#include "vmx.h"
#include "kvm.h"
-#define pgprintk(x...) do { } while (0)
+#define pgprintk(x...) do { printk(x); } while (0)
+#define rmap_printk(x...) do { printk(x); } while (0)
#define ASSERT(x) \
if (!(x)) { \
__FILE__, __LINE__, #x); \
}
-#define PT64_ENT_PER_PAGE 512
-#define PT32_ENT_PER_PAGE 1024
+#define PT64_PT_BITS 9
+#define PT64_ENT_PER_PAGE (1 << PT64_PT_BITS)
+#define PT32_PT_BITS 10
+#define PT32_ENT_PER_PAGE (1 << PT32_PT_BITS)
#define PT_WRITABLE_SHIFT 1
#define PT_DIRECTORY_LEVEL 2
#define PT_PAGE_TABLE_LEVEL 1
+#define RMAP_EXT 4
+
+struct kvm_rmap_desc {
+ u64 *shadow_ptes[RMAP_EXT];
+ struct kvm_rmap_desc *more;
+};
+
static int is_write_protection(struct kvm_vcpu *vcpu)
{
return vcpu->cr0 & CR0_WP_MASK;
return pte & PT_SHADOW_IO_MARK;
}
+static int is_rmap_pte(u64 pte)
+{
+ return (pte & (PT_WRITABLE_MASK | PT_PRESENT_MASK))
+ == (PT_WRITABLE_MASK | PT_PRESENT_MASK);
+}
+
+/*
+ * Reverse mapping data structures:
+ *
+ * If page->private bit zero is zero, then page->private points to the
+ * shadow page table entry that points to page_address(page).
+ *
+ * If page->private bit zero is one, (then page->private & ~1) points
+ * to a struct kvm_rmap_desc containing more mappings.
+ */
+static void rmap_add(struct kvm *kvm, u64 *spte)
+{
+ struct page *page;
+ struct kvm_rmap_desc *desc;
+ int i;
+
+ if (!is_rmap_pte(*spte))
+ return;
+ page = pfn_to_page((*spte & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT);
+ if (!page->private) {
+ rmap_printk("rmap_add: %p %llx 0->1\n", spte, *spte);
+ page->private = (unsigned long)spte;
+ } else if (!(page->private & 1)) {
+ rmap_printk("rmap_add: %p %llx 1->many\n", spte, *spte);
+ desc = kzalloc(sizeof *desc, GFP_NOWAIT);
+ if (!desc)
+ BUG(); /* FIXME: return error */
+ desc->shadow_ptes[0] = (u64 *)page->private;
+ desc->shadow_ptes[1] = spte;
+ page->private = (unsigned long)desc | 1;
+ } else {
+ rmap_printk("rmap_add: %p %llx many->many\n", spte, *spte);
+ desc = (struct kvm_rmap_desc *)(page->private & ~1ul);
+ while (desc->shadow_ptes[RMAP_EXT-1] && desc->more)
+ desc = desc->more;
+ if (desc->shadow_ptes[RMAP_EXT-1]) {
+ desc->more = kzalloc(sizeof *desc->more, GFP_NOWAIT);
+ if (!desc->more)
+ BUG(); /* FIXME: return error */
+ desc = desc->more;
+ }
+ for (i = 0; desc->shadow_ptes[i]; ++i)
+ ;
+ desc->shadow_ptes[i] = spte;
+ }
+}
+
+static void rmap_desc_remove_entry(struct page *page,
+ struct kvm_rmap_desc *desc,
+ int i,
+ struct kvm_rmap_desc *prev_desc)
+{
+ int j;
+
+ for (j = RMAP_EXT - 1; !desc->shadow_ptes[j] && j > i; --j)
+ ;
+ desc->shadow_ptes[i] = desc->shadow_ptes[j];
+ desc->shadow_ptes[j] = 0;
+ if (j != 0)
+ return;
+ if (!prev_desc && !desc->more)
+ page->private = (unsigned long)desc->shadow_ptes[0];
+ else
+ if (prev_desc)
+ prev_desc->more = desc->more;
+ else
+ page->private = (unsigned long)desc->more | 1;
+ kfree(desc);
+}
+
+static void rmap_remove(struct kvm *kvm, u64 *spte)
+{
+ struct page *page;
+ struct kvm_rmap_desc *desc;
+ struct kvm_rmap_desc *prev_desc;
+ int i;
+
+ if (!is_rmap_pte(*spte))
+ return;
