15:00:07 <pwhalen> #startmeeting Fedora ARM and AArch64 Status Meeting
15:00:07 <zodbot> Meeting started Tue Jun 21 15:00:07 2016 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:07 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:07 <zodbot> The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting'
15:00:08 <pwhalen> #chair pwhalen pbrobinson dgilmore hrw dmarlin yselkowitz jonmasters ahs3 msalter
15:00:08 <zodbot> Current chairs: ahs3 dgilmore dmarlin hrw jonmasters msalter pbrobinson pwhalen yselkowitz
15:00:13 <pwhalen> morning folks, whos here today?
15:00:25 * pbrobinson /o
15:00:28 <hrw> o/
15:00:30 <pbrobinson> o/
15:00:31 <pwhalen> #chair pwhalen pbrobinson dgilmore hrw dmarlin yselkowitz jonmasters ahs3 msalter
15:00:31 <zodbot> Current chairs: ahs3 dgilmore dmarlin hrw jonmasters msalter pbrobinson pwhalen yselkowitz
15:01:24 * cov is attending for the first time
15:01:27 <pwhalen> i think it should be a quick one this week.. i'll give it a couple of minutes for others to join
15:01:39 <pwhalen> welcome cov, thanks for joining
15:03:23 <pwhalen> alright. lets get started..
15:03:29 <pwhalen> #topic 0) ==== Fedora 24 Released today! ====
15:03:35 <pwhalen> #link https://fedoraproject.org/wiki/Architectures/ARM/F24/Installation
15:03:42 <pwhalen> #link https://fedoraproject.org/wiki/Architectures/AArch64/F24/Installation
15:04:14 <pwhalen> we released today! w00t w00t.. ive updated the install pages, of folks could have a look to see if i missed anything, or ...
15:04:38 * coremodule is here.
15:05:12 <pwhalen> not many known issues for either, the same encrypted partition bug that been around for a while on aarch64, but it now looks to have some traction
15:05:51 <pwhalen> ill also add Juno to the supported hw list for aarch64
15:06:37 <hrw> congratulations on release!
15:07:05 <pbrobinson> There will be disk images coming out from it shortly
15:07:16 <pwhalen> yes, congrats to all on the release! :)
15:07:40 <pwhalen> #info AArch64 disk images to follow shortly.
15:07:55 <pwhalen> cool.. anything else to mention or highlight?
15:08:01 <pbrobinson> #info untested minimal image available here based on GA https://arm.koji.fedoraproject.org/compose/24/test/
15:08:34 <hrw> yay! one stable url for fedora cloud for next hald year ;D
15:08:38 <hrw> half
15:09:06 <jlinton> BTW: For juno you probably should have this patch
15:09:07 <jlinton> https://patchwork.ozlabs.org/patch/636540/
15:09:56 <pwhalen> ah, thanks jlinton ..
15:10:36 <pwhalen> ok, moving on..
15:10:42 <pwhalen> #topic 1) ==== Package Status & Issues  ====
15:11:35 <pbrobinson> no major package issues, rawhide is pretty close to up to date. About 120 total behind
15:12:02 <pwhalen> nice, last week was about 170
15:12:24 <pwhalen> #info No major package issues, rawhide is pretty close to up to date. About 120 total behind primary.
15:12:35 <pwhalen> is anyone else aware of other issues?
15:12:53 <pbrobinson> statistics: {'older': 67, 'local_only': 1, 'remote_only': 295, 'same': 17774, 'newer': 2, 'total_missing_builds': 112}
15:13:11 <pbrobinson> sorry, my UX just crashed so having to log back into various sessions
15:13:34 <pwhalen> #info Rawhide - statistics: {'older': 67, 'local_only': 1, 'remote_only': 295, 'same': 17774, 'newer': 2, 'total_missing_builds': 112}
15:15:20 <pbrobinson> final F24 stats: statistics: {'older': 19, 'local_only': 1, 'remote_only': 294, 'same': 17747, 'newer': 2, 'total_missing_builds': 50}
15:15:30 <pwhalen> #info F24 - statistics: {'older': 19, 'local_only': 1, 'remote_only': 294, 'same': 17747, 'newer': 2, 'total_missing_builds': 50}
15:15:31 <pwhalen> thankls
15:16:00 <pwhalen> #topic 2) ==== Kernel Status ====
15:17:04 <jonmasters_> sorry I am late - on the phone with ARM
15:17:17 <pbrobinson> jonmasters_: you're always late!
15:17:28 <jonmasters_> that's what...nevermind
15:17:56 <pbrobinson> so from my side nothing to really report on kernel side
15:18:10 <pwhalen> im also not aware of any new kernel issues..
15:18:13 <pbrobinson> 4.5.7 will be the last 4.5 to go to F-24
15:18:31 <pbrobinson> 4.6.3 will likely be the first rebase, possibly next week, to F-24
15:18:41 <pwhalen> #info 4.5.7 will be the last 4.5 to go to F-24. 4.6.3 will likely be the first rebase, possibly next week, to F-24.
15:19:19 <pwhalen> ok, anyone have anything for kernel?
15:19:50 <jonmasters_> other than PCIe discussions, anything nasty that needs looking at in current kernels? (in case someone on the RH side has cycles)
15:20:15 <pbrobinson> jlinton: on the Juno patch, is that headed to stable too?
