15:00:30 <pwhalen> #startmeeting Fedora ARM and AArch64 Status Meeting
15:00:30 <zodbot> Meeting started Tue Apr  5 15:00:30 2022 UTC.
15:00:30 <zodbot> This meeting is logged and archived in a public location.
15:00:30 <zodbot> The chair is pwhalen. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
15:00:30 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:30 <zodbot> The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting'
15:00:30 <pwhalen> #chair pwhalen pbrobinson coremodule jlinton
15:00:30 <zodbot> Current chairs: coremodule jlinton pbrobinson pwhalen
15:00:30 <pwhalen> #topic roll call
15:00:39 <pwhalen> Good morning folks, who's here today?
15:00:53 <bcotton> .hello2
15:00:54 <zodbot> bcotton: Something blew up, please try again
15:00:56 <coremodule> hi pwhalen, good morning
15:00:57 <zodbot> bcotton: An error has occurred and has been logged. Please contact this bot's administrator for more information.
15:01:25 <bcotton> no, zodbot. you get one chance
15:02:47 <pwhalen> How are you fine folks doing today?
15:02:53 * pbrobinson is here
15:02:58 * jlinton waves
15:03:11 <pwhalen> OK, lets get started
15:03:13 <pwhalen> #topic 1) ==== Userspace Status  ====
15:03:26 <pwhalen> Any new issues?
15:03:36 <jlinton> I guess a compiler bug counts as userspace?
15:03:37 <jlinton> https://jira.arm.com/browse/GNUTOOLS-13382
15:04:02 <jlinton> no fedora bug yet, unless we want to count the rpi genet problem, but that is where it went.
15:04:21 <jlinton> Ops thats the wrong link
15:04:43 <jlinton> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105160
15:07:51 <pwhalen> thanks jlinton
15:08:08 <pbrobinson> jlinton: do we think this will cause other issues like the genet bug?
15:08:16 <pbrobinson> s/bug/issue
15:08:20 <jlinton> Yah, its likely the root of the genet bug
15:09:00 <jlinton> whatever it ends up being, the problem of course is that (say its directly the volatile asm) any package with inline assembly will need to be rebuilt.
15:10:00 <pbrobinson> could we see issues with other kernel modules for example?
15:10:12 <pbrobinson> I somehow doubt genet is the only outcome of this
15:10:15 <jlinton> Yah, or the core kernel itself (say not booting)
15:10:51 <pbrobinson> jlinton: could this be causing the issues with mustang/seatle?
15:11:08 <jlinton> yah, since that was a gcc 11/12 problem too AFAIK
15:11:21 <jlinton> and this bug appears to discard inline assembly, when it shouldn't be.
15:13:56 <pbrobinson> so is arm providing a fix for upstream?
15:14:50 <jlinton> I don't know who will provide the fix yet, the second/correct link I posted is the upstream report.
15:15:07 <pbrobinson> yes, I'd seen that
15:16:43 <pwhalen> any other issues?
15:17:00 <pwhalen> #topic 2) ==== Kernel Status ====
15:17:30 <pwhalen> #info kernel-5.17.1-300.fc36
15:17:30 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1939765
15:17:45 <pwhalen> oh, 5.18 too
15:18:33 <pwhalen> Has anyone has any issues with 5.17.1?
15:19:24 <pwhalen> #info 5.17.2 is expected to have the RPi4 ethernet fix, it should be available later this week.
15:19:41 <coremodule> woot! no issues seen here.
15:19:49 <pwhalen> Doesn't look like the mustang/seattle boot issue is going to be fixed
15:20:47 <pwhalen> #info kernel-5.18.0-0.rc1.18.fc37
15:20:47 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1942573
15:20:52 <jlinton> Unless the compiler is the problem, I need to pull the video card out of mine as nouveau isn't stable in it anymore.
15:21:05 <pwhalen> #info Please test and report any issues to the list or #fedora-arm.
15:22:19 <pwhalen> #topic 3) ==== Bootloader Status ====
15:22:28 <pwhalen> #info uboot-tools-2022.04-1.fc36
15:22:28 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1942618
15:22:36 <pwhalen> #info Please test and add karma - https://bodhi.fedoraproject.org/updates/FEDORA-2022-4aa9456e08
15:23:40 <pwhalen> Looks like its headed to stable, but please check your favourite hardware to make sure we don't have any regressions. Going to check the pinebook pro after the meeting. Will then test user switching (as requested by coremodule).
