15:00:31 #startmeeting Fedora ARM and AArch64 Status Meeting 15:00:31 Meeting started Tue Jun 11 15:00:31 2019 UTC. 15:00:31 This meeting is logged and archived in a public location. 15:00:31 The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:31 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:31 The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting' 15:00:31 #chair pwhalen pbrobinson 15:00:31 Current chairs: pbrobinson pwhalen 15:00:45 #topic roll call 15:00:53 good morning folks, who's here today? 15:02:08 * pbrobinson o/ 15:02:18 * jlinton waves 15:04:57 hey jlinton 15:05:10 Hi 15:05:51 I've still not fully gotten ld.gold working, my patch is functional the problem is that i'm slowly cleaning up the unit tests 15:07:25 * satellit listening 15:07:27 jlinton, sounds like progress 15:07:33 which is a can of worms because I think the latest gcc's are generating relocations/etc that the aarch64 portion doesn't understand 15:07:34 lets start 15:07:41 so most of the debugging is happening on x86 15:07:46 ok 15:07:58 #topic 1) ==== Userspace Status ==== 15:08:23 any issues to discuss? 15:09:28 I have one 15:09:38 https://bugzilla.redhat.com/show_bug.cgi?id=1705676 15:09:57 jlinton: I don't know if you have some cycles to look at the tiancore on qemu console issue 15:10:09 #info edk2 doesn't always enabled the graphical display on arm/aarch64 15:10:18 #link https://bugzilla.redhat.com/show_bug.cgi?id=1705676 15:11:53 Well, i'm full up, so any serious debugging is in the air, although this sounds like my kind of bug 15:11:57 there is a new stable tag for edk2 but I've not seen it imported/built yet 15:12:24 jlinton: it was intermittent, it seems to have dropped off some 15:12:34 What machine were you using? AKA, was it really slow and the spice/etc server wasn't up and running early enough? 15:12:53 jlinton: mustang generally 15:13:03 I tend to use virt-manager all the time when i'm running a gui and it seems consistent 15:13:20 ok, not a rpi.. 15:13:26 nope 15:13:28 heh 15:14:09 I found that it was just variable sometimes you'd get display, and if you had the virtmanager console open you'd even get the tianocore logo 15:14:26 and other times you get a "graphic display not available" style error 15:15:13 There is a boot delay in there, right, did you try turning that on? 15:15:57 jlinton: nope, not tried anything except the defaults :-/ 15:16:09 (maybe that is x86, have to check) 15:18:41 any other user space issues? 15:19:27 none I'm aware of 15:20:05 #topic 2) ==== Kernel Status ==== 15:20:16 #info F31: kernel-5.2.0-0.rc4.git0.1.fc31 15:20:16 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1285015 15:21:00 I think we're mostly good here, I pulled in a couple of small fixes for rc4 but I think 5.2 is finally on the straight and narrow. I'll be updating a lot more systems this week 15:21:02 5.2 rc4 looks pretty good so far, is anyone seeing any issues? 15:22:25 #info F30: kernel-5.1.8-300.fc30 15:22:25 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1283779 15:22:53 #info Please test and report any issues to the list or #fedora-arm. 15:24:17 #topic 3) ==== Bootloader Status ==== 15:24:41 #info New grub2 available for testing - grub2-2.02-87.fc31 15:24:42 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1285376 15:25:19 U-Boot 2019.07-rc4 came out today. I'm actually preparing to push a build, need to move things around for python3 so that's something I knew I had to do, probably not quite so soon 15:26:26 sounds good, thanks 15:27:21 #topic 4) ==== Fedora 31 ==== 15:27:34 #info SoaS initramfs fix added to fedora-kickstarts (not yet merged) 15:27:35 #link https://pagure.io/fedora-kickstarts/pull-request/531 15:27:47 satellit, ^ :) 15:28:08 * satellit : ) 15:28:30 any other issues in f31? 15:28:53 composes have been failing, looks like various reasons- none of which are arm (that I saw) 15:29:16 * satellit workstation on RpiB3+ ? why old build 15:29:33 satellit: see point on composes failing 15:29:42 l 15:29:47 k* 15:29:47 satellit, no new image yet. It will update when we have one 15:30:11 #info Last successful compose - Fedora-Rawhide-20190609.n.1 15:30:20 #link https://kojipkgs.fedoraproject.org/compose//rawhide/Fedora-Rawhide-20190609.n.1/ 15:31:25 seeing module conflicts on updates when I have none installed, not sure if that's reported yet 15:33:03 satellit, any other issues with the desktops you've tested? 15:33:46 KDE not starting on latest here..wayland vs classic option? 15:34:29 we're probably going to kill off kde and a bunch of the secondary desktop images 15:34:36 I must send that email out 15:34:46 might be a rpi specific issue, I think it worked ok in qemu 15:35:06 I've been bouncing between kde/gnome/lxqt on my thunderx 15:35:17 is wayland compositor supported (KDE)? 15:35:19 it seems to be dependent on what exact package version is installed 15:35:29 (i'm not doing wayland on kde, only gnome) 15:35:59 Last week or so I think all three were working 15:36:18 this is with a gt710 15:36:33 wayland has worked fine on a couple of devices I have 15:37:29 it even works on the rpi with gnome when there's enough memory to go around, which there's often not# 15:38:15 * satellit anyway to get classic option in gnome and KDE from KS? 15:39:39 satellit: in some cases wayland will work and classic won't depending on the device you have. RPi will work on either as it does xorg emulation 15:39:41 satellit: Post install script? 15:40:01 some driver don't support the glamour xorg option 15:40:28 satellit: Just add a dnf group install "XXXX" 15:40:52 there might be a better way 15:41:02 * satellit only a tester :/ 15:42:13 #topic 5) == Open Floor == 15:43:13 anything else for open floor? 15:44:03 nothing from me 15:47:37 #endmeeting