allow FSGroupPolicy and PodInfoOnMount to be mutable in CSIDriver.Spec

pkg/apis/storage/types.go

@@ -284,7 +284,7 @@ type CSIDriverSpec struct {
 	// permission of the volume before being mounted.
 	// Refer to the specific FSGroupPolicy values for additional details.
 	//
-	// This field is immutable.
+	// This field was immutable in Kubernetes < 1.29 and now is mutable.
 	//
 	// Defaults to ReadWriteOnceWithFSType, which will examine each volume
 	// to determine if Kubernetes should modify ownership and permissions of the volume.
@@ -318,7 +318,7 @@ type CSIDriverSpec struct {
 	// deployed on such a cluster and the deployment determines which mode that is, for example
 	// via a command line parameter of the driver.
 	//
-	// This field is immutable.
+	// This field was immutable in Kubernetes < 1.29 and now is mutable.
 	//
 	// +optional
 	PodInfoOnMount *bool

pkg/apis/storage/validation/validation.go

@@ -414,8 +414,6 @@ func ValidateCSIDriverUpdate(new, old *storage.CSIDriver) field.ErrorList {
 
 	// immutable fields should not be mutated.
 	allErrs = append(allErrs, apimachineryvalidation.ValidateImmutableField(new.Spec.AttachRequired, old.Spec.AttachRequired, field.NewPath("spec", "attachedRequired"))...)
-	allErrs = append(allErrs, apimachineryvalidation.ValidateImmutableField(new.Spec.FSGroupPolicy, old.Spec.FSGroupPolicy, field.NewPath("spec", "fsGroupPolicy"))...)
-	allErrs = append(allErrs, apimachineryvalidation.ValidateImmutableField(new.Spec.PodInfoOnMount, old.Spec.PodInfoOnMount, field.NewPath("spec", "podInfoOnMount"))...)
 	allErrs = append(allErrs, apimachineryvalidation.ValidateImmutableField(new.Spec.VolumeLifecycleModes, old.Spec.VolumeLifecycleModes, field.NewPath("spec", "volumeLifecycleModes"))...)
 
 	allErrs = append(allErrs, validateTokenRequests(new.Spec.TokenRequests, field.NewPath("spec", "tokenRequests"))...)

pkg/apis/storage/validation/validation_test.go

@@ -1715,6 +1715,17 @@ func TestCSIDriverValidationUpdate(t *testing.T) {
 		modify: func(new *storage.CSIDriver) {
 			new.Spec.SELinuxMount = &notSELinuxMount
 		},
+	}, {
+		name: "change PodInfoOnMount",
+		modify: func(new *storage.CSIDriver) {
+			new.Spec.PodInfoOnMount = &podInfoOnMount
+		},
+	}, {
+		name: "change FSGroupPolicy",
+		modify: func(new *storage.CSIDriver) {
+			fileFSGroupPolicy := storage.FileFSGroupPolicy
+			new.Spec.FSGroupPolicy = &fileFSGroupPolicy
+		},
 	}}
 	for _, test := range successCases {
 		t.Run(test.name, func(t *testing.T) {
@@ -1755,11 +1766,6 @@ func TestCSIDriverValidationUpdate(t *testing.T) {
 		modify: func(new *storage.CSIDriver) {
 			new.Spec.PodInfoOnMount = nil
 		},
-	}, {
-		name: "PodInfoOnMount changed",
-		modify: func(new *storage.CSIDriver) {
-			new.Spec.PodInfoOnMount = &podInfoOnMount
-		},
 	}, {
 		name: "invalid volume lifecycle mode",
 		modify: func(new *storage.CSIDriver) {
@@ -1792,12 +1798,6 @@ func TestCSIDriverValidationUpdate(t *testing.T) {
 			invalidFSGroupPolicy := storage.FSGroupPolicy("invalid")
 			new.Spec.FSGroupPolicy = &invalidFSGroupPolicy
 		},
-	}, {
-		name: "FSGroupPolicy changed",
-		modify: func(new *storage.CSIDriver) {
-			fileFSGroupPolicy := storage.FileFSGroupPolicy
-			new.Spec.FSGroupPolicy = &fileFSGroupPolicy
-		},
 	}, {
 		name: "TokenRequests invalidated",
 		modify: func(new *storage.CSIDriver) {

pkg/registry/storage/csidriver/strategy.go

@@ -83,10 +83,8 @@ func (csiDriverStrategy) PrepareForUpdate(ctx context.Context, obj, old runtime.
 		newCSIDriver.Spec.SELinuxMount = nil
 	}
 
