Skip to contentSkip to conversation

autofsng 0.3.1

announce Mike Waychison 3 participants 4 messages Nov 23, 2004 – Dec 7, 2004

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

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello All,

I'm pleased to announce the release of AutofsNG 0.3.1.

This automounter is intended to be 100% interoperable with autofs implementations on other Unix platforms. As such, it supports some feature that the current Linux automounters do not:

- - Direct Mounts (*)
- - /net (--hosts access) (*)
- - Lazy mounting of hierarchal / multimounts
- - Lazy unmounting of hierarchal / multimounts
- - Browsing (*)

(*) automount 4.1.3 supports these features but with limitations that impact interoperability.

Of course, autofsng also supports the usual indirect map support available elsewhere.

Maps are supported from:

- - flat files
- - executable maps
- - nis maps
- - nis+ maps
- - ldap maps (in not one, but 3! different flavours)
- - hesiod (dns) filsys namespace

Of course, all this name service stuff is handled by /etc/nsswitch.conf logic.

Also, as an added bonus, autofsng also supports the Linux specific 'namespace' paradigm (man clone(2)).

This release marks the first where we actually posted tarballs of the userspace bits and kernel patch in one place (at the same time as the announcement. They can be found at:

ftp://ftp-eng.cobalt.com/pub/users/ssmith/autofsng/0.3.1/

Also, we have bk trees at:

http://autofsng.bkbits.net

(ignore the -old tree)

There are a couple known issues with this release, all of which will get some attention in coming releases:

- - Default mount options in auto.master are currently ignored.
- - Default timeout / timeout on the command line are ignored (all mounts timeout after 10 seconds).
- - The parser may be a little too strict for some maps that use non-nfs(ipv4) mounts.
- - The first process to access an automounted filesystem that hangs is left in interruptible sleep.
- - Many applications are broken because they don't understand what a direct mount is (and thus trample into them).
- - The kernel code is hard-coded to use '/sbin/autofsng' to perform the mounts. (installing it anywhere else won't work).

- From the README:

Installation
- ------------

Setting up autofsng requires two main pieces of code to be built and set up on your machine, the kernel code and the userland bits.

Once you have the autofsng bits in your kernel tree, you will need to enabled CONFIG_AUTOFSNG_FS in the build. If set to 'y', then autofsng will be built into the kernel proper and provide the pseudo filesystem 'autofsng'. If built as a module, the module will be called 'autofsng' and it will provide the 'autofsng' filesystem.

Note that there currently exists two other autofs implementations in the kernel. CONFIG_AUTOFS_FS, which is the autofs3 implementation is no longer actively maintained and the favourable replacement for it has become autofs4, specified by CONFIG_AUTOFS4_FS. Both of these filesystems
are called 'autofs', which is why we call ourselves 'autofsng'.

You will need to run autoconf on the resulting pull.

To build the userspace tools , please run:

configure

to configure the system. See README.options for options that you can give configure.

After configuring, you can:

make
make install

Enjoy :)

- --
Mike Waychison
Sun Microsystems, Inc.
1 (650) 352-5299 voice
1 (416) 202-8336 voice

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTICE: The opinions expressed in this email are held by me, and may not represent the views of Sun Microsystems, Inc.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBopgJdQs4kOxk3/MRAnURAJ9lA8UnkQMqsEy55B+LshmGR+ARzACdGue5 TRSCaT3nZX9HpGRvbW1JxK0=
=oktk
-----END PGP SIGNATURE-----

maintainer
#1 lore raw
Quoted 11 lines · Mike Waychison

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello All,

I'm pleased to announce the release of AutofsNG 0.3.1.

This automounter is intended to be 100% interoperable with autofs implementations on other Unix platforms. As such, it supports some feature that the current Linux automounters do not:

Quoted 8 lines

- - Direct Mounts (*)
- - /net (--hosts access) (*)
- - Lazy mounting of hierarchal / multimounts
- - Lazy unmounting of hierarchal / multimounts
- - Browsing (*)

(*) automount 4.1.3 supports these features but with limitations that impact interoperability.

Yes. And in 4.x.x ...

Lazy multimounts would (will if ever) be extrordinarily difficult to implement.

