Re: [PATCH] Revert "KVM: MMU: Move kvm_mmu_free_some_pages() into kvm_mmu_alloc_page()"

[Gleb Natapov <gleb@redhat.com>]

On Fri, Mar 29, 2013 at 02:05:26PM +0900, Takuya Yoshikawa wrote:
> With the following commit, shadow pages can be zapped at random during
> a shadow page talbe walk:
>   KVM: MMU: Move kvm_mmu_free_some_pages() into kvm_mmu_alloc_page()
>   7ddca7e43c8f28f9419da81a0e7730b66aa60fe9
> 
> This patch reverts it and fixes __direct_map() and FNAME(fetch)().
> 
> Signed-off-by: Takuya Yoshikawa <yoshikawa_takuya_b1@lab.ntt.co.jp>
Applied, thanks.

> ---
>  arch/x86/kvm/mmu.c         |   11 +++++++----
>  arch/x86/kvm/paging_tmpl.h |    1 +
>  2 files changed, 8 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
> index 633e30c..004cc87 100644
> --- a/arch/x86/kvm/mmu.c
> +++ b/arch/x86/kvm/mmu.c
> @@ -1501,15 +1501,11 @@ static void drop_parent_pte(struct kvm_mmu_page *sp,
>  	mmu_spte_clear_no_track(parent_pte);
>  }
>  
> -static void make_mmu_pages_available(struct kvm_vcpu *vcpu);
> -
>  static struct kvm_mmu_page *kvm_mmu_alloc_page(struct kvm_vcpu *vcpu,
>  					       u64 *parent_pte, int direct)
>  {
>  	struct kvm_mmu_page *sp;
>  
> -	make_mmu_pages_available(vcpu);
> -
>  	sp = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_header_cache);
>  	sp->spt = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_cache);
>  	if (!direct)
> @@ -2806,6 +2802,7 @@ exit:
>  
>  static bool try_async_pf(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn,
>  			 gva_t gva, pfn_t *pfn, bool write, bool *writable);
> +static void make_mmu_pages_available(struct kvm_vcpu *vcpu);
>  
>  static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, u32 error_code,
>  			 gfn_t gfn, bool prefault)
> @@ -2847,6 +2844,7 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, u32 error_code,
>  	spin_lock(&vcpu->kvm->mmu_lock);
>  	if (mmu_notifier_retry(vcpu->kvm, mmu_seq))
>  		goto out_unlock;
> +	make_mmu_pages_available(vcpu);
>  	if (likely(!force_pt_level))
>  		transparent_hugepage_adjust(vcpu, &gfn, &pfn, &level);
>  	r = __direct_map(vcpu, v, write, map_writable, level, gfn, pfn,
> @@ -2924,6 +2922,7 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu *vcpu)
>  
>  	if (vcpu->arch.mmu.shadow_root_level == PT64_ROOT_LEVEL) {
>  		spin_lock(&vcpu->kvm->mmu_lock);
> +		make_mmu_pages_available(vcpu);
>  		sp = kvm_mmu_get_page(vcpu, 0, 0, PT64_ROOT_LEVEL,
>  				      1, ACC_ALL, NULL);
>  		++sp->root_count;
> @@ -2935,6 +2934,7 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu *vcpu)
>  
>  			ASSERT(!VALID_PAGE(root));
>  			spin_lock(&vcpu->kvm->mmu_lock);
> +			make_mmu_pages_available(vcpu);
>  			sp = kvm_mmu_get_page(vcpu, i << (30 - PAGE_SHIFT),
>  					      i << 30,
>  					      PT32_ROOT_LEVEL, 1, ACC_ALL,
> @@ -2973,6 +2973,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu *vcpu)
>  		ASSERT(!VALID_PAGE(root));
>  
>  		spin_lock(&vcpu->kvm->mmu_lock);
> +		make_mmu_pages_available(vcpu);
>  		sp = kvm_mmu_get_page(vcpu, root_gfn, 0, PT64_ROOT_LEVEL,
>  				      0, ACC_ALL, NULL);
>  		root = __pa(sp->spt);
> @@ -3006,6 +3007,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu *vcpu)
>  				return 1;
>  		}
>  		spin_lock(&vcpu->kvm->mmu_lock);
> +		make_mmu_pages_available(vcpu);
>  		sp = kvm_mmu_get_page(vcpu, root_gfn, i << 30,
>  				      PT32_ROOT_LEVEL, 0,
>  				      ACC_ALL, NULL);
> @@ -3311,6 +3313,7 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code,
>  	spin_lock(&vcpu->kvm->mmu_lock);
>  	if (mmu_notifier_retry(vcpu->kvm, mmu_seq))
>  		goto out_unlock;
> +	make_mmu_pages_available(vcpu);
>  	if (likely(!force_pt_level))
>  		transparent_hugepage_adjust(vcpu, &gfn, &pfn, &level);
>  	r = __direct_map(vcpu, gpa, write, map_writable,
> diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
> index af143f0..da20860 100644
> --- a/arch/x86/kvm/paging_tmpl.h
> +++ b/arch/x86/kvm/paging_tmpl.h
> @@ -627,6 +627,7 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, u32 error_code,
>  		goto out_unlock;
>  
>  	kvm_mmu_audit(vcpu, AUDIT_PRE_PAGE_FAULT);
> +	make_mmu_pages_available(vcpu);
>  	if (!force_pt_level)
>  		transparent_hugepage_adjust(vcpu, &walker.gfn, &pfn, &level);
>  	r = FNAME(fetch)(vcpu, addr, &walker, write_fault,
> -- 
> 1.7.5.4

--
			Gleb.