15:00:20 <pwhalen> #startmeeting Fedora ARM and AArch64 Status Meeting
15:00:20 <zodbot> Meeting started Tue Apr 30 15:00:20 2019 UTC.
15:00:20 <zodbot> This meeting is logged and archived in a public location.
15:00:20 <zodbot> The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:20 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:20 <zodbot> The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting'
15:00:21 <pwhalen> #chair pwhalen pbrobinson jlinton
15:00:21 <zodbot> Current chairs: jlinton pbrobinson pwhalen
15:00:29 <pwhalen> Good morning folks, who's here today?
15:01:02 * pbrobinson o/
15:01:12 * jlinton waves
15:01:57 * pwhalen sees the usual suspects
15:02:17 <pwhalen> #topic 0) ==== Fedora 30 Available now! ====
15:02:17 <pwhalen> #link https://fedoramagazine.org/announcing-fedora-30/
15:02:52 <pwhalen> Fedora is out today, anyone have any comments, thoughts?
15:03:07 <pbrobinson> mostly good from my perspective
15:03:18 <pbrobinson> seems fine across a bunch of Arm devices
15:03:33 <pbrobinson> #info please report any issues seen with the new Fedora 30 release
15:03:44 <pwhalen> same, some issues upgrading for folks, but that looks to be across all arches
15:04:45 <pbrobinson> pwhalen: got a link to common bugs we can link here?
15:05:07 <pwhalen> #link https://fedoraproject.org/wiki/Common_F30_bugs
15:05:38 <pbrobinson> thanks
15:06:04 <pwhalen> anything else for f30?
15:06:45 <pbrobinson> not that I can think of
15:06:50 <pwhalen> #topic 1) ==== Userspace Status  ====
15:07:18 <pbrobinson> I suppose the two are the ARMv7 issue
15:07:36 <pbrobinson> ARMv7 builder VM issue to be more precise
15:08:09 <jlinton> I had a m400 for testing, but got pulled into another fire.
15:08:12 <pwhalen> I saw your update on the bug, will test 4.18 with f30 userspace
15:08:31 <jlinton> but the seattle builds F29-armv7 no LPAE were working fine
15:08:48 <jlinton> (thats F29 host)
15:08:50 <pbrobinson> so I read through it this morning and reviewing again all the bits it looks like a regression in LPAE from 4.18 to 4.19
15:08:53 <jlinton> F30 guest
15:09:09 <pbrobinson> jlinton: sure, but we don't have them in the infra, and that's the problem we have sadly
15:09:33 <jlinton> yah, I just wanted to try the exact same combo on the m400 and see if I could get it failing
15:09:35 <pbrobinson> it's very likely to be a problem that probably comes back to or is more easily recreated on the X-Gene 1
15:09:38 <pwhalen> #info Fedorn 29 armv7 guests "pause" from time to time on rhel-alt 7.6 kernel-4.14.0-115.2.2.el7a.aarch64 kernel
15:09:46 <pwhalen> #link https://bugzilla.redhat.com/show_bug.cgi?id=1576593
15:09:48 <jlinton> then switch to LPAE in the guest
15:10:42 <jlinton> Are the infrastructure guys sold on RHEL or will they switch to fedora?
15:10:54 <pwhalen> I think they will use whatever works
15:10:55 <jlinton> I thought they were running RHEL because it worked on m400...
15:11:29 <pbrobinson> jlinton: well I believe pwhalen has recreated the problem on the mustang with Fedora aarch64 as the host too
15:12:01 <jlinton> Well mustang has a number of problems not on the B0+ cores
15:12:23 <jlinton> and i don't think fedora is carrying patches for those.
15:12:31 <pbrobinson> jlinton: I believe the B0+ is what is shipped in the moonshot
15:12:35 <jlinton> right
15:12:48 <jlinton> so it should be better than the mustang
15:13:10 <pwhalen> perhaps its worth trying Fedora host again, F30 GA
15:13:29 <nirik> I thought I tried fedora as host, but I could be misremembering (since this has been ongoing for so long)
15:14:01 <pwhalen> nirik, pretty sure we both did.. and right, its been a while
15:14:10 <jlinton> (i'm just trying to say we need to be careful with fedora as a host on a mustang)
15:16:28 <pwhalen> I will add some updates today/tomorrow.
15:16:33 <jlinton> BTW: Does anyone know if there is a public m400 somewhere?
15:17:36 <pwhalen> I dont know of one
15:19:18 <pbrobinson> jlinton: is there a specific reason for the public one over the one you mentioned you had?
15:20:13 <jlinton> Yah, I'm trying to find one so I can pass the buck to a kernel maintainer..
15:20:16 * jlinton chuckles
15:20:31 <jlinton> Preparing for the future an all that...
15:21:00 <pbrobinson> nirik: do we have a spare one in Fedora infra that could have temporary access?
15:21:32 <nirik> possibly, but not easily... they are all in use, could take a builder out I suppose...
