Add zfs allow and zfs unallow support

ZFS allows for specific permissions to be delegated to normal users
with the `zfs allow` and `zfs unallow` commands.  In addition, non-
privileged users should be able to run all of the follow commands:

  * zpool [list | iostat | status | get]
  * zfs [list | get]

Historically this functionality was unavailable on Linux.  In order
to add it the secpolicy_* functions need to be implemented and mapped
to the equivalent Linux capability.  Only then could the permissions on
the `/dev/zfs` be related and the internal ZFS permission checks used.

Even with this change some limitations remain.  Under Linux only the
root user is allowed to modify the namespace (unless it's a private
namespace).  This means the mount, mountpoint, canmount, unmount, and
remount delegations cannot be supported with the existing code.  It
may be possible to add this functionality in the future.

This functionality was validated with the cli_user and delegation test
cases from the ZFS Test Suite.  These tests exhaustively verify each
of the supported permissions which can be delegated and ensures only
an authorized user can perform it.

Two minor bug fixes were required for test-running.py.  First, the
Timer() object cannot be safely created in a `try:` block when there
is an unconditional `finally` block which references it.  Second,
when running as a normal user also check for scripts using the
both the .ksh and .sh suffixes.

Finally, existing users who are simulating delegations by setting
group permissions on the /dev/zfs device can revert that customization
when this change is applied.

Signed-off-by: Brian Behlendorf <[email protected]>
Issue #362

cmd/zfs/zfs_main.c

@@ -610,7 +610,12 @@ zfs_mount_and_share(libzfs_handle_t *hdl, const char *dataset, zfs_type_t type)
 	 */
 	if (zfs_prop_valid_for_type(ZFS_PROP_CANMOUNT, type, B_FALSE) &&
 	    zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT) == ZFS_CANMOUNT_ON) {
-		if (zfs_mount(zhp, NULL, 0) != 0) {
+		if (geteuid() != 0) {
+			(void) fprintf(stderr, gettext("filesystem "
+			    "successfully created, but it may only be "
+			    "mounted by root\n"));
+			ret = 1;
+		} else if (zfs_mount(zhp, NULL, 0) != 0) {
 			(void) fprintf(stderr, gettext("filesystem "
 			    "successfully created, but not mounted\n"));
 			ret = 1;

cmd/zpool/zpool_main.c

@@ -42,7 +42,6 @@
 #include <string.h>
 #include <strings.h>
 #include <unistd.h>
-#include <priv.h>
 #include <pwd.h>
 #include <zone.h>
 #include <zfs_prop.h>
@@ -2331,21 +2330,20 @@ zpool_do_import(int argc, char **argv)
 			(void) fprintf(stderr, gettext("too many arguments\n"));
 			usage(B_FALSE);
 		}
+	}
 
-		/*
-		 * Check for the SYS_CONFIG privilege.  We do this explicitly
-		 * here because otherwise any attempt to discover pools will
-		 * silently fail.
-		 */
-		if (argc == 0 && !priv_ineffect(PRIV_SYS_CONFIG)) {
-			(void) fprintf(stderr, gettext("cannot "
-			    "discover pools: permission denied\n"));
-			if (searchdirs != NULL)
-				free(searchdirs);
+	/*
+	 * Check for the effective uid.  We do this explicitly here because
+	 * otherwise any attempt to discover pools will silently fail.
+	 */
+	if (argc == 0 && geteuid() != 0) {
+		(void) fprintf(stderr, gettext("cannot "
+		    "discover pools: permission denied\n"));
+		if (searchdirs != NULL)
+			free(searchdirs);
 
-			nvlist_free(policy);
-			return (1);
-		}
+		nvlist_free(policy);
+		return (1);
 	}
 
 	/*

include/sys/Makefile.am

@@ -39,6 +39,7 @@ COMMON_H = \
 	$(top_srcdir)/include/sys/nvpair.h \
 	$(top_srcdir)/include/sys/nvpair_impl.h \
 	$(top_srcdir)/include/sys/pathname.h \
+	$(top_srcdir)/include/sys/policy.h \
 	$(top_srcdir)/include/sys/range_tree.h \
 	$(top_srcdir)/include/sys/refcount.h \
 	$(top_srcdir)/include/sys/rrwlock.h \

