#blocker-review:fedoraproject.org log

<@frantisekz:fedora.im>

16:02:26

!startmeeting F41-blocker-review

<@meetbot:fedora.im>

16:02:27

Meeting started at 2024-09-23 16:02:26 UTC

<@meetbot:fedora.im>

16:02:27

The Meeting name is 'F41-blocker-review'

<@frantisekz:fedora.im>

16:02:38

Hello, who's up for some fun today?

<@conan_kudo:matrix.org>

16:02:43

!hi

<@zodbot:fedora.im>

16:02:45

Neal Gompa (ngompa) - he / him / his

<@sgallagh:fedora.im>

16:03:17

!hi

<@zodbot:fedora.im>

16:03:18

Stephen Gallagher (sgallagh) - he / him / his

<@nielsenb:fedora.im>

16:03:27

!hi

<@zodbot:fedora.im>

16:03:28

Brandon Nielsen (nielsenb)

<@sgallagh:fedora.im>

16:03:42

I’m attending via phone while running errands.

<@derekenz:fedora.im>

16:03:53

!hi

<@zodbot:fedora.im>

16:03:54

Derek Enz (derekenz)

<@tablepc:fedora.im>

16:04:08

!hi

<@zodbot:fedora.im>

16:04:08

Pat Kelly (tablepc)

<@farribeiro:matrix.org>

16:06:13

!hi

<@zodbot:fedora.im>

16:06:16

Fábio Ribeiro (farribeiro) - he / him / his

<@frantisekz:fedora.im>

16:06:38

boilerplate incoming (I do hope neal's water boiler runs Fedora IoT :) )

<@frantisekz:fedora.im>

16:06:43

!link https://fedoraproject.org/wiki/Fedora_41_Final_Release_Criteria

<@frantisekz:fedora.im>

16:06:48

!info Our purpose in this meeting is to review proposed blocker and nice-to-have bugs and decide whether to accept them, and to monitor the progress of fixing existing accepted blocker and nice-to-have bugs.

<@frantisekz:fedora.im>

16:06:53

!link https://fedoraproject.org/wiki/Basic_Release_Criteria

<@frantisekz:fedora.im>

16:06:58

!info The criteria for release blocking bugs can be found at:

<@frantisekz:fedora.im>

16:07:02

!info We'll be following the process outlined at:

<@frantisekz:fedora.im>

16:07:07

!link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting

<@frantisekz:fedora.im>

16:07:12

!info The bugs up for review today are available at:

<@frantisekz:fedora.im>

16:07:15

!link http://qa.fedoraproject.org/blockerbugs/current

<@frantisekz:fedora.im>

16:07:20

!link https://fedoraproject.org/wiki/Fedora_41_Beta_Release_Criteria

<@frantisekz:fedora.im>

16:07:23

!topic Introduction

<@conan_kudo:matrix.org>

16:07:39

if I could only be so lucky

<@frantisekz:fedora.im>

16:07:58

!info For Final, we have:

<@frantisekz:fedora.im>

16:08:11

!info 1 Proposed Blocker

<@frantisekz:fedora.im>

16:08:25

!info 3 Accepted Blockers

<@frantisekz:fedora.im>

16:08:33

heh, not much, should be fast

<@frantisekz:fedora.im>

16:09:26

who wants to secretalize this one? (trying to ping bunch of my team: Kamil Páral Lukas Brabec )

<@frantisekz:fedora.im>

16:11:07

okey dokey, I'll secretalize myself, will fire them later...

<@frantisekz:fedora.im>

16:11:16

!info František Zatloukal will secretarialize

<@frantisekz:fedora.im>

16:11:26

!topic Proposed Final blockers

<@frantisekz:fedora.im>

16:12:03

!topic (2307226) f41 and rawhide should get the fix from wpa_supplicant-2.11-3.fc40

<@frantisekz:fedora.im>

16:12:07

!link https://bugzilla.redhat.com/show_bug.cgi?id=2307226

<@frantisekz:fedora.im>

16:12:12

!link https://pagure.io/fedora-qa/blocker-review/issue/1662

<@frantisekz:fedora.im>

16:12:16

!info Proposed Blocker, wpa_supplicant, NEW

<@frantisekz:fedora.im>

16:12:25

!info Ticket vote: FinalBlocker (+4,1,-1) (+cheese, +geraldosimiao, +nielsenb, +asciiwolf, jmaybaum, -sgallagh)

<@frantisekz:fedora.im>

16:12:57

We did punt this one last time as far as I remember

<@tablepc:fedora.im>

16:14:23

Need a good estimate as to when it can be available for Fedora testing

<@frantisekz:fedora.im>

16:14:44

there is some path wip upstream for this

<@sgallagh:fedora.im>

16:15:26

I put my thoughts on BZ; but to summarize: I think the number of people affected would be limited and fixing it is non-trivial. I’d happily break Freeze to land something, but it wouldn’t pass the “Last blocker at Go/No-Go” test for me.

