feat(mount): support --mount type=image

cmd/nerdctl/container/container_run_mount_image_linux_test.go

@@ -0,0 +1,137 @@
+/*
+   Copyright The containerd Authors.
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.
+*/
+
+package container
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/containerd/nerdctl/mod/tigron/expect"
+	"github.com/containerd/nerdctl/mod/tigron/require"
+	"github.com/containerd/nerdctl/mod/tigron/test"
+
+	"github.com/containerd/nerdctl/v2/pkg/testutil"
+	"github.com/containerd/nerdctl/v2/pkg/testutil/nerdtest"
+)
+
+// TestRunMountTypeImage verifies that `--mount type=image` mounts the source
+// image's filesystem into the container so its files are readable at the target.
+func TestRunMountTypeImage(t *testing.T) {
+	testCase := nerdtest.Setup()
+
+	testCase.Command = func(data test.Data, helpers test.Helpers) test.TestableCommand {
+		return helpers.Command("run", "--rm",
+			"--mount", fmt.Sprintf("type=image,source=%s,destination=/mnt/img", testutil.CommonImage),
+			testutil.CommonImage, "cat", "/mnt/img/etc/os-release")
+	}
+
+	testCase.Expected = func(data test.Data, helpers test.Helpers) *test.Expected {
+		return &test.Expected{
+			ExitCode: expect.ExitCodeSuccess,
+			Output:   expect.Contains("Alpine"),
+		}
+	}
+
+	testCase.Run(t)
+}
+
+// TestRunMountTypeImageMultipleDestinations verifies the same image can be
+// mounted at two destinations in one container.
+func TestRunMountTypeImageMultipleDestinations(t *testing.T) {
+	testCase := nerdtest.Setup()
+	// nerdctl-only: Docker keys an image mount by its source image and rejects
+	// mounting the same image twice ("mount already exists with name").
+	testCase.Require = require.Not(nerdtest.Docker)
+
+	testCase.Command = func(data test.Data, helpers test.Helpers) test.TestableCommand {
+		return helpers.Command("run", "--rm",
+			"--mount", fmt.Sprintf("type=image,source=%s,destination=/mnt/a", testutil.CommonImage),
+			"--mount", fmt.Sprintf("type=image,source=%s,destination=/mnt/b", testutil.CommonImage),
+			testutil.CommonImage, "cat", "/mnt/a/etc/os-release", "/mnt/b/etc/os-release")
+	}
+
+	testCase.Expected = func(data test.Data, helpers test.Helpers) *test.Expected {
+		return &test.Expected{
+			ExitCode: expect.ExitCodeSuccess,
+			Output:   expect.Contains("Alpine"),
+		}
+	}
+
+	testCase.Run(t)
+}
+
+// TestRunMountTypeImageReadOnly verifies an image mount is read-only (writing
+// fails). nerdctl-only: Docker mounts images read-write by default.
+func TestRunMountTypeImageReadOnly(t *testing.T) {
+	testCase := nerdtest.Setup()
+	testCase.Require = require.Not(nerdtest.Docker)
+
+	testCase.Command = func(data test.Data, helpers test.Helpers) test.TestableCommand {
+		return helpers.Command("run", "--rm",
+			"--mount", fmt.Sprintf("type=image,source=%s,destination=/mnt/img", testutil.CommonImage),
+			testutil.CommonImage, "touch", "/mnt/img/should-fail")
+	}
+
+	testCase.Expected = func(data test.Data, helpers test.Helpers) *test.Expected {
+		return &test.Expected{
+			ExitCode: expect.ExitCodeGenericFail,
+			Errors:   []error{fmt.Errorf("Read-only file system")},
+		}
+	}
+
+	testCase.Run(t)
+}
+
+// TestRunMountTypeImageErrors verifies that an image mount missing its source,
+// or using the not-yet-supported subpath option, is rejected. subpath is
+// nerdctl-specific behaviour here, so the test is not run against Docker.
+func TestRunMountTypeImageErrors(t *testing.T) {
+	testCase := nerdtest.Setup()
+	testCase.Require = require.Not(nerdtest.Docker)
+
+	testCase.SubTests = []*test.Case{
+		{
+			Description: "missing source",
+			Command: func(data test.Data, helpers test.Helpers) test.TestableCommand {
+				return helpers.Command("run", "--rm", "--mount", "type=image,destination=/mnt/img",
+					testutil.CommonImage, "true")
+			},
+			Expected: func(data test.Data, helpers test.Helpers) *test.Expected {
+				return &test.Expected{
+					ExitCode: expect.ExitCodeGenericFail,
+					Errors:   []error{fmt.Errorf("source")},
+				}
+			},
+		},
+		{
+			Description: "subpath not supported",
+			Command: func(data test.Data, helpers test.Helpers) test.TestableCommand {
+				return helpers.Command("run", "--rm",
+					"--mount", fmt.Sprintf("type=image,source=%s,destination=/mnt/img,subpath=etc", testutil.CommonImage),
+					testutil.CommonImage, "true")
+			},
+			Expected: func(data test.Data, helpers test.Helpers) *test.Expected {
+				return &test.Expected{
+					ExitCode: expect.ExitCodeGenericFail,
+					Errors:   []error{fmt.Errorf("subpath")},
+				}
+			},
+		},
+	}
+
+	testCase.Run(t)
+}