+ page = pfn_to_page((*spte & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT);
+ if (!page->private) {
+ printk(KERN_ERR "rmap_remove: %p %llx 0->BUG\n", spte, *spte);
+ BUG();
+ } else if (!(page->private & 1)) {
+ rmap_printk("rmap_remove: %p %llx 1->0\n", spte, *spte);
+ if ((u64 *)page->private != spte) {
+ printk(KERN_ERR "rmap_remove: %p %llx 1->BUG\n",
+ spte, *spte);
+ BUG();
+ }
+ page->private = 0;
+ } else {
+ rmap_printk("rmap_remove: %p %llx many->many\n", spte, *spte);
+ desc = (struct kvm_rmap_desc *)(page->private & ~1ul);
+ prev_desc = NULL;
+ while (desc) {
+ for (i = 0; i < RMAP_EXT && desc->shadow_ptes[i]; ++i)
+ if (desc->shadow_ptes[i] == spte) {
+ rmap_desc_remove_entry(page, desc, i,
+ prev_desc);
+ return;
+ }
+ prev_desc = desc;
+ desc = desc->more;
+ }
+ BUG();
+ }
+}
+
+static void rmap_write_protect(struct kvm *kvm, u64 gfn)
+{
+ struct page *page;
+ struct kvm_memory_slot *slot;
+ struct kvm_rmap_desc *desc;
+ u64 *spte;
+
+ slot = gfn_to_memslot(kvm, gfn);
+ BUG_ON(!slot);
+ page = gfn_to_page(slot, gfn);
+
+ while (page->private) {
+ if (!(page->private & 1))
+ spte = (u64 *)page->private;
+ else {
+ desc = (struct kvm_rmap_desc *)(page->private & ~1ul);
+ spte = desc->shadow_ptes[0];
+ }
+ BUG_ON(!spte);
+ BUG_ON((*spte & PT64_BASE_ADDR_MASK) !=
+ page_to_pfn(page) << PAGE_SHIFT);
+ BUG_ON(!(*spte & PT_PRESENT_MASK));
+ BUG_ON(!(*spte & PT_WRITABLE_MASK));
+ rmap_printk("rmap_write_protect: spte %p %llx\n", spte, *spte);
+ rmap_remove(kvm, spte);
+ *spte &= ~(u64)PT_WRITABLE_MASK;
+ }
+}
+
static void kvm_mmu_free_page(struct kvm_vcpu *vcpu, hpa_t page_hpa)
{
struct kvm_mmu_page *page_head = page_header(page_hpa);
list_del(&page_head->link);
page_head->page_hpa = page_hpa;
list_add(&page_head->link, &vcpu->free_pages);
+ ++vcpu->kvm->n_free_mmu_pages;
}
static int is_empty_shadow_page(hpa_t page_hpa)
return 1;
}
-static hpa_t kvm_mmu_alloc_page(struct kvm_vcpu *vcpu, u64 *parent_pte)
+static unsigned kvm_page_table_hashfn(gfn_t gfn)
+{
+ return gfn;
+}
+
+static struct kvm_mmu_page *kvm_mmu_alloc_page(struct kvm_vcpu *vcpu,
+ u64 *parent_pte)
{
struct kvm_mmu_page *page;
if (list_empty(&vcpu->free_pages))
- return INVALID_PAGE;
+ return NULL;
page = list_entry(vcpu->free_pages.next, struct kvm_mmu_page, link);
list_del(&page->link);
ASSERT(is_empty_shadow_page(page->page_hpa));
page->slot_bitmap = 0;
page->global = 1;
+ page->multimapped = 0;
page->parent_pte = parent_pte;
- return page->page_hpa;
+ --vcpu->kvm->n_free_mmu_pages;
+ return page;
+}
+
+static void mmu_page_add_parent_pte(struct kvm_mmu_page *page, u64 *parent_pte)
+{
+ struct kvm_pte_chain *pte_chain;
+ struct hlist_node *node;
+ int i;
+
+ if (!parent_pte)
+ return;
+ if (!page->multimapped) {
+ u64 *old = page->parent_pte;
+
+ if (!old) {
+ page->parent_pte = parent_pte;
+ return;
+ }
+ page->multimapped = 1;
+ pte_chain = kzalloc(sizeof(struct kvm_pte_chain), GFP_NOWAIT);
+ BUG_ON(!pte_chain);
+ INIT_HLIST_HEAD(&page->parent_ptes);
+ hlist_add_head(&pte_chain->link, &page->parent_ptes);
+ pte_chain->parent_ptes[0] = old;
+ }
+ hlist_for_each_entry(pte_chain, node, &page->parent_ptes, link) {