<@tablepc:fedora.im>

16:15:31

If it's too long it comes down to if the effected users are a large enough group etc. etc.

<@frantisekz:fedora.im>

16:15:35

I am afraid I personally don't have ton of ideas nor estimates of how bad it is :/

<@nielsenb:fedora.im>

16:16:49

"We broke your wifi" isn't a great look, especially since I'm not sure there is any workaround except hoping it gets fixed?

<@tablepc:fedora.im>

16:17:12

That's trye

<@frantisekz:fedora.im>

16:17:53

yeah, but on the other hand, we should weight it in with Stephen Gallagher 's perspective in the bz, if it's going to affect tiny amount of users, if it's going to be documented and resolved sometime(tm)

<@nielsenb:fedora.im>

16:18:26

But then there's the follow up comment saying they are affected

<@tablepc:fedora.im>

16:19:30

Yes there are apparently some effected corporates effected

<@sgallagh:fedora.im>

16:19:48

I’m sure the number is non-zero. I just can’t realistically believe that it’s sufficiently large that we can’t release-note it and ask for people to hold off until the fix is in.

<@nielsenb:fedora.im>

16:20:14

I'm pretty sure I have a laptop around here provisioned in such a way (financial sector), but it doesn't run Fedora, and never goes to the corporate campus, so it wouldn't ever actually be impacted

<@nielsenb:fedora.im>

16:20:41

But it shows there are non-government use cases

<@sgallagh:fedora.im>

16:21:10

I probably phrased that badly

<@frantisekz:fedora.im>

16:21:14

what I'd incline to here: -1 Final Blocker, +1 Final FE, and poking Davide how realistic is it to have fix for this in some sensible time-frame

<@conan_kudo:matrix.org>

16:21:25

so what, the fix is either to build it with engine api or port it to provider api

<@conan_kudo:matrix.org>

16:21:28

right?

<@frantisekz:fedora.im>

16:21:39

yeah, seems like these two options

<@sgallagh:fedora.im>

16:21:42

Correct

<@nhanlon:beeper.com>

16:21:42

!hi

<@zodbot:fedora.im>

16:21:44

Neil Hanlon (neil) - he / him / his

<@tablepc:fedora.im>

16:21:48

Yes need a good estimate

<@conan_kudo:matrix.org>

16:21:54

I assume this would be a bigger problem also for RHEL, where this is more common, so maybe someone is looking at porting the provider api?

<@nielsenb:fedora.im>

16:22:00

I do agree with the "the cross-section of "people likely to use this setup" and "people allowed to bring-their-own-device of Fedora" is likely to be extremely small." statement though

<@frantisekz:fedora.im>

16:22:09

I'd expect exactly that Neal

<@sgallagh:fedora.im>

16:22:20

An estimate doesn’t change the blocker discussion except if we were to declare it a blocker and then waive it as “too hard”

<@conan_kudo:matrix.org>

16:22:32

I do know of one major company that uses Fedora by default and in fact _does_ use hardware-based certificates for authentication even for Wi-Fi

<@conan_kudo:matrix.org>

16:22:57

so it would really suck if TPM based auth is completely broken

<@nielsenb:fedora.im>

16:22:59

Presumably any company using Fedora would test before a wide deployment

<@nielsenb:fedora.im>

16:23:10

Unfortunately I don't know how they would deploy any fix?

<@nielsenb:fedora.im>

16:23:25

They'd had to spin new install media, or install with a repo enabled

<@sgallagh:fedora.im>

16:23:38

Brandon Nielsen: I assume they would stay on F40 until a fix arrived on F41

<@conan_kudo:matrix.org>

16:23:51

we'd have to presume a fix would show up for F41

<@frantisekz:fedora.im>

16:23:53

the worst-case time window is pretty long - f40 is going to be supported till roughly ~el10 release where I'd expect to have it fixed by the latest

<@conan_kudo:matrix.org>

16:23:56

can we presume that?

