Skip to contentSkip to conversation

KVM: make kvm_mmu_zap_page() return the number of pages it actually freed.

PATCH In review Gui Jianfeng 2 participants 4 messages May 5, 2010 – May 6, 2010

Subsystems (3)
x86 KVM KVM x86
lore
Share & tools
lore

Gui Jianfeng

author
#0 lore raw

PATCH 1/1 KVM: make kvm_mmu_zap_page() return the number of pages it actually freed.

Currently, kvm_mmu_zap_page() returning the number of freed children sp. This might confuse the caller, because caller don't know the actual freed number. Let's make kvm_mmu_zap_page() return the number of pages it actually freed.

Signed-off-by Gui Jianfeng <[email protected]>
---
 arch/x86/kvm/mmu.c |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)
arch/x86/kvm/mmu.c +3 -2
index 51eb6d6..8ab6820 100644
@@ -1503,6 +1503,8 @@ static int kvm_mmu_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp)
 	if (sp->unsync)
 		kvm_unlink_unsync_page(kvm, sp);
 	if (!sp->root_count) {
+		/* Count self */
+		ret++;
 		hlist_del(&sp->hash_link);
 		kvm_mmu_free_page(kvm, sp);
 	} else {

@@ -1539,7 +1541,6 @@ void kvm_mmu_change_mmu_pages(struct kvm *kvm, unsigned int kvm_nr_mmu_pages)
 			page = container_of(kvm->arch.active_mmu_pages.prev,
 					    struct kvm_mmu_page, link);
 			used_pages -= kvm_mmu_zap_page(kvm, page);
-			used_pages--;
 		}
 		kvm_nr_mmu_pages = used_pages;
 		kvm->arch.n_free_mmu_pages = 0;

@@ -2908,7 +2909,7 @@ static int kvm_mmu_remove_some_alloc_mmu_pages(struct kvm *kvm)
 
 	page = container_of(kvm->arch.active_mmu_pages.prev,
 			    struct kvm_mmu_page, link);
-	return kvm_mmu_zap_page(kvm, page) + 1;
+	return kvm_mmu_zap_page(kvm, page);
 }
 
 static int mmu_shrink(int nr_to_scan, gfp_t gfp_mask)
Signature · 1 line
1.6.5.2
#1 lore raw
Quoted 6 lines · Gui Jianfeng

Currently, kvm_mmu_zap_page() returning the number of freed children sp. This might confuse the caller, because caller don't know the actual freed number. Let's make kvm_mmu_zap_page() return the number of pages it actually freed.

            if (kvm_mmu_zap_page(kvm, sp))
                goto restart;

Needs to be updated.

Signature · 1 line
error compiling committee.c: too many arguments to function
author
#2 lore raw
Quoted 9 lines · Avi Kivity
Quoted 7 lines · Gui Jianfeng

Currently, kvm_mmu_zap_page() returning the number of freed children sp. This might confuse the caller, because caller don't know the actual freed number. Let's make kvm_mmu_zap_page() return the number of pages it actually
freed.

Quoted 4 lines
            if (kvm_mmu_zap_page(kvm, sp))
                goto restart;

Needs to be updated.

Hi Avi,

if kvm_mmu_zap_page() returns 1, we don't know whether the freed sp is the one we passes into or the child. So here just restart hash walk as long as kvm_mmu_zap_page() returning a positive number, although sometimes un-needed hash walk will happen. This fix gets code simpler, the idea comes from Marcelo.

Thanks,
Gui

Quoted 1 line
#3 lore raw
Quoted 15 lines · Gui Jianfeng
Quoted 13 lines · Avi Kivity
Quoted 8 lines · Gui Jianfeng

Currently, kvm_mmu_zap_page() returning the number of freed children sp. This might confuse the caller, because caller don't know the actual freed number. Let's make kvm_mmu_zap_page() return the number of pages it actually
freed.

             if (kvm_mmu_zap_page(kvm, sp))
                 goto restart;
Quoted 9 lines
Quoted 2 lines

Needs to be updated.

Hi Avi,

if kvm_mmu_zap_page() returns 1, we don't know whether the freed sp is the one we passes into or the child. So here just restart hash walk as long as kvm_mmu_zap_page() returning a positive number, although sometimes un-needed hash walk will happen. This fix gets code simpler, the idea comes from Marcelo.

I see. Note this can invoke quadratic behaviour. I'll apply the patch, but we need to improve this area.

Signature · 1 line
error compiling committee.c: too many arguments to function