+ if (pte_chain->parent_ptes[NR_PTE_CHAIN_ENTRIES-1])
+ continue;
+ for (i = 0; i < NR_PTE_CHAIN_ENTRIES; ++i)
+ if (!pte_chain->parent_ptes[i]) {
+ pte_chain->parent_ptes[i] = parent_pte;
+ return;
+ }
+ }
+ pte_chain = kzalloc(sizeof(struct kvm_pte_chain), GFP_NOWAIT);
+ BUG_ON(!pte_chain);
+ hlist_add_head(&pte_chain->link, &page->parent_ptes);
+ pte_chain->parent_ptes[0] = parent_pte;
+}
+
+static void mmu_page_remove_parent_pte(struct kvm_mmu_page *page,
+ u64 *parent_pte)
+{
+ struct kvm_pte_chain *pte_chain;
+ struct hlist_node *node;
+ int i;
+
+ if (!page->multimapped) {
+ BUG_ON(page->parent_pte != parent_pte);
+ page->parent_pte = NULL;
+ return;
+ }
+ hlist_for_each_entry(pte_chain, node, &page->parent_ptes, link)
+ for (i = 0; i < NR_PTE_CHAIN_ENTRIES; ++i) {
+ if (!pte_chain->parent_ptes[i])
+ break;
+ if (pte_chain->parent_ptes[i] != parent_pte)
+ continue;
+ while (i + 1 < NR_PTE_CHAIN_ENTRIES
+ && pte_chain->parent_ptes[i + 1]) {
+ pte_chain->parent_ptes[i]
+ = pte_chain->parent_ptes[i + 1];
+ ++i;
+ }
+ pte_chain->parent_ptes[i] = NULL;
+ if (i == 0) {
+ hlist_del(&pte_chain->link);
+ kfree(pte_chain);
+ if (hlist_empty(&page->parent_ptes)) {
+ page->multimapped = 0;
+ page->parent_pte = NULL;
+ }
+ }
+ return;
+ }
+ BUG();
+}
+
+static struct kvm_mmu_page *kvm_mmu_lookup_page(struct kvm_vcpu *vcpu,
+ gfn_t gfn)
+{
+ unsigned index;
+ struct hlist_head *bucket;
+ struct kvm_mmu_page *page;
+ struct hlist_node *node;
+
+ pgprintk("%s: looking for gfn %lx\n", __FUNCTION__, gfn);
+ index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+ bucket = &vcpu->kvm->mmu_page_hash[index];
+ hlist_for_each_entry(page, node, bucket, hash_link)
+ if (page->gfn == gfn && !page->role.metaphysical) {
+ pgprintk("%s: found role %x\n",
+ __FUNCTION__, page->role.word);
+ return page;
+ }
+ return NULL;
+}
+
+static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu,
+ gfn_t gfn,
+ gva_t gaddr,
+ unsigned level,
+ int metaphysical,
+ u64 *parent_pte)
+{
+ union kvm_mmu_page_role role;
+ unsigned index;
+ unsigned quadrant;
+ struct hlist_head *bucket;
+ struct kvm_mmu_page *page;
+ struct hlist_node *node;
+
+ role.word = 0;
+ role.glevels = vcpu->mmu.root_level;
+ role.level = level;
+ role.metaphysical = metaphysical;
+ if (vcpu->mmu.root_level <= PT32_ROOT_LEVEL) {
+ quadrant = gaddr >> (PAGE_SHIFT + (PT64_PT_BITS * level));
+ quadrant &= (1 << ((PT32_PT_BITS - PT64_PT_BITS) * level)) - 1;
+ role.quadrant = quadrant;
+ }
+ pgprintk("%s: looking gfn %lx role %x\n", __FUNCTION__,
+ gfn, role.word);
+ index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+ bucket = &vcpu->kvm->mmu_page_hash[index];
+ hlist_for_each_entry(page, node, bucket, hash_link)
+ if (page->gfn == gfn && page->role.word == role.word) {
+ mmu_page_add_parent_pte(page, parent_pte);
+ pgprintk("%s: found\n", __FUNCTION__);
+ return page;
+ }
+ page = kvm_mmu_alloc_page(vcpu, parent_pte);
+ if (!page)
+ return page;
+ pgprintk("%s: adding gfn %lx role %x\n", __FUNCTION__, gfn, role.word);
+ page->gfn = gfn;
+ page->role = role;
+ hlist_add_head(&page->hash_link, bucket);
+ if (!metaphysical)
