Re: [PATCH v2] KVM: x86/mmu: Speed up slot_rmap_walk_next for sparsely populated rmaps

From: Paolo Bonzini
Date: Tue May 03 2022 - 05:32:54 EST


On 5/3/22 00:03, Vipin Sharma wrote:
Avoid calling handlers on empty rmap entries and skip to the next non
empty rmap entry.

Empty rmap entries are noop in handlers.

Signed-off-by: Vipin Sharma <vipinsh@xxxxxxxxxx>
Suggested-by: Sean Christopherson <seanjc@xxxxxxxxxx>
---

v2:
- Removed noinline from slot_rmap_walk_next signature

v1:
- https://lore.kernel.org/lkml/20220325233125.413634-1-vipinsh@xxxxxxxxxx

arch/x86/kvm/mmu/mmu.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 77785587332e..4e8d546431eb 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -1501,9 +1501,11 @@ static bool slot_rmap_walk_okay(struct slot_rmap_walk_iterator *iterator)
static void slot_rmap_walk_next(struct slot_rmap_walk_iterator *iterator)
{
- if (++iterator->rmap <= iterator->end_rmap) {
+ while (++iterator->rmap <= iterator->end_rmap) {
iterator->gfn += (1UL << KVM_HPAGE_GFN_SHIFT(iterator->level));
- return;
+
+ if (iterator->rmap->val)
+ return;
}
if (++iterator->level > iterator->end_level) {

base-commit: 71d7c575a673d42ad7175ad5fc27c85c80330311

Queued, thanks.

Paolo