16:00:22 #startmeeting Fedora ARM and AArch64 Status Meeting 16:00:22 Meeting started Tue Jan 31 16:00:22 2017 UTC. The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:22 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00:22 The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting' 16:00:22 #chair pwhalen pbrobinson dgilmore hrw dmarlin yselkowitz jonmasters ahs3 msalter 16:00:22 Current chairs: ahs3 dgilmore dmarlin hrw jonmasters msalter pbrobinson pwhalen yselkowitz 16:00:25 o/ 16:00:36 * pbrobinson o/ 16:00:37 morning folks, who's here today? 16:01:34 morning 16:02:15 * pwhalen gives it a minute for others to join 16:04:08 ok, lets get started .. 16:04:15 #topic 1) ==== Userspace Status ==== 16:04:27 morning 16:04:36 I think most of userspace is looking OK 16:05:01 The issue rob has with gnome is now fixed in the relevant mozjs31 16:05:28 there's going to be a mass rebuild that should kick off on Friday for gcc7 16:05:50 for the first time ever I won't need to deal with it from a koji-shadow PoV so \o/ 16:06:04 #info Mass rebuild to start on Friday Feb 3rd. 16:06:12 heh, win 16:06:37 pbrobinson: ;D 16:06:38 is anyone aware of any other userspace issues? 16:06:49 pbrobinson: even s390* joined koji now? 16:07:04 and mesa 17.0 RCs have started to land in rawhide, one thing of note it now has the etnaviv driver which should give accelerated wayland on imx6 and other vivante devices 16:07:27 hrw: no, but it might actually be done before mass rebuild (and I never did those ones) 16:07:37 ok 16:08:11 #info Mesa 17.0 RCs have started to land in rawhide, one thing of note it now has the etnaviv driver which should give accelerated wayland on imx6 and other vivante devices. 16:09:01 ok, moving on .. 16:09:02 #topic 2) ==== Kernel Status ==== 16:09:13 #info Latest kernel-4.10.0-0.rc6.git0.1.fc26 16:09:20 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=837098 16:09:49 I think we're mostly good here. There's a CMA patch headed upstream to address some issues we've seen in 4.9+ (worse in 4.10) so I'll push that shortly 16:11:11 there's also a LOT of fixes/changes headed upstream for the RPi so that should be nice for F-26 when they land 16:11:30 very nice. 16:11:39 is anyone seeing issues on 4.10 not yet discussed here? 16:11:49 I'm not aware of any other hot issues with kernel. jonmasters? 16:12:23 jlinton: you too? 16:12:36 Well, ACPI->KVM is nearly there. 16:12:44 for PCIe devices (for 4.11) 16:12:46 nothing specific - few things in flight, but not *bad* stuf 16:13:15 BTW: Somewhat related, is there interest in packaging kvmtool? 16:13:26 jlinton: yes, I've been watching those. I think 4.11+ is a safe bet for F-26 so that'll be nice 16:14:06 jlinton: there's no reason not to, but then it doesn't integrate with libvirt etc does it? 16:14:14 My ACPI->PMU patches are still up in the air... (eye rolling) 16:14:30 no, kvmtool stands (mostly) by itself 16:14:49 it avoids libvirt/qemu which is good for quick sanity checks 16:15:48 yes, so that's what I thought, I have no issues with it being packaged, likely useful 16:17:10 howdy rric, thanks for joining.. 16:17:15 hi 16:17:50 we're just discussing the kernel, are you seeing any issues in 4.10? 16:17:54 rric: any kernel issues on your radar? 16:18:25 this pfn_valid() issue is still not fixed 16:18:37 I will talk to Will D again 16:18:53 rric: there's a patch headed upstream, I'm about to pull it into 4.9/4.10 16:19:03 rric: http://www.spinics.net/lists/linux-mm/msg119932.html 16:19:05 oh, ok, need to check 16:19:23 rric: it's a 1 liner 16:19:54 rric: let me know if it's a different problem but I think that's it 16:19:57 I think this is a separate fix 16:20:03 OK 16:20:07 Yah, I think I saw the problem on the rpi3 last week too... So hopefully that fixes it. 16:20:34 well I'm pushing it already to fix issues on the RPi so it'll go out anyway so you can let me know 16:21:34 The only time I ever see it trigger is with X running 16:23:29 #topic 3) ==== Bootloader Status ==== 16:24:40 we should have a new uboot RC shortly to test, i'm not aware of any current issues. 16:25:10 is anyone seeing anything we should be concerned about? 16:25:10 jlinton: yep, exactly what I see 16:26:04 so I've done some reading about ATF and I think we should have decent upstream Allwinner A64 support shortly which should open up devices like Pine64... just FYI 16:26:32 Yah, the hikey support got upstreamed last week 16:26:35 but its still missing USB 16:26:47 jlinton: in u-boot? 16:26:51 edk2 16:26:55 ah, OK 16:27:31 I've got a 1/2 working version that i've been messing with since i solder the inductors back on the board.. 16:27:33 I have no urge to work out how to build that (we already ship it) but it might be interesting for Fedora 16:28:05 Yah, the point being that it should be able to _INSTALL_ from a dvd/net boot/ once the USB support is in place 16:29:29 that'd be nice. 16:29:37 any other bootloader news/issues? 16:29:56 #topic 4) ==== Rawhide Testing ==== 16:30:12 #info Testing so far(armhfp): https://fedoraproject.org/wiki/Test_Results:Fedora_26_Rawhide_20170125.n.0_Summary 16:30:12 #info Testing so far(aarch64): https://fedoraproject.org/wiki/Architectures/AArch64/QA/Fedora-Rawhide-20170125.n.0 16:30:47 the 25th was the last nominated compose due to ongoing issues in rawhide 16:31:13 no new bugs for us, but the current list is here 16:31:24 #link https://qa.fedoraproject.org/blockerbugs/milestone/26/alpha/buglist 16:31:28 will begin pulling these in regularly now I have a beaker setup :) 16:31:48 weekend plan is to automate syncing these+importing them, etc. 16:31:54 still seeing this 16:31:57 #info AArch64 systems fail to display the grub menu when PXE booted and instead drop to a grub prompt (BZ#1405208) 16:32:24 pwhalen: Sounces like a firmware bug 16:33:34 reproducible on mustang and seattle. I'm not aware of new firmware for either, but will inquire 16:34:31 Yah, that makes it less likely a firmware bug, but I fixed/tracked a couple similar things on juno early last year, both failures to read the grub.cfg following startup 16:36:50 any other issues encountered testing rawhide? 16:38:40 #topic 5) == Open Floor == 16:39:35 anything left for open floor? 16:39:53 just that we're going into mass rebuild 16:40:23 then that'll be followed by branching and straight into Alpha. F-26 is well under way 16:40:46 Other than the gnome rebase, no issues with 48bit VA right? 16:41:55 none that I've seen 16:44:37 thanks for joining! 16:44:39 #endmeeting