autofs-5.1.1

[Fernando de Oliveira]

​https://www.kernel.org/pub/linux/daemons/autofs/v5/autofs-5.1.1.tar.xz

​https://www.kernel.org/pub/linux/daemons/autofs/v5/autofs-5.1.1.tar.sign

​https://lwn.net/Articles/641188/

autofs 5.1.1 release
From: 	 	Ian Kent <raven@themaw.net>
To: 	 	autofs@vger.kernel.org
Subject: 	 	[ANNOUNCE] autofs 5.1.1 release
Date: 	 	Tue, 21 Apr 2015 18:20:04 +0800
Message-ID: 	 	<1429611604.2694.1.camel@pluto.fritz.box>
Cc: 	 	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
Archive-link: 	 	Article, Thread

Hi all,

The thing to watch out for in this release is a change made to program
map execution environments. The standard environment added at program
map execution introduced a security problem when interpreted languages
like python were used. By default, a prefix is now added to these names
to avoid the problem and, for those that wish to force the use of
standard names, a configuration option has been added to do that.

...

Here are the entries from the CHANGELOG which outline the updates:

21/04/2015 autofs-5.1.1
=======================
- fix compile error in defaults.c.
- add serialization to sasl init.
- dont allocate dev_ctl_ops too early.
- fix incorrect round robin host detection.
- fix race accessing qdn in get_query_dn().
- fix leak in cache_push_mapent().
- fix config entry read buffer not checked.
- fix FILE pointer check in defaults_read_config().
- fix memory leak in conf_amd_get_log_options().
- fix signed comparison in inet_fill_net().
- fix buffer size checks in get_network_proximity().
- fix leak in get_network_proximity().
- fix buffer size checks in merge_options().
- check amd lex buffer len before copy.
- add return check in ldap check_map_indirect().
- check host macro is set before use.
- check options length before use in parse_amd.c.
- fix some out of order evaluations in parse_amd.c.
- fix copy and paste error in dup_defaults_entry().
- fix leak in parse_mount().
- add mutex call return check in defaults.c.
- force disable browse mode for amd format maps.
- fix hosts map options check in lookup_amd_instance().
- fix memory leak in create_client().
- fix memory leak in get_exports().
- fix memory leak in get_defaults_entry().
- fix out of order clearing of options buffer.
- fix reset amd lexer scan buffer.
- ignore multiple commas in options strings.
- fix typo in flagdir configure option.
- clarify multiple mounts description.
- gaurd against incorrect umount return.
- update man page autofs(8) for systemd.
- dont pass sloppy option for other than nfs mounts.
- make service want network-online.
- fix fix master map type check.
- init qdn before use in get_query_dn().
- fix typo in update_hosts_mounts().
- fix hosts map update on reload.
- make negative cache update consistent for all lookup modules.
- ensure negative cache isn't updated on remount.
- dont add wildcard to negative cache.
- add a prefix to program map stdvars.
- add config option to force use of program map stdvars.
- fix incorrect check in parse_mount().
- handle duplicates in multi mounts.
- revert special case cifs escapes.
- fix map option parsing for 'strictatime'.
- fix showmount search in auto.net.
- remove obsolete comment in auto.net.
- fix macro usage in lookup_program.c.
- fix gcc5 complaints.
- remove unused offset handling code.
- fix mount as you go offset selection.
- link daemon with pthread library (Debian patch).
- manpage corrections (Debian patch).
- fix manpages hyphenation (Debian patch).

[Fernando de Oliveira]

Taking this, but will modify tags (lfs and gcc if any) for built only. If you object, please take it or change tag afterwards.

[Fernando de Oliveira]

Replying to fo:

If you object, please take it or change tag afterwards.

Or tell me that it works (if compiler is gcc5, this info is also important) then I will modify the book.

[bdubbs@…]

Please do not close the ticket when you are done and reassign it to me. I'll be able to test it next week.

[Fernando de Oliveira]

Thanks, Bruce.

Fixed at 15937. Left open to be tagged as checked for lfs77 and gcc5.

[bdubbs@…]

Checked both for gcc5 and LFS svn-20150523.