-	// Any changes to the mutable fields increment the generation number.
+	// Any changes to the spec increment the generation number.
-	if !apiequality.Semantic.DeepEqual(oldCSIDriver.Spec.TokenRequests, newCSIDriver.Spec.TokenRequests) ||
+	if !apiequality.Semantic.DeepEqual(oldCSIDriver.Spec, newCSIDriver.Spec) {
-		!apiequality.Semantic.DeepEqual(oldCSIDriver.Spec.RequiresRepublish, newCSIDriver.Spec.RequiresRepublish) ||
-		!apiequality.Semantic.DeepEqual(oldCSIDriver.Spec.SELinuxMount, newCSIDriver.Spec.SELinuxMount) {
 		newCSIDriver.Generation = oldCSIDriver.Generation + 1
 	}
 }

pkg/registry/storage/csidriver/strategy_test.go

@@ -88,7 +88,6 @@ func TestCSIDriverPrepareForUpdate(t *testing.T) {
 	})
 
 	attachRequired := true
-	podInfoOnMount := true
 	driverWithNothing := &storage.CSIDriver{
 		ObjectMeta: metav1.ObjectMeta{
 			Name: "foo",
@@ -100,7 +99,6 @@ func TestCSIDriverPrepareForUpdate(t *testing.T) {
 		},
 		Spec: storage.CSIDriverSpec{
 			AttachRequired: &attachRequired,
-			PodInfoOnMount: &podInfoOnMount,
 			VolumeLifecycleModes: []storage.VolumeLifecycleMode{
 				storage.VolumeLifecyclePersistent,
 			},
@@ -109,6 +107,49 @@ func TestCSIDriverPrepareForUpdate(t *testing.T) {
 	enabled := true
 	disabled := false
 	gcp := "gcp"
+	noneFsGroupPolicy := storage.NoneFSGroupPolicy
+	readWriteOnceWithFSTypeFSGroupPolicy := storage.ReadWriteOnceWithFSTypeFSGroupPolicy
+	fileFSGroupPolicy := storage.FileFSGroupPolicy
+	driverWithPodInfoOnMountEnabled := &storage.CSIDriver{
+		ObjectMeta: metav1.ObjectMeta{
+			Name: "foo",
+		},
+		Spec: storage.CSIDriverSpec{
+			PodInfoOnMount: &enabled,
+		},
+	}
+	driverWithPodInfoOnMountDisabled := &storage.CSIDriver{
+		ObjectMeta: metav1.ObjectMeta{
+			Name: "foo",
+		},
+		Spec: storage.CSIDriverSpec{
+			PodInfoOnMount: &disabled,
+		},
+	}
+	driverWithNoneFSGroupPolicy := &storage.CSIDriver{
+		ObjectMeta: metav1.ObjectMeta{
+			Name: "foo",
+		},
+		Spec: storage.CSIDriverSpec{
+			FSGroupPolicy: &noneFsGroupPolicy,
+		},
+	}
+	driverWithReadWriteOnceWithFSTypeFSGroupPolicy := &storage.CSIDriver{
+		ObjectMeta: metav1.ObjectMeta{
+			Name: "foo",
+		},
+		Spec: storage.CSIDriverSpec{
+			FSGroupPolicy: &readWriteOnceWithFSTypeFSGroupPolicy,
+		},
+	}
+	driverWithFileFSGroupPolicy := &storage.CSIDriver{
+		ObjectMeta: metav1.ObjectMeta{
+			Name: "foo",
+		},
+		Spec: storage.CSIDriverSpec{
+			FSGroupPolicy: &fileFSGroupPolicy,
+		},
+	}
 	driverWithCapacityEnabled := &storage.CSIDriver{
 		ObjectMeta: metav1.ObjectMeta{
 			Name: "foo",
@@ -165,22 +206,91 @@ func TestCSIDriverPrepareForUpdate(t *testing.T) {
 		wantSELinuxMount                    *bool
 	}{
 		{
-			name:         "capacity feature enabled, before: none, update: enabled",
+			name:           "podInfoOnMount feature enabled, before: none, update: enabled",
-			old:          driverWithNothing,
+			old:            driverWithNothing,
-			update:       driverWithCapacityEnabled,
+			update:         driverWithPodInfoOnMountEnabled,
-			wantCapacity: &enabled,
+			wantGeneration: 1,
 		},
 		{
-			name:         "capacity feature enabled, before: enabled, update: disabled",
+			name:           "podInfoOnMount feature enabled, before: enabled, update: disabled",
-			old:          driverWithCapacityEnabled,
+			old:            driverWithPodInfoOnMountEnabled,
-			update:       driverWithCapacityDisabled,
+			update:         driverWithPodInfoOnMountDisabled,
-			wantCapacity: &disabled,
+			wantGeneration: 1,
 		},
 		{
-			name:      "inline feature enabled, before: none, update: persistent",
