Fedora 17 Alpha status: open blockers, karma requests and blocker/NTH vote requests

[Adam Williamson]

Hey, folks. Here's another Alpha status report, focusing on open
blocker/NTH issues and updates that need karma.

As always, https://fedoraproject.org/wiki/Current_Release_Blockers lists
open blockers more or less in real time.

Here's a quick summary of requested actions for the impatient:

* anyone - please provide karma for:
https://admin.fedoraproject.org/updates/fedora-logos-17.0.0-1.fc17
https://admin.fedoraproject.org/updates/FEDORA-2012-1771/ceph-0.41-1.fc17...
https://admin.fedoraproject.org/updates/FEDORA-2012-1953/selinux-policy-3...
https://admin.fedoraproject.org/updates/FEDORA-2012-1948/kernel-3.3.0-0.r...
https://admin.fedoraproject.org/updates/FEDORA-2012-1766/dracut-016-1.fc17
https://admin.fedoraproject.org/updates/FEDORA-2012-1783/metacity-2.34.2-...

* people who usually join blocker reviews - please vote on blocker
status of #794899 and NTH status of #794690, #794886 and #783712

* mschwendt - can you double-check your -1 on
https://admin.fedoraproject.org/updates/FEDORA-2012-1783/metacity-2.34.2-... and adjust
it if the issue isn't a regression and doesn't cause any major problems?

* anaconda team - please push fix for #787461 and do a new anaconda
build, and keep an eye on #787744

* josef skladanka - please update #787744

* Jan Vcelak - if you provide a fix for
https://bugzilla.redhat.com/show_bug.cgi?id=771484 we can pull it into
the next compose

* desktop team - please look at #794690, decide which way you'd like to
fix it, and implement that

* Dan Williams - please look at #783712

Accepted blockers
-----------------

1. http://bugzilla.redhat.com/show_bug.cgi?id=787461 - "system halts at
first reboot after install"

This one's a bit messy, there could be multiple bugs here, but a big one
which may be the root cause of all the others is that anaconda cannot
shut down cleanly at the end of install. bcl has posted a patch for this
for review on the anaconda ML. Ideally we need a new anaconda build ASAP
so we can test and verify the fix in time for a new RC compose early
next week.

2. http://bugzilla.redhat.com/show_bug.cgi?id=787744 - "RuntimeError:
device is already mapped (F17 Alpha TC1)"

This was believed to be fixed in TC2, but Josef says the fix didn't
work. We're currently waiting on more info from Josef, I've pinged him
to provide this.

3. http://bugzilla.redhat.com/show_bug.cgi?id=787261 - "Fedora 17 Alpha
TC1 still has all F16 artwork"

This is fixed in RC2, but we need to have the update pushed stable -
https://admin.fedoraproject.org/updates/fedora-logos-17.0.0-1.fc17 - to
fully close it out (so network installs will get it, for e.g.) It has
enough karma for a manual push, but one more +1 would trigger an
autopush.

4. http://bugzilla.redhat.com/show_bug.cgi?id=791032 - "repoclosure
failure in 17-Alpha.RC2 DVDs (ceph-0.37-2.fc17)"

We have a fix for this one, thanks to Spot:
https://admin.fedoraproject.org/updates/FEDORA-2012-1771/ceph-0.41-1.fc17...
. We will pull it into the next compose. It also needs karma to be pushed stable, so
please test and karma it.

Proposed blockers
-----------------

1. http://bugzilla.redhat.com/show_bug.cgi?id=794899 - ""plymouth" not
installed in F-17 Alpha RC2 and causes boot to freeze"

We need votes on whether this should be a blocker (usual suspects,
please vote!) - my opinion is this is fairly likely to be accepted as a
blocker, it seems like a clear criterion infringement. There are several
possible ways this bug can be addressed, we need a developer to take
ownership and pick and implement one. Details are in the bug comments.

Accepted NTH
------------

1. http://bugzilla.redhat.com/show_bug.cgi?id=790607 - "F17 Alpha KDE
desktop background is blank (all black)"

We pulled the fix for this into RC2, so it should be good there. The
update has sufficient karma to be auto-pushed, so all we need is a
stable update push from releng for this one.

2. http://bugzilla.redhat.com/show_bug.cgi?id=787840 - "SELinux is
preventing /usr/libexec/postfix/local from 'sys_ptrace' accesses on the
None /var/spool/postfix/active/1C60C6EC7"

We pulled the fix for this (selinux-policy -88) into RC2. It needs
pushing stable, but it's now superseded by -89, which we'll also want to
pull in, and is discussed further down. 89 needs karma -
https://admin.fedoraproject.org/updates/FEDORA-2012-1953/selinux-policy-3...

3. http://bugzilla.redhat.com/show_bug.cgi?id=790537 - "[abrt] kernel:
WARNING: at arch/x86/kernel/traps.c:729 do_device_not_available
+0x33/0x40()"

There's a fix for this available as an update, the update needs karma:
https://admin.fedoraproject.org/updates/FEDORA-2012-1948/kernel-3.3.0-0.r...

4. http://bugzilla.redhat.com/show_bug.cgi?id=785737 - "update
mdadm/mdmon to work with systemd unrolling mounts to initramfs mount on
shutdown"

There's a fix for this that needs testing, too:
https://admin.fedoraproject.org/updates/FEDORA-2012-1766/dracut-016-1.fc17

5. http://bugzilla.redhat.com/show_bug.cgi?id=790639 - "Requires:
gsettings-desktop-schemas is missing from spec"

The fix for this was pulled into RC2, but it needs testing:
https://admin.fedoraproject.org/updates/FEDORA-2012-1783/metacity-2.34.2-... . Michael
Schwendt, I'm not sure about that -1 - is the %postun in question a regression
compared to the previous build? Does it cause some kind of critical problem? If not, it
may not be appropriate. If it *is* a critical problem, we need fmuellner to fix it and
update.

