle kernel paging request at ffffffffffffffff IP: [] srcu_read_lock+0x16/0x21 RIP: 0010:[] [] srcu_read_lock+0x16/0x21 Call Trace: [] kvm_arch_vcpu_uninit+0x1b/0x48 [kvm] [] kvm_vcpu_uninit+0x9/0x15 [kvm] [] vmx_free_vcpu+0x7f/0x8f [kvm_intel] [] kvm_arch_destroy_vm+0x78/0x111 [kvm] [] kvm_put_kvm+0xd4/0xfe [kvm] Move it to kvm_arch_destroy_vm. Signed-off-by: Marcelo Tosatti Reported-by: Jan Kiszka --- virt/kvm/kvm_main.c | 1 - 1 file changed, 1 deletion(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 2b7cd6c0d9ca..7c5c873812ff 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -506,7 +506,6 @@ static void kvm_destroy_vm(struct kvm *kvm) #else kvm_arch_flush_shadow(kvm); #endif - cleanup_srcu_struct(&kvm->srcu); kvm_arch_destroy_vm(kvm); hardware_disable_all(); mmdrop(mm); -- cgit v1.2.3 From f0f4b930900ffa8daddb5262522c3e5c67ee1835 Mon Sep 17 00:00:00 2001 From: Sheng Yang Date: Thu, 21 Jan 2010 16:20:04 +0800 Subject: KVM: Fix kvm_coalesced_mmio_ring duplicate allocation The commit 0953ca73 "KVM: Simplify coalesced mmio initialization" allocate kvm_coalesced_mmio_ring in the kvm_coalesced_mmio_init(), but didn't discard the original allocation... Signed-off-by: Sheng Yang Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 17 ----------------- 1 file changed, 17 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 7c5c873812ff..2b0974a14835 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -371,9 +371,6 @@ static struct kvm *kvm_create_vm(void) { int r = 0, i; struct kvm *kvm = kvm_arch_create_vm(); -#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET - struct page *page; -#endif if (IS_ERR(kvm)) goto out; @@ -402,23 +399,9 @@ static struct kvm *kvm_create_vm(void) } } -#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET - page = alloc_page(GFP_KERNEL | __GFP_ZERO); - if (!page) { - cleanup_srcu_struct(&kvm->srcu); - goto out_err; - } - - kvm->coalesced_mmio_ring = - (struct kvm_coalesced_mmio_ring *)page_address(page); -#endif - r = kvm_init_mmu_notifier(kvm); if (r) { cleanup_srcu_struct(&kvm->srcu); -#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET - put_page(page); -#endif goto out_err; } -- cgit v1.2.3 From 8f0b1ab6fb045a1324d9435ba00c2940783b0041 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 28 Jan 2010 12:37:56 +0100 Subject: KVM: Introduce kvm_host_page_size This patch introduces a generic function to find out the host page size for a given gfn. This function is needed by the kvm iommu code. This patch also simplifies the x86 host_mapping_level function. Signed-off-by: Joerg Roedel Signed-off-by: Avi Kivity --- virt/kvm/kvm_main.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 2b0974a14835..0a360c26cc34 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -45,6 +45,7 @@ #include #include #include +#include #include #include @@ -867,6 +868,30 @@ int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn) } EXPORT_SYMBOL_GPL(kvm_is_visible_gfn); +unsigned long kvm_host_page_size(struct kvm *kvm, gfn_t gfn) +{ + struct vm_area_struct *vma; + unsigned long addr, size; + + size = PAGE_SIZE; + + addr = gfn_to_hva(kvm, gfn); + if (kvm_is_error_hva(addr)) + return PAGE_SIZE; + + down_read(¤t->mm->mmap_sem); + vma = find_vma(current->mm, addr); + if (!vma) + goto out; + + size = vma_kernel_pagesize(vma); + +out: + up_read(¤t->mm->mmap_sem); + + return size; +} + int memslot_id(struct kvm *kvm, gfn_t gfn) { int i; -- cgit v1.2.3 From 70e335e16882df5b5d6971022e63c3603a1e8c23 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Thu, 18 Feb 2010 11:25:22 +0200 Subject: KVM: Convert kvm->requests_lock to raw_spinlock_t The code relies on kvm->requests_lock inhibiting preemption. Noted by Jan Kiszka. Signed-off-by: Avi Kivity --- virt/kvm/kvm_main.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'virt/kvm/kvm_main.c') diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 0a360c26cc34..548f9253c195 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -138,7 +138,7 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req) zalloc_cpumask_var(&cpus, GFP_ATOMIC); - spin_lock(&kvm->requests_lock); + raw_spin_lock(&kvm->requests_lock); me = smp_processor_id(); kvm_for_each_vcpu(i, vcpu, kvm) { if (test_and_set_bit(req, &vcpu->requests)) @@ -153,7 +153,7 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req) smp_call_function_many(cpus, ack_flush, NULL, 1); else called = false; - spin_unlock(&kvm->requests_lock); + raw_spin_unlock(&kvm->requests_lock); free_cpumask_var(cpus); return called; } @@ -409,7 +409,7 @@ static struct kvm *kvm_create_vm(void) kvm->mm = current->mm; atomic_inc(&kvm->mm->mm_count); spin_lock_init(&kvm->mmu_lock); - spin_lock_init(&kvm->requests_lock); + raw_spin_lock_init(&kvm->requests_lock); kvm_eventfd_init(kvm); mutex_init(&kvm->lock); mutex_init(&kvm->irq_lock); -- cgit v1.2.3