15:24:08 <pwhalen> I dont think anyone tested the nano yet
15:24:17 <coremodule> Thanks pwhalen
15:25:17 <pwhalen> #topic 4) ==== Fedora 36 ====
15:25:26 <pwhalen> #info Current testing results:
15:25:26 <pwhalen> #link https://fedoraproject.org/wiki/Test_Results:Fedora_36_Branched_20220401.n.0_Summary
15:25:49 <pwhalen> Looks pretty bare
15:26:00 <pbrobinson> I've tested the nano
15:26:49 <pwhalen> #info OpenQA Testing - https://openqa.fedoraproject.org/tests/overview?distri=fedora&version=36&build=Fedora-36-20220401.n.0&groupid=5
15:27:12 <pwhalen> pbrobinson: nice, thanks
15:27:32 <pwhalen> #info List of Blockers for Fedora Linux 36 (Final)
15:27:32 <pwhalen> #link https://qa.fedoraproject.org/blockerbugs/milestone/36/final/buglist
15:27:45 <pwhalen> #info Final Freeze - 2022-04-05 (Today)
15:27:45 <pwhalen> #info Fedora Linux 36 GA - Target Release 2022-04-19
15:29:05 <pwhalen> Looking at the list, the RPi4 ethernet freeze exception is the only arm specific issue
15:30:25 <pwhalen> The dnf update issue for the RPi zero 2, was rejected because we don't officially support the device, but I think the case could be made it should be accepted for cloud instances
15:31:29 <jlinton> I guess the question is, what happens if we have a critical compiler bug?
15:33:45 <pwhalen> And the extent isn't really known yet
15:34:28 <pbrobinson> pwhalen: yes, I was perplexed about the focus on the device and not the amount of memory
15:36:02 <coremodule> I was under the impression that the amount of memory is not what caused that bug...
15:36:06 <pwhalen> pbrobinson: I think we should test a small cloud instance
15:37:25 <pwhalen> I have to use a swapfile on the BeagleBone White, but I've been doing that for a while
15:42:12 <pwhalen> People have also hit the memory issues with cloud/dnf before - https://discussion.fedoraproject.org/t/cannot-install-updates-oom/32801
15:43:10 <jlinton> Well, we could turn a bunch of things which are unlikely on aarch64/workstation off by default, iscsi support comes to mind,
15:44:48 <jlinton> It would probably also decrease boot times if we can get rid of some of the stuff that is used by <1% of the users.
15:44:48 <pwhalen> definitely
15:46:00 <pwhalen> sounds like something for later releases or notes to add to the tweaks section for now
15:46:55 <pwhalen> be interested to see if cloud works with 512M now
15:47:12 <pwhalen> #topic 5)  == Open Floor ==
15:47:18 <pwhalen> Anything else for today?
15:47:41 * bcotton waves
15:48:09 <pwhalen> all yours bcotton
15:48:27 <bcotton> i wanted to give the team a chance to tell me https://pagure.io/fedora-docs/quick-docs/pull-request/442 is terrible before I go merge it :-)
15:48:47 <bcotton> and also I'm available if we want to discuss the location of the FAQ (currently in both the wiki and docs, which is Bad™)
15:51:28 <pwhalen> Is the FAQ section the Raspberry Pi wiki page? Or is it from somewhere else?
15:51:43 <bcotton> yes
15:52:08 <bcotton> specifically https://docs.fedoraproject.org/en-US/quick-docs/raspberry-pi/ and https://fedoraproject.org/wiki/Architectures/ARM/Raspberry_Pi#Raspberry_Pi_4
15:52:39 <bcotton> separately, Peter Boy raised a discussion of moving the docs out of the wiki entirely, but that's a matter for further discussion first
15:54:11 <pwhalen> I think its safe to say graphics work but not accelerated so performance is affected
15:54:34 <pwhalen> otherwise, thanks for fixing that bcotton
15:54:36 <jlinton> Yah, I was just going to say that.
15:54:49 <bcotton> ack, i'll update it to include that
15:55:26 <pwhalen> thanks
15:55:30 <SallyAhaj> pwhalen: It's affecting locks up more than performance
15:56:22 <pwhalen> the lock ups seemed better, but its been a while since I tested workstation.
15:56:23 <SallyAhaj> I'm a real GUI RPi4 user, there isn't a realibility on it to do anything
15:56:50 <jlinton> And the rpi3/4 actually don't need to use the arm image installer either, so...
15:56:51 <pwhalen> even with xorg?
15:56:52 <SallyAhaj> It's better, but not for real use with GUI
15:57:23 <SallyAhaj> pwhalen: yes,I tried gnome/XFCE,sway
15:57:35 <pwhalen> SallyAhaj: if you work out a reproducer when running xorg, please share. Even with wayland I had trouble last time I was testing
15:57:39 <jlinton> for the rpi3, since its DT only the edk2 port has at least as much functionality as the arm image installer, so the normal fedora boot/install media works.
15:57:55 <SallyAhaj> all the same locks up frequency
15:59:14 <SallyAhaj> pwhalen: There's nothing to see, I just can access to the system with mobile (ssh) to reboot,
16:00:54 <jlinton> Hmm, I don't see rpi4 lockups, I would worry that something like that is caused by powersupply/disk issues at this point.
16:00:58 <pwhalen> RIght, knowing how to reproduce it allows others to test without using the desktop for a long time
16:01:01 <SallyAhaj> Such locks up/freezes never happened while using downstream kernel (I'm on F36 with downstream RPMFusion)
16:01:40 <pwhalen> Thanks for coming folks, we're overtime
16:01:43 <pwhalen> #endmeeting