Re: [PATCH 6/10] KVM MMU: don't write-protect if have new mapping to unsync page

[Avi Kivity]

On 04/22/2010 09:13 AM, Xiao Guangrong wrote:

If have new mapping to the unsync page(i.e, add a new parent), just
update the page from sp-gfn but not write-protect gfn, and if need
create new shadow page form sp-gfn, we should sync it
   


Sorry, I don't understand this patch.  Can you clarify?  for example, 
the situation before adding the new parent, what the guest action was, 
and the new situation.


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

--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 6/10] KVM MMU: don't write-protect if have new mapping to unsync page

[Marcelo Tosatti]

On Thu, Apr 22, 2010 at 02:13:04PM +0800, Xiao Guangrong wrote:
 If have new mapping to the unsync page(i.e, add a new parent), just
 update the page from sp-gfn but not write-protect gfn, and if need
 create new shadow page form sp-gfn, we should sync it
 
 Signed-off-by: Xiao Guangrong xiaoguangr...@cn.fujitsu.com
 ---
  arch/x86/kvm/mmu.c |   27 +++
  1 files changed, 19 insertions(+), 8 deletions(-)
 
 diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
 index fd027a6..8607a64 100644
 --- a/arch/x86/kvm/mmu.c
 +++ b/arch/x86/kvm/mmu.c
 @@ -1196,16 +1196,20 @@ static void kvm_unlink_unsync_page(struct kvm *kvm, 
 struct kvm_mmu_page *sp)
  
  static int kvm_mmu_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp);
  
 -static int kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 +static int kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 +  bool clear_unsync)
  {
   if (sp-role.cr4_pae != !!is_pae(vcpu)) {
   kvm_mmu_zap_page(vcpu-kvm, sp);
   return 1;
   }
  
 - if (rmap_write_protect(vcpu-kvm, sp-gfn))
 - kvm_flush_remote_tlbs(vcpu-kvm);
 - kvm_unlink_unsync_page(vcpu-kvm, sp);
 + if (clear_unsync) {
 + if (rmap_write_protect(vcpu-kvm, sp-gfn))
 + kvm_flush_remote_tlbs(vcpu-kvm);
 + kvm_unlink_unsync_page(vcpu-kvm, sp);
 + }
 +
   if (vcpu-arch.mmu.sync_page(vcpu, sp)) {
   kvm_mmu_zap_page(vcpu-kvm, sp);
   return 1;
 @@ -1293,7 +1297,7 @@ static void mmu_sync_children(struct kvm_vcpu *vcpu,
   kvm_flush_remote_tlbs(vcpu-kvm);
  
   for_each_sp(pages, sp, parents, i) {
 - kvm_sync_page(vcpu, sp);
 + kvm_sync_page(vcpu, sp, true);
   mmu_pages_clear_parents(parents);
   }
   cond_resched_lock(vcpu-kvm-mmu_lock);
 @@ -1313,7 +1317,7 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct 
 kvm_vcpu *vcpu,
   unsigned index;
   unsigned quadrant;
   struct hlist_head *bucket;
 - struct kvm_mmu_page *sp;
 + struct kvm_mmu_page *sp, *unsync_sp = NULL;
   struct hlist_node *node, *tmp;
  
   role = vcpu-arch.mmu.base_role;
 @@ -1332,12 +1336,16 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct 
 kvm_vcpu *vcpu,
   hlist_for_each_entry_safe(sp, node, tmp, bucket, hash_link)
   if (sp-gfn == gfn) {
   if (sp-unsync)
 - if (kvm_sync_page(vcpu, sp))
 - continue;
 + unsync_sp = sp;

Xiao,

I don't see a reason why you can't create a new mapping to an unsync
page. The code already creates shadow pte entries using unsync
pagetables.

So all you need would be to kvm_sync_pages before write protecting.

Also make sure kvm_sync_pages is in place here before enabling multiple
unsync shadows, in the patch series.

  
   if (sp-role.word != role.word)
   continue;
  
 + if (unsync_sp  kvm_sync_page(vcpu, unsync_sp, false)) 
 {
 + unsync_sp = NULL;
 + continue;
 + }
 +
   mmu_page_add_parent_pte(vcpu, sp, parent_pte);
   if (sp-unsync_children) {
   set_bit(KVM_REQ_MMU_SYNC, vcpu-requests);
 @@ -1346,6 +1354,9 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct 
 kvm_vcpu *vcpu,
   trace_kvm_mmu_get_page(sp, false);
   return sp;
   }
 + if (unsync_sp)
 + kvm_sync_page(vcpu, unsync_sp, true);
 +
   ++vcpu-kvm-stat.mmu_cache_miss;
   sp = kvm_mmu_alloc_page(vcpu, parent_pte);
   if (!sp)
 -- 
 1.6.1.2
 
--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 6/10] KVM MMU: don't write-protect if have new mapping to unsync page

[Xiao Guangrong]

Marcelo Tosatti wrote:

  role = vcpu-arch.mmu.base_role;
 @@ -1332,12 +1336,16 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct 
 kvm_vcpu *vcpu,
  hlist_for_each_entry_safe(sp, node, tmp, bucket, hash_link)
  if (sp-gfn == gfn) {
  if (sp-unsync)
 -if (kvm_sync_page(vcpu, sp))
 -continue;
 +unsync_sp = sp;
 

Hi Marcelo,

Thanks for your comments, maybe the changlog is not clear, please allow
me explain here.

Two cases maybe happen in kvm_mmu_get_page() function:

- one case is, the goal sp is already in cache, if the sp is unsync,
  we only need update it to assure this mapping is valid, but not
  mark it sync and not write-protect sp-gfn since it not broke unsync
  rule(one shadow page for a gfn)

- another case is, the goal sp not existed, we need create a new sp
  for gfn, i.e, gfn (may)has another shadow page, to keep unsync rule,
  we should sync(mark sync and write-protect) gfn's unsync shadow page.
  After enabling multiple unsync shadows, we sync those shadow pages
  only when the new sp not allow to become unsync(also for the unsyc
  rule, the new rule is: allow all pte page become unsync)

 
 I don't see a reason why you can't create a new mapping to an unsync
 page. The code already creates shadow pte entries using unsync
 pagetables.

Do you means the case 2? In the original code, it unsync-ed gfn's unsync
page first regardless it's whether broke unsync rule:

|   hlist_for_each_entry_safe(sp, node, tmp, bucket, hash_link)
|   if (sp-gfn == gfn) {
|   if (sp-unsync)
|   if (kvm_sync_page(vcpu, sp))

And, my English is poor, sorry if i misunderstand your comment :-(

Xiao
--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html