docs/command-reference.md

@@ -295,10 +295,9 @@ Volume flags:
   Consists of multiple key-value pairs, separated by commas and each
   consisting of a `<key>=<value>` tuple.
   e.g., `-- mount type=bind,source=/src,target=/app,bind-propagation=shared`.
-  - :whale: `type`: Current supported mount types are `bind`, `volume`, `tmpfs`.
+  - :whale: `type`: Current supported mount types are `bind`, `volume`, `tmpfs`, `image`.
     The default type will be set to `volume` if not specified.
     i.e., `--mount src=vol-1,dst=/app,readonly` equals `--mount type=volume,src=vol-1,dst=/app,readonly`
-    - unimplemented type: `image`
   - Common Options:
     - :whale: `src`, `source`: Mount source spec for bind and volume. Mandatory for bind.
     - :whale: `dst`, `destination`, `target`: Mount destination spec.
@@ -313,6 +312,9 @@ Volume flags:
       Defaults to `1777` or world-writable.
   - Options specific to `volume`:
     - unimplemented options: `volume-nocopy`, `volume-label`, `volume-driver`, `volume-opt`
+  - Options specific to `image`:
+    - :whale: `src`, `source`: image reference (mandatory). The image filesystem is mounted read-only.
+    - unimplemented options: `subpath`
 - :whale: `--volumes-from`: Mount volumes from the specified container(s), e.g. "--volumes-from my-container".
 
 Rootfs flags:

pkg/cmd/container/create.go

@@ -70,7 +70,7 @@ import (
 )
 
 // Create will create a container.
-func Create(ctx context.Context, client *containerd.Client, args []string, netManager containerutil.NetworkOptionsManager, options types.ContainerCreateOptions) (containerd.Container, func(), error) {
+func Create(ctx context.Context, client *containerd.Client, args []string, netManager containerutil.NetworkOptionsManager, options types.ContainerCreateOptions) (_ containerd.Container, _ func(), retErr error) {
 	// Acquire an exclusive lock on the volume store until we are done to avoid being raced by any other
 	// volume operations (or any other operation involving volume manipulation)
 	volStore, err := volume.Store(options.GOptions.Namespace, options.GOptions.DataRoot, options.GOptions.Address)
@@ -94,6 +94,23 @@ func Create(ctx context.Context, client *containerd.Client, args []string, netMa
 	internalLabels.platform = options.Platform
 	internalLabels.namespace = options.GOptions.Namespace
 
+	// If creation fails after image-mount views are created, remove them so the
+	// snapshots do not leak (the cleanup label is only persisted on success).
+	defer func() {
+		if retErr == nil {
+			return
+		}
+		var keys []string
+		for _, mp := range internalLabels.mountPoints {
+			if mp.ImageMountSnapshot != "" {
+				keys = append(keys, mp.ImageMountSnapshot)
+			}
+		}
+		if len(keys) > 0 {
+			removeImageMountViews(ctx, client.SnapshotService(options.GOptions.Snapshotter), keys)
+		}
+	}()
+
 	var (
 		id    = idgen.GenerateID()
 		opts  []oci.SpecOpts
@@ -806,6 +823,22 @@ func withInternalLabels(internalLabels internalLabels) (containerd.NewContainerO
 		m[labels.AnonymousVolumes] = string(anonVolumeJSON)
 	}
 
