Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 10 Jun 2022 12:48:51 +0200
From:   Andrew Jones <drjones@redhat.com>
To:     Sean Christopherson <seanjc@google.com>
Cc:     Paolo Bonzini <pbonzini@redhat.com>, kvm@vger.kernel.org,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        David Matlack <dmatlack@google.com>,
        Ben Gardon <bgardon@google.com>,
        Oliver Upton <oupton@google.com>, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 126/144] KVM: selftests: Convert kvm_binary_stats_test
 away from vCPU IDs
Message-ID: <20220610104851.g2r6yzd6j22xod6m@gator>
References: <20220603004331.1523888-1-seanjc@google.com>
 <20220603004331.1523888-127-seanjc@google.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220603004331.1523888-127-seanjc@google.com>
Precedence: bulk

On Fri, Jun 03, 2022 at 12:43:13AM +0000, Sean Christopherson wrote:
> Track vCPUs by their 'struct kvm_vcpu' object in kvm_binary_stats_test,
> not by their ID.  The per-vCPU helpers will soon take a vCPU instead of a
> VM+vcpu_id pair.
> 
> Signed-off-by: Sean Christopherson <seanjc@google.com>
> ---
>  tools/testing/selftests/kvm/kvm_binary_stats_test.c | 13 +++++++++----
>  1 file changed, 9 insertions(+), 4 deletions(-)
> 
> diff --git a/tools/testing/selftests/kvm/kvm_binary_stats_test.c b/tools/testing/selftests/kvm/kvm_binary_stats_test.c
> index 407e9ea8e6f3..dfc3cf531ced 100644
> --- a/tools/testing/selftests/kvm/kvm_binary_stats_test.c
> +++ b/tools/testing/selftests/kvm/kvm_binary_stats_test.c
> @@ -172,9 +172,9 @@ static void vm_stats_test(struct kvm_vm *vm)
>  	TEST_ASSERT(fcntl(stats_fd, F_GETFD) == -1, "Stats fd not freed");
>  }
>  
> -static void vcpu_stats_test(struct kvm_vm *vm, int vcpu_id)
> +static void vcpu_stats_test(struct kvm_vcpu *vcpu)
>  {
> -	int stats_fd = vcpu_get_stats_fd(vm, vcpu_id);
> +	int stats_fd = vcpu_get_stats_fd(vcpu->vm, vcpu->id);
>  
>  	stats_test(stats_fd);
>  	close(stats_fd);
> @@ -195,6 +195,7 @@ static void vcpu_stats_test(struct kvm_vm *vm, int vcpu_id)
>  int main(int argc, char *argv[])
>  {
>  	int i, j;
> +	struct kvm_vcpu **vcpus;
>  	struct kvm_vm **vms;
>  	int max_vm = DEFAULT_NUM_VM;
>  	int max_vcpu = DEFAULT_NUM_VCPU;
> @@ -220,17 +221,21 @@ int main(int argc, char *argv[])
>  	/* Create VMs and VCPUs */
>  	vms = malloc(sizeof(vms[0]) * max_vm);
>  	TEST_ASSERT(vms, "Allocate memory for storing VM pointers");
> +
> +	vcpus = malloc(sizeof(struct kvm_vcpu *) * max_vm * max_vcpu);
> +	TEST_ASSERT(vcpus, "Allocate memory for storing vCPU pointers");
> +
>  	for (i = 0; i < max_vm; ++i) {
>  		vms[i] = vm_create_barebones();
>  		for (j = 0; j < max_vcpu; ++j)
> -			__vm_vcpu_add(vms[i], j);
> +			vcpus[j * max_vcpu + i] = __vm_vcpu_add(vms[i], j);

The expression for the index should be 'i * max_vcpu + j'. The swapped
i,j usage isn't causing problems now because
DEFAULT_NUM_VM == DEFAULT_NUM_VCPU, but that could change.

>  	}
>  
>  	/* Check stats read for every VM and VCPU */
>  	for (i = 0; i < max_vm; ++i) {
>  		vm_stats_test(vms[i]);
>  		for (j = 0; j < max_vcpu; ++j)
> -			vcpu_stats_test(vms[i], j);
> +			vcpu_stats_test(vcpus[j * max_vcpu + i]);

Same comment as above.

Thanks,
drew

>  	}
>  
>  	for (i = 0; i < max_vm; ++i)
> -- 
> 2.36.1.255.ge46751e96f-goog
>