Re: [PATCH v7] KVM: arm64: Fix confusion in documentation for pKVM SME assert

[Mark Rutland <mark.rutland@arm.com>]

On Thu, Feb 13, 2025 at 08:55:52AM +0000, Marc Zyngier wrote:
> On Wed, 12 Feb 2025 11:11:04 +0000,
> Mark Rutland <mark.rutland@arm.com> wrote:
> > On Wed, Feb 12, 2025 at 12:44:57AM +0000, Mark Brown wrote:
> > > diff --git a/arch/arm64/kvm/fpsimd.c b/arch/arm64/kvm/fpsimd.c
> > > index 4d3d1a2eb157047b4b2488e9c4ffaabc6f5a0818..e37e53883c357093ff4455f5afdaec90e662d744 100644
> > > --- a/arch/arm64/kvm/fpsimd.c
> > > +++ b/arch/arm64/kvm/fpsimd.c
> > > @@ -93,11 +93,14 @@ void kvm_arch_vcpu_load_fp(struct kvm_vcpu *vcpu)
> > >  	}
> > >  
> > >  	/*
> > > -	 * If normal guests gain SME support, maintain this behavior for pKVM
> > > -	 * guests, which don't support SME.
> > > +	 * Protected and non-protected KVM modes require that
> > > +	 * SVCR.{SM,ZA} == {0,0} when entering a guest so that no
> > > +	 * host/guest SME state needs to be saved/restored by hyp code.
> > > +	 *
> > > +	 * In protected mode, hyp code will verify this later.
> > >  	 */
> > > -	WARN_ON(is_protected_kvm_enabled() && system_supports_sme() &&
> > > -		read_sysreg_s(SYS_SVCR));
> > > +	WARN_ON_ONCE(is_protected_kvm_enabled() && system_supports_sme() &&
> > > +		     read_sysreg_s(SYS_SVCR));
> > 
> > As I mentioned on the last round, we can drop the is_protected_kvm_enabled()
> > check, i.e. have:
> > 
> > 	/*
> > 	 * Protected and non-protected KVM modes require that
> > 	 * SVCR.{SM,ZA} == {0,0} when entering a guest so that no
> > 	 * host/guest SME state needs to be saved/restored by hyp code.
> > 	 *
> > 	 * In protected mode, hyp code will verify this later.
> > 	 */
> > 	WARN_ON_ONCE(system_supports_sme() && read_sysreg_s(SYS_SVCR));
> > 
> > Either way:
> > 
> > Acked-by: Mark Rutland <mark.rutland@arm.com>
> > 
> > Marc, are you happy to queue this atop the recent fixes from me? Those
> > try to ensure SVCR.{SM,ZA} == {0,0} regardless of whether KVM is in
> > protected mode.
> 
> In all honesty, I find that at this stage, the comment just gets in
> the way and is over-describing what is at stake here.
> 
> The
> 
>  	WARN_ON_ONCE(system_supports_sme() && read_sysreg_s(SYS_SVCR));
> 
> is really the only thing that matters. It perfectly shows what we are
> checking for, and doesn't need an exegesis.
> 
> As for the Fixes: tag, and given the magnitude of the actual fixes
> that are already queued, I don't think we need it.

That's fair; if you haven't spun a patch for that already, I guess we're
after the following?

Mark.

---->8----
From 4d05f6dd6d39c747c175782b7b44daa775251994 Mon Sep 17 00:00:00 2001
From: Mark Rutland <mark.rutland@arm.com>
Date: Thu, 13 Feb 2025 09:15:31 +0000
Subject: [PATCH] KVM: arm64: Simplify warning in kvm_arch_vcpu_load_fp()

At the end of kvm_arch_vcpu_load_fp() we check that no bits are set in
SVCR. We only check this for protected mode despite this mattering
equally for non-protected mode, and the comment above this is confusing.

Remove the comment and simplify the check, moving from WARN_ON() to
WARN_ON_ONCE() to avoid spamming the log.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
---
 arch/arm64/kvm/fpsimd.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/arch/arm64/kvm/fpsimd.c b/arch/arm64/kvm/fpsimd.c
index 3cbb999419af7..7f6e43d256915 100644
--- a/arch/arm64/kvm/fpsimd.c
+++ b/arch/arm64/kvm/fpsimd.c
@@ -65,12 +65,7 @@ void kvm_arch_vcpu_load_fp(struct kvm_vcpu *vcpu)
 	fpsimd_save_and_flush_cpu_state();
 	*host_data_ptr(fp_owner) = FP_STATE_FREE;
 
-	/*
-	 * If normal guests gain SME support, maintain this behavior for pKVM
-	 * guests, which don't support SME.
-	 */
-	WARN_ON(is_protected_kvm_enabled() && system_supports_sme() &&
-		read_sysreg_s(SYS_SVCR));
+	WARN_ON_ONCE(system_supports_sme() && read_sysreg_s(SYS_SVCR));
 }
 
 /*
-- 
2.30.2