Received-SPF: pass (google.com: domain of linux-kernel+bounces-96496-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Date: Thu, 7 Mar 2024 20:13:03 -0800
In-Reply-To: <20240228101837.93642-4-vkuznets@redhat.com>
Precedence: bulk
Mime-Version: 1.0
References: <20240228101837.93642-1-vkuznets@redhat.com> <20240228101837.93642-4-vkuznets@redhat.com>
Message-ID: <ZeqQT1DAxnji75KZ@google.com>
Subject: Re: [PATCH 3/3] KVM: selftests: Check that KVM_FEATURE_PV_UNHALT is
 cleared with KVM_X86_DISABLE_EXITS_HLT
From: Sean Christopherson <seanjc@google.com>
To: Vitaly Kuznetsov <vkuznets@redhat.com>
Cc: kvm@vger.kernel.org, Paolo Bonzini <pbonzini@redhat.com>, 
	Li RongQing <lirongqing@baidu.com>, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="us-ascii"

Shortlog is a wee bit long, I went with:

  KVM: selftests: Check that PV_UNHALT is cleared when HLT exiting is disabled

On Wed, Feb 28, 2024, Vitaly Kuznetsov wrote:
> KVM_FEATURE_PV_UNHALT is expected to get cleared from KVM PV feature CPUID
> data when KVM_X86_DISABLE_EXITS_HLT is enabled. Add the corresponding test
> to kvm_pv_test.
> 
> Note, the newly added code doesn't actually test KVM_FEATURE_PV_UNHALT and
> KVM_X86_DISABLE_EXITS_HLT features.
> 
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> ---

..

> +	TEST_ASSERT(ent->eax & (1 << KVM_FEATURE_PV_UNHALT),
> +		    "Enabling X86_FEATURE_KVM_PV_UNHALT had no effect");
> +
> +	/* Make sure KVM clears vcpu->arch.kvm_cpuid */
> +	ent = vcpu_get_cpuid_entry(vcpu, KVM_CPUID_SIGNATURE);
> +	kvm_sig_old = ent->ebx;
> +	ent->ebx = 0xdeadbeef;
> +	vcpu_set_cpuid(vcpu);
> +
> +	vm_enable_cap(vm, KVM_CAP_X86_DISABLE_EXITS, KVM_X86_DISABLE_EXITS_HLT);
> +	ent = vcpu_get_cpuid_entry(vcpu, KVM_CPUID_SIGNATURE);
> +	ent->ebx = kvm_sig_old;
> +	vcpu_set_cpuid(vcpu);
> +	ent = vcpu_get_cpuid_entry(vcpu, KVM_CPUID_FEATURES);
> +
> +	TEST_ASSERT(!(ent->eax & (1 << KVM_FEATURE_PV_UNHALT)),

X86_FEATURE_KVM_PV_UNHALT already exists, all we're missing is a helper to get
a CPUID feature from host userspace given a vCPU.  I added this

	static inline bool vcpu_cpuid_has(struct kvm_vcpu *vcpu,
					  struct kvm_x86_cpu_feature feature)
	{
		struct kvm_cpuid_entry2 *entry;

		entry = __vcpu_get_cpuid_entry(vcpu, feature.function, feature.index);
		return *((&entry->eax) + feature.reg) & BIT(feature.bit);
	}

and used it in this test instead of open coding the reg+bit.