include/sys/policy.h

@@ -0,0 +1,60 @@
+/*
+ * 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
+ */
+
+/*
+ * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Copyright 2015, Joyent, Inc. All rights reserved.
+ * Copyright (c) 2016, Lawrence Livermore National Security, LLC.
+ */
+
+#ifndef _SYS_POLICY_H
+#define	_SYS_POLICY_H
+
+#ifdef _KERNEL
+
+#include <sys/cred.h>
+#include <sys/types.h>
+#include <sys/xvattr.h>
+#include <sys/zpl.h>
+
+int secpolicy_nfs(const cred_t *);
+int secpolicy_sys_config(const cred_t *, boolean_t);
+int secpolicy_vnode_access2(const cred_t *, struct inode *,
+    uid_t, mode_t, mode_t);
+int secpolicy_vnode_any_access(const cred_t *, struct inode *, uid_t);
+int secpolicy_vnode_chown(const cred_t *, uid_t);
+int secpolicy_vnode_create_gid(const cred_t *);
+int secpolicy_vnode_remove(const cred_t *);
+int secpolicy_vnode_setdac(const cred_t *, uid_t);
+int secpolicy_vnode_setid_retain(const cred_t *, boolean_t);
+int secpolicy_vnode_setids_setgids(const cred_t *, gid_t);
+int secpolicy_zinject(const cred_t *);
+int secpolicy_zfs(const cred_t *);
+void secpolicy_setid_clear(vattr_t *, cred_t *);
+int secpolicy_setid_setsticky_clear(struct inode *, vattr_t *,
+    const vattr_t *, cred_t *);
+int secpolicy_xvattr(xvattr_t *, uid_t, cred_t *, vtype_t);
+int secpolicy_vnode_setattr(cred_t *, struct inode *, struct vattr *,
+    const struct vattr *, int, int (void *, int, cred_t *), void *);
+int secpolicy_basic_link(const cred_t *);
+
+#endif /* _KERNEL */
+#endif /* _SYS_POLICY_H */

include/sys/zfs_context.h

@@ -738,6 +738,7 @@ extern int zfs_secpolicy_snapshot_perms(const char *name, cred_t *cr);
 extern int zfs_secpolicy_rename_perms(const char *from, const char *to,
     cred_t *cr);
 extern int zfs_secpolicy_destroy_perms(const char *name, cred_t *cr);
+extern int secpolicy_zfs(const cred_t *cr);
 extern zoneid_t getzoneid(void);
 
 /* SID stuff */

lib/libspl/include/Makefile.am

@@ -12,7 +12,6 @@ libspl_HEADERS = \
 	$(top_srcdir)/lib/libspl/include/limits.h \
 	$(top_srcdir)/lib/libspl/include/locale.h \
 	$(top_srcdir)/lib/libspl/include/note.h \
-	$(top_srcdir)/lib/libspl/include/priv.h \
 	$(top_srcdir)/lib/libspl/include/statcommon.h \
 	$(top_srcdir)/lib/libspl/include/stdio.h \
 	$(top_srcdir)/lib/libspl/include/stdlib.h \

lib/libzfs/libzfs_dataset.c

@@ -43,7 +43,6 @@
 #include <fcntl.h>
 #include <sys/mntent.h>
 #include <sys/mount.h>
-#include <priv.h>
 #include <pwd.h>
 #include <grp.h>
 #include <stddef.h>

lib/libzpool/kernel.c

@@ -1302,6 +1302,12 @@ zfs_secpolicy_destroy_perms(const char *name, cred_t *cr)
 	return (0);
 }
 
+int
+secpolicy_zfs(const cred_t *cr)
+{
+	return (0);
+}
+
 ksiddomain_t *
 ksid_lookupdomain(const char *dom)
 {

module/zfs/Makefile.in

@@ -43,6 +43,7 @@ $(MODULE)-objs += lz4.o
 $(MODULE)-objs += metaslab.o
 $(MODULE)-objs += multilist.o
 $(MODULE)-objs += pathname.o
+$(MODULE)-objs += policy.o
 $(MODULE)-objs += range_tree.o
 $(MODULE)-objs += refcount.o
 $(MODULE)-objs += rrwlock.o

module/zfs/dmu_objset.c

@@ -52,6 +52,7 @@
 #include <sys/zfs_onexit.h>
 #include <sys/dsl_destroy.h>
 #include <sys/vdev.h>
+#include <sys/policy.h>
 
 /*
  * Needed to close a window in dnode_move() that allows the objset to be freed

module/zfs/dmu_send.c

@@ -56,6 +56,7 @@
 #include <sys/zfeature.h>
 #include <sys/bqueue.h>
 #include <sys/zvol.h>
+#include <sys/policy.h>
 
 /* Set this tunable to TRUE to replace corrupt data with 0x2f5baddb10c */
 int zfs_send_corrupt_data = B_FALSE;

module/zfs/dsl_dataset.c

@@ -51,6 +51,7 @@
 #include <sys/dsl_destroy.h>
 #include <sys/dsl_userhold.h>
 #include <sys/dsl_bookmark.h>
+#include <sys/policy.h>
 
 /*
  * The SPA supports block sizes up to 16MB.  However, very large blocks