<@sgallagh:fedora.im>

16:24:05

Or skip F40 entirely

<@sgallagh:fedora.im>

16:24:12

Sorry F41

<@conan_kudo:matrix.org>

16:24:22

and note that we don't do respins for non-x86 yet

<@sgallagh:fedora.im>

16:24:50

Conan Kudo: Given that RHEL is likely to hit this issue with their customers, it’s likely that it’ll be fixed in the next few months indeed

<@conan_kudo:matrix.org>

16:25:07

is there a RHEL jira we can link the fedora bz to?

<@conan_kudo:matrix.org>

16:25:13

because it's a fairly serious problem

<@sgallagh:fedora.im>

16:25:17

I haven’t looked

<@sgallagh:fedora.im>

16:25:24

And I’m not at my computer

<@frantisekz:fedora.im>

16:25:26

I am not aware of one and Jira scares me :D

<@tablepc:fedora.im>

16:25:27

we're guessing a lot here and this has potential, maybe small to me at least a big image problem.

<@conan_kudo:matrix.org>

16:25:30

lol

<@conan_kudo:matrix.org>

16:25:38

jira scares everyone

<@geraldosimiao:matrix.org>

16:25:54

!hi

<@zodbot:fedora.im>

16:25:56

Geraldo S. Simião Kutz (geraldosimiao) - he / him / his

<@frantisekz:fedora.im>

16:26:26

so, I am inclined to -1 and having it documented in Common Issues, what bout other's inclinations?

<@nielsenb:fedora.im>

16:27:07

I'm still FinalBlocker +1

<@tablepc:fedora.im>

16:27:09

-1 Need more data

<@conan_kudo:matrix.org>

16:27:15

FinalBlocker +1

<@conan_kudo:matrix.org>

16:27:29

I think it's a seriously bad thing if we don't do something to fix it

<@tablepc:fedora.im>

16:27:38

szorry met +1

<@derekenz:fedora.im>

16:27:44

agree

<@conan_kudo:matrix.org>

16:27:53

if no port to provider API is in sight, we can fix it by enabling the engine API for it

<@nielsenb:fedora.im>

16:27:53

Unfortunately I don't think convincing data will be aggregated

<@nielsenb:fedora.im>

16:28:21

So as long as I'm guessing, I'm guessing on the cautious side

<@derekenz:fedora.im>

16:28:27

considering too its enterprise level

<@geraldosimiao:matrix.org>

16:28:34

I'm still +1 FB by now

<@tablepc:fedora.im>

16:28:37

Best in my opinion

<@frantisekz:fedora.im>

16:30:42

okey dokey, we have enough +1 votes, gimme a few mins to do the wording somehow :)

<@sgallagh:fedora.im>

16:31:36

What criterion does this violate, specifically?

<@conan_kudo:matrix.org>

16:31:44

we should definitely make sure a rhel jira exists to track this regardless

<@sgallagh:fedora.im>

16:31:50

Just for posterity

<@frantisekz:fedora.im>

16:32:42

yeah, I am looking through the criterions - we probably could pick any that requires working network (eg. package installation, netinst image, etc. but I am not sure if I like that one), and I'd be glad for better ideas

<@sgallagh:fedora.im>

16:33:04

Except that “wired dongle” is a workaround

<@nielsenb:fedora.im>

16:33:18

Yeah, kinda interesting

<@nielsenb:fedora.im>

16:33:24

There's a working sound criteria

<@nielsenb:fedora.im>

16:33:30

But not networking that I see

<@frantisekz:fedora.im>

16:33:41

Kamil Páral: can you look for something better since you are more versed in criterions?

<@kparal:matrix.org>

16:34:28

I'm sorry, I'm not familiar with this bug, so it will take me some time to read through it

<@frantisekz:fedora.im>

16:34:45

"some wifi setups don't work" would tldr it good

<@frantisekz:fedora.im>

16:34:54

we have enough +1 votes

<@kparal:matrix.org>

16:34:59

ok, will look

<@sgallagh:fedora.im>

16:35:28

We have a lot of +1s, but no one seems to know what criterion is being cited…

<@nielsenb:fedora.im>

16:35:30

https://fedoraproject.org/wiki/Basic_Release_Criteria#Basic_networking

<@nielsenb:fedora.im>

16:35:31

<@kparal:matrix.org>

16:35:33

does it affect anaconda netinst as well?