15:22:58 <pbrobinson> I suppose it depends on how quickly jlinton thinks it might accelerate the debug
15:24:00 <jlinton> Well we need a clean reproduce, which is my first goal, but once that happens I can try and catch someone attention, the problem is that if its only m400 the desire to debug it might be low
15:26:35 <pwhalen> I planned to try to reproduce else where, I can try seattle
15:27:47 <jlinton> An a72 platform would get the most interest
15:28:01 <jlinton> but seattle is good too
15:28:36 <pwhalen> ok, will put anything I can find in the bug.
15:28:47 <pwhalen> shall we move on?
15:30:11 <pwhalen> #topic 2) ==== Kernel Status ====
15:30:24 <pwhalen> #info F30: kernel-5.0.9-301.fc30
15:30:24 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1253909
15:30:41 <pwhalen> not aware of any issues with 5.0.9
15:31:30 <pbrobinson> I've not seen anything of note
15:31:48 <pwhalen> #info F31: kernel-5.1.0-0.rc7.git0.1.fc31
15:31:48 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1260063
15:32:36 <pwhalen> Also not aware of any issues with 5.1
15:32:48 <pwhalen> when is that supposed to go to f30?
15:35:25 <pwhalen> guess it will be a while yet.
15:35:30 <pwhalen> anything else for kernel news?
15:35:40 <pwhalen> #info Please test and report any issues to the list or #fedora-arm.
15:36:06 <pbrobinson> so we should have 5.1 GA next week
15:36:12 <pbrobinson> barring any issues
15:36:36 <pwhalen> then another week or so til it lands in f30?
15:36:39 <pbrobinson> and it will go to stable around the 5.1.3 time frame, so probably around 4 weeks time
15:36:47 <pwhalen> ok, so a fair amount
15:36:48 <pwhalen> thanks
15:37:28 <pwhalen> #topic 3) ==== Bootloader Status ====
15:37:32 <pbrobinson> the usual rebase formula is the .3 release, sometimes it's .2 or .4 and the stable releases are usually but not always weekly
15:37:38 <pwhalen> #undo
15:37:38 <zodbot> Removing item from minutes: <MeetBot.items.Topic object at 0x7f7ee018c490>
15:37:49 <pbrobinson> that's it
15:37:51 <jlinton> Is anyone else noticing that grub/F30 doesn't seem to save the last boot selection?
15:38:17 <jlinton> I'm not sure its a arm, machine, or f30 problem
15:38:22 <pbrobinson> jlinton: was that an upgrade?
15:38:44 <pbrobinson> it's likely to be a BLS problem
15:40:18 <pbrobinson> #topic 3) ==== Bootloader Status ====
15:40:53 <pbrobinson> jlinton: make sure you have the latest grub and then reinstall the kernel, there was a few issues, I think most should now be resolved
15:40:56 <pwhalen> I havent noticed it myself, but will look now
15:41:17 <pbrobinson> the maintainer for that is pretty responsive
15:42:56 <pwhalen> there were definitely issues with upgrades on all arches
15:43:21 <pwhalen> has anyone had any problems with uboot?
15:43:23 <pbrobinson> there's a new u-boot 2019.07-rc1 release but it might be a while before that lands
15:43:43 <pbrobinson> I have one issue remaining to test that I saw
15:43:52 <jlinton> (Oh, yah this was one of the RC's from a couple weeks ago, clean install)
15:44:00 <pbrobinson> but the vast majority issues I saw have been resolved
15:44:37 <pbrobinson> jlinton: I  would update to the latest grub2 before the next kernel and see if the problem goes away, I suspect it's fixed
15:47:13 <pwhalen> #topic 4) ==== Fedora 31 ====
15:47:24 <pwhalen> #link https://fedoraproject.org/wiki/Releases/31/ChangeSet
15:47:42 <pwhalen> #link https://fedoraproject.org/wiki/Releases/31/Schedule
15:47:50 <pbrobinson> we have until end of August until the first (Beta freeze) so there's a summer of fun!
15:48:20 <jlinton> I had the last min F30 32-bit container feature sitting on ice, but I didn't fix it for f31, should we consider it for F30?
15:48:30 <jlinton> for F32
15:49:30 <pbrobinson> jlinton: what is it? Any reason it couldn't be fixed for f31 (the current rawhide)
15:49:50 <jlinton> Maybe, I think the rpm issue has been picked up
15:50:48 <pbrobinson> ah, that one, yep we'll keep an eye on the rpm side of things. There may be other tweaks needed there but I think we should be OK for f31 for that
15:52:19 <jlinton> Yah, I will retest it on on my rawhide install when I get a moment
15:53:40 <jlinton> Can we turn config compat on, just for rawhide?
15:54:14 <jlinton> and then make the decision whether to leave it on when F31 forks?
15:56:06 <pbrobinson> jlinton: you mean in the kernel?
15:56:49 <jlinton> yes
15:57:36 <pbrobinson> it's in the F-30 kernel (and possibly f29) already
15:57:49 <pbrobinson> so I think all that bit is there and I've had reports it works fine
15:58:55 <jlinton> Oh, hehe
15:59:26 <pwhalen> oops, almost 12
15:59:33 <pwhalen> #topic 5)  == Open Floor ==
16:00:01 <pwhalen> lets move over to #fedora-arm
16:00:06 <pwhalen> #endmeeting