On Tue, Sep 08, 2015 at 12:34:05PM -0700, Steve Ellcey wrote: > I am building qemu on an old system with python 2.4.3. The latest checkin: > > commit 9d34158a5af734e8de0b42b0a7228200c426a8d0 > Merge: 8f1ed5f bd80a8a > Author: Peter Maydell <peter.mayd...@linaro.org> > Date: Mon Sep 7 16:07:47 2015 +0100 > > Merge remote-tracking branch 'remotes/cohuck/tags/s390x-20150907' into > staging > > caused my qemu build to fail with: > > File "/scratch/gcc/nightly/src/qemu/scripts/qapi.py", line 947, in cgen > raw = re.subn("^.", indent + r'\g<0>', raw, 0, re.MULTILINE) > raw = re.subn("^.", indent + r'\g<0>', raw, 0, re.MULTILINE) > TypeError: subn() : subn() takes at most 4 arguments (5 given) > > Now the recent patch did not introduce the use of subn, it just seems > to have changed something so that I call it now whereas before I did not. > The actual use of subn was introduced with this qemu patch: > > commit 77e703b861d34bb2879f3e845482d5cf0a3a0ad1 > Author: Markus Armbruster <arm...@redhat.com> > Date: Wed Jun 24 19:27:32 2015 +0200 > > qapi: Clean up cgen() and mcgen() > > It looks like the 5th (optional) argument to subn in python was added > with python 2.7 and I am using python 2.4.3. > > Is the use of the 5th argument on subn needed? And is it intentional > that qemu cannot be built without a python 2.7 or later?
No, that is a bug. The current intention is for QEMU to be buildable with Python >= 2.6.0, in order to spport RHEL6 as a min OS version 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 :|