<@frantisekz:fedora.im>

16:35:54

hmm, yeah

<@frantisekz:fedora.im>

16:36:01

just looked at it Brandon Nielsen :D

<@nielsenb:fedora.im>

16:36:03

There's a carve out for exotic wifi configurations

<@frantisekz:fedora.im>

16:36:03

"Common wireless network configurations using supported hardware as defined above are covered by this criterion. This includes access to home and enterprise wireless networks using 802.11 series connection protocols and WPA2 and WPA3 personal and enterprise security protocols"

<@frantisekz:fedora.im>

16:36:19

we could say we see it as non-exotic...

<@kparal:matrix.org>

16:36:25

what about this? See the footnote for wireless: https://fedoraproject.org/wiki/Basic_Release_Criteria#Basic_networking

<@frantisekz:fedora.im>

16:38:45

proposed !agreed 2307226 - AcceptedBlocker (Final) - this is accepted under "It must be possible to establish both IPv4 and IPv6 network connections using both typical router-provided addressing systems (e.g. DHCP on IPv4 or SLAAC or IPv6) and static addressing." at https://fedoraproject.org/wiki/Basic_Release_Criteria#Basic_networking . We concluded that the affected configuration falls under "Common wireless network configurations using supported hardware".

<@geraldosimiao:matrix.org>

16:39:20

Ack

<@farribeiro:matrix.org>

16:39:27

Ack

<@derekenz:fedora.im>

16:39:29

ack

<@nielsenb:fedora.im>

16:39:31

ack

<@conan_kudo:matrix.org>

16:39:35

ack

<@tablepc:fedora.im>

16:39:39

ack

<@frantisekz:fedora.im>

16:39:46

!agreed 2307226 - AcceptedBlocker (Final) - this is accepted under "It must be possible to establish both IPv4 and IPv6 network connections using both typical router-provided addressing systems (e.g. DHCP on IPv4 or SLAAC or IPv6) and static addressing." at https://fedoraproject.org/wiki/Basic_Release_Criteria#Basic_networking . We concluded that the affected configuration falls under "Common wireless network configurations using supported hardware".

<@frantisekz:fedora.im>

16:40:25

okay, that was all proposed for final, let's get a quick look at accepted stuff :)

<@frantisekz:fedora.im>

16:40:27

!topic Accepted Final blockers

<@frantisekz:fedora.im>

16:40:48

!topic (2311936) pyanaconda.modules.common.errors.storage.UnknownDeviceError: Volume0_0

<@frantisekz:fedora.im>

16:40:53

!link https://bugzilla.redhat.com/show_bug.cgi?id=2311936

<@frantisekz:fedora.im>

16:40:57

!link https://pagure.io/fedora-qa/blocker-review/issue/1658

<@frantisekz:fedora.im>

16:41:01

!info Accepted Blocker, anaconda, POST

<@frantisekz:fedora.im>

16:42:33

we have a proposed fix for this one, it remains to be tested (probably in our Brno office where it was reproduced first), we have plenty of time till final, seems alright

<@frantisekz:fedora.im>

16:43:42

okay, moving on

<@frantisekz:fedora.im>

16:43:44

!topic (2282171) gsk: vulkan renderer causes gtk4 apps to crash on resize operations on Raspberry Pi 4 and 400

<@frantisekz:fedora.im>

16:43:49

!link https://bugzilla.redhat.com/show_bug.cgi?id=2282171

<@frantisekz:fedora.im>

16:43:53

!link https://pagure.io/fedora-qa/blocker-review/issue/1605

<@frantisekz:fedora.im>

16:43:58

!info Accepted Blocker, bcm283x-firmware, ASSIGNED

<@frantisekz:fedora.im>

16:44:25

so, here, we have a workaround, we just need proper fix for final, which, as far as I know, Peter Robinson is working on

<@frantisekz:fedora.im>

16:45:08

do we have some estimate of how does it look with that? (here is me kinda hoping Peter's gonna show up :) )

<@frantisekz:fedora.im>

16:47:05

nevermind, I'll ask him off-meeting how does it look like

<@frantisekz:fedora.im>

16:47:21

!topic (2278845) gnome-initial-setup: Choosing avatar results in "SetIconFile call failed" "unknown reason" due to SELinux denial