+			name:           "fSGroupPolicy feature enabled, before: nil, update: none",
-			old:       driverWithNothing,
+			old:            driverWithNothing,
-			update:    driverWithPersistent,
+			update:         driverWithNoneFSGroupPolicy,
-			wantModes: resultPersistent,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: nil, update: readWriteOnceWithFSType",
+			old:            driverWithNothing,
+			update:         driverWithReadWriteOnceWithFSTypeFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: nil, update: file",
+			old:            driverWithNothing,
+			update:         driverWithFileFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: none, update: readWriteOnceWithFSType",
+			old:            driverWithNoneFSGroupPolicy,
+			update:         driverWithReadWriteOnceWithFSTypeFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: none, update: file",
+			old:            driverWithNoneFSGroupPolicy,
+			update:         driverWithFileFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: readWriteOnceWithFSType, update: none",
+			old:            driverWithReadWriteOnceWithFSTypeFSGroupPolicy,
+			update:         driverWithNoneFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: readWriteOnceWithFSType, update: file",
+			old:            driverWithReadWriteOnceWithFSTypeFSGroupPolicy,
+			update:         driverWithFileFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: file, update: none",
+			old:            driverWithFileFSGroupPolicy,
+			update:         driverWithNoneFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "fSGroupPolicy feature enabled, before: file, update: readWriteOnceWithFSType",
+			old:            driverWithFileFSGroupPolicy,
+			update:         driverWithReadWriteOnceWithFSTypeFSGroupPolicy,
+			wantGeneration: 1,
+		},
+		{
+			name:           "capacity feature enabled, before: none, update: enabled",
+			old:            driverWithNothing,
+			update:         driverWithCapacityEnabled,
+			wantCapacity:   &enabled,
+			wantGeneration: 1,
+		},
+		{
+			name:           "capacity feature enabled, before: enabled, update: disabled",
+			old:            driverWithCapacityEnabled,
+			update:         driverWithCapacityDisabled,
+			wantCapacity:   &disabled,
+			wantGeneration: 1,
+		},
+		{
+			name:           "inline feature enabled, before: none, update: persistent",
+			old:            driverWithNothing,
+			update:         driverWithPersistent,
+			wantModes:      resultPersistent,
+			wantGeneration: 1,
 		},
 		{
 			name:                  "service account token feature enabled, before: none, update: audience=gcp",

staging/src/k8s.io/api/storage/v1/types.go

@@ -306,7 +306,7 @@ type CSIDriverSpec struct {
 	// deployed on such a cluster and the deployment determines which mode that is, for example
 	// via a command line parameter of the driver.
 	//
-	// This field is immutable.
+	// This field was immutable in Kubernetes < 1.29 and now is mutable.
 	//
 	// +optional
 	PodInfoOnMount *bool `json:"podInfoOnMount,omitempty" protobuf:"bytes,2,opt,name=podInfoOnMount"`
@@ -353,7 +353,7 @@ type CSIDriverSpec struct {
 	// permission of the volume before being mounted.
 	// Refer to the specific FSGroupPolicy values for additional details.
 	//
-	// This field is immutable.
+	// This field was immutable in Kubernetes < 1.29 and now is mutable.
 	//
 	// Defaults to ReadWriteOnceWithFSType, which will examine each volume
 	// to determine if Kubernetes should modify ownership and permissions of the volume.

