linux-kernel - Re: [PATCH v3] KVM: x86/pmu: Support full width counting

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, 8 May 2020 11:50:56 +0200
From:   Paolo Bonzini <pbonzini@...hat.com>
To:     Like Xu <like.xu@...ux.intel.com>
Cc:     Sean Christopherson <sean.j.christopherson@...el.com>,
        Vitaly Kuznetsov <vkuznets@...hat.com>,
        Wanpeng Li <wanpengli@...cent.com>,
        Jim Mattson <jmattson@...gle.com>,
        Joerg Roedel <joro@...tes.org>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] KVM: x86/pmu: Support full width counting

I would just do small changes to the validity checks for MSRs.

On 08/05/20 10:32, Like Xu wrote:
>  		return 0;
> +	case MSR_IA32_PERF_CAPABILITIES:
> +		*data = vcpu->arch.perf_capabilities;
> +		return 0;

This should be:

		if (!msr_info->host_initiated &&
		    !guest_cpuid_has(vcpu, X86_FEATURE_PDCM))
			return 1;

>  	default:
> -		if ((pmc = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0))) {
> +		if ((pmc = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0)) ||
> +			(pmc = get_gp_pmc(pmu, msr, MSR_IA32_PMC0))) {
>  			u64 val = pmc_read_counter(pmc);
>  			*data = val & pmu->counter_bitmask[KVM_PMC_GP];
>  			return 0;
> @@ -258,9 +277,21 @@ static int intel_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
>  			return 0;
>  		}
>  		break;
> +	case MSR_IA32_PERF_CAPABILITIES:
> +		if (msr_info->host_initiated &&
> +			!(data & ~vmx_get_perf_capabilities())) {

Likewise:

		if (!msr->info->host_initiated)
			return 1;
		if (guest_cpuid_has(vcpu, X86_FEATURE_PDCM))
		    ? data & ~vmx_get_perf_capabilities()
		    : data)
			return 1;

Otherwise looks good, I'm going to queue this.

Paolo