Hi folks!
We have a 'situation' for F27 Final that needs resolving.
One blocker bug - https://bugzilla.redhat.com/show_bug.cgi?id=1164492 -
definitely requires a change to gnome-boxes. That change isn't under
dispute, and is present in an update, gnome-boxes-3.26.1-4.fc27 .
So what's the problem? The problem is that update also includes
*another* change, and I'm not at all sure we actually still want that
change.
gnome-boxes-3.26.1-3.fc27 changed the default video adapter for new VMs
to 'virtio', as an attempt to workaround
https://bugzilla.redhat.com/show_bug.cgi?id=1491320 . However, that bug
now appears to be addressed on the kernel side. Changing the default
video adapter for new VMs could only ever be an incomplete workaround
(it wouldn't help VMs created on other OSes, or existing VMs, or
various other situations), and I believe was a downstream-only change.
So do we still actually want to change Boxes' default video adapter for
F27? If not, can someone please build and submit a 3.26.1-5.fc27 update
which includes the dependency change for #1164492, but reverts the
virtio default video adapter change? Thanks!
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
# F27 Blocker Review meeting
# Date: 2017-10-30
# Time: 16:00 UTC
# Location: #fedora-blocker-review on irc.freenode.net
Hi folks! We currently have 2 proposed Final blockers, 1 proposed
Final freeze exception, and 1 proposed Server Beta blocker, so let's
have a Fedora 27 blocker review meeting.
If you have time tonight, you can take a look at the proposed or
accepted blockers before the meeting - the full lists can be found
here: https://qa.fedoraproject.org/blockerbugs/ .
We'll be evaluating these bugs to see if they violate any of the
Release Criteria and warrant the blocking of a release if they're not
fixed. Information on the release criteria for F27 can be found on the
wiki [0].
For more information about the Blocker and Freeze exception process,
check out these links:
- https://fedoraproject.org/wiki/QA:SOP_blocker_bug_process
- https://fedoraproject.org/wiki/QA:SOP_freeze_exception_bug_process
And for those of you who are curious how a Blocker Review Meeting
works - or how it's supposed to go and you want to run one - check out
the SOP on the wiki:
- https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting
Note that for the special case of the Server release being split from
the main release stream for Fedora 27, we have created separate Server
blocker tracking bugs and will be following the same basic process for
the Server release dates using those trackers.
Have a good night and see you tomorrow!
[0] https://fedoraproject.org/wiki/Fedora_Release_Criteria
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
IIRC, the background image (the default Fedora background for a release) we
use as the background for the upgrading slide in GNOME Software was not
updated properly until after the F26 Release.
Is there anyway we can check that this is showing the correct background
before release? If i follow the directions here on the magazine [1], it
still shows the flat blue fallback background.
cheers,
ryanlerch
[1] https://fedoramagazine.org/upgrade-fedora-workstation-fedora-27-beta/
Hi Workstation WG,
as the F27 Final release is quickly approaching, it would be nice if
the WG would start working on a Fedora Magazine article "What's New in
F27 Workstation". For an inspiration you can check an older article
for F26 [1]. The article should be finished prior Readiness meeting
for F27 Final, planned on 2017-Oct-26.
[1] https://fedoramagazine.org/whats-new-fedora-26-workstation/
Regards,
Jan
--
Jan Kuřík
Platform & Fedora Program Manager
Red Hat Czech s.r.o., Purkynova 99/71, 612 45 Brno, Czech Republic
Hi folks! As we're now frozen for Final, here's an update on blocker
bug status.
