linux-kernel - Re: [PATCH] KVM: x86: Use cpu_feature_enabled() for PKU instead of #ifdef

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 2 Jun 2023 14:49:46 -0700
From:   Sean Christopherson <seanjc@...gle.com>
To:     Jim Mattson <jmattson@...gle.com>
Cc:     Jon Kohler <jon@...anix.com>, Mingwei Zhang <mizhang@...gle.com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        "kvm@...r.kernel.org" <kvm@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] KVM: x86: Use cpu_feature_enabled() for PKU instead of #ifdef

On Fri, Jun 02, 2023, Jim Mattson wrote:
> On Fri, Jun 2, 2023 at 8:51 AM Jon Kohler <jon@...anix.com> wrote:
> > > On Jun 1, 2023, at 9:05 PM, Sean Christopherson <seanjc@...gle.com> wrote:
> > > @@ -1032,15 +1030,13 @@ void kvm_load_host_xsave_state(struct kvm_vcpu *vcpu)
> > >       if (vcpu->arch.guest_state_protected)
> > >               return;
> > >
> > > -#ifdef CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS
> > > -     if (static_cpu_has(X86_FEATURE_PKU) &&
> > > +     if (cpu_feature_enabled(X86_FEATURE_PKU) &&
> > >           ((vcpu->arch.xcr0 & XFEATURE_MASK_PKRU) ||
> > >            kvm_is_cr4_bit_set(vcpu, X86_CR4_PKE))) {
> > >               vcpu->arch.pkru = rdpkru();
> > >               if (vcpu->arch.pkru != vcpu->arch.host_pkru)
> > >                       write_pkru(vcpu->arch.host_pkru);
> > >       }
> > > -#endif /* CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS */
> > >
> > >       if (kvm_is_cr4_bit_set(vcpu, X86_CR4_OSXSAVE)) {
> > >
> > >
> > > base-commit: a053a0e4a9f8c52f3acf8a9d2520c4bf39077a7e
> > > --
> > > 2.41.0.rc2.161.g9c6817b8e7-goog
> > >
> >
> > Thanks for the cleanup!
> >
> > Reviewed-by: Jon Kohler <jon@...anix.com>
> 
> +Mingwei Zhang
> 
> As we move towards enabling PKRU on the host, due to some customer
> requests, I have to wonder if PKRU-disabled is the norm.
> 
> In other words, is this a likely() or unlikely() optimization?

Neither?  I don't see any reason to speculate on guest state.  I'll bet dollars
to donuts that adding (un)likely() is negligible in terms of performance.