We currently WARN() if the host attempts to share a page that is not in
an acceptable state with a guest. This isn't strictly necessary and
makes testing much harder, so drop the WARN and fix the error code.

Signed-off-by: Quentin Perret <qperret@google.com>
---
 arch/arm64/kvm/hyp/nvhe/mem_protect.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
index 19c3c631708c..ae39abc7e604 100644
--- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
+++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
@@ -912,7 +912,6 @@ int __pkvm_host_share_guest(u64 pfn, u64 gfn, struct pkvm_hyp_vcpu *vcpu,
 		if (page->host_share_guest_count)
 			break;
 		/* Only host to np-guest multi-sharing is tolerated */
-		WARN_ON(1);
 		fallthrough;
 	default:
 		ret = -EPERM;
-- 
2.48.1.658.g4767266eb4-goog

On Tue, 25 Feb 2025 01:53:25 +0000,
Quentin Perret <qperret@google.com> wrote:
> 
> We currently WARN() if the host attempts to share a page that is not in
> an acceptable state with a guest. This isn't strictly necessary and
> makes testing much harder, so drop the WARN and fix the error code.

Are you really fixing the error code? You still seem to return a
-EPERM. I guess this was never reachable thanks to WARN() being a
panic with pKVM?

	M.

> 
> Signed-off-by: Quentin Perret <qperret@google.com>
> ---
>  arch/arm64/kvm/hyp/nvhe/mem_protect.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> index 19c3c631708c..ae39abc7e604 100644
> --- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> +++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> @@ -912,7 +912,6 @@ int __pkvm_host_share_guest(u64 pfn, u64 gfn, struct pkvm_hyp_vcpu *vcpu,
>  		if (page->host_share_guest_count)
>  			break;
>  		/* Only host to np-guest multi-sharing is tolerated */
> -		WARN_ON(1);
>  		fallthrough;
>  	default:
>  		ret = -EPERM;
> -- 
> 2.48.1.658.g4767266eb4-goog
> 
> 

-- 
Without deviation from the norm, progress is not possible.

On Tuesday 25 Feb 2025 at 18:02:02 (+0000), Marc Zyngier wrote:
> On Tue, 25 Feb 2025 01:53:25 +0000,
> Quentin Perret <qperret@google.com> wrote:
> > 
> > We currently WARN() if the host attempts to share a page that is not in
> > an acceptable state with a guest. This isn't strictly necessary and
> > makes testing much harder, so drop the WARN and fix the error code.
> 
> Are you really fixing the error code? You still seem to return a
> -EPERM. I guess this was never reachable thanks to WARN() being a
> panic with pKVM?

Exactly, this is really poor wording in the commit message. 'Fix the
error code' in this case was intended to mean 'make sure to return the
error code properly instead outright crashing the device'.

Happy to send out a v3 with a better commit message.

Thanks!
Quentin

On Tue, 25 Feb 2025 19:49:43 +0000,
Quentin Perret <qperret@google.com> wrote:
> 
> On Tuesday 25 Feb 2025 at 18:02:02 (+0000), Marc Zyngier wrote:
> > On Tue, 25 Feb 2025 01:53:25 +0000,
> > Quentin Perret <qperret@google.com> wrote:
> > > 
> > > We currently WARN() if the host attempts to share a page that is not in
> > > an acceptable state with a guest. This isn't strictly necessary and
> > > makes testing much harder, so drop the WARN and fix the error code.
> > 
> > Are you really fixing the error code? You still seem to return a
> > -EPERM. I guess this was never reachable thanks to WARN() being a
> > panic with pKVM?
> 
> Exactly, this is really poor wording in the commit message. 'Fix the
> error code' in this case was intended to mean 'make sure to return the
> error code properly instead outright crashing the device'.
> 
> Happy to send out a v3 with a better commit message.

Nah, that's probably something Oliver can fixup when picking up the
patch.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.