Thunar fails to build with debug
Enrico Tröger
enrico.troeger at uvena.de
Sun Jan 18 23:03:03 CET 2009
On Sun, 18 Jan 2009 12:00:48 -0800, "Brian J. Tarricone"
<bjt23 at cornell.edu> wrote:
>
>
>On Sun, 18 Jan 2009 20:23:55 +0100 Stephan Arts wrote:
>
>> On Sun, Jan 18, 2009 at 8:20 PM, Enrico Tröger
>> <enrico.troeger at uvena.de> wrote:
>> > On Sun, 18 Jan 2009 20:13:04 +0100, David Gustafsson
>> > <david.s.gustafsson at gmail.com> wrote:
>> >
>> >>On Sun, Jan 18, 2009 at 12:31 AM, Mike Massonnet
>> >><mmassonnet at gmail.com> wrote:
>> >>> Le Sat, 17 Jan 2009 23:29:36 +0100,
>> >>> Mike Massonnet <mmassonnet at gmail.com> a écrit :
>> >>>
>> >>>> Hi,
>> >>>>
>> >>>> I was building thunar and hit a warning. Frankly I don't
>> >>>> understand it, it's why I'm fowarding it here (and not as a bug
>> >>>> as I don't experiment it on my other box except the glib
>> >>>> versions).
>> >>> are different
>> >>>>
>> >>>> cc1: warnings being treated as errors
>> >>>> thunar-vfs-io-local.c: In function
>> >>>> '_thunar_vfs_io_local_listdir': thunar-vfs-io-local.c:647:
>> >>>> error: dereferencing type-punned pointer will break
>> >>>> strict-aliasing rules thunar-vfs-io-local.c:686: error:
>> >>>> dereferencing type-punned pointer will break strict-aliasing
>> >>>> rules
>> >>>>
>> >>>> It fails with the G_LOCK_* macros.
>> >>>>
>> >>>> This warning happens with glib-2.18.3.
>> >>>>
>> >>>> Cheers
>> >>>> Mike
>> >>> _______________________________________________
>> >>> Xfce4-dev mailing list
>> >>> Xfce4-dev at xfce.org
>> >>> http://foo-projects.org/mailman/listinfo/xfce4-dev
>> >>
>> >>Hi,
>> >>
>> >>This also happens in libexo when compiling with debug. The problem
>> >>is a known bug in glib[1].
>> >>
>> >>exo-mount-point.c: In function 'IA__exo_mount_point_list_matched':
>> >>exo-mount-point.c:561: error: dereferencing type-punned pointer
>> >>will break strict-aliasing rules
>> >>exo-mount-point.c:575: error: dereferencing type-punned pointer
>> >>will break strict-aliasing rules
>> >
>> > -Werror just should never appear in any Makefiles, regardless of
>> > debug builds or not.
>> > You just cannot know which warnings a particular system will
>> > generate. Developers who see any advantage in -Werror should
>> > specify it in their personal CFLAGS environment variable but not in
>> > general purpose Makefiles.
>> >
>> > I personally don't see any use in -Werror as I can see compiler
>> > warnings while compiling and so I have no need to break my
>> > compilation. But that's just me :).
>> >
>> >
>> > Regards,
>> > Enrico
>>
>> Perhaps we should remove -Werror from debug=full, and add a
>> debug=paranoid option ;-)
>
>No, it's fine as-is. If people don't care about warnings, they
>shouldn't build with --debug=full.
Compiling without any --debug flags from SVN is enough to get a broken
build. It would be cool if a normal build (*) from SVN would just work.
(*) normal build = Without fiddling with debug options just to get it
compiled, similar for the --enable-maintainer-mode stuff which is
necessary for certain packages.
Sure, the SVN version is not for productive use and for average users
but I think it would be not that bad if also the SVN version would
be a bit more user-friendly for those like me who want to get the
latest code and test it to give early feedback.
I'll hit the sack for now, maybe my problems with this aren't that
worse after getting some sleep :).
Regards,
Enrico
--
Get my GPG key from http://www.uvena.de/pub.asc
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://mail.xfce.org/pipermail/xfce4-dev/attachments/20090118/822f8f3b/attachment.pgp>
More information about the Xfce4-dev
mailing list