KVM: MMU: Push clean gpte write protection out of gpte_access()

gpte_access() computes the access permissions of a guest pte and also
write-protects clean gptes.  This is wrong when we are servicing a
write fault (since we'll be setting the dirty bit momentarily) but
correct when instantiating a speculative spte, or when servicing a
read fault (since we'll want to trap a following write in order to
set the dirty bit).

It doesn't seem to hurt in practice, but in order to make the code
readable, push the write protection out of gpte_access() and into
a new protect_clean_gpte() which is called explicitly when needed.

Reviewed-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>

3 files changed, 26 insertions, 13 deletions

diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index aa0b469ee07d..54c9cb4fdfa4 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -3408,6 +3408,18 @@ static bool is_rsvd_bits_set(struct kvm_mmu *mmu, u64 gpte, int level)
         return (gpte & mmu->rsvd_bits_mask[bit7][level-1]) != 0;
 }
 
+static inline void protect_clean_gpte(unsigned *access, unsigned gpte)
+{
+        unsigned mask;
+
+        BUILD_BUG_ON(PT_WRITABLE_MASK != ACC_WRITE_MASK);
+
+        mask = (unsigned)~ACC_WRITE_MASK;
+        /* Allow write access to dirty gptes */
+        mask |= (gpte >> (PT_DIRTY_SHIFT - PT_WRITABLE_SHIFT)) & PT_WRITABLE_MASK;
+        *access &= mask;
+}
+
 static bool sync_mmio_spte(u64 *sptep, gfn_t gfn, unsigned access,
                            int *nr_present)
 {
diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
index e374db9af021..2832081e9b2e 100644
--- a/arch/x86/kvm/mmu.h
+++ b/arch/x86/kvm/mmu.h
@@ -18,7 +18,8 @@
 #define PT_PCD_MASK (1ULL << 4)
 #define PT_ACCESSED_SHIFT 5
 #define PT_ACCESSED_MASK (1ULL << PT_ACCESSED_SHIFT)
-#define PT_DIRTY_MASK (1ULL << 6)
+#define PT_DIRTY_SHIFT 6
+#define PT_DIRTY_MASK (1ULL << PT_DIRTY_SHIFT)
 #define PT_PAGE_SIZE_MASK (1ULL << 7)
 #define PT_PAT_MASK (1ULL << 7)
 #define PT_GLOBAL_MASK (1ULL << 8)
diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index bf8c42bf50fe..bf7b4ffafab8 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -101,14 +101,11 @@ static int FNAME(cmpxchg_gpte)(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
         return (ret != orig_pte);
 }
 
-static unsigned FNAME(gpte_access)(struct kvm_vcpu *vcpu, pt_element_t gpte,
-                                   bool last)
+static unsigned FNAME(gpte_access)(struct kvm_vcpu *vcpu, pt_element_t gpte)
 {
         unsigned access;
 
         access = (gpte & (PT_WRITABLE_MASK | PT_USER_MASK)) | ACC_EXEC_MASK;
-        if (last && !is_dirty_gpte(gpte))
-                access &= ~ACC_WRITE_MASK;
 
 #if PTTYPE == 64
         if (vcpu->arch.mmu.nx)
@@ -222,8 +219,7 @@ retry_walk:
 
                 last_gpte = FNAME(is_last_gpte)(walker, vcpu, mmu, pte);
                 if (last_gpte) {
-                        pte_access = pt_access &
-                                     FNAME(gpte_access)(vcpu, pte, true);
+                        pte_access = pt_access & FNAME(gpte_access)(vcpu, pte);
                         /* check if the kernel is fetching from user page */
                         if (unlikely(pte_access & PT_USER_MASK) &&
                             kvm_read_cr4_bits(vcpu, X86_CR4_SMEP))
@@ -274,7 +270,7 @@ retry_walk:
                         break;
                 }
 
-                pt_access &= FNAME(gpte_access)(vcpu, pte, false);
+                pt_access &= FNAME(gpte_access)(vcpu, pte);
                 --walker->level;
         }
 
@@ -283,7 +279,9 @@ retry_walk:
                 goto error;
         }
 
-        if (write_fault && unlikely(!is_dirty_gpte(pte))) {
+        if (!write_fault)
+                protect_clean_gpte(&pte_access, pte);
+        else if (unlikely(!is_dirty_gpte(pte))) {
                 int ret;
 
                 trace_kvm_mmu_set_dirty_bit(table_gfn, index, sizeof(pte));
@@ -368,7 +366,8 @@ static void FNAME(update_pte)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
                 return;
 
         pgprintk("%s: gpte %llx spte %p\n", __func__, (u64)gpte, spte);
-        pte_access = sp->role.access & FNAME(gpte_access)(vcpu, gpte, true);
+        pte_access = sp->role.access & FNAME(gpte_access)(vcpu, gpte);
+        protect_clean_gpte(&pte_access, gpte);
         pfn = gfn_to_pfn_atomic(vcpu->kvm, gpte_to_gfn(gpte));
         if (mmu_invalid_pfn(pfn))
                 return;
@@ -441,8 +440,8 @@ static void FNAME(pte_prefetch)(struct kvm_vcpu *vcpu, struct guest_walker *gw,
                 if (FNAME(prefetch_invalid_gpte)(vcpu, sp, spte, gpte))
                         continue;
 
-                pte_access = sp->role.access & FNAME(gpte_access)(vcpu, gpte,
-                                                                  true);
+                pte_access = sp->role.access & FNAME(gpte_access)(vcpu, gpte);
+                protect_clean_gpte(&pte_access, gpte);
                 gfn = gpte_to_gfn(gpte);
                 pfn = pte_prefetch_gfn_to_pfn(vcpu, gfn,
                                       pte_access & ACC_WRITE_MASK);
@@ -794,7 +793,8 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 
                 gfn = gpte_to_gfn(gpte);
                 pte_access = sp->role.access;
-                pte_access &= FNAME(gpte_access)(vcpu, gpte, true);
+                pte_access &= FNAME(gpte_access)(vcpu, gpte);
+                protect_clean_gpte(&pte_access, gpte);
 
                 if (sync_mmio_spte(&sp->spt[i], gfn, pte_access, &nr_present))
                         continue;