Browse code
Only chown network files within container metadata
If the user specifies a mountpath from the host, we should not be
attempting to chown files outside the daemon's metadata directory
(represented by `daemon.repository` at init time).
This forces users who want to use user namespaces to handle the
ownership needs of any external files mounted as network files
(/etc/resolv.conf, /etc/hosts, /etc/hostname) separately from the
daemon. In all other volume/bind mount situations we have taken this
same line--we don't chown host file content.
Docker-DCO-1.1-Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
Phil Estes authored on 2017/07/23 11:11:09Showing 3 changed files
- daemon/volumes_unix.go
- integration-cli/docker_api_containers_unix_test.go
- integration-cli/docker_cli_run_test.go
| ... | ... |
@@ -86,8 +86,13 @@ func (daemon *Daemon) setupMounts(c *container.Container) ([]container.Mount, er |
| 86 | 86 |
// remapped root (user namespaces) |
| 87 | 87 |
rootIDs := daemon.idMappings.RootPair() |
| 88 | 88 |
for _, mount := range netMounts {
|
| 89 |
- if err := os.Chown(mount.Source, rootIDs.UID, rootIDs.GID); err != nil {
|
|
| 90 |
- return nil, err |
|
| 89 |
+ // we should only modify ownership of network files within our own container |
|
| 90 |
+ // metadata repository. If the user specifies a mount path external, it is |
|
| 91 |
+ // up to the user to make sure the file has proper ownership for userns |
|
| 92 |
+ if strings.Index(mount.Source, daemon.repository) == 0 {
|
|
| 93 |
+ if err := os.Chown(mount.Source, rootIDs.UID, rootIDs.GID); err != nil {
|
|
| 94 |
+ return nil, err |
|
| 95 |
+ } |
|
| 91 | 96 |
} |
| 92 | 97 |
} |
| 93 | 98 |
return append(mounts, netMounts...), nil |
| 94 | 99 |
new file mode 100644 |
| ... | ... |
@@ -0,0 +1,77 @@ |
| 0 |
+// +build !windows |
|
| 1 |
+ |
|
| 2 |
+package main |
|
| 3 |
+ |
|
| 4 |
+import ( |
|
| 5 |
+ "io/ioutil" |
|
| 6 |
+ "os" |
|
| 7 |
+ "path/filepath" |
|
| 8 |
+ |
|
| 9 |
+ "github.com/docker/docker/api/types" |
|
| 10 |
+ containertypes "github.com/docker/docker/api/types/container" |
|
| 11 |
+ mounttypes "github.com/docker/docker/api/types/mount" |
|
| 12 |
+ networktypes "github.com/docker/docker/api/types/network" |
|
| 13 |
+ "github.com/docker/docker/client" |
|
| 14 |
+ "github.com/docker/docker/integration-cli/checker" |
|
| 15 |
+ "github.com/docker/docker/pkg/ioutils" |
|
| 16 |
+ "github.com/docker/docker/pkg/system" |
|
| 17 |
+ "github.com/go-check/check" |
|
| 18 |
+ "github.com/stretchr/testify/assert" |
|
| 19 |
+ "golang.org/x/net/context" |
|
| 20 |
+) |
|
| 21 |
+ |
|
| 22 |
+func (s *DockerSuite) TestContainersAPINetworkMountsNoChown(c *check.C) {
|
|
| 23 |
+ // chown only applies to Linux bind mounted volumes; must be same host to verify |
|
| 24 |
+ testRequires(c, DaemonIsLinux, SameHostDaemon) |
|
| 25 |
+ |
|
| 26 |
+ tmpDir, err := ioutils.TempDir("", "test-network-mounts")
|
|
| 27 |
+ c.Assert(err, checker.IsNil) |
|
| 28 |
+ defer os.RemoveAll(tmpDir) |
|
| 29 |
+ |
|
| 30 |
+ // make tmp dir readable by anyone to allow userns process to mount from |
|
| 31 |
+ err = os.Chmod(tmpDir, 0755) |
|
| 32 |
+ c.Assert(err, checker.IsNil) |
|
| 33 |
+ // create temp files to use as network mounts |
|
| 34 |
+ tmpNWFileMount := filepath.Join(tmpDir, "nwfile") |
|
| 35 |
+ |
|
| 36 |
+ err = ioutil.WriteFile(tmpNWFileMount, []byte("network file bind mount"), 0644)
|
|
| 37 |
+ c.Assert(err, checker.IsNil) |
|
| 38 |
+ |
|
| 39 |
+ config := containertypes.Config{
