[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230526210340.2799158-3-seanjc@google.com>
Date: Fri, 26 May 2023 14:03:40 -0700
From: Sean Christopherson <seanjc@...gle.com>
To: Sean Christopherson <seanjc@...gle.com>,
Paolo Bonzini <pbonzini@...hat.com>
Cc: kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
Takahiro Itazuri <itazur@...zon.com>
Subject: [PATCH v3 2/2] KVM: selftests: Extend cpuid_test to verify
KVM_GET_CPUID2 "nent" updates
Verify that KVM reports the actual number of CPUID entries on success, but
doesn't touch the userspace struct on failure (which for better or worse,
is KVM's ABI).
Signed-off-by: Sean Christopherson <seanjc@...gle.com>
---
.../testing/selftests/kvm/x86_64/cpuid_test.c | 21 +++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/tools/testing/selftests/kvm/x86_64/cpuid_test.c b/tools/testing/selftests/kvm/x86_64/cpuid_test.c
index 2fc3ad9c887e..d3c3aa93f090 100644
--- a/tools/testing/selftests/kvm/x86_64/cpuid_test.c
+++ b/tools/testing/selftests/kvm/x86_64/cpuid_test.c
@@ -163,6 +163,25 @@ static void set_cpuid_after_run(struct kvm_vcpu *vcpu)
ent->eax = eax;
}
+static void test_get_cpuid2(struct kvm_vcpu *vcpu)
+{
+ struct kvm_cpuid2 *cpuid = allocate_kvm_cpuid2(vcpu->cpuid->nent + 1);
+ int i, r;
+
+ vcpu_ioctl(vcpu, KVM_GET_CPUID2, cpuid);
+ TEST_ASSERT(cpuid->nent == vcpu->cpuid->nent,
+ "KVM didn't update nent on success, wanted %u, got %u\n",
+ vcpu->cpuid->nent, cpuid->nent);
+
+ for (i = 0; i < vcpu->cpuid->nent; i++) {
+ cpuid->nent = i;
+ r = __vcpu_ioctl(vcpu, KVM_GET_CPUID2, cpuid);
+ TEST_ASSERT(r && errno == E2BIG, KVM_IOCTL_ERROR(KVM_GET_CPUID2, r));
+ TEST_ASSERT(cpuid->nent == i, "KVM modified nent on failure");
+ }
+ free(cpuid);
+}
+
int main(void)
{
struct kvm_vcpu *vcpu;
@@ -183,5 +202,7 @@ int main(void)
set_cpuid_after_run(vcpu);
+ test_get_cpuid2(vcpu);
+
kvm_vm_free(vm);
}
--
2.41.0.rc0.172.g3f132b7071-goog
Powered by blists - more mailing lists