1. Patchew
  2. linux
  3. [PATCH v9 00/10] support FEAT_MTE_TAGGED_FAR feature
  4. View patch

[PATCH v9 03/10] arm64/kvm: expose FEAT_MTE_TAGGED_FAR feature to guest

Yeoreum Yun posted 10 patches 7 months, 3 weeks ago
[PATCH v9 03/10] arm64/kvm: expose FEAT_MTE_TAGGED_FAR feature to guest
Posted by Yeoreum Yun 7 months, 3 weeks ago 
expose FEAT_MTE_TAGGED_FAR feature to guest.

Signed-off-by: Yeoreum Yun <yeoreum.yun@arm.com>
---
 arch/arm64/kvm/sys_regs.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index 76c2f0da821f..c8c92cb9da01 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -1586,7 +1586,7 @@ static u64 __kvm_read_sanitised_id_reg(const struct kvm_vcpu *vcpu,
 				       const struct sys_reg_desc *r)
 {
 	u32 id = reg_to_encoding(r);
-	u64 val;
+	u64 val, mask;
 
 	if (sysreg_visible_as_raz(vcpu, r))
 		return 0;
@@ -1617,8 +1617,12 @@ static u64 __kvm_read_sanitised_id_reg(const struct kvm_vcpu *vcpu,
 		val &= ~ARM64_FEATURE_MASK(ID_AA64PFR1_EL1_MPAM_frac);
 		break;
 	case SYS_ID_AA64PFR2_EL1:
-		/* We only expose FPMR */
-		val &= ID_AA64PFR2_EL1_FPMR;
+		mask = ID_AA64PFR2_EL1_FPMR;
+
+		if (kvm_has_mte(vcpu->kvm))
+			mask |= ID_AA64PFR2_EL1_MTEFAR;
+
+		val &= mask;
 		break;
 	case SYS_ID_AA64ISAR1_EL1:
 		if (!vcpu_has_ptrauth(vcpu))
@@ -2878,7 +2882,9 @@ static const struct sys_reg_desc sys_reg_descs[] = {
 				       ID_AA64PFR1_EL1_MPAM_frac |
 				       ID_AA64PFR1_EL1_RAS_frac |
 				       ID_AA64PFR1_EL1_MTE)),
-	ID_WRITABLE(ID_AA64PFR2_EL1, ID_AA64PFR2_EL1_FPMR),
+	ID_WRITABLE(ID_AA64PFR2_EL1,
+		    ID_AA64PFR2_EL1_FPMR |
+		    ID_AA64PFR2_EL1_MTEFAR),
 	ID_UNALLOCATED(4,3),
 	ID_WRITABLE(ID_AA64ZFR0_EL1, ~ID_AA64ZFR0_EL1_RES0),
 	ID_HIDDEN(ID_AA64SMFR0_EL1),
-- 
LEVI:{C3F47F37-75D8-414A-A8BA-3980EC8A46D7}
Re: [PATCH v9 03/10] arm64/kvm: expose FEAT_MTE_TAGGED_FAR feature to guest
Posted by Marc Zyngier 7 months, 3 weeks ago 
In general, please use a patch title format that matches the one used
for the subsystem. For KVM, that'd be "KVM: arm64: Expose ..."/

On Wed, 18 Jun 2025 09:45:06 +0100,
Yeoreum Yun <yeoreum.yun@arm.com> wrote:
> 
> expose FEAT_MTE_TAGGED_FAR feature to guest.
> 
> Signed-off-by: Yeoreum Yun <yeoreum.yun@arm.com>
> ---
>  arch/arm64/kvm/sys_regs.c | 14 ++++++++++----
>  1 file changed, 10 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
> index 76c2f0da821f..c8c92cb9da01 100644
> --- a/arch/arm64/kvm/sys_regs.c
> +++ b/arch/arm64/kvm/sys_regs.c
> @@ -1586,7 +1586,7 @@ static u64 __kvm_read_sanitised_id_reg(const struct kvm_vcpu *vcpu,
>  				       const struct sys_reg_desc *r)
>  {
>  	u32 id = reg_to_encoding(r);
> -	u64 val;
> +	u64 val, mask;
>  
>  	if (sysreg_visible_as_raz(vcpu, r))
>  		return 0;
> @@ -1617,8 +1617,12 @@ static u64 __kvm_read_sanitised_id_reg(const struct kvm_vcpu *vcpu,
>  		val &= ~ARM64_FEATURE_MASK(ID_AA64PFR1_EL1_MPAM_frac);
>  		break;
>  	case SYS_ID_AA64PFR2_EL1:
> -		/* We only expose FPMR */
> -		val &= ID_AA64PFR2_EL1_FPMR;
> +		mask = ID_AA64PFR2_EL1_FPMR;
> +
> +		if (kvm_has_mte(vcpu->kvm))
> +			mask |= ID_AA64PFR2_EL1_MTEFAR;
> +
> +		val &= mask;

I don't think there is a need for an extra variable, and you could
follow the pattern established in this file by writing this as:

	val &= (ID_AA64PFR2_EL1_FPMR |
		(kvm_has_mte(vcpu->kvm) ? ID_AA64PFR2_EL1_MTEFAR : 0));

Not a big deal though.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.
Re: [PATCH v9 03/10] arm64/kvm: expose FEAT_MTE_TAGGED_FAR feature to guest
Posted by Catalin Marinas 7 months, 2 weeks ago 
Hi Marc,