+ rmap_write_protect(vcpu->kvm, gfn);
+ return page;
+}
+
+static void kvm_mmu_page_unlink_children(struct kvm_vcpu *vcpu,
+ struct kvm_mmu_page *page)
+{
+ unsigned i;
+ u64 *pt;
+ u64 ent;
+
+ pt = __va(page->page_hpa);
+
+ if (page->role.level == PT_PAGE_TABLE_LEVEL) {
+ for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
+ if (pt[i] & PT_PRESENT_MASK)
+ rmap_remove(vcpu->kvm, &pt[i]);
+ pt[i] = 0;
+ }
+ return;
+ }
+
+ for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
+ ent = pt[i];
+
+ pt[i] = 0;
+ if (!(ent & PT_PRESENT_MASK))
+ continue;
+ ent &= PT64_BASE_ADDR_MASK;
+ mmu_page_remove_parent_pte(page_header(ent), &pt[i]);
+ }
+}
+
+static void kvm_mmu_put_page(struct kvm_vcpu *vcpu,
+ struct kvm_mmu_page *page,
+ u64 *parent_pte)
+{
+ mmu_page_remove_parent_pte(page, parent_pte);
+}
+
+static void kvm_mmu_zap_page(struct kvm_vcpu *vcpu,
+ struct kvm_mmu_page *page)
+{
+ u64 *parent_pte;
+
+ while (page->multimapped || page->parent_pte) {
+ if (!page->multimapped)
+ parent_pte = page->parent_pte;
+ else {
+ struct kvm_pte_chain *chain;
+
+ chain = container_of(page->parent_ptes.first,
+ struct kvm_pte_chain, link);
+ parent_pte = chain->parent_ptes[0];
+ }
+ BUG_ON(!parent_pte);
+ kvm_mmu_put_page(vcpu, page, parent_pte);
+ *parent_pte = 0;
+ }
+ kvm_mmu_page_unlink_children(vcpu, page);
+ hlist_del(&page->hash_link);
+ kvm_mmu_free_page(vcpu, page->page_hpa);
+}
+
+static int kvm_mmu_unprotect_page(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+ unsigned index;
+ struct hlist_head *bucket;
+ struct kvm_mmu_page *page;
+ struct hlist_node *node, *n;
+ int r;
+
+ pgprintk("%s: looking for gfn %lx\n", __FUNCTION__, gfn);
+ r = 0;
+ index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+ bucket = &vcpu->kvm->mmu_page_hash[index];
+ hlist_for_each_entry_safe(page, node, n, bucket, hash_link)
+ if (page->gfn == gfn && !page->role.metaphysical) {
+ pgprintk("%s: gfn %lx role %x\n", __FUNCTION__, gfn,
+ page->role.word);
+ kvm_mmu_zap_page(vcpu, page);
+ r = 1;
+ }
+ return r;
}
static void page_header_update_slot(struct kvm *kvm, void *pte, gpa_t gpa)
static void release_pt_page_64(struct kvm_vcpu *vcpu, hpa_t page_hpa,
int level)
{
+ u64 *pos;
+ u64 *end;
+
ASSERT(vcpu);
ASSERT(VALID_PAGE(page_hpa));
ASSERT(level <= PT64_ROOT_LEVEL && level > 0);
- if (level == 1)
- memset(__va(page_hpa), 0, PAGE_SIZE);
- else {
- u64 *pos;
- u64 *end;
+ for (pos = __va(page_hpa), end = pos + PT64_ENT_PER_PAGE;
+ pos != end; pos++) {
+ u64 current_ent = *pos;
- for (pos = __va(page_hpa), end = pos + PT64_ENT_PER_PAGE;
- pos != end; pos++) {
- u64 current_ent = *pos;
-
- *pos = 0;
- if (is_present_pte(current_ent))
+ if (is_present_pte(current_ent)) {
+ if (level != 1)
release_pt_page_64(vcpu,
current_ent &
PT64_BASE_ADDR_MASK,
level - 1);
+ else
+ rmap_remove(vcpu->kvm, pos);
}
+ *pos = 0;
}
kvm_mmu_free_page(vcpu, page_hpa);
}
for (; ; level--) {
u32 index = PT64_INDEX(v, level);
u64 *table;
+ u64 pte;
ASSERT(VALID_PAGE(table_addr));
table = __va(table_addr);
if (level == 1) {
+ pte = table[index];
+ if (is_present_pte(pte) && is_writeble_pte(pte))
