15:06:00 <pwhalen> #startmeeting Fedora ARM and AArch64 Status Meeting 15:06:00 <zodbot> Meeting started Tue Apr 21 15:06:00 2020 UTC. 15:06:00 <zodbot> This meeting is logged and archived in a public location. 15:06:00 <zodbot> The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:06:00 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:06:00 <zodbot> The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting' 15:06:00 <pwhalen> #chair pwhalen pbrobinson jlinton 15:06:00 <zodbot> Current chairs: jlinton pbrobinson pwhalen 15:06:00 <pwhalen> #topic roll call 15:06:37 <pwhalen> Sorry I'm late, had a kids pool flying around outside I had to catch. Really high winds here 15:07:09 * pbrobinson is here 15:07:10 <pwhalen> heh :) 15:07:28 <paulisaacs> New attendee here 15:07:48 <pwhalen> howdy paulisaacs, welcome! 15:08:13 * jlinton waaves 15:08:23 <paulisaacs> Thanks - representing 'Linaro' 15:08:52 <pbrobinson> paulisaacs: that's a broad sweeping statement, any particular bit? Welcome 15:09:29 <paulisaacs> All things LDCG, HPC-SIG and AI on Servers...so quite broad but not a broad as the whole Linaro 15:10:15 <pbrobinson> cool 15:10:32 <pwhalen> good to have you. lets get started 15:10:34 <pwhalen> #topic 1) ==== Userspace Status ==== 15:10:45 <pwhalen> any new issues? 15:10:57 <pwhalen> I really hope not :) 15:11:10 <pbrobinson> nothing of note 15:11:21 <jlinton> I've tracked down the odd cursor behavior on remote X'ing to not sourcing /etc/X11/Xresources on ssh login 15:11:24 <pbrobinson> paulisaacs: btw just chip in where ever, ask questions etc 15:11:38 <jlinton> Although, i'm not clear it its intentional 15:11:41 <paulisaacs> Thanks, I will 15:11:52 <pwhalen> jlinton, is this on f32? 15:12:07 <pbrobinson> jlinton: OK, I would probably file a bug, that would likely be the quickest way to find out intention 15:12:10 <jlinton> Yah, its actually been there since at least F31, it appeared somewhere in that timeframe 15:12:30 <jlinton> Its been on my, list of generic fedora problems to follow up on. 15:12:51 <pwhalen> right, agree, bug would great. 15:13:59 <jlinton> Anway, it seems the display dpi isn't getting set 15:14:47 <pwhalen> ok, thanks 15:14:50 <pwhalen> #topic 2) ==== Kernel Status ==== 15:15:03 <pwhalen> so we have a new kernel for F32, included in the latest RC 15:15:10 <pwhalen> #info Latest kernel-5.6.5-300.fc32 15:15:11 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1495011 15:15:30 <pbrobinson> it seems fine with my testing 15:15:40 <pbrobinson> there's a 5.6.6 being built ATM 15:16:25 <pwhalen> no issues here either, if folks could test it all around and make sure there no regression, it would be greatly appreciated 15:17:36 <pwhalen> Any other kernel news? 15:17:49 <pwhalen> #info Please test and report any issues to the list or #fedora-arm. 15:18:03 <pwhalen> #topic 3) ==== Bootloader Status ==== 15:18:04 <pbrobinson> the 5.7-rc2 is built in rawhide too 15:18:17 <pwhalen> #undo 15:18:17 <zodbot> Removing item from minutes: <MeetBot.items.Topic object at 0x7fdff2ad0e90> 15:18:44 <pbrobinson> just continue, it was FYI 15:18:49 <pwhalen> #info kernel-5.7.0-0.rc2.1.fc33 available for testing 15:19:00 <pwhalen> #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1496190 15:19:05 <pwhalen> sure 15:19:16 <pwhalen> #topic 3) ==== Bootloader Status ==== 15:19:24 <pwhalen> #info New U-Boot available for testing. Please test and add karma. 