daemon.WithCommonOptions() fix detection of user-namespaces

Commit dae652e2e5 added support for non-privileged
containers to use ICMP_PROTO (used for `ping`). This option cannot be set for
containers that have user-namespaces enabled.

However, the detection looks to be incorrect; HostConfig.UsernsMode was added
in 6993e891d1 / ee2183881b,
and the property only has meaning if the daemon is running with user namespaces
enabled. In other situations, the property has no meaning.
As a result of the above, the sysctl would only be set for containers running
with UsernsMode=host on a daemon running with user-namespaces enabled.

This patch adds a check if the daemon has user-namespaces enabled (RemappedRoot
having a non-empty value), or if the daemon is running inside a user namespace
(e.g. rootless mode) to fix the detection.

Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
(cherry picked from commit a826ca3aef)

---
The cherry-pick was almost clean but `userns.RunningInUserNS()` -> `sys.RunningInUserNS()`.

Fix docker/buildx issue 561
---

Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>

daemon/oci_linux.go

@@ -767,7 +767,8 @@ func WithCommonOptions(daemon *Daemon, c *container.Container) coci.SpecOpts {
 		// joining an existing namespace, only if we create a new net namespace.
 		if c.HostConfig.NetworkMode.IsPrivate() {
 			// We cannot set up ping socket support in a user namespace
-			if !c.HostConfig.UsernsMode.IsPrivate() && sysctlExists("net.ipv4.ping_group_range") {
+			userNS := daemon.configStore.RemappedRoot != "" && c.HostConfig.UsernsMode.IsPrivate()
+			if !userNS && !sys.RunningInUserNS() && sysctlExists("net.ipv4.ping_group_range") {
 				// allow unprivileged ICMP echo sockets without CAP_NET_RAW
 				s.Linux.Sysctl["net.ipv4.ping_group_range"] = "0 2147483647"
 			}

daemon/oci_linux_test.go

@@ -120,7 +120,6 @@ func TestSysctlOverride(t *testing.T) {
 		HostConfig: &containertypes.HostConfig{
 			NetworkMode: "bridge",
 			Sysctls:     map[string]string{},
-			UsernsMode:  "host",
 		},
 	}
 	d := setupFakeDaemon(t, c)
@@ -148,6 +147,20 @@ func TestSysctlOverride(t *testing.T) {
 	assert.Equal(t, s.Hostname, "foobar")
 	assert.Equal(t, s.Linux.Sysctl["kernel.domainname"], c.HostConfig.Sysctls["kernel.domainname"])
 	assert.Equal(t, s.Linux.Sysctl["net.ipv4.ip_unprivileged_port_start"], c.HostConfig.Sysctls["net.ipv4.ip_unprivileged_port_start"])
+
+	// Ensure the ping_group_range is not set on a daemon with user-namespaces enabled
+	d.configStore.RemappedRoot = "dummy:dummy"
+	s, err = d.createSpec(c)
+	assert.NilError(t, err)
+	_, ok := s.Linux.Sysctl["net.ipv4.ping_group_range"]
+	assert.Assert(t, !ok)
+
+	// Ensure the ping_group_range is set on a container in "host" userns mode
+	// on a daemon with user-namespaces enabled
+	c.HostConfig.UsernsMode = "host"
+	s, err = d.createSpec(c)
+	assert.NilError(t, err)
+	assert.Equal(t, s.Linux.Sysctl["net.ipv4.ping_group_range"], "0 2147483647")
 }
 
 // TestSysctlOverrideHost ensures that any implicit network sysctls are not set
@@ -159,7 +172,6 @@ func TestSysctlOverrideHost(t *testing.T) {
 		HostConfig: &containertypes.HostConfig{
 			NetworkMode: "host",
 			Sysctls:     map[string]string{},
-			UsernsMode:  "host",
 		},
 	}
 	d := setupFakeDaemon(t, c)

integration/container/run_linux_test.go

@@ -99,3 +99,32 @@ func TestHostnameDnsResolution(t *testing.T) {
 	assert.Check(t, is.Equal("", res.Stderr()))
 	assert.Equal(t, 0, res.ExitCode)
 }
+
+func TestUnprivilegedPortsAndPing(t *testing.T) {
+	skip.If(t, testEnv.DaemonInfo.OSType != "linux")
+	skip.If(t, testEnv.IsRootless, "rootless mode doesn't support setting net.ipv4.ping_group_range and net.ipv4.ip_unprivileged_port_start")
+
+	defer setupTest(t)()
+	client := testEnv.APIClient()
+	ctx := context.Background()
+
+	cID := container.Run(ctx, t, client, func(c *container.TestContainerConfig) {
+		c.Config.User = "1000:1000"
+	})
+
+	poll.WaitOn(t, container.IsInState(ctx, client, cID, "running"), poll.WithDelay(100*time.Millisecond))
+
+	// Check net.ipv4.ping_group_range.
+	res, err := container.Exec(ctx, client, cID, []string{"cat", "/proc/sys/net/ipv4/ping_group_range"})
+	assert.NilError(t, err)
+	assert.Assert(t, is.Len(res.Stderr(), 0))
+	assert.Equal(t, 0, res.ExitCode)
+	assert.Equal(t, `0	2147483647`, strings.TrimSpace(res.Stdout()))
+
+	// Check net.ipv4.ip_unprivileged_port_start.
+	res, err = container.Exec(ctx, client, cID, []string{"cat", "/proc/sys/net/ipv4/ip_unprivileged_port_start"})
+	assert.NilError(t, err)
+	assert.Assert(t, is.Len(res.Stderr(), 0))
+	assert.Equal(t, 0, res.ExitCode)
+	assert.Equal(t, "0", strings.TrimSpace(res.Stdout()))
+}