15:00:07 #startmeeting Fedora ARM Working Group Meeting 15:00:07 Meeting started Tue Aug 23 15:00:07 2022 UTC. 15:00:07 This meeting is logged and archived in a public location. 15:00:07 The chair is pwhalen. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 15:00:07 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:07 The meeting name has been set to 'fedora_arm_working_group_meeting' 15:00:07 #chair pwhalen pbrobinson jlinton coremodule 15:00:07 #topic roll call 15:00:07 Current chairs: coremodule jlinton pbrobinson pwhalen 15:00:21 * jlinton waves 15:02:10 good morning 15:02:17 Howdy jlinton, how're you doing? 15:02:45 * pwhalen is setting up Minecraft on a RPi4, thanks jforbes! 15:02:48 * pbrobinson o/ 15:02:51 * coremodule is here 15:02:53 hi pwhalen 15:03:03 heh, cool 15:03:25 I remember setting up minecraft once to chase down a kernel bug, it's been a while though 15:03:38 My son can't get enough of it 15:03:45 #topic 1) ==== Stable Releases ==== 15:04:06 Any issues in our stable releases? 15:05:39 Minecraft is part of the default rpi image isn't it? IIRC my kids were playing with the pi400 too, and that is about the only thing they wanted to do with it. 15:06:06 TBH I have no idea 15:06:41 jlinton: perhaps it is, but this is Fedora! :) 15:06:58 #info Bug 2117817 - removal of grubby-deprecated is incorrectly handled 15:06:58 I don't know that I have ever booted a default rpi image 15:07:07 No, new stable issues on my side, the only problematic thing was the serial console in virt-manager, but I haven't had a chance to track it down, and i don't think that defect is moved. 15:08:11 thanks jlinton, did it affect F37 that you know of? 15:08:41 Yes it was in rawhide too, right before the branch, that is where i initially saw it. 15:09:36 Ok, thanks. 15:09:51 jlinton: there was a edk2 bug around that I thought that is now fixed? 15:11:00 jlinton: was it this bug? https://bugzilla.redhat.com/show_bug.cgi?id=2116534 15:11:20 or is that a different but related bug? 15:12:00 Yes that is the one. 15:12:54 yes, so it's progressing, I'm not sure if builds were done but I think they were testing it 15:12:58 so, I guess I missed "Builds done, rpms should be in rawhide and f36 updates testing soon." 15:14:37 Ok, we should keep an eye on that one and nominate it if we can find appropriate criteria 15:15:13 Any other issues? 15:15:59 pwhalen: so the first build landed for f37 yesterday so it's in stable there now 15:16:54 even better, just needs some testing 15:17:02 #topic 2) ==== Kernel Status = 15:17:15 #info kernel-5.19.3-300.fc37 15:17:15 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=2049703 15:17:37 5.19.4 is expected in a couple of days, and I plan to push that to f35/f36 as well 15:18:14 Upgrades in progress, no issues so far. Well, nothing terrible anyways. A few backtraces on the pi3 but everything is working. 15:18:49 there was a bunch of splats for rpi vc4 driver due to the 20 odd patches that landed in 5.19.2, it works fine still though, but I've reached out to the maintainer 15:18:54 The 5.19 test kernel seems to have passed all the fedora tests I ran last week. 15:18:58 thanks jforbes, we'll need to test armv7 when its available. 15:19:25 pwhalen: 5.19.2 had armv7 builds available for test week 15:19:29 excellent, thanks for testing jlinton. I was out on PTO last week 15:19:45 Ok. I'll check for those. 15:20:25 Less and arm, thing but maybe the fedora kernel tests should have the kernel self test now that they tend to pass on aarch64. 