Skip to contentSkip to conversation

Allow default options in /etc/exports

PATCH In review Steinar H. Gunderson 4 participants 6 messages Dec 26, 2006 – Feb 7, 2007

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Allow default options in /etc/exports

[Cc-on-reply, please]

Merry Christmas, to those of you who celebrate that. :-) I found some time to write a patch for a two-year-old feature request which was slowly getting more and more annoying; hope it can go into upstream, even though it won't reach our upcoming Debian release.

Patch: Allow default options in /etc/exports

Implement default options in /etc/exports, to fix a long-standing wishlist bug in Debian. (The user claims the syntax matches that of OpenBSD.) This makes it possible to write "/srv/www -sync,no_subtree_check host1 host2 host3" instead of having to write (sync,no_subtree_check) over and over and over again, driving the administrator slowly mad. Such option lines can be placed anywhere on the line, and affects anything after them (I do not know if OpenBSD allows this). The patch is slightly convoluted in order to avoid triggering spurious warnings; for instance, we want "/srv/www -sync host1" to trigger a warning, but not "/srv/www -sync,no_subtree_check host1" or "/srv/www -sync host1(no_subtree_check)".

There was also a suggestion for a truly global (ie. per-file) option list, but this seemed like the safest bet, given that it matches that of other implementations.

Also, the man page is updated with information on the new possibilities, and an example.

support/nfs/exports.c +54 -28
Index: nfs-utils-1.0.10/support/nfs/exports.c
===================================================================
@@ -39,12 +39,13 @@ int export_errno;
 static char	*efname = NULL;
 static XFILE	*efp = NULL;
 static int	first;
+static int	has_default_opts, has_default_subtree_opts;
 static int	*squids = NULL, nsquids = 0,
 		*sqgids = NULL, nsqgids = 0;
 
 static int	getexport(char *exp, int len);
 static int	getpath(char *path, int len);
-static int	parseopts(char *cp, struct exportent *ep, int warn);
+static int	parseopts(char *cp, struct exportent *ep, int warn, int *had_subtree_opt_ptr);
 static int	parsesquash(char *list, int **idp, int *lenp, char **ep);
 static int	parsenum(char **cpp);
 static int	parsemaptype(char *type);
@@ -68,7 +69,7 @@ setexportent(char *fname, char *type)
 struct exportent *
 getexportent(int fromkernel, int fromexports)
 {
-	static struct exportent	ee;
+	static struct exportent	ee, def_ee;
 	char		exp[512], *hostname;
 	char		rpath[MAXPATHLEN+1];
 	char		*opt, *sp;
@@ -78,31 +79,36 @@ getexportent(int fromkernel, int fromexp
 		return NULL;
 
 	freesquash();
-	ee.e_flags = EXPORT_DEFAULT_FLAGS;
-	/* some kernels assume the default is sync rather than
-	 * async.  More recent kernels always report one or other,
-	 * but this test makes sure we assume same as kernel
-	 * Ditto for wgather
-	 */
-	if (fromkernel) {
-		ee.e_flags &= ~NFSEXP_ASYNC;
-		ee.e_flags &= ~NFSEXP_GATHERED_WRITES;
-	}
-	ee.e_maptype = CLE_MAP_IDENT;
-	ee.e_anonuid = 65534;
-	ee.e_anongid = 65534;
-	ee.e_squids = NULL;
-	ee.e_sqgids = NULL;
-	ee.e_mountpoint = NULL;
-	ee.e_nsquids = 0;
-	ee.e_nsqgids = 0;
 
