glib2.0 2.25.8-1ubuntu1 FTBFS on armel: expected 'struct va_list *' but argument is of type 'va_list'

Bug #593516 reported by Steve Langasek
12
This bug affects 2 people
Affects Status Importance Assigned to Milestone
glib2.0 (Ubuntu)
Fix Released
High
Unassigned
Maverick
Fix Released
High
Unassigned

Bug Description

glib2.0 is failing to build on armel with the following error:

/bin/bash ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I/build/buildd/glib2.0-2.25.8/glib -I.. -I/build/buildd/glib2.0-2.25.8 -DG_LOG_DOMAIN=\"GLib\" -DG_ENABLE_DEBUG -DG_DISABLE_DEPRECATED -DGLIB_COMPILATION -DPCRE_STATIC -DG_DISABLE_SINGLE_INCLUDES -pthread -g -O2 -Wall -g -O2 -MT gvariant-parser.lo -MD -MP -MF .deps/gvariant-parser.Tpo -c -o gvariant-parser.lo /build/buildd/glib2.0-2.25.8/glib/gvariant-parser.c
libtool: compile: gcc -DHAVE_CONFIG_H -I. -I/build/buildd/glib2.0-2.25.8/glib -I.. -I/build/buildd/glib2.0-2.25.8 -DG_LOG_DOMAIN=\"GLib\" -DG_ENABLE_DEBUG -DG_DISABLE_DEPRECATED -DGLIB_COMPILATION -DPCRE_STATIC -DG_DISABLE_SINGLE_INCLUDES -pthread -g -O2 -Wall -g -O2 -MT gvariant-parser.lo -MD -MP -MF .deps/gvariant-parser.Tpo -c /build/buildd/glib2.0-2.25.8/glib/gvariant-parser.c -fPIC -DPIC -o .libs/gvariant-parser.o
/build/buildd/glib2.0-2.25.8/glib/gvariant-parser.c: In function 'IA__g_variant_builder_add_parsed':
/build/buildd/glib2.0-2.25.8/glib/gvariant-parser.c:2290: error: incompatible type for argument 2 of 'IA__g_variant_new_parsed_va'
/build/buildd/glib2.0-2.25.8/glib/gvariant-parser.c:2174: note: expected 'struct va_list *' but argument is of type 'va_list'
make[5]: *** [gvariant-parser.lo] Error 1

Full build log at <http://launchpadlibrarian.net/50019388/buildlog_ubuntu-maverick-armel.glib2.0_2.25.8-1ubuntu1_FAILEDTOBUILD.txt.gz>.

The line numbers are correct, though the function names are confusing due to some symbol aliases being defined in the included galiasdef.c.

Note that this is a cross-platform bug, it just winds up as a compiler warning on archs where va_list happens to be a pointer type instead of causing a build failure - but at runtime this function is going to fail consistently.

Revision history for this message
Steve Langasek (vorlon) wrote :

Marcin, could you have a look at fixing this?

Changed in glib2.0 (Ubuntu Maverick):
status: New → Triaged
importance: Undecided → High
milestone: none → maverick-alpha-2
assignee: nobody → Marcin Juszkiewicz (hrw)
milestone: maverick-alpha-2 → none
Revision history for this message
Sebastien Bacher (seb128) wrote :

the bug has already been fixed some days ago in the GNOME git

Changed in glib2.0 (Ubuntu Maverick):
assignee: Marcin Juszkiewicz (hrw) → nobody
status: Triaged → Fix Committed
Revision history for this message
Sebastien Bacher (seb128) wrote :

I will backport the commit if we don't get a new tarball:
http://git.gnome.org/browse/glib/commit/?id=653921e17a00d161a25b21d61ff12b8d03208202

Revision history for this message
David Sugar (dyfet-deactivatedaccount) wrote :

This is actually not an armel specific bug, but effects all architectures. It's just not visible elsewhere because most architectures still effectively seem to define va_list itself as a pointer. I also have a patch for this, but if there is a pending backport, I will wait for that to appear.

Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package glib2.0 - 2.25.8-1ubuntu2

---------------
glib2.0 (2.25.8-1ubuntu2) maverick; urgency=low

  * debian/patches/90_git_fix_typo.patch:
    - git change to fix a typo, should fix the build on armel as well
      (lp: #593516)
  * debian/libglib2.0-dev.install:
    - include the python gdb macro there as well
 -- Sebastien Bacher <email address hidden> Tue, 15 Jun 2010 14:49:57 +0200

Changed in glib2.0 (Ubuntu Maverick):
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.