+ return 0;
mark_page_dirty(vcpu->kvm, v >> PAGE_SHIFT);
page_header_update_slot(vcpu->kvm, table, v);
table[index] = p | PT_PRESENT_MASK | PT_WRITABLE_MASK |
PT_USER_MASK;
+ rmap_add(vcpu->kvm, &table[index]);
return 0;
}
if (table[index] == 0) {
- hpa_t new_table = kvm_mmu_alloc_page(vcpu,
- &table[index]);
-
- if (!VALID_PAGE(new_table)) {
+ struct kvm_mmu_page *new_table;
+ gfn_t pseudo_gfn;
+
+ pseudo_gfn = (v & PT64_DIR_BASE_ADDR_MASK)
+ >> PAGE_SHIFT;
+ new_table = kvm_mmu_get_page(vcpu, pseudo_gfn,
+ v, level - 1,
+ 1, &table[index]);
+ if (!new_table) {
pgprintk("nonpaging_map: ENOMEM\n");
return -ENOMEM;
}
- if (level == PT32E_ROOT_LEVEL)
- table[index] = new_table | PT_PRESENT_MASK;
- else
- table[index] = new_table | PT_PRESENT_MASK |
- PT_WRITABLE_MASK | PT_USER_MASK;
+ table[index] = new_table->page_hpa | PT_PRESENT_MASK
+ | PT_WRITABLE_MASK | PT_USER_MASK;
}
table_addr = table[index] & PT64_BASE_ADDR_MASK;
}
}
-static void nonpaging_flush(struct kvm_vcpu *vcpu)
+static void mmu_free_roots(struct kvm_vcpu *vcpu)
{
- hpa_t root = vcpu->mmu.root_hpa;
+ int i;
- ++kvm_stat.tlb_flush;
- pgprintk("nonpaging_flush\n");
- ASSERT(VALID_PAGE(root));
- release_pt_page_64(vcpu, root, vcpu->mmu.shadow_root_level);
- root = kvm_mmu_alloc_page(vcpu, NULL);
- ASSERT(VALID_PAGE(root));
- vcpu->mmu.root_hpa = root;
- if (is_paging(vcpu))
- root |= (vcpu->cr3 & (CR3_PCD_MASK | CR3_WPT_MASK));
- kvm_arch_ops->set_cr3(vcpu, root);
- kvm_arch_ops->tlb_flush(vcpu);
+#ifdef CONFIG_X86_64
+ if (vcpu->mmu.shadow_root_level == PT64_ROOT_LEVEL) {
+ hpa_t root = vcpu->mmu.root_hpa;
+
+ ASSERT(VALID_PAGE(root));
+ vcpu->mmu.root_hpa = INVALID_PAGE;
+ return;
+ }
+#endif
+ for (i = 0; i < 4; ++i) {
+ hpa_t root = vcpu->mmu.pae_root[i];
+
+ ASSERT(VALID_PAGE(root));
+ root &= PT64_BASE_ADDR_MASK;
+ vcpu->mmu.pae_root[i] = INVALID_PAGE;
+ }
+ vcpu->mmu.root_hpa = INVALID_PAGE;
+}
+
+static void mmu_alloc_roots(struct kvm_vcpu *vcpu)
+{
+ int i;
+ gfn_t root_gfn;
+ root_gfn = vcpu->cr3 >> PAGE_SHIFT;
+
+#ifdef CONFIG_X86_64
+ if (vcpu->mmu.shadow_root_level == PT64_ROOT_LEVEL) {
+ hpa_t root = vcpu->mmu.root_hpa;
+
+ ASSERT(!VALID_PAGE(root));
+ root = kvm_mmu_get_page(vcpu, root_gfn, 0,
+ PT64_ROOT_LEVEL, 0, NULL)->page_hpa;
+ vcpu->mmu.root_hpa = root;
+ return;
+ }
+#endif
+ for (i = 0; i < 4; ++i) {
+ hpa_t root = vcpu->mmu.pae_root[i];
+
+ ASSERT(!VALID_PAGE(root));
+ if (vcpu->mmu.root_level == PT32E_ROOT_LEVEL)
+ root_gfn = vcpu->pdptrs[i] >> PAGE_SHIFT;
+ else if (vcpu->mmu.root_level == 0)
+ root_gfn = 0;
+ root = kvm_mmu_get_page(vcpu, root_gfn, i << 30,
+ PT32_ROOT_LEVEL, !is_paging(vcpu),
+ NULL)->page_hpa;
+ vcpu->mmu.pae_root[i] = root | PT_PRESENT_MASK;
+ }
+ vcpu->mmu.root_hpa = __pa(vcpu->mmu.pae_root);
}
static gpa_t nonpaging_gva_to_gpa(struct kvm_vcpu *vcpu, gva_t vaddr)
static int nonpaging_page_fault(struct kvm_vcpu *vcpu, gva_t gva,
u32 error_code)
{
- int ret;
gpa_t addr = gva;
+ hpa_t paddr;
ASSERT(vcpu);
ASSERT(VALID_PAGE(vcpu->mmu.root_hpa));
- for (;;) {
- hpa_t paddr;
- paddr = gpa_to_hpa(vcpu , addr & PT64_BASE_ADDR_MASK);
