KVM: MMU: increase per-vcpu rmap cache alloc size

KVM: MMU: increase per-vcpu rmap cache alloc size

[Marcelo Tosatti]

The page fault path can use two rmap_desc structures, if:

- walk_addr's dirty pte update allocates one rmap_desc.
- mmu_lock is dropped, sptes are zapped resulting in rmap_desc being
freed.
- fetch->mmu_set_spte allocates another rmap_desc.

Increase to 4 for safety.

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


diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 79cb4a9..2477a24 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -316,7 +316,7 @@ static int mmu_topup_memory_caches(struct kvm_vcpu *vcpu)
 	if (r)
 		goto out;
 	r = mmu_topup_memory_cache(&vcpu->arch.mmu_rmap_desc_cache,
-				   rmap_desc_cache, 1);
+				   rmap_desc_cache, 4);
 	if (r)
 		goto out;
 	r = mmu_topup_memory_cache_page(&vcpu->arch.mmu_page_cache, 8);

Re: KVM: MMU: increase per-vcpu rmap cache alloc size

[Avi Kivity]

Marcelo Tosatti wrote:
> The page fault path can use two rmap_desc structures, if:
>
> - walk_addr's dirty pte update allocates one rmap_desc.
> - mmu_lock is dropped, sptes are zapped resulting in rmap_desc being
> freed.
> - fetch->mmu_set_spte allocates another rmap_desc.
>
> Increase to 4 for safety.
>   

Applied and queued for 2.6.28, thanks.

-- 
error compiling committee.c: too many arguments to function