test/e2e/storage/csi_mock/base.go

@@ -248,6 +248,17 @@ func (m *mockDriverSetup) cleanup(ctx context.Context) {
 	framework.ExpectNoError(err, "while cleaning up after test")
 }
 
+func (m *mockDriverSetup) update(o utils.PatchCSIOptions) {
+	item, err := m.cs.StorageV1().CSIDrivers().Get(context.TODO(), m.config.GetUniqueDriverName(), metav1.GetOptions{})
+	framework.ExpectNoError(err, "Failed to get CSIDriver %v", m.config.GetUniqueDriverName())
+
+	err = utils.PatchCSIDeployment(nil, o, item)
+	framework.ExpectNoError(err, "Failed to apply %v to CSIDriver object %v", o, m.config.GetUniqueDriverName())
+
+	_, err = m.cs.StorageV1().CSIDrivers().Update(context.TODO(), item, metav1.UpdateOptions{})
+	framework.ExpectNoError(err, "Failed to update CSIDriver %v", m.config.GetUniqueDriverName())
+}
+
 func (m *mockDriverSetup) createPod(ctx context.Context, withVolume volumeType) (class *storagev1.StorageClass, claim *v1.PersistentVolumeClaim, pod *v1.Pod) {
 	ginkgo.By("Creating pod")
 	f := m.f
@@ -722,8 +733,8 @@ func checkPodLogs(ctx context.Context, getCalls func(ctx context.Context) ([]dri
 		switch call.Method {
 		case "NodePublishVolume":
 			numNodePublishVolume++
-			if numNodePublishVolume == 1 {
+			if numNodePublishVolume == expectedNumNodePublish {
-				// Check that NodePublish had expected attributes for first volume
+				// Check that NodePublish had expected attributes for last of expected volume
 				for k, v := range expectedAttributes {
 					vv, found := call.Request.VolumeContext[k]
 					if found && (v == vv || (v == "<nonempty>" && len(vv) != 0)) {

test/e2e/storage/csi_mock/csi_fsgroup_policy.go

@@ -21,9 +21,9 @@ import (
 	"fmt"
 	"math/rand"
 	"strconv"
-	"time"
 
 	"github.com/onsi/ginkgo/v2"
+	"github.com/onsi/gomega"
 	storagev1 "k8s.io/api/storage/v1"
 	"k8s.io/kubernetes/test/e2e/framework"
 	e2epod "k8s.io/kubernetes/test/e2e/framework/pod"
@@ -74,47 +74,67 @@ var _ = utils.SIGDescribe("CSI Mock volume fsgroup policies", func() {
 				})
 				ginkgo.DeferCleanup(m.cleanup)
 
-				// kube-scheduler may need some time before it gets the CSIDriver object.
+				waitUtilFSGroupInPod(ctx, m, test.modified)
-				// Without them, scheduling doesn't run as expected by the test.
-				syncDelay := 5 * time.Second
-				time.Sleep(syncDelay)
 
-				fsGroupVal := int64(rand.Int63n(20000) + 1024)
+				// The created resources will be removed by the cleanup() function,
-				fsGroup := &fsGroupVal
+				// so need to delete anything here.
+			})
+		}
+	})
 
