Specify all capabilities in TestGPUCheckpointRestore.

This test was broken by 5e6589e0b7 ("Update CUDA test compatibility to keep
up with added gVisor support.") which requires all
images/gpu/cuda-tests/run_sample.go users to specify "all" driver capabilities.

PiperOrigin-RevId: 713170590
This commit is contained in:
Ayush Ranjan
2025-01-07 23:28:03 -08:00
committed by gVisor bot
parent 8b7b69c978
commit 2edcee37d9
5 changed files with 13 additions and 12 deletions
+5 -6
View File
@@ -41,13 +41,12 @@ const (
)
const (
// AllGPUCapabilities is the environment variable that enables all NVIDIA
// AllGPUCapabilitiesEnv is the environment variable that enables all NVIDIA
// GPU capabilities within a container.
AllGPUCapabilities = "NVIDIA_DRIVER_CAPABILITIES=all"
AllGPUCapabilitiesEnv = "NVIDIA_DRIVER_CAPABILITIES=all"
// DefaultGPUCapabilities is the environment variable that enables default
// NVIDIA GPU capabilities within a container.
DefaultGPUCapabilities = "NVIDIA_DRIVER_CAPABILITIES=compute,utility"
// DefaultGPUCapabilities are the driver capabilities enabled by default.
DefaultGPUCapabilities = "compute,utility"
)
// GPURunOpts returns Docker run options with GPU support enabled.
@@ -79,7 +78,7 @@ func GPURunOpts(sniffGPUOpts SniffGPUOpts) (RunOpts, error) {
ReadOnly: true,
})
}
gpuEnv := []string{sniffGPUOpts.GPUCapabilities()}
gpuEnv := []string{"NVIDIA_DRIVER_CAPABILITIES=" + sniffGPUOpts.GPUCapabilities()}
if !*setCOSGPU {
return RunOpts{
+2 -2
View File
@@ -433,7 +433,7 @@ func (*FullyCompatible) IsExpectedFailure(ctx context.Context, env *TestEnvironm
// getContainerOpts returns the container run options to run CUDA tests.
func getContainerOpts() (dockerutil.RunOpts, error) {
opts, err := dockerutil.GPURunOpts(dockerutil.SniffGPUOpts{
Capabilities: dockerutil.AllGPUCapabilities,
Capabilities: "all",
})
if err != nil {
return dockerutil.RunOpts{}, fmt.Errorf("failed to get GPU run options: %w", err)
@@ -885,7 +885,7 @@ func TestCUDA(t *testing.T) {
t.Errorf(
" $ docker run --runtime=%s --gpus=all -e %s --rm %s /run_sample %s",
dockerutil.Runtime(),
dockerutil.AllGPUCapabilities,
dockerutil.AllGPUCapabilitiesEnv,
runOpts.Image,
failedTests[0],
)
+2 -2
View File
@@ -34,7 +34,7 @@ func TestFffmpegEncodeGPU(t *testing.T) {
container := dockerutil.MakeContainer(ctx, t)
defer container.CleanUp(ctx)
opts, err := dockerutil.GPURunOpts(dockerutil.SniffGPUOpts{
Capabilities: "NVIDIA_DRIVER_CAPABILITIES=video",
Capabilities: "video",
AllowIncompatibleIoctl: true, // TODO(gvisor.dev/issue/9452): Remove once supported in gVisor.
})
if err != nil {
@@ -60,7 +60,7 @@ func TestFffmpegDecodeGPU(t *testing.T) {
container := dockerutil.MakeContainer(ctx, t)
defer container.CleanUp(ctx)
opts, err := dockerutil.GPURunOpts(dockerutil.SniffGPUOpts{
Capabilities: "NVIDIA_DRIVER_CAPABILITIES=video",
Capabilities: "video",
AllowIncompatibleIoctl: true, // TODO(gvisor.dev/issue/9452): Remove once supported in gVisor.
})
if err != nil {
+1 -1
View File
@@ -28,7 +28,7 @@ func TestGPUHello(t *testing.T) {
defer c.CleanUp(ctx)
opts, err := dockerutil.GPURunOpts(dockerutil.SniffGPUOpts{
Capabilities: dockerutil.AllGPUCapabilities,
Capabilities: "all",
})
if err != nil {
t.Fatalf("failed to get GPU run options: %v", err)
+3 -1
View File
@@ -35,7 +35,9 @@ func TestGPUCheckpointRestore(t *testing.T) {
c := dockerutil.MakeContainer(ctx, t)
defer c.CleanUp(ctx)
opts, err := dockerutil.GPURunOpts(dockerutil.SniffGPUOpts{})
opts, err := dockerutil.GPURunOpts(dockerutil.SniffGPUOpts{
Capabilities: "all",
})
if err != nil {
t.Fatalf("failed to get GPU run options: %v", err)
}