 	if (first || (ok = getexport(exp, sizeof(exp))) == 0) {
-		ok = getpath(ee.e_path, sizeof(ee.e_path));
+		has_default_opts = 0;
+		has_default_subtree_opts = 0;
+	
+		def_ee.e_flags = EXPORT_DEFAULT_FLAGS;
+		/* some kernels assume the default is sync rather than
+		 * async.  More recent kernels always report one or other,
+		 * but this test makes sure we assume same as kernel
+		 * Ditto for wgather
+		 */
+		if (fromkernel) {
+			def_ee.e_flags &= ~NFSEXP_ASYNC;
+			def_ee.e_flags &= ~NFSEXP_GATHERED_WRITES;
+		}
+		def_ee.e_maptype = CLE_MAP_IDENT;
+		def_ee.e_anonuid = 65534;
+		def_ee.e_anongid = 65534;
+		def_ee.e_squids = NULL;
+		def_ee.e_sqgids = NULL;
+		def_ee.e_mountpoint = NULL;
+		def_ee.e_nsquids = 0;
+		def_ee.e_nsqgids = 0;
+
+		ok = getpath(def_ee.e_path, sizeof(def_ee.e_path));
 		if (ok <= 0)
 			return NULL;
-		strncpy (ee.m_path, ee.e_path, sizeof (ee.m_path) - 1);
-		ee.m_path [sizeof (ee.m_path) - 1] = '\0';
+
+		strncpy (def_ee.m_path, def_ee.e_path, sizeof (def_ee.m_path) - 1);
+		def_ee.m_path [sizeof (def_ee.m_path) - 1] = '\0';
 		ok = getexport(exp, sizeof(exp));
 	}
 	if (ok < 0) {
@@ -111,6 +117,23 @@ getexportent(int fromkernel, int fromexp
 		return NULL;
 	}
 	first = 0;
+		
+	/* Check for default options */
+	if (exp[0] == '-') {
+		if (parseopts(exp + 1, &def_ee, 0, &has_default_subtree_opts) < 0)
+			return NULL;
+		
+		has_default_opts = 1;
+
+		ok = getexport(exp, sizeof(exp));
+		if (ok < 0) {
+			xlog(L_ERROR, "expected client(options...)");
+			export_errno = EINVAL;
+			return NULL;
+		}
+	}
+
+	ee = def_ee;
 
 	/* Check for default client */
 	if (ok == 0)
@@ -130,7 +153,8 @@ getexportent(int fromkernel, int fromexp
 		}
 		*sp = '\0';
 	} else {
-	    xlog(L_WARNING, "No options for %s %s: suggest %s(sync) to avoid warning", ee.e_path, exp, exp);
+		if (!has_default_opts)
+			xlog(L_WARNING, "No options for %s %s: suggest %s(sync) to avoid warning", ee.e_path, exp, exp);
 	}
 	if (strlen(hostname) >= sizeof(ee.e_hostname)) {
 		syntaxerr("client name too long");
@@ -140,7 +164,7 @@ getexportent(int fromkernel, int fromexp
 	strncpy(ee.e_hostname, hostname, sizeof (ee.e_hostname) - 1);
 	ee.e_hostname[sizeof (ee.e_hostname) - 1] = '\0';
 
-	if (parseopts(opt, &ee, fromexports) < 0)
+	if (parseopts(opt, &ee, fromexports && !has_default_subtree_opts, NULL) < 0)
 		return NULL;
 
 	/* resolve symlinks */
@@ -293,7 +317,7 @@ mkexportent(char *hname, char *path, cha
 	ee.e_path[sizeof (ee.e_path) - 1] = '\0';
 	strncpy (ee.m_path, ee.e_path, sizeof (ee.m_path) - 1);
 	ee.m_path [sizeof (ee.m_path) - 1] = '\0';
-	if (parseopts(options, &ee, 0) < 0)
+	if (parseopts(options, &ee, 0, NULL) < 0)
 		return NULL;
 	return &ee;
 }
@@ -301,7 +325,7 @@ mkexportent(char *hname, char *path, cha
 int
 updateexportent(struct exportent *eep, char *options)
 {
-	if (parseopts(options, eep, 0) < 0)
+	if (parseopts(options, eep, 0, NULL) < 0)
 		return 0;
 	return 1;
 }
@@ -310,7 +334,7 @@ updateexportent(struct exportent *eep, c
  * Parse option string pointed to by cp and set mount options accordingly.
  */
 static int
-parseopts(char *cp, struct exportent *ep, int warn)
+parseopts(char *cp, struct exportent *ep, int warn, int *had_subtree_opt_ptr)
 {
 	int	had_subtree_opt = 0;
 	char 	*flname = efname?efname:"command line";
@@ -461,6 +485,8 @@ out:
 