+ paddr = gpa_to_hpa(vcpu , addr & PT64_BASE_ADDR_MASK);
- if (is_error_hpa(paddr))
- return 1;
+ if (is_error_hpa(paddr))
+ return 1;
- ret = nonpaging_map(vcpu, addr & PAGE_MASK, paddr);
- if (ret) {
- nonpaging_flush(vcpu);
- continue;
- }
- break;
- }
- return ret;
-}
-
-static void nonpaging_inval_page(struct kvm_vcpu *vcpu, gva_t addr)
-{
+ return nonpaging_map(vcpu, addr & PAGE_MASK, paddr);
}
static void nonpaging_free(struct kvm_vcpu *vcpu)
{
- hpa_t root;
-
- ASSERT(vcpu);
- root = vcpu->mmu.root_hpa;
- if (VALID_PAGE(root))
- release_pt_page_64(vcpu, root, vcpu->mmu.shadow_root_level);
- vcpu->mmu.root_hpa = INVALID_PAGE;
+ mmu_free_roots(vcpu);
}
static int nonpaging_init_context(struct kvm_vcpu *vcpu)
context->new_cr3 = nonpaging_new_cr3;
context->page_fault = nonpaging_page_fault;
- context->inval_page = nonpaging_inval_page;
context->gva_to_gpa = nonpaging_gva_to_gpa;
context->free = nonpaging_free;
- context->root_level = PT32E_ROOT_LEVEL;
+ context->root_level = 0;
context->shadow_root_level = PT32E_ROOT_LEVEL;
- context->root_hpa = kvm_mmu_alloc_page(vcpu, NULL);
+ mmu_alloc_roots(vcpu);
ASSERT(VALID_PAGE(context->root_hpa));
kvm_arch_ops->set_cr3(vcpu, context->root_hpa);
return 0;
}
-
static void kvm_mmu_flush_tlb(struct kvm_vcpu *vcpu)
{
- struct kvm_mmu_page *page, *npage;
-
- list_for_each_entry_safe(page, npage, &vcpu->kvm->active_mmu_pages,
- link) {
- if (page->global)
- continue;
-
- if (!page->parent_pte)
- continue;
-
- *page->parent_pte = 0;
- release_pt_page_64(vcpu, page->page_hpa, 1);
- }
++kvm_stat.tlb_flush;
kvm_arch_ops->tlb_flush(vcpu);
}
static void paging_new_cr3(struct kvm_vcpu *vcpu)
{
+ pgprintk("%s: cr3 %lx\n", __FUNCTION__, vcpu->cr3);
+ mmu_free_roots(vcpu);
+ mmu_alloc_roots(vcpu);
kvm_mmu_flush_tlb(vcpu);
+ kvm_arch_ops->set_cr3(vcpu, vcpu->mmu.root_hpa);
}
static void mark_pagetable_nonglobal(void *shadow_pte)
u64 *shadow_pte,
gpa_t gaddr,
int dirty,
- u64 access_bits)
+ u64 access_bits,
+ gfn_t gfn)
{
hpa_t paddr;
if (!dirty)
access_bits &= ~PT_WRITABLE_MASK;
- if (access_bits & PT_WRITABLE_MASK)
- mark_page_dirty(vcpu->kvm, gaddr >> PAGE_SHIFT);
+ paddr = gpa_to_hpa(vcpu, gaddr & PT64_BASE_ADDR_MASK);
*shadow_pte |= access_bits;
- paddr = gpa_to_hpa(vcpu, gaddr & PT64_BASE_ADDR_MASK);
-
if (!(*shadow_pte & PT_GLOBAL_MASK))
mark_pagetable_nonglobal(shadow_pte);
*shadow_pte |= gaddr;
*shadow_pte |= PT_SHADOW_IO_MARK;
*shadow_pte &= ~PT_PRESENT_MASK;
- } else {
- *shadow_pte |= paddr;
- page_header_update_slot(vcpu->kvm, shadow_pte, gaddr);
+ return;
}
+
+ *shadow_pte |= paddr;
+
+ if (access_bits & PT_WRITABLE_MASK) {
+ struct kvm_mmu_page *shadow;
+
+ shadow = kvm_mmu_lookup_page(vcpu, gfn);
+ if (shadow) {
+ pgprintk("%s: found shadow page for %lx, marking ro\n",
+ __FUNCTION__, gfn);
+ access_bits &= ~PT_WRITABLE_MASK;
+ *shadow_pte &= ~PT_WRITABLE_MASK;
+ }
+ }
+
+ if (access_bits & PT_WRITABLE_MASK)
+ mark_page_dirty(vcpu->kvm, gaddr >> PAGE_SHIFT);
+
+ page_header_update_slot(vcpu->kvm, shadow_pte, gaddr);
+ rmap_add(vcpu->kvm, shadow_pte);
}
static void inject_page_fault(struct kvm_vcpu *vcpu,
return 1;
}
-/*
- * Remove a shadow pte.
