Bug 351387 – dev-libs/glib-2.26.1 fix for compilers with no builtin atomic

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Alan Hourihane 2011-01-11 19:23:22 UTC]

Update ebuild with this, pulled from later glib git sources.

[Comment 1 Alan Hourihane 2011-01-11 19:23:39 UTC]

Created attachment 259561 [details, diff]
atomic compiler fix

[Comment 2 Alan Hourihane 2011-01-12 21:09:33 UTC]

Oh, and this hunk in the ebuild can now be removed.

        if [[ ${CHOST} == *-mint* ]] ; then
                myconf="${myconf} --disable-threads"
        else
                myconf="${myconf} --with-threads=${mythreads}"
        fi

and --with-threads=${mythreads} can be put directly into the econf command.

Thanks.

[Comment 3 Alan Hourihane 2011-01-19 11:39:46 UTC]

ping.

[Comment 4 Raúl Porcel (RETIRED) 2011-01-22 15:32:52 UTC]

I'm stealing this bug report since it happens on Linux as well.

According the upstream bug report it should affect:
- sh4
- mips64 (big/little endian)
- vax
- hp pa-risc

In my case is sh4

[Comment 5 Pacho Ramos 2011-01-22 15:51:32 UTC]

(In reply to comment #2)
> Oh, and this hunk in the ebuild can now be removed.
> 
>         if [[ ${CHOST} == *-mint* ]] ; then
>                 myconf="${myconf} --disable-threads"
>         else
>                 myconf="${myconf} --with-threads=${mythreads}"
>         fi
> 
> and --with-threads=${mythreads} can be put directly into the econf command.
> 
> Thanks.
> 

Where this "${CHOST} == *-mint*" stuff come from? I cannot see it in neither glib ebuild :-/

I will probably try to apply the patch compile and try it on my amd64 box

[Comment 6 Fabian Groffen 2011-01-22 15:55:32 UTC]

this bug was originally reported against the prefix team, glib is in our overlay...

[Comment 7 Pacho Ramos 2011-01-23 11:29:27 UTC]

Sadly, I am suffering some problems running glib tests (http://bugs.gentoo.org/show_bug.cgi?id=352451#c5) and won't have much time for looking into these :-(, then, if any other gnome team member wants to take over this... thanks :-)

[Comment 8 Pacho Ramos 2011-01-24 13:56:38 UTC]

+  24 Jan 2011; Pacho Ramos <[e-mail]> glib-2.26.1-r1.ebuild,
+  +files/glib-2.26.1-gatomic-header.patch,
+  +files/glib-2.26.1-gdatetime-test.patch:
+  Add openbsd patch to fix build on several arches (bug #351387 by Alan
+  Hourihane and Raúl Porcel) once I confirmed it doesn't cause any test
+  failure regression. Remove test that seems to fail depending on time of day
+  as done by upstream (was causing one of the failures in bug #352451) and
+  unset GSETTINGS_BACKEND (bug #352451) until upstream takes care (upstream bug
+  #640412).
+