Re: [PATCH v3 8/9] KVM: x86: Expose Architectural LBR CPUID leaf

From: Sean Christopherson <seanjc@google.com>
To: Like Xu <like.xu@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Vitaly Kuznetsov <vkuznets@redhat.com>,
	Wanpeng Li <wanpengli@tencent.com>,
	Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>,
	Kan Liang <kan.liang@linux.intel.com>,
	Dave Hansen <dave.hansen@intel.com>,
	wei.w.wang@intel.com, Borislav Petkov <bp@alien8.de>,
	kvm@vger.kernel.org, x86@kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 8/9] KVM: x86: Expose Architectural LBR CPUID leaf
Date: Wed, 3 Mar 2021 10:01:43 -0800	[thread overview]
Message-ID: <YD/PB18qLqS7noKH@google.com> (raw)
In-Reply-To: <YD/IeTdqbK9kEDNp@google.com>

On Wed, Mar 03, 2021, Sean Christopherson wrote:
> On Wed, Mar 03, 2021, Like Xu wrote:
> > If CPUID.(EAX=07H, ECX=0):EDX[19] is set to 1, then KVM supports Arch
> > LBRs and CPUID leaf 01CH indicates details of the Arch LBRs capabilities.
> > Currently, KVM only supports the current host LBR depth for guests,
> > which is also the maximum supported depth on the host.
> > 
> > Signed-off-by: Like Xu <like.xu@linux.intel.com>
> > ---
> >  arch/x86/kvm/cpuid.c   | 25 ++++++++++++++++++++++++-
> >  arch/x86/kvm/vmx/vmx.c |  2 ++
> >  2 files changed, 26 insertions(+), 1 deletion(-)
> > 
> > diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c
> > index b4247f821277..4473324fe7be 100644
> > --- a/arch/x86/kvm/cpuid.c
> > +++ b/arch/x86/kvm/cpuid.c
> > @@ -450,7 +450,7 @@ void kvm_set_cpu_caps(void)
> >  		F(AVX512_4VNNIW) | F(AVX512_4FMAPS) | F(SPEC_CTRL) |
> >  		F(SPEC_CTRL_SSBD) | F(ARCH_CAPABILITIES) | F(INTEL_STIBP) |
> >  		F(MD_CLEAR) | F(AVX512_VP2INTERSECT) | F(FSRM) |
> > -		F(SERIALIZE) | F(TSXLDTRK) | F(AVX512_FP16)
> > +		F(SERIALIZE) | F(TSXLDTRK) | F(AVX512_FP16) | F(ARCH_LBR)
> >  	);
> >  
> >  	/* TSC_ADJUST and ARCH_CAPABILITIES are emulated in software. */
> 
> ...
> 
> > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> > index 2f307689a14b..034708a3df20 100644
> > --- a/arch/x86/kvm/vmx/vmx.c
> > +++ b/arch/x86/kvm/vmx/vmx.c
> > @@ -7258,6 +7258,8 @@ static __init void vmx_set_cpu_caps(void)
> >  		kvm_cpu_cap_clear(X86_FEATURE_INVPCID);
> >  	if (vmx_pt_mode_is_host_guest())
> >  		kvm_cpu_cap_check_and_set(X86_FEATURE_INTEL_PT);
> > +	if (cpu_has_vmx_arch_lbr())
> > +		kvm_cpu_cap_check_and_set(X86_FEATURE_ARCH_LBR);
> 
> Using kvm_cpu_cap_check_and_set(), which queries boot_cpu_has(), is only
> necessary if a feature is not exposed by default in kvm_set_cpu_caps().  That's
> why INTEL_PT uses it.  ARCH_LBR on the other hand is set in the "enable by
> default" mask.
> 
> That being said, it's probably a bad idea to advertise ARCH_LBR by default.  In
> the unlikely case that AMD adds support for ARCH_LBR, enable-by-default means
> guest will be able to use ARCH_LBR on old KVMs that presumably would lack support
> for ARCH_LBR on SVM.
> 
> TL;DR: omit F(ARCH_LBR) or replace it with "0 /* ARCH_LBR */".

Actually, I take that back.  It'll require changing SVM, but due to the XSS
interaction it's probably cleaner to leaf F(ARCH_LBR) as is, and do:

	if (!cpu_has_vmx_arch_lbr())
		kvm_cpu_cap_clear(X86_FEATURE_ARCH_LBR);

and then unconditionally clear the cap for SVM.  In a way, that's arguably
better documentation as it explicitly shows that SVM lacks supports.

More thoughts in the next patch...