- */
-static void paging_inval_page(struct kvm_vcpu *vcpu, gva_t addr)
-{
- hpa_t page_addr = vcpu->mmu.root_hpa;
- int level = vcpu->mmu.shadow_root_level;
-
- ++kvm_stat.invlpg;
-
- for (; ; level--) {
- u32 index = PT64_INDEX(addr, level);
- u64 *table = __va(page_addr);
-
- if (level == PT_PAGE_TABLE_LEVEL ) {
- table[index] = 0;
- return;
- }
-
- if (!is_present_pte(table[index]))
- return;
-
- page_addr = table[index] & PT64_BASE_ADDR_MASK;
-
- if (level == PT_DIRECTORY_LEVEL &&
- (table[index] & PT_SHADOW_PS_MARK)) {
- table[index] = 0;
- release_pt_page_64(vcpu, page_addr, PT_PAGE_TABLE_LEVEL);
-
- kvm_arch_ops->tlb_flush(vcpu);
- return;
- }
- }
-}
-
static void paging_free(struct kvm_vcpu *vcpu)
{
nonpaging_free(vcpu);
#include "paging_tmpl.h"
#undef PTTYPE
-static int paging64_init_context(struct kvm_vcpu *vcpu)
+static int paging64_init_context_common(struct kvm_vcpu *vcpu, int level)
{
struct kvm_mmu *context = &vcpu->mmu;
ASSERT(is_pae(vcpu));
context->new_cr3 = paging_new_cr3;
context->page_fault = paging64_page_fault;
- context->inval_page = paging_inval_page;
context->gva_to_gpa = paging64_gva_to_gpa;
context->free = paging_free;
- context->root_level = PT64_ROOT_LEVEL;
- context->shadow_root_level = PT64_ROOT_LEVEL;
- context->root_hpa = kvm_mmu_alloc_page(vcpu, NULL);
+ context->root_level = level;
+ context->shadow_root_level = level;
+ mmu_alloc_roots(vcpu);
ASSERT(VALID_PAGE(context->root_hpa));
kvm_arch_ops->set_cr3(vcpu, context->root_hpa |
(vcpu->cr3 & (CR3_PCD_MASK | CR3_WPT_MASK)));
return 0;
}
+static int paging64_init_context(struct kvm_vcpu *vcpu)
+{
+ return paging64_init_context_common(vcpu, PT64_ROOT_LEVEL);
+}
+
static int paging32_init_context(struct kvm_vcpu *vcpu)
{
struct kvm_mmu *context = &vcpu->mmu;
context->new_cr3 = paging_new_cr3;
context->page_fault = paging32_page_fault;
- context->inval_page = paging_inval_page;
context->gva_to_gpa = paging32_gva_to_gpa;
context->free = paging_free;
context->root_level = PT32_ROOT_LEVEL;
context->shadow_root_level = PT32E_ROOT_LEVEL;
- context->root_hpa = kvm_mmu_alloc_page(vcpu, NULL);
+ mmu_alloc_roots(vcpu);
ASSERT(VALID_PAGE(context->root_hpa));
kvm_arch_ops->set_cr3(vcpu, context->root_hpa |
(vcpu->cr3 & (CR3_PCD_MASK | CR3_WPT_MASK)));
static int paging32E_init_context(struct kvm_vcpu *vcpu)
{
- int ret;
-
- if ((ret = paging64_init_context(vcpu)))
- return ret;
-
- vcpu->mmu.root_level = PT32E_ROOT_LEVEL;
- vcpu->mmu.shadow_root_level = PT32E_ROOT_LEVEL;
- return 0;
+ return paging64_init_context_common(vcpu, PT32E_ROOT_LEVEL);
}
static int init_kvm_mmu(struct kvm_vcpu *vcpu)
if (!is_paging(vcpu))
return nonpaging_init_context(vcpu);
- else if (kvm_arch_ops->is_long_mode(vcpu))
+ else if (is_long_mode(vcpu))
return paging64_init_context(vcpu);
else if (is_pae(vcpu))
return paging32E_init_context(vcpu);
return init_kvm_mmu(vcpu);
}
+void kvm_mmu_pre_write(struct kvm_vcpu *vcpu, gpa_t gpa, int bytes)
+{
+ gfn_t gfn = gpa >> PAGE_SHIFT;
+ struct kvm_mmu_page *page;
+ struct kvm_mmu_page *child;
+ struct hlist_node *node;
+ struct hlist_head *bucket;
+ unsigned index;
+ u64 *spte;
+ u64 pte;
+ unsigned offset = offset_in_page(gpa);
+ unsigned page_offset;
+ int level;
+
+ pgprintk("%s: gpa %llx bytes %d\n", __FUNCTION__, gpa, bytes);
+ index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+ bucket = &vcpu->kvm->mmu_page_hash[index];