+	// Record the snapshot keys of any type=image mount views so they can be
+	// removed when the container is deleted.
+	var imageMountSnapshots []string
+	for _, mp := range internalLabels.mountPoints {
+		if mp.ImageMountSnapshot != "" {
+			imageMountSnapshots = append(imageMountSnapshots, mp.ImageMountSnapshot)
+		}
+	}
+	if len(imageMountSnapshots) > 0 {
+		b, err := json.Marshal(imageMountSnapshots)
+		if err != nil {
+			return nil, err
+		}
+		m[labels.ImageMountSnapshots] = string(b)
+	}
+
 	if internalLabels.pidFile != "" {
 		m[labels.PIDFile] = internalLabels.pidFile
 	}

pkg/cmd/container/remove.go

@@ -109,6 +109,13 @@ func RemoveContainer(ctx context.Context, c containerd.Container, globalOptions
 		return err
 	}
 
+	// Capture the container's snapshotter before deletion: image-mount views were
+	// created against it, which may differ from the current --snapshotter flag.
+	imageMountSnapshotter := globalOptions.Snapshotter
+	if info, err := c.Info(ctx); err == nil && info.Snapshotter != "" {
+		imageMountSnapshotter = info.Snapshotter
+	}
+
 	// Get datastore
 	dataStore, err := clientutil.DataStore(globalOptions.DataRoot, globalOptions.Address)
 	if err != nil {
@@ -275,6 +282,16 @@ func RemoveContainer(ctx context.Context, c containerd.Container, globalOptions
 				}
 			}
 		}
+
+		// Remove the read-only views backing type=image mounts - soft failure.
+		if snapshotsJSON, ok := containerLabels[labels.ImageMountSnapshots]; ok {
+			var keys []string
+			if err = json.Unmarshal([]byte(snapshotsJSON), &keys); err != nil {
+				log.G(ctx).WithError(err).Warnf("failed to unmarshal image-mount snapshots for container %q", id)
+			} else {
+				removeImageMountViews(ctx, client.SnapshotService(imageMountSnapshotter), keys)
+			}
+		}
 	}()
 
 	// Get the task.

pkg/cmd/container/run_mount.go

@@ -37,6 +37,7 @@ import (
 	"github.com/containerd/containerd/v2/core/containers"
 	"github.com/containerd/containerd/v2/core/leases"
 	"github.com/containerd/containerd/v2/core/mount"
+	"github.com/containerd/containerd/v2/core/snapshots"
 	"github.com/containerd/containerd/v2/pkg/oci"
 	"github.com/containerd/continuity/fs"
 	"github.com/containerd/errdefs"
@@ -122,17 +123,83 @@ func parseMountFlags(volStore volumestore.VolumeStore, options types.ContainerCr
 	return parsed, nil
 }
 