 				flname, flline,
 				ep->e_hostname, ep->e_path);
+	if (had_subtree_opt_ptr)
+		*had_subtree_opt_ptr = had_subtree_opt;
 
 	return 1;
 }
utils/exportfs/exports.man +9 -0
Index: nfs-utils-1.0.10/utils/exportfs/exports.man
===================================================================
@@ -22,6 +22,11 @@ client may be immediately followed by a 
 list of export options for that client. No whitespace is permitted
 between a client and its option list.
 .PP
+Also, each line may have one or more specifications for default options
+after the path name, in the form of a dash ("\-") followed by an option
+list. The option list is used for all subsequent exports on that line
+only.
+.PP
 Blank lines are ignored.  A pound sign ("#") introduces a comment to the
 end of the line. Entries may be continued across newlines using a
 backslash. If an export name contains spaces it should be quoted using
@@ -502,6 +507,7 @@ is supposedly that of user joe).
 /usr            *.local.domain(ro) @trusted(rw)
 /home/joe       pc001(rw,all_squash,anonuid=150,anongid=100)
 /pub            (ro,insecure,all_squash)
+/srv/www        -sync,rw server @trusted @external(ro)
 '''/pub/private    (noaccess)
 .fi
 .PP
@@ -515,6 +521,9 @@ under the nobody account. The
 .I insecure 
 option in this entry also allows clients with NFS implementations that
 don't use a reserved port for NFS.
+The sixth line exports a directory read-write to the machine 'server'
+as well as the `@trusted' netgroup, and read-only to netgroup `@external',
+all three mounts with the `sync' option enabled.
 ''' The last line denies all NFS clients
 '''access to the private directory.
 '''.SH CAVEATS

/* Steinar */

Signature · 10 lines
Homepage: http://www.sesse.net/

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist  -  [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
#1 lore raw
Quoted 1 line · Steinar H. Gunderson

+/srv/www -sync,rw server @trusted @external(ro)

FWIW, the - here should be a \-.

/* Steinar */

Signature · 10 lines
Homepage: http://www.sesse.net/

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist  -  [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
#2 lore raw
In reply to Steinar H. Gunderson
Quoted 1 line · On Tuesday December 26, [email protected] wrote:

[Cc-on-reply, please]

Always.... doesn't everybody?

Quoted 5 lines

Merry Christmas, to those of you who celebrate that. :-) I found some time to write a patch for a two-year-old feature request which was slowly getting more and more annoying; hope it can go into upstream, even though it won't reach our upcoming Debian release.

Yes, I like it. I've applied it.

I would like to have system-wide defaults too (as you mention). Any suggestions on syntax? Having

-sync,no_root_squash,no_subtree_check

would be syntactically consistent (which is good) but is kind-a ugly.

Thanks,

NeilBrown

Quoted 14 lines

Patch: Allow default options in /etc/exports

Implement default options in /etc/exports, to fix a long-standing wishlist bug in Debian. (The user claims the syntax matches that of OpenBSD.) This makes it possible to write "/srv/www -sync,no_subtree_check host1 host2 host3" instead of having to write (sync,no_subtree_check) over and over and over again, driving the administrator slowly mad. Such option lines can be placed anywhere on the line, and affects anything after them (I do not know if OpenBSD allows this). The patch is slightly convoluted in order to avoid triggering spurious warnings; for instance, we want "/srv/www -sync host1" to trigger a warning, but not "/srv/www -sync,no_subtree_check host1" or "/srv/www -sync host1(no_subtree_check)".