<@frantisekz:fedora.im>

16:47:25

!link https://bugzilla.redhat.com/show_bug.cgi?id=2278845

<@frantisekz:fedora.im>

16:47:30

!link https://pagure.io/fedora-qa/blocker-review/issue/1603

<@frantisekz:fedora.im>

16:47:39

!info Accepted Blocker, selinux-policy, POST

<@frantisekz:fedora.im>

16:48:19

so, we have fix for this one submitted upstream, looking at selinux-policy release cadence, we'll get at least some rebase before final

<@frantisekz:fedora.im>

16:49:34

so, all in all, everything looks pretty calm so far for F41, we'll see how long it lasts

<@frantisekz:fedora.im>

16:49:35

!topic Open floor

<@frantisekz:fedora.im>

16:49:47

Does anybody have something on mind for f41 business?

<@tablepc:fedora.im>

16:50:15

Otherwise F41 is look great to me.

<@frantisekz:fedora.im>

16:50:52

I'll just note that we finally have GNOME 41 Final rebase submitted in bodhi: https://bodhi.fedoraproject.org/updates/FEDORA-2024-d43fddfda6 (and giving it testing hell would be much appreciated)

<@frantisekz:fedora.im>

16:51:10

I assume KDE is going to get one more rebase Conan Kudo ?

<@tablepc:fedora.im>

16:51:17

Looking forward to it.

<@conan_kudo:matrix.org>

16:51:21

yes

<@conan_kudo:matrix.org>

16:51:28

we're going to rebase to 6.2

<@frantisekz:fedora.im>

16:51:45

okey dokey, that fits before the final freeze, right? or after?

<@conan_kudo:matrix.org>

16:51:51

yes

<@conan_kudo:matrix.org>

16:52:13

I believe 6.2 final releases October 3

<@conan_kudo:matrix.org>

16:52:21

!link https://community.kde.org/Schedules/Plasma_6

<@frantisekz:fedora.im>

16:52:32

right, Final freeze is on 15th Oct

<@conan_kudo:matrix.org>

16:52:58

I think we'll probably merge 6.2 beta into F41 this week so we can be prepared for that

<@frantisekz:fedora.im>

16:53:14

yeah, that makes sense, to get some early exposure, imo :)

<@geraldosimiao:matrix.org>

16:53:17

That's good

<@conan_kudo:matrix.org>

16:53:18

right

<@conan_kudo:matrix.org>

16:53:28

we already caught a few things from landing in Rawhide :)

<@conan_kudo:matrix.org>

16:53:35

so I'm pretty pleased how that's coming along

<@geraldosimiao:matrix.org>

16:53:47

I'll be glad to test it on baremetal

<@derekenz:fedora.im>

16:54:00

6.2 is looking good in F42

<@conan_kudo:matrix.org>

16:54:14

I'm still hoping that I can convince upstream to shift the Plasma cadence to be more in alignment with us

<@conan_kudo:matrix.org>

16:54:28

we've gotten lucky in F39 and F41 to have schedule alignment (sort of)

<@conan_kudo:matrix.org>

16:54:34

but I don't want to count on that :)

<@frantisekz:fedora.im>

16:55:00

:) , (don't) look over at LLVM.... which is still coming to f42, f41

<@frantisekz:fedora.im>

16:56:47

I'll also have two personal notes (regarded mainly at Intel iGPU users) :

<@frantisekz:fedora.im>

16:56:47

- I'll be rebasing media-driver in F41 tomorrow, so testing vaapi there would be welcome

<@frantisekz:fedora.im>

16:56:47

- I'll be rebasing compute-runtime (opencl and L0) later this week (also as the last rebase for f41, f42 as Intel axed support for anything released before 2020, will be filing change proposal for F43 for that)

<@conan_kudo:matrix.org>

16:57:13

oh interesting

<@conan_kudo:matrix.org>

16:57:25

F43 and not F42?

<@frantisekz:fedora.im>

16:57:57

yeah, I'd try to give users of older gpus one more release with compute support before moving to Xe-Only world with that

<@frantisekz:fedora.im>

16:58:56

but anyhow, thanks everybody for coming, I'll secretalize the one change in a bit, and see ya all in a week!

<@frantisekz:fedora.im>

16:58:57

!endmeeting