15:19:24 <pwhalen> #link https://bodhi.fedoraproject.org/updates/FEDORA-2020-e0ad98f651 15:19:36 <jlinton> So, its 5.7 that adds the image_base check and pops 15:19:37 <jlinton> https://bugzilla.redhat.com/show_bug.cgi?id=1825411 15:19:41 <pwhalen> Just started to test this 15:20:17 <pwhalen> thanks for the bz, jlinton 15:20:18 <jlinton> It might actually be a grub problem on aarch32, since I've seen it on two machines, (rpi & honeycomb) although It could be edk2 15:21:25 <pbrobinson> jlinton: which 5.7? There was some fixes in rc2 that Linus called out in his announcement 15:22:03 <pbrobinson> to quote "The more noticeable one might be fixing the regression that was 15:22:03 <pbrobinson> due to the EFI loaders not clearing the BSS section and us having lost 15:22:04 <pbrobinson> the workaround for that. That caused boot problems for some." 15:22:04 <jlinton> I was on rc1, I have an rc2 I just booted, let me see if the patch got tweaked 15:22:14 <pbrobinson> https://lwn.net/Articles/817924/ 15:23:09 <jlinton> That looks like it might be the x86 version of the aarch fix that causes the warning (looking at it quickly) 15:23:57 <pbrobinson> the fixes came from Ard I think so I would hope it would be universal 15:24:59 <jlinton> No, the warning is still in rc2 15:25:19 <jlinton> I think he applied the same fix to x86 that was already applied to aarch64, and its that fix that causes the warning 15:25:52 <pbrobinson> ah 15:25:54 <pbrobinson> joy 15:26:41 <pwhalen> #topic 4) ==== Fedora 32 Final ==== 15:26:49 <pwhalen> #info Fedora 32 Final Blockers 15:26:49 <pwhalen> #link https://qa.fedoraproject.org/blockerbugs/milestone/32/final/buglist 15:26:58 <pwhalen> #info Latest nominated nightly 15:26:58 <pwhalen> #link https://fedoraproject.org/wiki/Test_Results:Current_Installation_Test 15:28:34 <pwhalen> Looks like we might need another RC, but we should have plenty of time for testing before the Go/NoGo 15:28:40 <pwhalen> any issues not filed? 15:29:02 <pbrobinson> none I'm aware of, I filed a couple of FE requests for U-Boot 15:29:31 <pbrobinson> I might substitute that built for a new one so might need to beg for karma again but I'm not sure yet 15:30:46 <pwhalen> no worries, let us know. 15:30:52 <pwhalen> Anything else for F32? 15:31:20 <pwhalen> #info F32 Final release(target) - 2020-04-28 15:31:39 <pwhalen> #info F32 Go/NoGo - 2020-04-23 15:31:57 <pwhalen> #topic 5) == Open Floor == 15:32:07 <pwhalen> That is all I have, anything for open floor? 15:32:14 <pbrobinson> I think over all we're looking pretty reasonable 15:32:37 <pwhalen> me too, I was ok with shipping this week.. alas 15:32:48 <pbrobinson> paulisaacs: was there anything in particular you wanted to discuss/bring up, or just seeing what happens? 15:33:14 <paulisaacs> Just an open comment from Linus' statement about x86 last year.... my challenge is to promote from a single compute module through to supercomputers that ARM is here (Linus - https://www.theregister.co.uk/2019/02/23/linus_torvalds_arm_x86_servers/) 15:33:46 <paulisaacs> When I get the chance I will be loading F32 on a Neoverse N1 platform 15:34:22 <pbrobinson> cool, I think jlinton tests on that, I've seen F-30 run on a N1 ~ 1 year ago 15:34:27 <jlinton> It works, you just need to apply the pci quirk 15:35:10 <jlinton> I have a semi updated image I distribute in house for people who don't want to keep applying the quirk 15:35:17 <paulisaacs> (I also personally own a Pine64 Clusterboard and enjoy trying to get distros booting on it, including Fedora) - so I'm involved in all scales 15:36:08 <paulisaacs> Happy to try with pci quirk 15:37:03 <pwhalen> let us know how it goes :) 15:37:21 <jlinton> Although, I guess if your "N1" platform is graviton/etc then you probably don't need the quirk 15:37:32 <paulisaacs> I will :) end goal is running a new/novel AI on it across CCIX 15:38:27 <jlinton> BTW: The bug reported on fedora-arm earlier with the rpi4 was the same thing I was seeing last week 15:38:33 <jlinton> (setup menu, no keyboard/etc) 15:38:43 <jlinton> Which set me off hacking the DT for a serial port 15:39:01 <jlinton> I have a possible upstream patch for the DT which sets up uart0/uart1 aliases 15:39:28 <jlinton> so the proprietary start4 firmware sets up the uarts meaninfully for ATF 15:39:55 <pbrobinson> OK 15:40:45 <jlinton> I need to clear/post another set though before I go back and verify it fixes the uboot serial problem too 15:42:13 <pwhalen> lets continue in #fedora-arm, I'm going to close this to grab some lunch 15:42:16 <pwhalen> #endmeeting