|
|
| 40 |
+ Image: "busybox", |
|
| 41 |
+ } |
|
| 42 |
+ hostConfig := containertypes.HostConfig{
|
|
| 43 |
+ Mounts: []mounttypes.Mount{
|
|
| 44 |
+ {
|
|
| 45 |
+ Type: "bind", |
|
| 46 |
+ Source: tmpNWFileMount, |
|
| 47 |
+ Target: "/etc/resolv.conf", |
|
| 48 |
+ }, |
|
| 49 |
+ {
|
|
| 50 |
+ Type: "bind", |
|
| 51 |
+ Source: tmpNWFileMount, |
|
| 52 |
+ Target: "/etc/hostname", |
|
| 53 |
+ }, |
|
| 54 |
+ {
|
|
| 55 |
+ Type: "bind", |
|
| 56 |
+ Source: tmpNWFileMount, |
|
| 57 |
+ Target: "/etc/hosts", |
|
| 58 |
+ }, |
|
| 59 |
+ }, |
|
| 60 |
+ } |
|
| 61 |
+ |
|
| 62 |
+ cli, err := client.NewEnvClient() |
|
| 63 |
+ c.Assert(err, checker.IsNil) |
|
| 64 |
+ defer cli.Close() |
|
| 65 |
+ |
|
| 66 |
+ ctrCreate, err := cli.ContainerCreate(context.Background(), &config, &hostConfig, &networktypes.NetworkingConfig{}, "")
|
|
| 67 |
+ c.Assert(err, checker.IsNil) |
|
| 68 |
+ // container will exit immediately because of no tty, but we only need the start sequence to test the condition |
|
| 69 |
+ err = cli.ContainerStart(context.Background(), ctrCreate.ID, types.ContainerStartOptions{})
|
|
| 70 |
+ c.Assert(err, checker.IsNil) |
|
| 71 |
+ |
|
| 72 |
+ // check that host-located bind mount network file did not change ownership when the container was started |
|
| 73 |
+ statT, err := system.Stat(tmpNWFileMount) |
|
| 74 |
+ c.Assert(err, checker.IsNil) |
|
| 75 |
+ assert.Equal(c, uint32(0), statT.UID(), "bind mounted network file should not change ownership from root") |
|
| 76 |
+} |
| ... | ... |
@@ -3115,6 +3115,11 @@ func (s *DockerSuite) TestRunNetworkFilesBindMount(c *check.C) {
|
| 3115 | 3115 |
filename := createTmpFile(c, expected) |
| 3116 | 3116 |
defer os.Remove(filename) |
| 3117 | 3117 |
|
| 3118 |
+ // for user namespaced test runs, the temp file must be accessible to unprivileged root |
|
| 3119 |
+ if err := os.Chmod(filename, 0646); err != nil {
|
|
| 3120 |
+ c.Fatalf("error modifying permissions of %s: %v", filename, err)
|
|
| 3121 |
+ } |
|
| 3122 |
+ |
|
| 3118 | 3123 |
nwfiles := []string{"/etc/resolv.conf", "/etc/hosts", "/etc/hostname"}
|
| 3119 | 3124 |
|
| 3120 | 3125 |
for i := range nwfiles {
|
| ... | ... |
@@ -3132,6 +3137,11 @@ func (s *DockerSuite) TestRunNetworkFilesBindMountRO(c *check.C) {
|
| 3132 | 3132 |
filename := createTmpFile(c, "test123") |
| 3133 | 3133 |
defer os.Remove(filename) |
| 3134 | 3134 |
|
| 3135 |
+ // for user namespaced test runs, the temp file must be accessible to unprivileged root |
|
| 3136 |
+ if err := os.Chmod(filename, 0646); err != nil {
|
|
| 3137 |
+ c.Fatalf("error modifying permissions of %s: %v", filename, err)
|
|
| 3138 |
+ } |
|
| 3139 |
+ |
|
| 3135 | 3140 |
nwfiles := []string{"/etc/resolv.conf", "/etc/hosts", "/etc/hostname"}
|
| 3136 | 3141 |
|
| 3137 | 3142 |
for i := range nwfiles {
|
| ... | ... |
@@ -3149,6 +3159,11 @@ func (s *DockerSuite) TestRunNetworkFilesBindMountROFilesystem(c *check.C) {
|
| 3149 | 3149 |
filename := createTmpFile(c, "test123") |
| 3150 | 3150 |
defer os.Remove(filename) |
| 3151 | 3151 |
|
| 3152 |
+ // for user namespaced test runs, the temp file must be accessible to unprivileged root |
|
| 3153 |
+ if err := os.Chmod(filename, 0646); err != nil {
|
|
| 3154 |
+ c.Fatalf("error modifying permissions of %s: %v", filename, err)
|
|
| 3155 |
+ } |
|
| 3156 |
+ |
|
| 3152 | 3157 |
nwfiles := []string{"/etc/resolv.conf", "/etc/hosts", "/etc/hostname"}
|
| 3153 | 3158 |
|
| 3154 | 3159 |
for i := range nwfiles {
|