-				_, _, pod := m.createPodWithFSGroup(ctx, fsGroup) /* persistent volume */
+	ginkgo.Context("CSI FSGroupPolicy Update [LinuxOnly]", func() {
-
+		tests := []struct {
-				mountPath := pod.Spec.Containers[0].VolumeMounts[0].MountPath
+			name             string
-				dirName := mountPath + "/" + f.UniqueName
+			oldFSGroupPolicy storagev1.FSGroupPolicy
-				fileName := dirName + "/" + f.UniqueName
+			newFSGroupPolicy storagev1.FSGroupPolicy
-
+		}{
-				err := e2epod.WaitForPodNameRunningInNamespace(ctx, m.cs, pod.Name, pod.Namespace)
+			{
-				framework.ExpectNoError(err, "failed to start pod")
+				name:             "should update fsGroup if update from None to File",
-
+				oldFSGroupPolicy: storagev1.NoneFSGroupPolicy,
-				// Create the subdirectory to ensure that fsGroup propagates
+				newFSGroupPolicy: storagev1.FileFSGroupPolicy,
-				createDirectory := fmt.Sprintf("mkdir %s", dirName)
+			},
-				_, _, err = e2evolume.PodExec(f, pod, createDirectory)
+			{
-				framework.ExpectNoError(err, "failed: creating the directory: %s", err)
+				name:             "should update fsGroup if update from None to default",
-
+				oldFSGroupPolicy: storagev1.NoneFSGroupPolicy,
-				// Inject the contents onto the mount
+				newFSGroupPolicy: storagev1.ReadWriteOnceWithFSTypeFSGroupPolicy,
-				createFile := fmt.Sprintf("echo '%s' > '%s'; sync", "filecontents", fileName)
+			},
-				_, _, err = e2evolume.PodExec(f, pod, createFile)
+			{
-				framework.ExpectNoError(err, "failed: writing the contents: %s", err)
+				name:             "should not update fsGroup if update from File to None",
-
+				oldFSGroupPolicy: storagev1.FileFSGroupPolicy,
-				// Delete the created file. This step is mandatory, as the mock driver
+				newFSGroupPolicy: storagev1.NoneFSGroupPolicy,
-				// won't clean up the contents automatically.
+			},
-				defer func() {
+			{
-					deleteDir := fmt.Sprintf("rm -fr %s", dirName)
+				name:             "should update fsGroup if update from File to default",
-					_, _, err = e2evolume.PodExec(f, pod, deleteDir)
+				oldFSGroupPolicy: storagev1.FileFSGroupPolicy,
-					framework.ExpectNoError(err, "failed: deleting the directory: %s", err)
+				newFSGroupPolicy: storagev1.ReadWriteOnceWithFSTypeFSGroupPolicy,
-				}()
+			},
-
+			{
-				// Ensure that the fsGroup matches what we expect
+				name:             "should not update fsGroup if update from detault to None",
-				if test.modified {
+				oldFSGroupPolicy: storagev1.ReadWriteOnceWithFSTypeFSGroupPolicy,
-					utils.VerifyFSGroupInPod(f, fileName, strconv.FormatInt(*fsGroup, 10), pod)
+				newFSGroupPolicy: storagev1.NoneFSGroupPolicy,
-				} else {
+			},
-					utils.VerifyFSGroupInPod(f, fileName, "root", pod)
+			{
+				name:             "should update fsGroup if update from detault to File",
+				oldFSGroupPolicy: storagev1.ReadWriteOnceWithFSTypeFSGroupPolicy,
+				newFSGroupPolicy: storagev1.FileFSGroupPolicy,
+			},
+		}
+		for _, t := range tests {
+			test := t
+			ginkgo.It(test.name, func(ctx context.Context) {
+				if framework.NodeOSDistroIs("windows") {
+					e2eskipper.Skipf("FSGroupPolicy is only applied on linux nodes -- skipping")
 				}
+				m.init(ctx, testParameters{
+					disableAttach:  true,
+					registerDriver: true,
+					fsGroupPolicy:  &test.oldFSGroupPolicy,
+				})
+				ginkgo.DeferCleanup(m.cleanup)
+
+				waitUtilFSGroupInPod(ctx, m, test.oldFSGroupPolicy != storagev1.NoneFSGroupPolicy)
+				m.update(utils.PatchCSIOptions{FSGroupPolicy: &test.newFSGroupPolicy})
+				waitUtilFSGroupInPod(ctx, m, test.newFSGroupPolicy != storagev1.NoneFSGroupPolicy)
 