tl;dr action summary
====================
Accepted blockers
-----------------
1. https://bugzilla.redhat.com/show_bug.cgi?id=1164492
ACTION: desktop team to make change and submit update
2. https://bugzilla.redhat.com/show_bug.cgi?id=1486002
ACTION: Konrad and pjones to continue collaborating and ship a fix
3. https://bugzilla.redhat.com/show_bug.cgi?id=1491320
ACTION: kernel team to ship official update with fix, QA to test it
4. https://bugzilla.redhat.com/show_bug.cgi?id=1494061
ACTION: desktop team to investigate and fix
5. https://bugzilla.redhat.com/show_bug.cgi?id=1498207
ACTION: dnf team to clarify need for second update
Proposed blockers
-----------------
1. https://bugzilla.redhat.com/show_bug.cgi?id=1503496
ACTION: pjones / anaconda team to review, merge and ship proposed fix
2. https://bugzilla.redhat.com/show_bug.cgi?id=1489942
ACTION: pjones to continue trying to fix (but likely not blocker)
3. https://bugzilla.redhat.com/show_bug.cgi?id=1504059
ACTION: none specific needed (just blocker review to review bug)
4. https://bugzilla.redhat.com/show_bug.cgi?id=1497691
ACTION: none specific needed (update is tested and ready)
5. https://bugzilla.redhat.com/show_bug.cgi?id=1500834
ACTION: blocker review team to decide blocker/FE status, kernel team
to send out fix if accepted as blocker/FE
6. https://bugzilla.redhat.com/show_bug.cgi?id=1501762
ACTION: blocker review team to review, otherwise none needed
7. https://bugzilla.redhat.com/show_bug.cgi?id=1501362
ACTION: anton to update microcode bundle (but likely not a blocker)
Bug-by-bug detail
=================
Accepted blockers
-----------------
1. https://bugzilla.redhat.com/show_bug.cgi?id=1164492 - gnome-boxes - ASSIGNED
Please drop libvirt 'default' network dependency for F27 GA, disrupts livecd networking
This is a 'greatest hit' we've had to deal with for several releases now.
Everyone involved more or less knows what the deal is, we just need to
make the change and send out an update.
2. https://bugzilla.redhat.com/show_bug.cgi?id=1486002 - grub2 - NEW
grub2-mkconfig does not work if xen.gz is installed.
This one seems to be well understood and being worked on by pjones and
Konrad (our Xen contact). We just need the proposed fix to be verified
by Konrad and then sent out as an update.
3. https://bugzilla.redhat.com/show_bug.cgi?id=1491320 - kernel - NEW
heavy screen flicker with latest kernels in qxl+spice VM
We've been kicking this one around for a while, but as of today it seems
like a kernel-side fix for it may be possible. We now need folks to test
and confirm the proposed kernel fix, and kernel team to send out an
update containing it (assuming it works).
4. https://bugzilla.redhat.com/show_bug.cgi?id=1494061 - gnome-software - NEW
gnome-software doesn't show F25->F27 upgrade even though it should
This one seems quite straightforward - Fedora 25 doesn't show F27 as
an option for upgrade, even when the gsettings setting to show development
releases is toggled. We need the devs to take a look at this and fix it.
5. https://bugzilla.redhat.com/show_bug.cgi?id=1498207 - libdnf - VERIFIED
DNF crash during upgrade installation F26 -> F27
This one seems to be pretty much addressed; in fact an update which two
reporters claimed already fixed the issue went stable a week ago. There's
a newer update now marked as fixing the bug, so I am asking in the bug
for clarification as to whether the newer update is really required to
fully fix this bug.
Proposed blockers
-----------------
1. https://bugzilla.redhat.com/show_bug.cgi?id=1503496 - anaconda - POST
FileNotFoundError: [Errno 2] No such file or directory: 'grub2-mkconfig'
This looks like a fairly clear-cut blocker (the bug entirely prevents
Mac installs from working, I think). I have proposed a fix for this; need
pjones / anaconda folks to review and apply it, and send out an update.
2. https://bugzilla.redhat.com/show_bug.cgi?id=1489942 - dbxtool - MODIFIED
dbxtool fails at boot 'Could not apply database update "DBXUpdate-2016-08-09-13-16-00.bin": Permission denied'
AFAICS this is aarch64-specific, so cannot be accepted as a blocker.
pjones keeps trying to fix it, but without any success so far.
3. https://bugzilla.redhat.com/show_bug.cgi?id=1504059 - distribution - NEW
Include Firefox 57 at compose
This is basically a 'make a choice' bug, there's no difficult work needed
here as the package is already available (and tested and karmaed).
4. https://bugzilla.redhat.com/show_bug.cgi?id=1497691 - dnsmasq - ON_QA
CVE-2017-14491 CVE-2017-14492 CVE-2017-14493 CVE-2017-14494 CVE-2017-14495 CVE-2017-14496 dnsmasq: various flaws [fedora-all]
This is already accepted as an FE, and the update is available and
karma'ed, so there's no work needed here, it'll just get closed when we
do another stable push.
