[libvirt] Status of gnulib/local/m4/ssize_t.m4.diff ?

Way back in 2012, we added the ssize_t.m4.diff file because we didn't want to risk pulling in a gnulib update at time of release. ommit 1012dc2933378640ac0ab1025a4fec2fc8c6f84c Author: Eric Blake <eblake@redhat.com> Date: Fri Mar 30 10:28:03 2012 -0600 build: fix mingw ssize_t, syntax check We are so close to a release that we don't want to pull in a gnulib submodule update and risk regressions, since there has been a lot of other gnulib churn upstream. However, there are a couple of gnulib issues that are worth fixing in isolation, by applying local patches to gnulib. There was an upstream gnulib bug in maint.mk that rendered most of our syntax checks ineffective (and fixing it flushed out a minor bug in our code): https://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00194.html There is still an upstream bug where gnulib uses the wrong type for ssize_t on mingw; we need the fix now even though it has not yet been accepted into gnulib: https://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00188.html Fast forward 4 years, and I'm wondering why we stsill have a ssize_t.m4.diff file that we're updating. It doesn't look like gnulib ever merged it upstream. Can we kill it from libvirt too yet, or does gnulib really need to accept it ? Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|

On 07/04/2016 08:17 AM, Daniel P. Berrange wrote:
Way back in 2012, we added the ssize_t.m4.diff file because we didn't want to risk pulling in a gnulib update at time of release.
There is still an upstream bug where gnulib uses the wrong type for ssize_t on mingw; we need the fix now even though it has not yet been accepted into gnulib: https://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00188.html
Fast forward 4 years, and I'm wondering why we stsill have a ssize_t.m4.diff file that we're updating. It doesn't look like gnulib ever merged it upstream. Can we kill it from libvirt too yet, or does gnulib really need to accept it ?
I'll investigate what happens if we drop it, and have a decision later today on the best course forward. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org

On 07/04/2016 08:17 AM, Daniel P. Berrange wrote:
Way back in 2012, we added the ssize_t.m4.diff file because we didn't want to risk pulling in a gnulib update at time of release.
ommit 1012dc2933378640ac0ab1025a4fec2fc8c6f84c Author: Eric Blake <eblake@redhat.com> Date: Fri Mar 30 10:28:03 2012 -0600
There is still an upstream bug where gnulib uses the wrong type for ssize_t on mingw; we need the fix now even though it has not yet been accepted into gnulib: https://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00188.html
Fast forward 4 years, and I'm wondering why we stsill have a ssize_t.m4.diff file that we're updating. It doesn't look like gnulib ever merged it upstream. Can we kill it from libvirt too yet, or does gnulib really need to accept it ?
I think this is the answer: https://lists.gnu.org/archive/html/bug-gnulib/2012-04/msg00021.html The bug is only present in the (now-defunct) mingw (the 32-bit-only) project; the mingw64 32- and 64-bit headers are correct, so it no longer buys us anything to work around it. I'm testing a patch now. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
participants (2)
-
Daniel P. Berrange
-
Eric Blake