Fedora 20 Final blocker bug status report (ACTIONS REQUIRED: karma from QA, fixes from devs)

Thu Nov 28 02:25:31 UTC 2013

Hi, folks. We have just over one week to the Fedora 20 Go/No-Go meeting,
and a big happy pile of blocker bugs - everyone's having fun! In order
to be able to complete testing in time for the go/no-go meeting
comprehensively and comfortably, we would need a build with all these
bugs addressed by the end of next Tuesday (2013-12-03). To have a chance
at just making it by the skin of our teeth, we'd need a build with these
bugs addressed by the end of next Wednesday (2013-12-04). Anything later
than that means the release will be delayed.

Here's my round-up of the status of all the currently open blockers, and
the action needed to move forwards on all of them. If you're personally
CCed on this mail, it means you have an outstanding blocker bug assigned
to you or you are clearly involved in working on one: search this mail
for your name or email identifier to find the bits relevant to you.

The full list of proposed and accepted blocker and freeze exception bugs
can always be viewed at
http://qa.fedoraproject.org/blockerbugs/current .

QA: karma required
------------------

QA folks, we need karma on the following updates. The blocker bug fixes
have all been confirmed, but we need a few more people to check and
provide +1 karma before these can go stable:

* https://admin.fedoraproject.org/updates/FEDORA-2013-21928 (anaconda,
blivet, pykickstart: +1 if TC3 install is basically OK)

* https://admin.fedoraproject.org/updates/FEDORA-2013-22183 (sqlite:
check things that use sqlite are OK, the specific bugfix to check is
https://bugzilla.redhat.com/show_bug.cgi?id=1034714 )

* https://admin.fedoraproject.org/updates/FEDORA-2013-21347
(system-config-services, the bug is
https://bugzilla.redhat.com/show_bug.cgi?id=790339 : +1 if it runs and
works)

It would also be good to get karma on the following updates for freeze
exception issues:

* https://admin.fedoraproject.org/updates/glib2-2.38.2-2.fc20 for
https://bugzilla.redhat.com/show_bug.cgi?id=1035423 "Use after free
causes critical warnings in g-s-d, potential crasher"

* https://admin.fedoraproject.org/updates/FEDORA-2013-21747 for
https://bugzilla.redhat.com/show_bug.cgi?id=1035066 "Volume gets
restored to 100% after each knotify event"

* https://admin.fedoraproject.org/updates/FEDORA-2013-22176
(kde-settings) for https://bugzilla.redhat.com/show_bug.cgi?id=1034828
"Enable kwallet Silently Create Initial Wallet=true feature by default"

* https://admin.fedoraproject.org/updates/FEDORA-2013-22263
(xorg-x11-drv-qxl) for
https://bugzilla.redhat.com/show_bug.cgi?id=1020393 "Very sluggish
graphical performance in qxl/spice guests with spice streaming enabled
(default configuration)"

* https://admin.fedoraproject.org/updates/FEDORA-2013-22185 (taglib) for
https://bugzilla.redhat.com/show_bug.cgi?id=1026080 "[abrt]
amarok-2.8.0-3.fc19.1: getDisplay: Process /usr/bin/amarok was killed by
signal 11 (SIGSEGV)"

Developers: fixes required
--------------------------

These remaining outstanding blockers are mostly still unaddressed and
require fixes from developers.

anaconda/python-blivet/installer stack
--------------------------------------

* https://bugzilla.redhat.com/show_bug.cgi?id=1020974 "incorrectly
treats a disk with partially corrupt GPT as having no partition at all"
- this one is clearly defined and simply awaiting developer attention.
Assigned to bcl.

* https://bugzilla.redhat.com/show_bug.cgi?id=1027947 "Cannot change a
partition's size, then return it to the original size" - this was opened
for a crasher. The crash was fixed, but the behaviour is still incorrect
and requires further developer attention. Assigned to dlehman.

* https://bugzilla.redhat.com/show_bug.cgi?id=1027965 "CreateException:
Can't have a partition outside the disk!" - this one again seems clearly
defined and waiting on developer attention. Not yet assigned to a
specific developer.

* https://bugzilla.redhat.com/show_bug.cgi?id=1028367 "Invalid resize
operation crashes" - very similar to the previous bug, but different
crash. Again waiting for a developer.

* https://bugzilla.redhat.com/show_bug.cgi?id=1035462 "
initial-setup-text only displays timezone spoke" - seems like this
regressed in Final TC3, for some reason. Critical issue for ARM text
deployments. Assigned to vpodzime, waiting for his attention.

systemd
-------

* https://bugzilla.redhat.com/show_bug.cgi?id=1006386 "Journal flushing
often slow, can prevent system booting correctly" - there's active work
going on on this one, but we're still short a definite fix. Michal
Schmidt and Lennart seem to be working on it.

* https://bugzilla.redhat.com/show_bug.cgi?id=1026860 "Instantiated
service is not run, it stays in inactive state (and systemd debug log
does not state why)" - this is a somewhat complex issue which appears to
sometimes (and sometimes not) affect activation of LVM volumes on top of
mdraid. Again, devs (Peter Rajnoha and Lennart) seem to be actively
working on it, but pinning it down is proving tricky.

others
------

* https://bugzilla.redhat.com/show_bug.cgi?id=864198 "grubby fatal error
updating grub.cfg when /boot is btrfs" - grubby - the problem here is
clearly defined. We are waiting on pjones to decide how to move forward
with this. Gene Czarcinski has posted a proposed change to
new-kernel-pkg to use grub2-mkconfig instead of grubby for updating
bootloader config, which would address this bug but be a late, major
change in our behaviour. The alternative would be to fix the problem in
grubby. Either way, this is waiting on Peter as the bootloader guy.

* https://bugzilla.redhat.com/show_bug.cgi?id=1032921 "KDE f20 TC2
x86_64 fails to shutdown from menu bar" - kde-workspace - this may well
be a dupe of an earlier report,
https://bugzilla.redhat.com/show_bug.cgi?id=983110 . It is possible that
switching back to KDM by default may help with this. Otherwise, we're
waiting on the KDE developers to investigate and come up with a fix.
Both bugs are assigned to Ngo Than, CCing Rex Dieter and Kevin Kofler as
our lead KDE maintainers.

* https://bugzilla.redhat.com/show_bug.cgi?id=969524 "qml-based systray
plasma widgets unclickable on arm" - this issue is under active
investigation, but still waiting for a fix. Again, assigned to Ngo,
Kevin and Rex are actively involved in looking for the fix. Rex posted
this nugget to IRC earlier: "<rdieter> bconoboy: it is arm specific, we
*suspect* there's still some qt qreal != float assumptions somewhere"

* https://bugzilla.redhat.com/show_bug.cgi?id=1000893 "Desktop Live is
oversized (larger than 1 GB)" - we got it undersize from Beta through
Final TC2, but a combination of things pushed it slightly over in TC3.
Matthias and I are both looking at this and trying to figure out the
best fix. It's always possible to 'fix' these size issues somehow, just
a question of how to do so as optimally as possible.

Workload notes: not actually too bad. anaconda has the most issues as
always, but five to fix with a week to go isn't awful. Four of them are
in storage, which is dlehman territory, so he is probably the most
heavily loaded person: anything we can do to lighten his load will help.
After that, systemd and KDE are the identifiable areas with the most
bugs, at two each.

Misc: I haven't filed bureaucracy bugs for them yet, but we do still
need to make sure the final artwork, release notes, fedora-release (and
generic-release) and spin-kickstarts are all in place prior to ship.
Jreznik may be best placed to try and co-ordinate that effort.

Thanks, everyone!
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net