6. http://bugzilla.redhat.com/show_bug.cgi?id=771484 - "ldap_result does
not succeed for sssd"

We're waiting on a fix for this from the maintainer.

7. http://bugzilla.redhat.com/show_bug.cgi?id=790522 - "Private /tmp are
broken"

We pulled a fix for this into RC2, it has sufficient karma so we just
need a stable update push from releng to close this one.

8. http://bugzilla.redhat.com/show_bug.cgi?id=725219 - "anaconda should
run in clone not span mode"

This is a Greatest Hit, just waiting on some kind of a code fix from
anaconda / X developers.

Proposed NTH
------------

All these need votes from the usual NTH/blocker status voters,
obviously. (In case anyone's wondering, really anyone can vote on
NTH/blocker status, and usually a pretty solid consensus emerges).

1. http://bugzilla.redhat.com/show_bug.cgi?id=794690 - "PulseAudio fails
to run if ConsoleKit is not present, and CK is not included currently in
F17 desktop live image"

This is a pretty serious consequence of the ConsoleKit removal feature.
It could be fixed either by re-introducing ConsoleKit to the desktop
spin at least temporarily, or by adjusting PA's default config to remove
the CK references. We need desktop team to make a choice here and
execute it.

2. http://bugzilla.redhat.com/show_bug.cgi?id=794886 - "Cannot shut down
from GNOME with selinux-policy-3.10.0-88.fc17"

This is another fairly serious but not-quite-blocking bug: selinux
prevents shutdown via GNOME's menus with selinux-policy -88. -89 fixes
this, so it just needs karma, as mentioned above:
https://admin.fedoraproject.org/updates/FEDORA-2012-1953/selinux-policy-3...

3. http://bugzilla.redhat.com/show_bug.cgi?id=783712 - "No wireless
Access points - Fedora-17-Nightly-20120120.10-i686-Live-soas.iso"

Apparently, wireless networking just isn't working in current nightlies
and RC2 desktop live (not just Sugar as the summary states). I didn't
verify this myself yet but it's obviously a pretty serious bug, it might
even be blocker rather than NTH. We need NM team to look into this one.
Seems like all you need to do to reproduce is to boot Alpha RC2 desktop
live and try to connect to a wireless AP.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net

[Ian Pilcher]

On 02/17/2012 10:44 PM, Adam Williamson wrote:
>  8. http://bugzilla.redhat.com/show_bug.cgi?id=725219 - "anaconda
> should
>  run in clone not span mode"
>  
>  This is a Greatest Hit, just waiting on some kind of a code fix from
>  anaconda / X developers. 
This actually wasn't a problem during my (very brief) attempt to
install RC2.  X was pretty obviously spanning both displays, but
anaconda was only using a smaller region (1024x768?) of one of the
displays.

-- 
========================================================================
Ian Pilcher                                         arequipeno(a)gmail.com
"If you're going to shift my paradigm ... at least buy me dinner first."
========================================================================

[Adam Williamson]

On Sat, 2012-02-18 at 19:54 -0600, Ian Pilcher wrote:
>  On 02/17/2012 10:44 PM, Adam Williamson wrote:
>  > 8. http://bugzilla.redhat.com/show_bug.cgi?id=725219 - "anaconda should
>  > run in clone not span mode"
>  > 
>  > This is a Greatest Hit, just waiting on some kind of a code fix from
>  > anaconda / X developers.
>  
>  This actually wasn't a problem during my (very brief) attempt to
>  install RC2.  X was pretty obviously spanning both displays, but
>  anaconda was only using a smaller region (1024x768?) of one of the
>  displays. 
Could you see the Next buttons, though? For many people, anaconda
*mostly* renders on one screen or the other, but the buttons wind up on
the _other_ screen, or not present on either screen.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net

[Ian Pilcher]

On 02/19/2012 01:32 AM, Adam Williamson wrote:
>  On Sat, 2012-02-18 at 19:54 -0600, Ian Pilcher wrote:
> > This actually wasn't a problem during my (very brief) attempt to
> > install RC2.  X was pretty obviously spanning both displays, but
> > anaconda was only using a smaller region (1024x768?) of one of the
> > displays.
>  
>  Could you see the Next buttons, though? For many people, anaconda
>  *mostly* renders on one screen or the other, but the buttons wind up on
>  the _other_ screen, or not present on either screen. 
Yup.  The entire anaconda display area, buttons and all, didn't even
take up all of one 1680x1050 screen.

-- 
========================================================================
Ian Pilcher                                         arequipeno(a)gmail.com
"If you're going to shift my paradigm ... at least buy me dinner first."
========================================================================

Fedora 17: NTH fix for openldap available

[Jan Vcelak]

Hello,

I just pushed fixed openldap to updates-testing:
https://admin.fedoraproject.org/updates/openldap-2.4.29-3.fc17

Please, test it and give it some karma! Especially SSSD users.

Thanks & Regards,
Jan

>  
>  Accepted NTH
>  ------------
>  
>  6. http://bugzilla.redhat.com/show_bug.cgi?id=771484 - "ldap_result
>  does
>  not succeed for sssd"
>  
>  We're waiting on a fix for this from the maintainer.
>