KVM: MMU: Optimize guest page table walk
This patch optimizes the guest page table walk by using get_user()
instead of copy_from_user().
With this patch applied, paging64_walk_addr_generic() has become
about 0.5us to 1.0us faster on my Phenom II machine with NPT on.
Signed-off-by: Takuya Yoshikawa <yoshikawa.takuya@oss.ntt.co.jp>
Signed-off-by: Avi Kivity <avi@redhat.com>
diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index 1b68990..a32a1c8 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -123,6 +123,7 @@
gva_t addr, u32 access)
{
pt_element_t pte;
+ pt_element_t __user *ptep_user;
gfn_t table_gfn;
unsigned index, pt_access, uninitialized_var(pte_access);
gpa_t pte_gpa;
@@ -158,6 +159,9 @@
pt_access = ACC_ALL;
for (;;) {
+ gfn_t real_gfn;
+ unsigned long host_addr;
+
index = PT_INDEX(addr, walker->level);
table_gfn = gpte_to_gfn(pte);
@@ -166,9 +170,22 @@
walker->table_gfn[walker->level - 1] = table_gfn;
walker->pte_gpa[walker->level - 1] = pte_gpa;
- if (kvm_read_guest_page_mmu(vcpu, mmu, table_gfn, &pte,
- offset, sizeof(pte),
- PFERR_USER_MASK|PFERR_WRITE_MASK)) {
+ real_gfn = mmu->translate_gpa(vcpu, gfn_to_gpa(table_gfn),
+ PFERR_USER_MASK|PFERR_WRITE_MASK);
+ if (real_gfn == UNMAPPED_GVA) {
+ present = false;
+ break;
+ }
+ real_gfn = gpa_to_gfn(real_gfn);
+
+ host_addr = gfn_to_hva(vcpu->kvm, real_gfn);
+ if (kvm_is_error_hva(host_addr)) {
+ present = false;
+ break;
+ }
+
+ ptep_user = (pt_element_t __user *)((void *)host_addr + offset);
+ if (get_user(pte, ptep_user)) {
present = false;
break;
}