+// gcRootLabel marks a snapshot as a GC root so containerd does not reclaim it.
+const gcRootLabel = "containerd.io/gc.root"
+
+// setupImageMount ensures and unpacks ref, then creates a read-only GC-rooted
+// snapshot view of its rootfs. It returns the OCI mount for destination and the
+// view's snapshot key.
+func setupImageMount(ctx context.Context, client *containerd.Client, options types.ContainerCreateOptions, ref, destination string) (specs.Mount, string, error) {
+	ensured, err := imgutil.EnsureImage(ctx, client, ref, options.ImagePullOpt)
+	if err != nil {
+		return specs.Mount{}, "", fmt.Errorf("failed to ensure image %q for image mount: %w", ref, err)
+	}
+	if err := ensured.Image.Unpack(ctx, options.GOptions.Snapshotter); err != nil {
+		return specs.Mount{}, "", fmt.Errorf("failed to unpack image %q for image mount: %w", ref, err)
+	}
+	diffIDs, err := ensured.Image.RootFS(ctx)
+	if err != nil {
+		return specs.Mount{}, "", fmt.Errorf("failed to get rootfs of image %q for image mount: %w", ref, err)
+	}
+	chainID := identity.ChainID(diffIDs).String()
+
+	snapshotKey := idgen.GenerateID() + "-image-mount"
+	s := client.SnapshotService(options.GOptions.Snapshotter)
+	mounts, err := s.View(ctx, snapshotKey, chainID, snapshots.WithLabels(map[string]string{
+		gcRootLabel: time.Now().UTC().Format(time.RFC3339),
+	}))
+	if err != nil {
+		return specs.Mount{}, "", fmt.Errorf("failed to create read-only view of image %q: %w", ref, err)
+	}
+	// overlayfs and native snapshotters each yield a single mount for a view.
+	if len(mounts) != 1 {
+		if rmErr := s.Remove(ctx, snapshotKey); rmErr != nil && !errdefs.IsNotFound(rmErr) {
+			log.G(ctx).WithError(rmErr).Warnf("failed to remove image-mount snapshot %q", snapshotKey)
+		}
+		return specs.Mount{}, "", fmt.Errorf("image mount expects exactly one mount from the snapshotter, got %d", len(mounts))
+	}
+
+	m := mounts[0]
+	opts := m.Options
+	// A view without an upper dir is already read-only; make it explicit for
+	// bind-backed snapshotters.
+	if !strutil.InStringSlice(opts, "ro") {
+		opts = append(opts, "ro")
+	}
+	return specs.Mount{
+		Type:        m.Type,
+		Source:      m.Source,
+		Destination: destination,
+		Options:     opts,
+	}, snapshotKey, nil
+}
+
+// removeImageMountViews removes the snapshotter views created for type=image
+// mounts. NotFound is ignored; other failures are logged but not fatal.
+func removeImageMountViews(ctx context.Context, s snapshots.Snapshotter, keys []string) {
+	for _, k := range keys {
+		if err := s.Remove(ctx, k); err != nil && !errdefs.IsNotFound(err) {
+			log.G(ctx).WithError(err).Warnf("failed to remove image-mount snapshot %q", k)
+		}
+	}
+}
+
 // generateMountOpts generates volume-related mount opts.
 // Other mounts such as procfs mount are not handled here.
 func generateMountOpts(ctx context.Context, client *containerd.Client, ensuredImage *imgutil.EnsuredImage,
-	volStore volumestore.VolumeStore, options types.ContainerCreateOptions) ([]oci.SpecOpts, []string, []*mountutil.Processed, error) {
+	volStore volumestore.VolumeStore, options types.ContainerCreateOptions) (opts []oci.SpecOpts, anonVolumes []string, mountPoints []*mountutil.Processed, retErr error) {
 	//nolint:prealloc
 	var (
-		opts        []oci.SpecOpts
-		anonVolumes []string
-		userMounts  []specs.Mount
-		mountPoints []*mountutil.Processed
+		userMounts      []specs.Mount
+		imageMountViews []string
 	)
+	// Remove any image-mount views created here if this function fails, so a
+	// partial setup does not leak snapshots.
+	defer func() {
+		if retErr != nil && len(imageMountViews) > 0 {
+			removeImageMountViews(ctx, client.SnapshotService(options.GOptions.Snapshotter), imageMountViews)
+		}
+	}()
 	mounted := make(map[string]struct{})
 	var imageVolumes map[string]struct{}
 	var tempDir string
@@ -229,6 +296,20 @@ func generateMountOpts(ctx context.Context, client *containerd.Client, ensuredIm
 	} else if len(parsed) > 0 {
 		ociMounts := make([]specs.Mount, len(parsed))
 		for i, x := range parsed {
+			// type=image: build the read-only view now and record its snapshot
+			// key for cleanup on container removal.
+			if x.Type == mountutil.Image {
+				m, snapshotKey, err := setupImageMount(ctx, client, options, x.Mount.Source, x.Mount.Destination)
+				if err != nil {
+					return nil, nil, nil, err
+				}
+				imageMountViews = append(imageMountViews, snapshotKey)
+				ociMounts[i] = m
+				x.ImageMountSnapshot = snapshotKey
+				mounted[filepath.Clean(x.Mount.Destination)] = struct{}{}
+				continue
+			}
+
 			ociMounts[i] = x.Mount
 			mounted[filepath.Clean(x.Mount.Destination)] = struct{}{}
 

pkg/labels/labels.go

@@ -80,6 +80,10 @@ const (
 	// AnonymousVolumes is a JSON-marshalled string of []string
 	AnonymousVolumes = Prefix + "anonymous-volumes"
 
+	// ImageMountSnapshots is a JSON-marshalled []string of snapshotter keys for
+	// the read-only views backing `--mount type=image`, removed on container deletion.
+	ImageMountSnapshots = Prefix + "image-mount-snapshots"
+
 	// Platform is the normalized platform string like "linux/ppc64le".
 	Platform = Prefix + "platform"
 

pkg/mountutil/mountutil.go

@@ -39,6 +39,7 @@ const (
 	Bind          = "bind"
 	Volume        = "volume"
 	Tmpfs         = "tmpfs"
+	Image         = "image"
 	Npipe         = "npipe"
 	pathSeparator = string(os.PathSeparator)
 )
@@ -50,6 +51,9 @@ type Processed struct {
 	AnonymousVolume string // anonymous volume name
 	Mode            string
 	Opts            []oci.SpecOpts
+	// ImageMountSnapshot is the snapshotter key of the read-only view for a
+	// type=image mount; empty for other mount types.
+	ImageMountSnapshot string
 }
 
 type volumeSpec struct {