15:21:02 We do build them, just haven't been running them 15:21:02 jforbes: thanks 15:21:29 I didn't see the v7 ones but I'll grab and test on my devices 15:21:52 #info kernel-6.0.0-0.rc2.19.fc38 15:21:52 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=2050106 15:21:56 pbrobinson: 5.19.2-300.fc36 was built as an official for securei boto signing 15:22:04 err secure boot signing even 15:22:18 #info Please test and report any issues to #fedora-arm or the mailing list. 15:22:49 Also of note, pbrobinson is reviewing an MR to turn on apple silicon. Once that is ready, I will likely bring it back to 5.19 as well 15:23:21 Yah, I saw that. Doesn't it require a firmware shim in there too? 15:24:05 yep 15:24:14 it's basically step 1 15:24:16 Are we going to ship it, how will that work? 15:24:51 jlinton: TBH ATM I don't know, there's a few different people looking at it and most of those don't seem to be bothering to communicate 15:25:32 Don't know if we can or will at this point, but that part is a bit easier to grab from elsewhere. A working kernel gives us something to experiment with 15:25:33 The x13 requires a shim now too doesn't it (to enable el2 IIRC) 15:26:30 I don't think the M1 will be a "supported" platform quite so soon. Would love to see the X13 as one though. 15:28:11 Me too. Thanks for the updates. 15:28:16 #topic 3) ==== Bootloader Status ==== 15:28:24 #info uboot-tools-2022.10-0.2.rc1.fc37 15:28:24 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=2047365 15:29:51 Believe I tested the RPi3/4 , need to double check 15:30:58 I see we have RC3 building in Rawhide, are we planning to ship RC1 in Beta? 15:33:54 #topic 4) ==== Fedora 37 status === 15:33:54 there's a rc3 build happening ATM 15:34:05 #undo 15:34:05 Removing item from minutes: 15:34:07 pwhalen: depends on testing 15:34:17 ok, thanks 15:34:20 #topic 4) ==== Fedora 37 status === 15:35:03 #info OpenQA Testing - https://openqa.fedoraproject.org/tests/overview?distri=fedora&version=37&build=Fedora-37-20220823.n.0&groupid=5 15:35:51 coremodule: how does F37 look so far? 15:37:43 I have Workstation up on a RPi4, will check some of those. Be great if others could assist too. 15:38:26 noticed we have some spins failing for aarch64, xfce just needs more space. Didnt look at any others 15:40:12 Anything else for F37? 15:40:13 pwhalen, I haven't seen any major issues at all so far, so I think it's looking good. It seems to be stable so far. We had the first blocker review meeting of this cycle yesterday and there were a few accepted, but mostly as Final blockers, so we still have some time. 15:41:21 ok, thanks coremodule! 15:41:30 #topic 5) ==== Open Floor ==== 15:41:34 Anything else for today? 15:42:00 the main one will be that F37 probably won't have a working dnf for any of the devices with <= 1Gb RAM 15:43:10 was that one punted? 15:43:38 * pwhalen will check the logs, thought it was 15:43:52 The systemd-boot thing is still WIP.. , but the noteworthy thing here, is that I think we should remove grub from the base configs, and depend on anaconda putting it back. 15:45:18 https://pagure.io/fedora-comps/blob/main/f/comps-f37.xml.in#_481 15:45:38 I just wanted to mention armv7 builders are happy again on 5.19.x 15:45:45 yeah, it was delayed for onow 15:46:07 nirik \0/ 15:46:09 #link https://bugzilla.redhat.com/show_bug.cgi?id=1907030 15:46:53 jlinton: interesting. would that need an official change? or would that happen in F37? 15:47:07 Probably F38 at the earliest 15:47:37 Sounds right, thanks 15:47:48 and i'm not sure about the official change, likely for the comps that should be noted, but at the moment the option is sorta hidden and only available via kickstarts (aka I didn't add a GUI option yet) 15:48:22 the remaining gap is really around grubby and commandline tweakage, as well as a few missing bits, like the compressed kernel work Ard is doing. 15:48:47 The official change, is something I think would happen once we want it to be "officially" supported/etc. 15:48:53 vs just an option that can be tested. 15:49:19 nirik: excellent 15:50:59 Cool. 15:51:05 Anythign else for open floor? 15:52:08 #endmeeting