Quoted 8 lines

There was also a suggestion for a truly global (ie. per-file) option list, but this seemed like the safest bet, given that it matches that of other implementations.

Also, the man page is updated with information on the new possibilities, and an example.

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier. Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#3 lore raw
Quoted 2 lines · Neil Brown
Quoted 1 line

[Cc-on-reply, please]

Always.... doesn't everybody?

Not always. I tried subscribing to the list, but it doesn't seem like it worked, for some reason -- odd. SourceForge is not always my friend, although you'd believe it would be hard to go wrong with Mailman...

Quoted 6 lines

I would like to have system-wide defaults too (as you mention). Any suggestions on syntax? Having

-sync,no_root_squash,no_subtree_check

would be syntactically consistent (which is good) but is kind-a ugly.

My first hunch is to put it on the command line (so one could have it in /etc/default/nfs-common), but I realize that would be confusing (and possibly different across distributions, and less flexible). I'm a bit short on good syntax suggestions apart from that, really -- adding a different kind of line sounds a bit like opening a can of worms syntax-wise...

/* Steinar */

Signature · 10 lines
Homepage: http://www.sesse.net/

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
NFS maillist  -  [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
#4 lore raw
Quoted 6 lines · Steinar H. Gunderson
Quoted 5 lines · Neil Brown

I would like to have system-wide defaults too (as you mention). Any suggestions on syntax? Having

-sync,no_root_squash,no_subtree_check

Quoted 7 lines
Quoted 1 line

would be syntactically consistent (which is good) but is kind-a ugly.

My first hunch is to put it on the command line (so one could have it in /etc/default/nfs-common), but I realize that would be confusing (and possibly different across distributions, and less flexible). I'm a bit short on good syntax suggestions apart from that, really -- adding a different kind of line sounds a bit like opening a can of worms syntax-wise...

How about making the line the exact same syntax as normal lines but a pseudo export point string which is not a valid absolute pathname, something like "defaults". For example:

defaults *(rw,sync,no_subtree_check)
/foo
/bar *(insecure)

could mean the same as

/foo *(rw,sync,no_subtree_check)
/bar *(rw,sync,no_subtree_check,insecure)

There are export options which cannot usefully appear in such a defaults line (fsid certainly, perhaps also mountpoint=/somewhere).

Greg.

Signature · 11 lines
Greg (not Dave) Banks, R&D Software Engineer, SGI Australian Software Group.
I don't speak for SGI.

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
NFS maillist  -  [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
#5 lore raw
Quoted 5 lines · On Wednesday February 7, [email protected] wrote:

How about making the line the exact same syntax as normal lines but a pseudo export point string which is not a valid absolute pathname, something like "defaults". For example:

Quoted 8 lines

defaults *(rw,sync,no_subtree_check)
/foo
/bar *(insecure)

could mean the same as

/foo *(rw,sync,no_subtree_check)
/bar *(rw,sync,no_subtree_check,insecure)

Thats probably heading in the right direction.
That example you have given sets defaults for a particular client (or client set). I was imagining setting defaults for everything. I guess both could be useful.
Global defaults could be set with

defaults (rw,sync)

i.e. no client at all.
So if you have

default (rw,sync,no_subtree_check)
default @friends(no_root_squash)

/foo -ro,root_squash @friends() @enemies(all_squash)

Then what options apply to a mount of /foo from a friend? Both root_squash and no_root_squash are provided as defaults, one for the filesystem, one for the client...

I suspect the filesystem default should rule as it is textually closer.

Quoted 3 lines

There are export options which cannot usefully appear in such a defaults line (fsid certainly, perhaps also mountpoint=/somewhere).

Yes, we would need to be careful about that.

NeilBrown

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier. Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs