2021-10-17 03:21:01

by Andrei Vagin

[permalink] [raw]
Subject: [PATCH] KVM: x86/mmu: kvm_faultin_pfn has to return false if pfh is returned

This looks like a typo in 8f32d5e563cb. This change didn't intend to do
any functional changes.

The problem was caught by gVisor tests.

Fixes: 8f32d5e563cb ("KVM: x86/mmu: allow kvm_faultin_pfn to return page fault handling code")
Cc: Maxim Levitsky <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Signed-off-by: Andrei Vagin <[email protected]>
---
arch/x86/kvm/mmu/mmu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 1a64ba5b9437..5dce77b45476 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -3956,7 +3956,7 @@ static bool kvm_faultin_pfn(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn,

*pfn = __gfn_to_pfn_memslot(slot, gfn, false, NULL,
write, writable, hva);
-
+ return false;
out_retry:
*r = RET_PF_RETRY;
return true;
--
2.31.1


2021-10-17 08:39:32

by Paolo Bonzini

[permalink] [raw]
Subject: Re: [PATCH] KVM: x86/mmu: kvm_faultin_pfn has to return false if pfh is returned

On 15/10/21 18:32, Andrei Vagin wrote:
> This looks like a typo in 8f32d5e563cb. This change didn't intend to do
> any functional changes.
>
> The problem was caught by gVisor tests.
>
> Fixes: 8f32d5e563cb ("KVM: x86/mmu: allow kvm_faultin_pfn to return page fault handling code")
> Cc: Maxim Levitsky <[email protected]>
> Cc: Paolo Bonzini <[email protected]>
> Signed-off-by: Andrei Vagin <[email protected]>
> ---
> arch/x86/kvm/mmu/mmu.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> index 1a64ba5b9437..5dce77b45476 100644
> --- a/arch/x86/kvm/mmu/mmu.c
> +++ b/arch/x86/kvm/mmu/mmu.c
> @@ -3956,7 +3956,7 @@ static bool kvm_faultin_pfn(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn,
>
> *pfn = __gfn_to_pfn_memslot(slot, gfn, false, NULL,
> write, writable, hva);
> -
> + return false;
> out_retry:
> *r = RET_PF_RETRY;
> return true;
>

Ouch! Queued, thanks.

Paolo