KVM: MMU: Fix rare oops on guest context switch

A guest context switch to an uncached cr3 can require allocation of
shadow pages, but we only recycle shadow pages in kvm_mmu_page_fault().

Move shadow page recycling to mmu_topup_memory_caches(), which is called
from both the page fault handler and from guest cr3 reload.

Signed-off-by: Avi Kivity <avi@qumranet.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index 3ac9cbce3369f31c7ae48bde1cd9a593be1393bd..336be86c6f5a9c023556a4b8703152213df094fb 100644 (file)
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -619,7 +619,7 @@ unsigned long segment_base(u16 selector);
 void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
                       const u8 *old, const u8 *new, int bytes);
 int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva);
-void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu);
+void __kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu);
 int kvm_mmu_load(struct kvm_vcpu *vcpu);
 void kvm_mmu_unload(struct kvm_vcpu *vcpu);
 
@@ -628,11 +628,15 @@ int kvm_hypercall(struct kvm_vcpu *vcpu, struct kvm_run *run);
 static inline int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gva_t gva,
                                     u32 error_code)
 {
-       if (unlikely(vcpu->kvm->n_free_mmu_pages < KVM_MIN_FREE_MMU_PAGES))
-               kvm_mmu_free_some_pages(vcpu);
        return vcpu->mmu.page_fault(vcpu, gva, error_code);
 }
 
+static inline void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu)
+{
+       if (unlikely(vcpu->kvm->n_free_mmu_pages < KVM_MIN_FREE_MMU_PAGES))
+               __kvm_mmu_free_some_pages(vcpu);
+}
+
 static inline int kvm_mmu_reload(struct kvm_vcpu *vcpu)
 {
        if (likely(vcpu->mmu.root_hpa != INVALID_PAGE))

diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
index 1a87ba9d515620986e079d64fd83ff546054a9b7..23965aa5ee7807315aebecf5b03d0970fbd1efbe 100644 (file)
--- a/drivers/kvm/mmu.c
+++ b/drivers/kvm/mmu.c
@@ -273,12 +273,14 @@ static int mmu_topup_memory_caches(struct kvm_vcpu *vcpu)
        int r;
 
        r = __mmu_topup_memory_caches(vcpu, GFP_NOWAIT);
+       kvm_mmu_free_some_pages(vcpu);
        if (r < 0) {
                spin_unlock(&vcpu->kvm->lock);
                kvm_arch_ops->vcpu_put(vcpu);
                r = __mmu_topup_memory_caches(vcpu, GFP_KERNEL);
                kvm_arch_ops->vcpu_load(vcpu);
                spin_lock(&vcpu->kvm->lock);
+               kvm_mmu_free_some_pages(vcpu);
        }
        return r;
 }
@@ -1208,7 +1210,7 @@ int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva)
        return kvm_mmu_unprotect_page(vcpu, gpa >> PAGE_SHIFT);
 }
 
-void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu)
+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;
@@ -1218,7 +1220,6 @@ void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu)
                kvm_mmu_zap_page(vcpu->kvm, page);
        }
 }
-EXPORT_SYMBOL_GPL(kvm_mmu_free_some_pages);
 
 static void free_mmu_pages(struct kvm_vcpu *vcpu)
 {