Direct mount rework is likely to cause 2.4 kernels to choke! But then autofsng is 2.6+ specific also.

Ian

#2 lore raw
In reply to Mike Waychison
Quoted 1 line · Mike Waychison

I'm pleased to announce the release of AutofsNG 0.3.1.

Has anyone successfully built on an amd64 host?

On a 64bit, sles9 host, I get:

make[1]: Entering directory `/local/mnt/workspace/builds/autofsng-0.3.1/modules'
gcc -shared -O3 -fomit-frame-pointer -Wall -I../include -fpic -DAUTOFS_LIB_DIR=\"/lib/autofsng\" -o map_hosts.so map_hosts.c -lrpcsvc /usr/lib64/gcc-lib/x86_64-suse-linux/3.3.3/../../../../x86_64-suse-linux/bin/ld:
/usr/lib64/gcc-lib/x86_64-suse-linux/3.3.3/../../../../lib64/librpcsvc.a(xmount.o):
relocation R_X86_64_32 can not be used when making a shared object; recompile with -fPIC /usr/lib64/gcc-lib/x86_64-suse-linux/3.3.3/../../../../lib64/librpcsvc.a:
could not read symbols: Bad value
collect2: ld returned 1 exit status
make[1]: *** [map_hosts.so] Error 1

Tried changing flags in Makefile from -fpic to -fPIC and it didn't help.

Compile works fine on a i686 Suse 9.2 install.

Signature · 4 lines
Mike Marion-Unix SysAdmin/Staff Engineer-http://www.qualcomm.com
Bob: "Ok, I'm goin'... but a swift kick in the ass isn't the solution to
everything you know."
Red: "I gotta disagree with you there Bob!" -- That 70's show
#3 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 23 lines · [email protected] wrote:
Quoted 1 line · Mike Waychison

I'm pleased to announce the release of AutofsNG 0.3.1.

Has anyone successfully built on an amd64 host?

On a 64bit, sles9 host, I get:

make[1]: Entering directory `/local/mnt/workspace/builds/autofsng-0.3.1/modules'
gcc -shared -O3 -fomit-frame-pointer -Wall -I../include -fpic -DAUTOFS_LIB_DIR=\"/lib/autofsng\" -o map_hosts.so map_hosts.c -lrpcsvc /usr/lib64/gcc-lib/x86_64-suse-linux/3.3.3/../../../../x86_64-suse-linux/bin/ld:
/usr/lib64/gcc-lib/x86_64-suse-linux/3.3.3/../../../../lib64/librpcsvc.a(xmount.o):
relocation R_X86_64_32 can not be used when making a shared object; recompile with -fPIC /usr/lib64/gcc-lib/x86_64-suse-linux/3.3.3/../../../../lib64/librpcsvc.a:
could not read symbols: Bad value
collect2: ld returned 1 exit status
make[1]: *** [map_hosts.so] Error 1

Quoted 4 lines

Tried changing flags in Makefile from -fpic to -fPIC and it didn't help.

Compile works fine on a i686 Suse 9.2 install.

I've attached a patch I will push if it works for you. What it does is generate the xdr routines in the autofsng build instead of trying to link in librpcsvc.

The problem you are seeing is probably due to that bit of glibc not being compiled with -fPIC, which means the .a in question can't be linked into a .so on x86-64.

Another heads up for trying out autofsng on x86-64: I haven't yet added the system call table entry for sys_mountfd. If anyone gets to it before I do, patches are welcome ;)

- --
Mike Waychison
Sun Microsystems, Inc.
1 (650) 352-5299 voice
1 (416) 202-8336 voice

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTICE: The opinions expressed in this email are held by me, and may not represent the views of Sun Microsystems, Inc.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBtUHvdQs4kOxk3/MRAjZRAKCBbQ1SGgv0GYVy+vFQ06+2s79NxACdFgsG RXrzAv53caHfRHXkI0PQWBM=
=xr9S
-----END PGP SIGNATURE-----

# This is a BitKeeper generated diff -Nru style patch. #
# ChangeSet

#   2004/12/06 21:59:29-08:00 [email protected] 
#   This patch adds mount service xdr generation to the autofsng build.  This is
#   done because we can't pull xdr_exports reliably on some platforms.  For
#   instance, on sles9 x86-64 librpcsvr is built -fpic, which means we can't link it
#   into a .so.

