When we tested KVM under memory pressure, with THP enabled on the host,
we noticed that MMU notifier took a long time to invalidate huge pages.
Since the invalidation was done with mmu_lock held, it not only wasted
the CPU but also made the host harder to respond.
This patch mitigates this by using kvm_handle_hva_range().
Signed-off-by: Takuya Yoshikawa <yoshikawa.takuya@xxxxxxxxxxxxx>
Cc: Alexander Graf <agraf@xxxxxxx>
Cc: Paul Mackerras <paulus@xxxxxxxxx>
---
arch/powerpc/include/asm/kvm_host.h | 2 ++
arch/powerpc/kvm/book3s_64_mmu_hv.c | 7 +++++++
arch/x86/include/asm/kvm_host.h | 1 +
arch/x86/kvm/mmu.c | 5 +++++
virt/kvm/kvm_main.c | 3 +--
5 files changed, 16 insertions(+), 2 deletions(-)
diff --git a/arch/powerpc/include/asm/kvm_host.h b/arch/powerpc/include/asm/kvm_host.h
index 50ea12f..572ad01 100644
--- a/arch/powerpc/include/asm/kvm_host.h
+++ b/arch/powerpc/include/asm/kvm_host.h
@@ -52,6 +52,8 @@
struct kvm;
extern int kvm_unmap_hva(struct kvm *kvm, unsigned long hva);
+extern int kvm_unmap_hva_range(struct kvm *kvm,
+ unsigned long start, unsigned long end);
extern int kvm_age_hva(struct kvm *kvm, unsigned long hva);
extern int kvm_test_age_hva(struct kvm *kvm, unsigned long hva);
extern void kvm_set_spte_hva(struct kvm *kvm, unsigned long hva, pte_t pte);
diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
index aec8046..33a6dc2 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -865,6 +865,13 @@ int kvm_unmap_hva(struct kvm *kvm, unsigned long hva)
return 0;
}
+int kvm_unmap_hva_range(struct kvm *kvm, unsigned long start, unsigned long end)
+{
+ if (kvm->arch.using_mmu_notifiers)
+ kvm_handle_hva_range(kvm, start, end, kvm_unmap_rmapp);
+ return 0;
+}
+
static int kvm_age_rmapp(struct kvm *kvm, unsigned long *rmapp,
unsigned long gfn)
{
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index db7c1f2..d6ad50d 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -930,6 +930,7 @@ extern bool kvm_rebooting;
#define KVM_ARCH_WANT_MMU_NOTIFIER
int kvm_unmap_hva(struct kvm *kvm, unsigned long hva);
+int kvm_unmap_hva_range(struct kvm *kvm, unsigned long start, unsigned long end);
int kvm_age_hva(struct kvm *kvm, unsigned long hva);
int kvm_test_age_hva(struct kvm *kvm, unsigned long hva);
void kvm_set_spte_hva(struct kvm *kvm, unsigned long hva, pte_t pte);
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index f995b8a..0659376 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -1243,6 +1243,11 @@ int kvm_unmap_hva(struct kvm *kvm, unsigned long hva)
return kvm_handle_hva(kvm, hva, 0, kvm_unmap_rmapp);
}
+int kvm_unmap_hva_range(struct kvm *kvm, unsigned long start, unsigned long end)
+{
+ return kvm_handle_hva_range(kvm, start, end, 0, kvm_unmap_rmapp);
+}
+
void kvm_set_spte_hva(struct kvm *kvm, unsigned long hva, pte_t pte)
{
kvm_handle_hva(kvm, hva, (unsigned long)&pte, kvm_set_pte_rmapp);
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 02cb440..44e44c0 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -332,8 +332,7 @@ static void kvm_mmu_notifier_invalidate_range_start(struct mmu_notifier *mn,
* count is also read inside the mmu_lock critical section.
*/
kvm->mmu_notifier_count++;
- for (; start < end; start += PAGE_SIZE)
- need_tlb_flush |= kvm_unmap_hva(kvm, start);
+ need_tlb_flush = kvm_unmap_hva_range(kvm, start, end);
need_tlb_flush |= kvm->tlbs_dirty;
/* we've to flush the tlb before the pages can be freed */
if (need_tlb_flush)
--
1.7.5.4
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