5. https://bugzilla.redhat.com/show_bug.cgi?id=1500834 - kernel - MODIFIED
VirtualBox: black screen on fedora 27
I'd say it's relatively unlikely we'd take this as a blocker, but in
any case, a fix appears to be available and in a kernel build already,
so we'll likely pull it in along with other required kernel fixes. We'll
just need people to verify the fix once it's in an official update.
6. https://bugzilla.redhat.com/show_bug.cgi?id=1501762 - kernel - ON_QA
CVE-2017-5123 kernel: Missing access_ok() checks in waitid() [fedora-all]
This bug is kinda representing several kernel CVEs the team wants fixed
as part of the release. Again the update is already there, so the only
work needed here is for QA to verify the kernel works OK.
7. https://bugzilla.redhat.com/show_bug.cgi?id=1501362 - microcode_ctl - NEW
Intel microcode not updating --> [Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x25 (or later)
I would say this is unlikely to be a blocker (as it doesn't seem to have
any really bad consequences, just a scary warning message). It may be
worth an FE. Fixing it shouldn't be a lot of work, the packager only needs
to update our bundle of Intel microcode updates (it hasn't been updated
since July).
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
Hi folks! So I just spent an hour or two debugging an unexpected change
in behaviour in the openQA base images. I've figured out what's going
on, but I'm unsure if it's intended.
There was a change to anaconda between Fedora 25 and Fedora 26 which
causes it to write a line:
post_install_tools_disabled=1
to /etc/sysconfig/anaconda for all kickstarted installs where there is
no explicit 'firstboot' directive in the kickstart. (Prior to Fedora
26, anaconda would disable 'initial-setup.service' by default for
kickstarted installs, but it didn't write that line, or disable gnome-
initial-setup - either mode - in any way.)
It seems that gnome-initial-setup reads that file on startup, and if it
sees that post_install_tools_disabled is set, it will kinda short-
circuit - it won't display its UI at all, it'll just do its shutdown
tasks (create ~/.config/gnome-initial-setup-done and so on) and then
quit. It doesn't only do this in its system-wide, 'run before gdm if no
user accounts exist' mode - it also does it when running in its 'first
login to GNOME as any user' mode.
So an effect of this change - which may not have been intended - is
that on most kickstarted installs of Fedora 26+, gnome-initial-setup
does not run *even in 'first login to a user account' mode*. (This is
the change that affected openQA). I can't immediately see that it makes
logical sense to disable this g-i-s mode on kickstarted installs by
default, but maybe someone disagrees.
Anyway, just wanted to raise this issue for discussion between desktop
and anaconda folks (who are CCed on this), before filing bugs anywhere.
Thanks!
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
# F27 Blocker Review meeting
# Date: 2017-10-23
# Time: 16:00 UTC
# Location: #fedora-blocker-review on irc.freenode.net
Hi folks! We currently have 8 proposed Final blockers, 9 proposed
Final freeze exceptions, 5 proposed Server Beta blockers, and 1
proposed Server Beta freeze exception to review, so let's have a
Fedora 27 blocker review meeting.
If you have time this weekend, you can take a look at the proposed or
accepted blockers before the meeting - the full lists can be found
here: https://qa.fedoraproject.org/blockerbugs/ .
We'll be evaluating these bugs to see if they violate any of the
Release Criteria and warrant the blocking of a release if they're not
fixed. Information on the release criteria for F27 can be found on the
wiki [0].
For more information about the Blocker and Freeze exception process,
check out these links:
- https://fedoraproject.org/wiki/QA:SOP_blocker_bug_process
- https://fedoraproject.org/wiki/QA:SOP_freeze_exception_bug_process
And for those of you who are curious how a Blocker Review Meeting
works - or how it's supposed to go and you want to run one - check out
the SOP on the wiki:
- https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting
Note that for the special case of the Server release being split from
the main release stream for Fedora 27, we have created separate Server
blocker tracking bugs and will be following the same basic process for
the Server release dates using those trackers.
Have a good weekend and see you on Monday!
[0] https://fedoraproject.org/wiki/Fedora_Release_Criteria
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net