#
# autofsng/modules/Makefile
# 2004/12/06 21:57:02-08:00 [email protected] +10 -3
# - Add targets for mount_xdr and use it instead of trying to link in librpcsvr.a
#

# autofsng/modules/mount.x
#   2004/12/06 21:43:07-08:00 [email protected] +0 -0
#   Rename: autofsng/modules/sunrpc/mount.x -> autofsng/modules/mount.x
# 
# autofsng/modules/sunrpc/mount.x
#   2004/12/06 21:34:56-08:00 [email protected] +0 -0
#   Rename: autofsng/modules/mount.x -> autofsng/modules/sunrpc/mount.x
# 
# autofsng/modules/mount.x
#   2004/12/06 21:32:35-08:00 [email protected] +161 -0
# 
# autofsng/modules/mount.x
#   2004/12/06 21:32:35-08:00 [email protected] +0 -0
#   BitKeeper file /home/autofsng-dev/autofsng/modules/mount.x
# 
diff -Nru a/autofsng/modules/Makefile b/autofsng/modules/Makefile
--- a/autofsng/modules/Makefile	2004-12-06 22:00:35 -08:00
+++ b/autofsng/modules/Makefile	2004-12-06 22:00:35 -08:00
@@ -39,7 +39,7 @@
 all: $(MODS)

clean:
- rm -f *.o *.s *.so
+ rm -f *.o *.s *.so mount.h mount_xdr.c

map_hesiod.so: map_hesiod.c
$(CC) $(SOLDFLAGS) $(CFLAGS) $(HESIOD_FLAGS) -o map_hesiod.so \
@@ -50,8 +50,15 @@
$(CC) $(SOLDFLAGS) $(CFLAGS) -o map_ldap.so map_ldap.c -lldap $(STRIP) map_ldap.so

-map_hosts.so: map_hosts.c
- $(CC) $(SOLDFLAGS) $(CFLAGS) -o map_hosts.so map_hosts.c -lrpcsvc
+mount_xdr.o: CFLAGS+=-I.
+mount_xdr.c: mount.x mount.h
+ rm -f mount_xdr.c; rpcgen -c mount.x | grep -v '^ register int32_t \*buf;$$' > mount_xdr.c
+
+mount.h: mount.x
+ rm -f mount.h; rpcgen -o mount.h -h mount.x
+
+map_hosts.so: map_hosts.c mount_xdr.o
+ $(CC) $(SOLDFLAGS) $(CFLAGS) -o map_hosts.so map_hosts.c mount_xdr.o
$(STRIP) map_hosts.so

