[PATCH] KVM: no need to test bit before setting dirty bit these days

[PATCH] KVM: no need to test bit before setting dirty bit these days

[Takuya Yoshikawa]

As Avi pointed out, testing bit part in mark_page_dirty() was important
in the days of shadow paging, but currently EPT and NPT has already become
common and the chance of faulting a page more that once per iteration is
small. So let's remove the test bit to avoid extra access.

Signed-off-by: Takuya Yoshikawa <yoshikawa.takuya@oss.ntt.co.jp>
---
 virt/kvm/kvm_main.c |    4 +---
 1 files changed, 1 insertions(+), 3 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 6dc9404..9ab1a77 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1192,9 +1192,7 @@ void mark_page_dirty(struct kvm *kvm, gfn_t gfn)
 	if (memslot && memslot->dirty_bitmap) {
 		unsigned long rel_gfn = gfn - memslot->base_gfn;
 
-		/* avoid RMW */
-		if (!generic_test_le_bit(rel_gfn, memslot->dirty_bitmap))
-			generic___set_le_bit(rel_gfn, memslot->dirty_bitmap);
+		generic___set_le_bit(rel_gfn, memslot->dirty_bitmap);
 	}
 }
 
-- 
1.6.3.3

Re: [PATCH] KVM: no need to test bit before setting dirty bit these days

[Avi Kivity]

On 04/23/2010 11:48 AM, Takuya Yoshikawa wrote:
> As Avi pointed out, testing bit part in mark_page_dirty() was important
> in the days of shadow paging, but currently EPT and NPT has already become
> common and the chance of faulting a page more that once per iteration is
> small. So let's remove the test bit to avoid extra access.
>    

Applied, thanks.

-- 
Do not meddle in the internals of kernels, for they are subtle and quick to panic.