+ hlist_for_each_entry(page, node, bucket, hash_link) {
+ if (page->gfn != gfn || page->role.metaphysical)
+ continue;
+ page_offset = offset;
+ level = page->role.level;
+ if (page->role.glevels == PT32_ROOT_LEVEL) {
+ page_offset <<= 1; /* 32->64 */
+ page_offset &= ~PAGE_MASK;
+ }
+ spte = __va(page->page_hpa);
+ spte += page_offset / sizeof(*spte);
+ pte = *spte;
+ if (is_present_pte(pte)) {
+ if (level == PT_PAGE_TABLE_LEVEL)
+ rmap_remove(vcpu->kvm, spte);
+ else {
+ child = page_header(pte & PT64_BASE_ADDR_MASK);
+ mmu_page_remove_parent_pte(child, spte);
+ }
+ }
+ *spte = 0;
+ }
+}
+
+void kvm_mmu_post_write(struct kvm_vcpu *vcpu, gpa_t gpa, int bytes)
+{
+}
+
+int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva)
+{
+ gpa_t gpa = vcpu->mmu.gva_to_gpa(vcpu, gva);
+
+ return kvm_mmu_unprotect_page(vcpu, gpa >> PAGE_SHIFT);
+}
+
+void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu)
+{
+ while (vcpu->kvm->n_free_mmu_pages < KVM_REFILL_PAGES) {
+ struct kvm_mmu_page *page;
+
+ page = container_of(vcpu->kvm->active_mmu_pages.prev,
+ struct kvm_mmu_page, link);
+ kvm_mmu_zap_page(vcpu, page);
+ }
+}
+EXPORT_SYMBOL_GPL(kvm_mmu_free_some_pages);
+
static void free_mmu_pages(struct kvm_vcpu *vcpu)
{
while (!list_empty(&vcpu->free_pages)) {
__free_page(pfn_to_page(page->page_hpa >> PAGE_SHIFT));
page->page_hpa = INVALID_PAGE;
}
+ free_page((unsigned long)vcpu->mmu.pae_root);
}
static int alloc_mmu_pages(struct kvm_vcpu *vcpu)
{
+ struct page *page;
int i;
ASSERT(vcpu);
for (i = 0; i < KVM_NUM_MMU_PAGES; i++) {
- struct page *page;
struct kvm_mmu_page *page_header = &vcpu->page_header_buf[i];
INIT_LIST_HEAD(&page_header->link);
- if ((page = alloc_page(GFP_KVM_MMU)) == NULL)
+ if ((page = alloc_page(GFP_KERNEL)) == NULL)
goto error_1;
page->private = (unsigned long)page_header;
page_header->page_hpa = (hpa_t)page_to_pfn(page) << PAGE_SHIFT;
memset(__va(page_header->page_hpa), 0, PAGE_SIZE);
list_add(&page_header->link, &vcpu->free_pages);
+ ++vcpu->kvm->n_free_mmu_pages;
}
+
+ /*
+ * When emulating 32-bit mode, cr3 is only 32 bits even on x86_64.
+ * Therefore we need to allocate shadow page tables in the first
+ * 4GB of memory, which happens to fit the DMA32 zone.
+ */
+ page = alloc_page(GFP_KERNEL | __GFP_DMA32);
+ if (!page)
+ goto error_1;
+ vcpu->mmu.pae_root = page_address(page);
+ for (i = 0; i < 4; ++i)
+ vcpu->mmu.pae_root[i] = INVALID_PAGE;
+
return 0;
error_1:
return -ENOMEM;
}
-int kvm_mmu_init(struct kvm_vcpu *vcpu)
+int kvm_mmu_create(struct kvm_vcpu *vcpu)
{
- int r;
-
ASSERT(vcpu);
ASSERT(!VALID_PAGE(vcpu->mmu.root_hpa));
ASSERT(list_empty(&vcpu->free_pages));
- if ((r = alloc_mmu_pages(vcpu)))
- return r;
+ return alloc_mmu_pages(vcpu);
+}
- if ((r = init_kvm_mmu(vcpu))) {
- free_mmu_pages(vcpu);
- return r;
- }
- return 0;
+int kvm_mmu_setup(struct kvm_vcpu *vcpu)
+{
+ ASSERT(vcpu);
+ ASSERT(!VALID_PAGE(vcpu->mmu.root_hpa));
+ ASSERT(!list_empty(&vcpu->free_pages));
+
+ return init_kvm_mmu(vcpu);
}
void kvm_mmu_destroy(struct kvm_vcpu *vcpu)
pt = __va(page->page_hpa);
for (i = 0; i < PT64_ENT_PER_PAGE; ++i)
/* avoid RMW */
- if (pt[i] & PT_WRITABLE_MASK)
+ if (pt[i] & PT_WRITABLE_MASK) {
+ rmap_remove(kvm, &pt[i]);
pt[i] &= ~PT_WRITABLE_MASK;
-
+ }
}
}
git.openpandora.org / pandora-kernel.git / blobdiff