map_files.so: map_files.c

autofsng/modules/mount.x new file +161 -0
diff -Nru a/autofsng/modules/mount.x b/autofsng/modules/mount.x
@@ -0,0 +1,161 @@
+/* @(#)mount.x	2.1 88/08/01 4.0 RPCSRC */
+/* @(#)mount.x 1.2 87/09/18 Copyr 1987 Sun Micro */
+
+/*
+ * Sun RPC is a product of Sun Microsystems, Inc. and is provided for
+ * unrestricted use provided that this legend is included on all tape
+ * media and as a part of the software program in whole or part.  Users
+ * may copy or modify Sun RPC without charge, but are not authorized
+ * to license or distribute it to anyone else except as part of a product or
+ * program developed by the user.
+ *
+ * SUN RPC IS PROVIDED AS IS WITH NO WARRANTIES OF ANY KIND INCLUDING THE
+ * WARRANTIES OF DESIGN, MERCHANTIBILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE, OR ARISING FROM A COURSE OF DEALING, USAGE OR TRADE PRACTICE.
+ *
+ * Sun RPC is provided with no support and without any obligation on the
+ * part of Sun Microsystems, Inc. to assist in its use, correction,
+ * modification or enhancement.
+ *
+ * SUN MICROSYSTEMS, INC. SHALL HAVE NO LIABILITY WITH RESPECT TO THE
+ * INFRINGEMENT OF COPYRIGHTS, TRADE SECRETS OR ANY PATENTS BY SUN RPC
+ * OR ANY PART THEREOF.
+ *
+ * In no event will Sun Microsystems, Inc. be liable for any lost revenue
+ * or profits or other special, indirect and consequential damages, even if
+ * Sun has been advised of the possibility of such damages.
+ *
+ * Sun Microsystems, Inc.
+ * 2550 Garcia Avenue
+ * Mountain View, California  94043
+ */
+
+/*
+ * Protocol description for the mount program
+ */
+
+
+const MNTPATHLEN = 1024;	/* maximum bytes in a pathname argument */
+const MNTNAMLEN = 255;		/* maximum bytes in a name argument */
+const FHSIZE = 32;		/* size in bytes of a file handle */
+
+/*
+ * The fhandle is the file handle that the server passes to the client.
+ * All file operations are done using the file handles to refer to a file
+ * or a directory. The file handle can contain whatever information the
+ * server needs to distinguish an individual file.
+ */
+typedef opaque fhandle[FHSIZE];
+
+/*
+ * If a status of zero is returned, the call completed successfully, and
+ * a file handle for the directory follows. A non-zero status indicates
+ * some sort of error. The status corresponds with UNIX error numbers.
+ */
+union fhstatus switch (unsigned fhs_status) {
+case 0:
+	fhandle fhs_fhandle;
+default:
+	void;
+};
+
+/*
+ * The type dirpath is the pathname of a directory
+ */
+typedef string dirpath<MNTPATHLEN>;
+
+/*
+ * The type name is used for arbitrary names (hostnames, groupnames)
+ */
+typedef string name<MNTNAMLEN>;
+
+/*
+ * A list of who has what mounted
+ */
+typedef struct mountbody *mountlist;
+struct mountbody {
+	name ml_hostname;
+	dirpath ml_directory;
+	mountlist ml_next;
+};
+
+/*
+ * A list of netgroups
+ */
+typedef struct groupnode *groups;
+struct groupnode {
+	name gr_name;
+	groups gr_next;
+};
+
+/*
+ * A list of what is exported and to whom
+ */
+typedef struct exportnode *exports;
+struct exportnode {
+	dirpath ex_dir;
+	groups ex_groups;
+	exports ex_next;
+};
+
+program MOUNTPROG {
+	/*
+	 * Version one of the mount protocol communicates with version two
+	 * of the NFS protocol. The only connecting point is the fhandle
+	 * structure, which is the same for both protocols.
+	 */
+	version MOUNTVERS {
+		/*
+		 * Does no work. It is made available in all RPC services
+		 * to allow server response testing and timing
+		 */
+		void
+		MOUNTPROC_NULL(void) = 0;
+
+		/*
+		 * If fhs_status is 0, then fhs_fhandle contains the
+	 	 * file handle for the directory. This file handle may
+		 * be used in the NFS protocol. This procedure also adds
+		 * a new entry to the mount list for this client mounting
+		 * the directory.
+		 * Unix authentication required.
+		 */
+		fhstatus
+		MOUNTPROC_MNT(dirpath) = 1;
+
+		/*
+		 * Returns the list of remotely mounted filesystems. The
+		 * mountlist contains one entry for each hostname and
+		 * directory pair.
+		 */
+		mountlist
+		MOUNTPROC_DUMP(void) = 2;
+
+		/*
+		 * Removes the mount list entry for the directory
+		 * Unix authentication required.
+		 */
+		void
+		MOUNTPROC_UMNT(dirpath) = 3;
+
+		/*
+		 * Removes all of the mount list entries for this client
+		 * Unix authentication required.
+		 */
+		void
+		MOUNTPROC_UMNTALL(void) = 4;
+
+		/*
+		 * Returns a list of all the exported filesystems, and which
+		 * machines are allowed to import it.
+		 */
+		exports
+		MOUNTPROC_EXPORT(void)  = 5;
+
+		/*
+		 * Identical to MOUNTPROC_EXPORT above
+		 */
+		exports
+		MOUNTPROC_EXPORTALL(void) = 6;
+	} = 1;
+} = 100005;

_______________________________________________
autofs mailing list
[email protected]
http://linux.kernel.org/mailman/listinfo/autofs