4:48 p.m.
Hi folks! We've got quite a lot of outstanding blockers for Fedora 24
Final which are going to need dealing with, so here's a roundup. If
you're CC'ed on this mail and you're in a hurry, search for your
name/nick and you'll probably find the bits you need to look at.
tl;dr summary
-------------
QA folks: test fixes for:
* https://bugzilla.redhat.com/show_bug.cgi?id=1330766
* https://bugzilla.redhat.com/show_bug.cgi?id=1314637
anaconda folks: look at:
* https://bugzilla.redhat.com/show_bug.cgi?id=1333998
* https://bugzilla.redhat.com/show_bug.cgi?id=1332266
other devs to look at bugs:
mclasen/cfergeau - https://bugzilla.redhat.com/show_bug.cgi?id=1164492
mfabian - https://bugzilla.redhat.com/show_bug.cgi?id=1333998 and maybe
https://bugzilla.redhat.com/show_bug.cgi?id=1318045
pjones - https://bugzilla.redhat.com/show_bug.cgi?id=1320273
zbyszek - https://bugzilla.redhat.com/show_bug.cgi?id=1318045
anaconda/zbyszek/harald/jwb/mclasen/pfrields
- https://bugzilla.redhat.com/show_bug.cgi?id=1332266
(i.e. I don't care who does it but can someone *please* bang some heads
together and decide what the heck we want to do about that mess)
Detailed info
-------------
1. https://bugzilla.redhat.com/show_bug.cgi?id=1333998 - anaconda - NEW
After Russian install, console keymap is 'us', not 'ru'
This is a type of bug that's always a pain to track down, but we're
gonna need someone to look into it. It may involve localed and/or
langtable behaviour, and I'm CCing mfabian as he may be able to help.
2. https://bugzilla.redhat.com/show_bug.cgi?id=1164492 - gnome-boxes - NEW
Please drop libvirt 'default' network dependency for F24 GA, disrupts livecd
networking
This is the latest edition of a bug that's been around for years; if
you boot in a libvirt VM and the libvirt default network starts up in
the VM, there can be a conflict which results in the VM failing to have
any network connectivity. To work around this in the past we've dropped
gnome-boxes' dependency on the libvirt default network package for the
final release, and the bug is a request to do that again, we need
desktop team to do that.
3. https://bugzilla.redhat.com/show_bug.cgi?id=1320273 - grub2 - NEW
chainloading bootmgr.efi on UEFI results in error: out of memory
This is a failure in dual-boot alongside Windows on UEFI (Windows fails
to boot from the Fedora grub menu). We need someone (i.e. pjones) to
take a look at it ASAP.
4. https://bugzilla.redhat.com/show_bug.cgi?id=1293167 - kf5-kinit - NEW
[abrt] kf5-kinit: qt_message_fatal(): kdeinit5 killed by SIGABRT
This is an issue in KDE user switching. Recent discussion seems to
suggest it may not affect all users, and there's probably scope to re-
evaluate the blockeriness of this one, so we might want to consider
that. A fix would be best, of course.
5. https://bugzilla.redhat.com/show_bug.cgi?id=1325471 - libsolv - ON_QA
resolving Supplements: dependencies pull in multilib packages
The fix for this is pending stable, so shouldn't be much more needed
here.
6. https://bugzilla.redhat.com/show_bug.cgi?id=1330766 - realmd - MODIFIED
[abrt] realmd: g_cancellable_is_cancelled(): realmd killed by SIGSEGV
This is an issue in FreeIPA client enrolment which I hit in manual
testing, a potential fix just appeared so we need to test that.
7. https://bugzilla.redhat.com/show_bug.cgi?id=1314637 - selinux-policy - ON_QA
SELinux is preventing fwupd from 'write' accesses on the directory
0000:00:02.0.
This is an OOTB SELinux denial. It should actually have been fixed for
some time already; we just need to verify that and close the bug if so.
8. https://bugzilla.redhat.com/show_bug.cgi?id=1333106 - selinux-policy - ASSIGNED
Server deployment fails on current Fedora Rawhide (2016-05-04)
FreeIPA and SELinux folks are working on this one, we'll need to test a
fix once it's available.
9. https://bugzilla.redhat.com/show_bug.cgi?id=1318045 - systemd - NEW
Incorrect keymap when decrypting encrypted partitions
This seems to be a live image-specific issue with the keyboard layout
set in /etc/vconsole.conf not being used at the time of encrypted
device passphrase entry on boot of the installed system. We kinda need
the systemd devs to have a look at it.
10. https://bugzilla.redhat.com/show_bug.cgi?id=1332266 - systemd - NEW
systemd-logind does not verify if system has resume device defined when checking if
can.hibernate or can.hybridsleep
This is currently proposed, but mentioning it here as it's a messy one
with lots of discussion. Basically it's been pointed out this cycle
that there's an awful lot of issues with hibernate/resume, it's
entirely broken. We've usually never considered hibernate issues to be
release blocking, but this bug points out a case where hibernate is
triggered automatically (critical battery life on Workstation) and
resume is not even attempted, which may cause data loss in apps that
were open at the time of hibernation. Whether this is really release-
blocking is still questionable, but it sure would be nice if we could
resolve this hibernate mess *somehow*. There seems to be an ongoing
disagreement between anaconda/dracut/kernel/systemd folks on exactly
how resume should be handled, it would be nice to get that resolved.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net