 				// The created resources will be removed by the cleanup() function,
 				// so need to delete anything here.
@@ -122,3 +142,56 @@ var _ = utils.SIGDescribe("CSI Mock volume fsgroup policies", func() {
 		}
 	})
 })
+
+func waitUtilFSGroupInPod(ctx context.Context, m *mockDriverSetup, modified bool) {
+	var err error
+
+	utils.WaitUntil(framework.Poll, framework.PodStartTimeout, func() bool {
+		err = gomega.InterceptGomegaFailure(func() {
+			fsGroupVal := int64(rand.Int63n(20000) + 1024)
+			fsGroup := &fsGroupVal
+
+			_, _, pod := m.createPodWithFSGroup(ctx, fsGroup) /* persistent volume */
+			defer func() {
+				err = e2epod.DeletePodWithWait(context.TODO(), m.f.ClientSet, pod)
+				framework.ExpectNoError(err, "failed: deleting the pod: %s", err)
+			}()
+
+			mountPath := pod.Spec.Containers[0].VolumeMounts[0].MountPath
+			dirName := mountPath + "/" + m.f.UniqueName
+			fileName := dirName + "/" + m.f.UniqueName
+
+			err = e2epod.WaitForPodNameRunningInNamespace(ctx, m.cs, pod.Name, pod.Namespace)
+			framework.ExpectNoError(err, "failed to start pod")
+
+			// Create the subdirectory to ensure that fsGroup propagates
+			createDirectory := fmt.Sprintf("mkdir %s", dirName)
+			_, _, err = e2evolume.PodExec(m.f, pod, createDirectory)
+			framework.ExpectNoError(err, "failed: creating the directory: %s", err)
+
+			// Inject the contents onto the mount
+			createFile := fmt.Sprintf("echo '%s' > '%s'; sync", "filecontents", fileName)
+			_, _, err = e2evolume.PodExec(m.f, pod, createFile)
+			framework.ExpectNoError(err, "failed: writing the contents: %s", err)
+
+			// Delete the created file. This step is mandatory, as the mock driver
+			// won't clean up the contents automatically.
+			defer func() {
+				deleteDir := fmt.Sprintf("rm -fr %s", dirName)
+				_, _, err = e2evolume.PodExec(m.f, pod, deleteDir)
+				framework.ExpectNoError(err, "failed: deleting the directory: %s", err)
+			}()
+
+			// Ensure that the fsGroup matches what we expect
+			if modified {
+				utils.VerifyFSGroupInPod(m.f, fileName, strconv.FormatInt(*fsGroup, 10), pod)
+			} else {
+				utils.VerifyFSGroupInPod(m.f, fileName, "root", pod)
+			}
+		})
+
+		return err == nil
+	})
+
+	framework.ExpectNoError(err, "failed: verifing fsgroup in pod: %s", err)
+}

test/e2e/storage/csi_mock/csi_workload.go

@@ -20,6 +20,7 @@ import (
 	"context"
 
 	"github.com/onsi/ginkgo/v2"
+	"github.com/onsi/gomega"
 	"k8s.io/kubernetes/test/e2e/framework"
 	e2epod "k8s.io/kubernetes/test/e2e/framework/pod"
 	"k8s.io/kubernetes/test/e2e/storage/testsuites"
@@ -34,7 +35,6 @@ var _ = utils.SIGDescribe("CSI Mock workload info", func() {
 	m := newMockDriverSetup(f)
 	ginkgo.Context("CSI workload information using mock driver", func() {
 		var (
-			err          error
 			podInfoTrue  = true
 			podInfoFalse = false
 		)
@@ -89,36 +89,99 @@ var _ = utils.SIGDescribe("CSI Mock workload info", func() {
 
 				ginkgo.DeferCleanup(m.cleanup)
 
-				withVolume := pvcReference
+				waitUntilPodInfoInLog(ctx, m, test.expectPodInfo, test.expectEphemeral, 1)
-				if test.expectEphemeral {
-					withVolume = csiEphemeral
-				}
-				_, _, pod := m.createPod(ctx, withVolume)
-				if pod == nil {
-					return
-				}
-				err = e2epod.WaitForPodNameRunningInNamespace(ctx, m.cs, pod.Name, pod.Namespace)
-				framework.ExpectNoError(err, "Failed to start pod: %v", err)
 
-				// If we expect an ephemeral volume, the feature has to be enabled.
+			})
-				// Otherwise need to check if we expect pod info, because the content
+		}
-				// of that depends on whether the feature is enabled or not.
+	})
-				csiInlineVolumesEnabled := test.expectEphemeral
-				if test.expectPodInfo {
-					ginkgo.By("checking for CSIInlineVolumes feature")
-					csiInlineVolumesEnabled, err = testsuites.CSIInlineVolumesEnabled(ctx, m.cs, f.Timeouts, f.Namespace.Name)
-					framework.ExpectNoError(err, "failed to test for CSIInlineVolumes")
-				}
 