On Wed, Jun 18, 2025 at 05:43:07PM +0100, Marc Zyngier wrote:
> In general, please use a patch title format that matches the one used
> for the subsystem. For KVM, that'd be "KVM: arm64: Expose ..."/
[...]
> >  	case SYS_ID_AA64PFR2_EL1:
> > -		/* We only expose FPMR */
> > -		val &= ID_AA64PFR2_EL1_FPMR;
> > +		mask = ID_AA64PFR2_EL1_FPMR;
> > +
> > +		if (kvm_has_mte(vcpu->kvm))
> > +			mask |= ID_AA64PFR2_EL1_MTEFAR;
> > +
> > +		val &= mask;
> 
> I don't think there is a need for an extra variable, and you could
> follow the pattern established in this file by writing this as:
> 
> 	val &= (ID_AA64PFR2_EL1_FPMR |
> 		(kvm_has_mte(vcpu->kvm) ? ID_AA64PFR2_EL1_MTEFAR : 0));
> 
> Not a big deal though.

I can make the changes locally. Are you ok with the patch otherwise?

-- 
Catalin
Re: [PATCH v9 03/10] arm64/kvm: expose FEAT_MTE_TAGGED_FAR feature to guest
Posted by Marc Zyngier 7 months, 2 weeks ago 
On Tue, 24 Jun 2025 17:33:52 +0100,
Catalin Marinas <catalin.marinas@arm.com> wrote:
> 
> Hi Marc,
> 
> On Wed, Jun 18, 2025 at 05:43:07PM +0100, Marc Zyngier wrote:
> > In general, please use a patch title format that matches the one used
> > for the subsystem. For KVM, that'd be "KVM: arm64: Expose ..."/
> [...]
> > >  	case SYS_ID_AA64PFR2_EL1:
> > > -		/* We only expose FPMR */
> > > -		val &= ID_AA64PFR2_EL1_FPMR;
> > > +		mask = ID_AA64PFR2_EL1_FPMR;
> > > +
> > > +		if (kvm_has_mte(vcpu->kvm))
> > > +			mask |= ID_AA64PFR2_EL1_MTEFAR;
> > > +
> > > +		val &= mask;
> > 
> > I don't think there is a need for an extra variable, and you could
> > follow the pattern established in this file by writing this as:
> > 
> > 	val &= (ID_AA64PFR2_EL1_FPMR |
> > 		(kvm_has_mte(vcpu->kvm) ? ID_AA64PFR2_EL1_MTEFAR : 0));
> > 
> > Not a big deal though.
> 
> I can make the changes locally. Are you ok with the patch otherwise?

Yup, that'd fine by me. With this fixed:

Acked-by: Marc Zyngier <maz@kernel.org>

	M.

-- 
Jazz isn't dead. It just smells funny.
Re: [PATCH v9 03/10] arm64/kvm: expose FEAT_MTE_TAGGED_FAR feature to guest
Posted by Yeoreum Yun 7 months, 3 weeks ago 
Hi Marc,

> In general, please use a patch title format that matches the one used
> for the subsystem. For KVM, that'd be "KVM: arm64: Expose ..."/
>
> On Wed, 18 Jun 2025 09:45:06 +0100,
> Yeoreum Yun <yeoreum.yun@arm.com> wrote:
> >
> > expose FEAT_MTE_TAGGED_FAR feature to guest.
> >
> > Signed-off-by: Yeoreum Yun <yeoreum.yun@arm.com>
> > ---
> >  arch/arm64/kvm/sys_regs.c | 14 ++++++++++----
> >  1 file changed, 10 insertions(+), 4 deletions(-)
> >
> > diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
> > index 76c2f0da821f..c8c92cb9da01 100644
> > --- a/arch/arm64/kvm/sys_regs.c
> > +++ b/arch/arm64/kvm/sys_regs.c
> > @@ -1586,7 +1586,7 @@ static u64 __kvm_read_sanitised_id_reg(const struct kvm_vcpu *vcpu,
> >  				       const struct sys_reg_desc *r)
> >  {
> >  	u32 id = reg_to_encoding(r);
> > -	u64 val;
> > +	u64 val, mask;
> >
> >  	if (sysreg_visible_as_raz(vcpu, r))
> >  		return 0;
> > @@ -1617,8 +1617,12 @@ static u64 __kvm_read_sanitised_id_reg(const struct kvm_vcpu *vcpu,
> >  		val &= ~ARM64_FEATURE_MASK(ID_AA64PFR1_EL1_MPAM_frac);
> >  		break;
> >  	case SYS_ID_AA64PFR2_EL1:
> > -		/* We only expose FPMR */
> > -		val &= ID_AA64PFR2_EL1_FPMR;
> > +		mask = ID_AA64PFR2_EL1_FPMR;
> > +
> > +		if (kvm_has_mte(vcpu->kvm))
> > +			mask |= ID_AA64PFR2_EL1_MTEFAR;
> > +
> > +		val &= mask;
>
> I don't think there is a need for an extra variable, and you could
> follow the pattern established in this file by writing this as:
>
> 	val &= (ID_AA64PFR2_EL1_FPMR |
> 		(kvm_has_mte(vcpu->kvm) ? ID_AA64PFR2_EL1_MTEFAR : 0));
>
> Not a big deal though.

Thanks for your suggestion. I'll apply this with STORE_ONLY patch too
in end of this day.

--
Sincerely,
Yeoreum Yun

Patchew 2.1-devel-b67e4c2

© 2016 - 2026 Red Hat, Inc.