KVM: switch to get_user_pages_fast

Convert gfn_to_pfn to use get_user_pages_fast, which can do lockless
pagetable lookups on x86. Kernel compilation on 4-way guest is 3.7%
faster on VMX.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>

1 files changed, 1 insertions, 7 deletions

diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index b671f61be41e..6dd08e096e24 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -102,14 +102,10 @@ static bool FNAME(cmpxchg_gpte)(struct kvm *kvm,
         pt_element_t *table;
         struct page *page;
 
-        down_read(&current->mm->mmap_sem);
         page = gfn_to_page(kvm, table_gfn);
-        up_read(&current->mm->mmap_sem);
 
         table = kmap_atomic(page, KM_USER0);
-
         ret = CMPXCHG(&table[index], orig_pte, new_pte);
-
         kunmap_atomic(table, KM_USER0);
 
         kvm_release_page_dirty(page);
@@ -418,7 +414,6 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
                 return 0;
         }
 
-        down_read(&current->mm->mmap_sem);
         if (walker.level == PT_DIRECTORY_LEVEL) {
                 gfn_t large_gfn;
                 large_gfn = walker.gfn & ~(KVM_PAGES_PER_HPAGE-1);
@@ -428,9 +423,8 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
                 }
         }
         mmu_seq = vcpu->kvm->mmu_notifier_seq;
-        /* implicit mb(), we'll read before PT lock is unlocked */
+        smp_rmb();
         pfn = gfn_to_pfn(vcpu->kvm, walker.gfn);
-        up_read(&current->mm->mmap_sem);
 
         /* mmio */
         if (is_error_pfn(pfn)) {