-				ginkgo.By("Deleting the previously created pod")
+	ginkgo.Context("CSI PodInfoOnMount Update", func() {
-				err = e2epod.DeletePodWithWait(ctx, m.cs, pod)
+		var (
-				framework.ExpectNoError(err, "while deleting")
+			podInfoTrue  = true
+			podInfoFalse = false
+		)
+		tests := []struct {
+			name              string
+			oldPodInfoOnMount *bool
+			newPodInfoOnMount *bool
+		}{
+			{
+				name:              "should not be passed when update from true to false",
+				oldPodInfoOnMount: &podInfoTrue,
+				newPodInfoOnMount: &podInfoFalse,
+			},
+			{
+				name:              "should not be passed when update from true to nil",
+				oldPodInfoOnMount: &podInfoTrue,
+				newPodInfoOnMount: nil,
+			},
+			{
+				name:              "should be passed when update from false to true",
+				oldPodInfoOnMount: &podInfoFalse,
+				newPodInfoOnMount: &podInfoTrue,
+			},
+			{
+				name:              "should be passed when update from nil to true",
+				oldPodInfoOnMount: nil,
+				newPodInfoOnMount: &podInfoTrue,
+			},
+		}
+		for _, t := range tests {
+			test := t
+			ginkgo.It(t.name, func(ctx context.Context) {
+				m.init(ctx, testParameters{
+					registerDriver: true,
+					podInfo:        test.oldPodInfoOnMount})
+
+				ginkgo.DeferCleanup(m.cleanup)
+
+				waitUntilPodInfoInLog(ctx, m, test.oldPodInfoOnMount != nil && *test.oldPodInfoOnMount, false, 1)
+				m.update(utils.PatchCSIOptions{PodInfo: test.newPodInfoOnMount})
+				waitUntilPodInfoInLog(ctx, m, test.newPodInfoOnMount != nil && *test.newPodInfoOnMount, false, 2)
 
-				ginkgo.By("Checking CSI driver logs")
-				err = checkPodLogs(ctx, m.driver.GetCalls, pod, test.expectPodInfo, test.expectEphemeral, csiInlineVolumesEnabled, false, 1)
-				framework.ExpectNoError(err)
 			})
 		}
 	})
 
 })
+
+func waitUntilPodInfoInLog(ctx context.Context, m *mockDriverSetup, expectPodInfo, expectEphemeral bool, expectedNumNodePublish int) {
+	var err error
+
+	utils.WaitUntil(framework.Poll, framework.PodStartTimeout, func() bool {
+		err = gomega.InterceptGomegaFailure(func() {
+			withVolume := pvcReference
+			if expectEphemeral {
+				withVolume = csiEphemeral
+			}
+			_, _, pod := m.createPod(ctx, withVolume)
+			if pod == nil {
+				return
+			}
+			err = e2epod.WaitForPodNameRunningInNamespace(ctx, m.cs, pod.Name, pod.Namespace)
+			framework.ExpectNoError(err, "Failed to start pod: %v", err)
+
+			// If we expect an ephemeral volume, the feature has to be enabled.
+			// Otherwise need to check if we expect pod info, because the content
+			// of that depends on whether the feature is enabled or not.
+			csiInlineVolumesEnabled := expectEphemeral
+			if expectPodInfo {
+				ginkgo.By("checking for CSIInlineVolumes feature")
+				csiInlineVolumesEnabled, err = testsuites.CSIInlineVolumesEnabled(ctx, m.cs, m.f.Timeouts, m.f.Namespace.Name)
+				framework.ExpectNoError(err, "failed to test for CSIInlineVolumes")
+			}
+
+			ginkgo.By("Deleting the previously created pod")
+			err = e2epod.DeletePodWithWait(ctx, m.cs, pod)
+			framework.ExpectNoError(err, "while deleting")
+
+			ginkgo.By("Checking CSI driver logs")
+			err = checkPodLogs(ctx, m.driver.GetCalls, pod, expectPodInfo, expectEphemeral, csiInlineVolumesEnabled, false, expectedNumNodePublish)
+			framework.ExpectNoError(err)
+		})
+
+		return err == nil
+	})
+
+	framework.ExpectNoError(err, "failed: verifing PodInfo: %s", err)
+}