Your message dated Fri, 20 Jan 2017 10:14:35 +0100
with message-id <[email protected]>
and subject line Re: Bug#717604: git-buildpackage: support for --binary-arch
has caused the Debian Bug report #717604,
regarding git-buildpackage: support for --binary-arch
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
717604: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=717604
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: git-buildpackage
Version: 0.6.2
Severity: wishlist
Now that the buildds properly distinguish between arch-specific builds
and arch-independent builds and support using build-arch, it would be
nice to have a simple way of testing that such package builds work
properly while still using git-buildpackage. However, at least when
using git-buildpackage in conjunction with git-pbuilder, this doesn't
seem to work. I just get various versions of invalid flag errors.
pbuilder itself supports a --binary-arch flag that does the right things
(calls dpkg-buildpackage with -B but also doesn't install things in
Build-Depends-Indep, etc.), but that doesn't seem to make it down to
pbuilder in the right fashion currently.
I'm not sure if this is a problem with git-pbuilder or with something
else, and haven't had a chance to investigate further. Filing this bug
partly as a placeholder and partly to see if anyone else has already
looked into this in more detail.
-- System Information:
Debian Release: jessie/sid
APT prefers testing
APT policy: (990, 'testing'), (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Kernel: Linux 3.8-2-686-pae (SMP w/4 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages git-buildpackage depends on:
ii devscripts 2.13.2
ii git 1:1.7.10.4-2
ii man-db 2.6.5-2
ii python 2.7.5-2
ii python-dateutil 1.5+dfsg-0.1
ii python-pkg-resources 0.6.37-1
Versions of packages git-buildpackage recommends:
ii cowbuilder 0.70
ii pristine-tar 1.28
Versions of packages git-buildpackage suggests:
pn python-notify <none>
ii unzip 6.0-9
-- no debconf information
--- End Message ---
--- Begin Message ---
Hi Russ,
On Thu, Aug 22, 2013 at 08:56:36AM -0700, Russ Allbery wrote:
> Guido Günther <[email protected]> writes:
>
> > Since this is a option that needs to be passed to pbuilder and not to
> > dpkg-buildpackage one needs to use the --git-pbuilder-options command
> > line option:
>
> > # gbp buildpackage --git-pbuilder-options='--binary-arch'
>
> > -> Invoking pbuilder
> > forking: pbuilder build --debbuildopts --debbuildopts --binary-arch
> > --buildplace /var/scratch/pbuilder/build//cow.31843 --buildresult
> > /var/scratch/src/git-buildpackage --debbuildopts --binary-arch --no-targz
> > --internal-chrootexec chroot /var/scratch/pbuilder/build//cow.31843
> > cow-shell /var/scratch/src/git-buildpackage/git-buildpackage_0.6.2.dsc
> >
> > The current situation for prefixing gbp options with --git- , not
> > prefixing dpkg-buildpackage options and passing pbuilder options via
> > --git-pbuilder is confusing but I'm not sure how we can make it much
> > simpler.
>
> Ah! Thank you. I had missed the correct escape hatch.
Since it works (although the argument passing is a bit bizarre) I'm
closing this one. I wonder if we should do s.th. like
gbp builpackage <gbp-options> -- <debbuildopts> --- <git-pbuilder-options>
This would do away with the --git- prefix and other weird things.
and clarify the situation (not only for this case but for other cases as
well)? I've filed #851944 for this.
Cheers,
-- Guido
--- End Message ---
