Take user namespaces into account in policy checks

Change file related checks to use user namespaces and make
sure involved uids/gids are mappable in the current
namespace.

Note that checks without file ownership information will
still not take user namespaces into account, as some of
these should be handled via 'zfs allow' (otherwise root in a
user namespace could issue commands such as `zpool export`).

This also adds an initial user namespace regression test
for the setgid bit loss, with a user_ns_exec helper usable
in further tests.

Additionally, configure checks for the required user
namespace related features are added for:
  * ns_capable
  * kuid/kgid_has_mapping()
  * user_ns in cred_t

Reviewed-by: Brian Behlendorf <[email protected]>
Signed-off-by: Wolfgang Bumiller <[email protected]>
Closes #6800 
Closes #7270 

4 files changed, 187 insertions, 0 deletions

diff --git a/tests/zfs-tests/cmd/Makefile.am b/tests/zfs-tests/cmd/Makefile.am
index 123a79849..bc19f184c 100644
--- a/tests/zfs-tests/cmd/Makefile.am
+++ b/tests/zfs-tests/cmd/Makefile.am
@@ -2,6 +2,7 @@ EXTRA_DIST = file_common.h
 
 SUBDIRS = \
 	chg_usr_exec \
+	user_ns_exec \
 	devname2devid \
 	dir_rd_update \
 	file_check \
diff --git a/tests/zfs-tests/cmd/user_ns_exec/.gitignore b/tests/zfs-tests/cmd/user_ns_exec/.gitignore
new file mode 100644
index 000000000..655867a64
--- /dev/null
+++ b/tests/zfs-tests/cmd/user_ns_exec/.gitignore
@@ -0,0 +1 @@
+/user_ns_exec
diff --git a/tests/zfs-tests/cmd/user_ns_exec/Makefile.am b/tests/zfs-tests/cmd/user_ns_exec/Makefile.am
new file mode 100644
index 000000000..5b4bc9aaa
--- /dev/null
+++ b/tests/zfs-tests/cmd/user_ns_exec/Makefile.am
@@ -0,0 +1,6 @@
+include $(top_srcdir)/config/Rules.am
+
+pkgexecdir = $(datadir)/@PACKAGE@/zfs-tests/bin
+
+pkgexec_PROGRAMS = user_ns_exec
+user_ns_exec_SOURCES = user_ns_exec.c
diff --git a/tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c b/tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c
new file mode 100644
index 000000000..cd46738bd
--- /dev/null
+++ b/tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c
@@ -0,0 +1,179 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+
+#include <stdio.h>
+#include <unistd.h>
+#include <string.h>
+#include <limits.h>
+#include <sys/types.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/wait.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <signal.h>
+#include <sched.h>
+
+#define	EXECSHELL	"/bin/sh"
+#define	UIDMAP		"0 100000 65536"
+
+static int
+child_main(int argc, char *argv[], int sync_pipe)
+{
+	char sync_buf;
+	char cmds[BUFSIZ] = { 0 };
+	char sep[] = " ";
+	int i, len;
+
+	if (unshare(CLONE_NEWUSER | CLONE_NEWNS) != 0) {
+		perror("unshare");
+		return (1);
+	}
+
+	/* tell parent we entered the new namespace */
+	if (write(sync_pipe, "1", 1) != 1) {
+		perror("write");
+		return (1);
+	}
+
+	/* wait for parent to setup the uid mapping */
+	if (read(sync_pipe, &sync_buf, 1) != 1) {
+		(void) fprintf(stderr, "user namespace setup failed\n");
+		return (1);
+	}
+
+	close(sync_pipe);
+
+	if (setuid(0) != 0) {
+		perror("setuid");
+		return (1);
+	}
+	if (setgid(0) != 0) {
+		perror("setgid");
+		return (1);
+	}
+
+	len = 0;
+	for (i = 1; i < argc; i++) {
+		(void) snprintf(cmds+len, sizeof (cmds)-len,
+		    "%s%s", argv[i], sep);
+		len += strlen(argv[i]) + strlen(sep);
+	}
+
+	if (execl(EXECSHELL, "sh",  "-c", cmds, (char *)NULL) != 0) {
+		perror("execl: " EXECSHELL);
+		return (1);
+	}
+
+	return (0);
+}
+
+static int
+set_idmap(pid_t pid, const char *file)
+{
+	int result = 0;
+	int mapfd;
+	char path[PATH_MAX];
+
+	(void) snprintf(path, sizeof (path), "/proc/%d/%s", (int)pid, file);
+
+	mapfd = open(path, O_WRONLY);
+	if (mapfd < 0) {
+		result = errno;
+		perror("open");
+		return (errno);
+	}
+
+	if (write(mapfd, UIDMAP, sizeof (UIDMAP)-1) != sizeof (UIDMAP)-1) {
+		perror("write");
+		result = (errno);
+	}
+
+	close(mapfd);
+
+	return (result);
+}
+
+int
+main(int argc, char *argv[])
+{
+	char sync_buf;
+	int result, wstatus;
+	int syncfd[2];
+	pid_t child;
+
+	if (argc < 2 || strlen(argv[1]) == 0) {
+		(void) printf("\tUsage: %s <commands> ...\n", argv[0]);
+		return (1);
+	}
+
+	if (socketpair(AF_UNIX, SOCK_STREAM, 0, syncfd) != 0) {
+		perror("socketpair");
+		return (1);
+	}
+
+	child = fork();
+	if (child == (pid_t)-1) {
+		perror("fork");
+		return (1);
+	}
+
+	if (child == 0) {
+		close(syncfd[0]);
+		return (child_main(argc, argv, syncfd[1]));
+	}
+
+	close(syncfd[1]);
+
+	result = 0;
+	/* wait for the child to have unshared its namespaces */
+	if (read(syncfd[0], &sync_buf, 1) != 1) {
+		perror("read");
+		kill(child, SIGKILL);
+		result = 1;
+		goto reap;
+	}
+
+	/* write uid mapping */
+	if (set_idmap(child, "uid_map") != 0 ||
+	    set_idmap(child, "gid_map") != 0) {
+		result = 1;
+		kill(child, SIGKILL);
+		goto reap;
+	}
+
+	/* tell the child to proceed */
+	if (write(syncfd[0], "1", 1) != 1) {
+		perror("write");
+		kill(child, SIGKILL);
+		result = 1;
+		goto reap;
+	}
+	close(syncfd[0]);
+
+reap:
+	while (waitpid(child, &wstatus, 0) != child)
+		kill(child, SIGKILL);
+	if (result == 0)
+		result = WEXITSTATUS(wstatus);
+
+	return (result);
+}