2021-06-22 07:26:29

by Sean Christopherson

Subject: [PATCH] KVM: x86/mmu: Don't WARN on a NULL shadow page in TDP MMU check

Treat a NULL shadow page in the "is a TDP MMU" check as valid, non-TDP
root. KVM uses a "direct" PAE paging MMU when TDP is disabled and the
guest is running with paging disabled. In that case, root_hpa points at
the pae_root page (of which only 32 bytes are used), not a standard
shadow page, and the WARN fires (a lot).

Fixes: 0b873fd7fb53 ("KVM: x86/mmu: Remove redundant is_tdp_mmu_enabled check")
Cc: David Matlack <[email protected]>
Signed-off-by: Sean Christopherson <[email protected]>
---
arch/x86/kvm/mmu/tdp_mmu.h | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kvm/mmu/tdp_mmu.h b/arch/x86/kvm/mmu/tdp_mmu.h
index b981a044ab55..1cae4485b3bc 100644
--- a/arch/x86/kvm/mmu/tdp_mmu.h
+++ b/arch/x86/kvm/mmu/tdp_mmu.h
@@ -94,11 +94,13 @@ static inline bool is_tdp_mmu(struct kvm_mmu *mmu)
if (WARN_ON(!VALID_PAGE(hpa)))
return false;

+ /*
+ * A NULL shadow page is legal when shadowing a non-paging guest with
+ * PAE paging, as the MMU will be direct with root_hpa pointing at the
+ * pae_root page, not a shadow page.
+ */
sp = to_shadow_page(hpa);
- if (WARN_ON(!sp))
- return false;
-
- return is_tdp_mmu_page(sp) && sp->root_count;
+ return sp && is_tdp_mmu_page(sp) && sp->root_count;
}
#else
static inline bool kvm_mmu_init_tdp_mmu(struct kvm *kvm) { return false; }
--
2.32.0.288.g62a8d224e6-goog

2021-06-22 22:41:07

by David Matlack

Subject: Re: [PATCH] KVM: x86/mmu: Don't WARN on a NULL shadow page in TDP MMU check

On Tue, Jun 22, 2021 at 12:24 AM Sean Christopherson <[email protected]> wrote:
>
> Treat a NULL shadow page in the "is a TDP MMU" check as valid, non-TDP
> root. KVM uses a "direct" PAE paging MMU when TDP is disabled and the
> guest is running with paging disabled. In that case, root_hpa points at
> the pae_root page (of which only 32 bytes are used), not a standard
> shadow page, and the WARN fires (a lot).
>
> Fixes: 0b873fd7fb53 ("KVM: x86/mmu: Remove redundant is_tdp_mmu_enabled check")
> Cc: David Matlack <[email protected]>
> Signed-off-by: Sean Christopherson <[email protected]>

Thanks for the fix. I was able to reproduce the issue by running a
kvm-unit-test with EPT=N. I'll add that to my "pre-send-email"
workflow in the future.

Reviewed-by: David Matlack <[email protected]>

> ---
> arch/x86/kvm/mmu/tdp_mmu.h | 10 ++++++----
> 1 file changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/arch/x86/kvm/mmu/tdp_mmu.h b/arch/x86/kvm/mmu/tdp_mmu.h
> index b981a044ab55..1cae4485b3bc 100644
> --- a/arch/x86/kvm/mmu/tdp_mmu.h
> +++ b/arch/x86/kvm/mmu/tdp_mmu.h
> @@ -94,11 +94,13 @@ static inline bool is_tdp_mmu(struct kvm_mmu *mmu)
> if (WARN_ON(!VALID_PAGE(hpa)))
> return false;
>
> + /*
> + * A NULL shadow page is legal when shadowing a non-paging guest with
> + * PAE paging, as the MMU will be direct with root_hpa pointing at the
> + * pae_root page, not a shadow page.
> + */
> sp = to_shadow_page(hpa);
> - if (WARN_ON(!sp))
> - return false;
> -
> - return is_tdp_mmu_page(sp) && sp->root_count;
> + return sp && is_tdp_mmu_page(sp) && sp->root_count;
> }
> #else
> static inline bool kvm_mmu_init_tdp_mmu(struct kvm *kvm) { return false; }
> --
> 2.32.0.288.g62a8d224e6-goog
>

2021-06-23 13:29:04

by Paolo Bonzini

Subject: Re: [PATCH] KVM: x86/mmu: Don't WARN on a NULL shadow page in TDP MMU check

On 22/06/21 09:24, Sean Christopherson wrote:
> Treat a NULL shadow page in the "is a TDP MMU" check as valid, non-TDP
> root. KVM uses a "direct" PAE paging MMU when TDP is disabled and the
> guest is running with paging disabled. In that case, root_hpa points at
> the pae_root page (of which only 32 bytes are used), not a standard
> shadow page, and the WARN fires (a lot).
>
> Fixes: 0b873fd7fb53 ("KVM: x86/mmu: Remove redundant is_tdp_mmu_enabled check")
> Cc: David Matlack <[email protected]>
> Signed-off-by: Sean Christopherson <[email protected]>
> ---
> arch/x86/kvm/mmu/tdp_mmu.h | 10 ++++++----
> 1 file changed, 6 insertions(+), 4 deletions(-)

Queued, thanks.

Paolo

>
> diff --git a/arch/x86/kvm/mmu/tdp_mmu.h b/arch/x86/kvm/mmu/tdp_mmu.h
> index b981a044ab55..1cae4485b3bc 100644
> --- a/arch/x86/kvm/mmu/tdp_mmu.h
> +++ b/arch/x86/kvm/mmu/tdp_mmu.h
> @@ -94,11 +94,13 @@ static inline bool is_tdp_mmu(struct kvm_mmu *mmu)
> if (WARN_ON(!VALID_PAGE(hpa)))
> return false;
>
> + /*
> + * A NULL shadow page is legal when shadowing a non-paging guest with
> + * PAE paging, as the MMU will be direct with root_hpa pointing at the
> + * pae_root page, not a shadow page.
> + */
> sp = to_shadow_page(hpa);
> - if (WARN_ON(!sp))
> - return false;
> -
> - return is_tdp_mmu_page(sp) && sp->root_count;
> + return sp && is_tdp_mmu_page(sp) && sp->root_count;
> }
> #else
> static inline bool kvm_mmu_init_tdp_mmu(struct kvm *kvm) { return false; }
>