Your message dated Sun, 30 Nov 2014 12:42:09 +0100
with message-id <20141130114209.gb18...@ugent.be>
and subject line Re: Bug#771075: unblock: systemd/215-7
has caused the Debian Bug report #771075,
regarding unblock: systemd/215-7
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 ow...@bugs.debian.org
immediately.)
--
771075: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=771075
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Package: release.debian.org
Severity: normal
User: release.debian....@packages.debian.org
Usertags: unblock
systemd 215-6 made it into testing today, so I'd like to ask for
landing another set of small fixes.
Note that I did *not* upload -7 to unstable yet, I'd like to get the
release team's pre-approval to avoid reverting changes and messing up
the package history.
I attach the full debdiff between 215-6 and current git (which is to
become 215-7, modulo the dch -r changes); alternatively you might want
to review the individual changes since the debian/215-6 tag on
http://anonscm.debian.org/cgit/pkg-systemd/systemd.git/log/
Annotated changelog:
| [ Martin Pitt ]
| * Add myself to Uploaders.
| * Add boot-and-services autopkgtest: Check booting with systemd-sysv and
| that the most crucial services behave as expected.
| * logind autopkgtest: Fix stderr output in waiting loop for scsi_debug.
| * Add nspawn test to boot-and-services autopkgtest.
This just affects autopkgtests, thus nothing at runtime. We can't run
them on ci.debian.net yet as they require better isolation than
schroot (so they will continue to be skipped), but we do run them on
Ubuntu and thus can catch regressions efficiently.
| * Make systemd-nspawn@.service work out of the box: (Closes: #770275)
| - Pre-create /var/lib/container with a secure mode (0700) via tmpfiles.d.
| - Add new try-{guest,host} modes for --link-journal to silently skip
| setting up the guest journal if the host has no persistant journal.
| - Extend boot-and-services autopkgtest to cover systemd-nspawn@.service.
This is a comparatively big fix in LoC, although conceptually the
patch is quite simple. It is covered by an autopkgtest now (which
reproduces the original problem). I'd say it's low risk since this is
an entirely new feature in jessie compared to wheezy, and it is
currently completely broken with our default install.
| * Cherry-pick upstream patch to fix SELinux unit access check (regression
| in 215).
Trivial fix for an obvious typo in the original commit:
http://cgit.freedesktop.org/systemd/systemd/commit/?id=ebcf1f97de4
(Note how it accidentally changes _u->source_path to
_unit->fragment_path, this patch fixes it back). It doesn't affect our
default install, but people who enable SELinux.
| [ Jon Severinsson]
| * Add /run/shm -> /dev/shm symlink in debian/tmpfiles.d/debian.conf. This
| avoids breakage in Jessie for packages which still refer to /run/shm, and
| while https://wiki.debian.org/ReleaseGoals/RunDirectory is still official.
| (LP: #1320534, Closes: #674755).
We've had this in Ubuntu for some time, but kind of forgot to apply it
in Debian. Originally the plan was to fix everything in Debian which
refers to /run/shm/ and use the upstream glibc location /dev/shm/
instead, but with the freeze now is not the time any more to do that.
So this is a safety measure to avoid running into regressions like the
one Keith Packard mentioned in libxshmfence in the Debian bug.
There's one more thing which I'd like to land, but didn't commit yet:
the patch set in https://bugs.debian.org/769734 which fixes a
regression under systemd with the "discard" mount option for swap
partitions. The patches are backported from upstream, and Aurelien
Jarno tested them in Debian with our version. Not trimming swap
partitions when you expect it causes quite a noticeable performance
degradation. However, the patches aren't trivial, so I'd rather only
apply them once the RT says they are ok for the freeze.
If this looks fine to you, I'll do the -7 upload to unstable ASAP,
with or without the discard fix above.
Thanks for considering,
Martin
unblock systemd/215-7
--
Martin Pitt | http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
--- End Message ---
--- Begin Message ---
Hi,
On Sun, Nov 30, 2014 at 11:25:46AM +0100, Martin Pitt wrote:
> Niels Thykier [2014-11-29 21:02 +0100]:
> > Martin, I see #771287 appeared in the mean time suggesting that 215-7
> > introduces a regression.
>
> This has been there forever, and isn't a regression at all. It's a
> workaround for packaged display managers which don't ship a systemd
> unit yet. As the release team apparently nack'ed adding units
> post-freeze, we filed that to track this workaround in systemd itself.
>
> > On the other hand, if 215-6 is also affected by #771287, please
> > update the "found" version of the bug. In this case, please remove
> > the "moreinfo" tag once you have fixed the bug metadata.
I added the unblock-udeb (based on the ack from kibi). Niels already added the
unblock.
Cheers,
Ivo
--- End Message ---