15:20:51 <pbrobinson> jonmasters_: I'm looking at pci-e stuff, but had other problems that took priority this last week
15:21:02 <jlinton> I would hope so, but I haven't heard anything about it, I will probably resubmit it with the minor tweaks to the commit message/ack's etc
15:21:07 <pwhalen> #info Juno patch needed - https://patchwork.ozlabs.org/patch/636540/
15:21:11 <jlinton> at that point I will request a backport to stable
15:21:54 <pbrobinson> jlinton: and cc: stable@ if you didn't already, feel free to add pbrobinson@gmail on the cc: to it too
15:23:01 <pwhalen> #topic 3) ==== Bootloader Status ====
15:23:27 <pwhalen> #info uboot-tools-2016.07-0.2.rc2.fc25 now available in rawhide for testing.
15:24:53 <pwhalen> #info edk2-20160418gita8c39ba-1 now available in all releases, includes 'edk2-aarch64'
15:25:49 <pbrobinson> so jonmasters_ and others lurking, I'm aware I dropped the ball on qemu u-boot stuff, it's on my list for today/tomorrow at the latest
15:26:12 <jonmasters_> okey doke
15:26:20 <jonmasters_> you're busy, as we know :)
15:26:31 <pbrobinson> I think we're looking fairly good in F-24 with 2016.05, there might be a few bits I push as aarch64 fixes
15:27:21 <pbrobinson> there was a new libvirt pushed last week (it's in updates-testing) that fixed issues with edk2-aarch64 as shipped in Fedora. So people will want that for aarch64 virt
15:28:12 <pbrobinson> #info libvirt-1.3.3.1-3.fc24 fixes a minor issues with edk2-aarch64 in F-24, you'll want it for aarch64 virt (on x86 or aarch64 HW)
15:28:55 <pwhalen> #topic 4) == Open Floor ==
15:29:40 <hrw> are there plans to build arm32 uefi? checked Linaro builds and they work in virt
15:30:21 <hrw> maybe some changes in libvirt will be needed to get arm32 recognized as valid uefi platform but it is doable
15:31:58 <pbrobinson> hrw: not currently, we'd also need to have grub2 and other components, it hasn't been ruled out but it would be a different code path and support bits over all the physical HW
15:32:26 <hrw> I know
15:33:06 <pwhalen> anything else for today?
15:33:14 <pbrobinson> jlinton: how you getting on with mozjs bits, I was revisiting mozjs45 with the 47/48 bit mapping stuff
15:33:35 <pbrobinson> with the path (and without) there's some tests that fail that you might be able to assist with
15:34:19 <jlinton> I haven't spent any significant time on mozjs, my focus has been some ACPI/PMU issues
15:34:35 <cov> IMO the best long-term solution is to create a new task size auxiliary vector
15:34:52 <jlinton> Once that clears up (or it gets late in the kernel cycle) I will focus on mozjs again
15:35:11 <pbrobinson> jlinton: OK, what I'll do is get a bug filed for it and add you on cc for when you get some cycles, the mongodb maintainer is actively looking to move to it so that might be one crossed off your list
15:35:38 <jlinton> Sure, you can cc jeremy.linton@arm.com
15:36:33 <cov> jlinton, probinson: what's your current favorite fix for the mozjs issue? munmap/similar probing?
15:37:03 <jlinton> I think the mmap change that went upstream has the advantage of not requiring a full recompile
15:37:09 <jlinton> of the whole world
15:37:23 <pbrobinson> jlinton: will do
15:37:58 <jlinton> the problem with the previous changes, was that there are header files included by everyone, that effectivly broke the ABI
15:38:52 <jlinton> (personally I find the mmap change ugly, but thats what went upstream, and it has the advantage of potentially working with >48bit VA's)
15:39:38 <pbrobinson> upstream is good, even if it's not the best solution. The recompile all deps isn't a major issue from our perspective
15:39:48 <cov> jlinton: do you know who to talk to about updates to the AArch64 ELF specification?
15:40:42 <jlinton> cov: Not directly, but you might try Charles Garcia-Tobin
15:40:49 <jlinton> if he isn't the person he will know who is.
15:41:16 <cov> I'm not sure the one current arch-specifc auxv entry is documented (AT_SYSINFO_EHDR), but if it is, it'd be best to add to the list if folks agree on the change
15:41:44 <cov> AT_SYSINFO_EHDR being the VDSO address
15:41:56 <jonmasters_> cov: for the ELF spec it's the toolchain group at ARM in combination with ATG (architecture group)
15:42:15 <jonmasters_> let me find out who owns it, but it was richard
15:43:38 <jonmasters_> Earnshaw that is
15:44:16 <cov> first instinct was to call it AT_TASKSZ, but I wonder if there's more ELF-specification-compatible naming
15:44:37 <cov> "task" might be linux vocabulary, idk
15:45:49 <jonmasters_> sent email to richard
15:46:18 <jonmasters_> will connect cov and jlinton with whoever he says owns it now
15:46:54 <pwhalen> thanks jonmasters_
15:46:58 <jlinton> I'm poking around inhouse to see if I can find out too..
15:48:39 <pwhalen> any other business for today?
15:48:58 <jonmasters_> jlinton: copied you on mail to Richard E
15:49:08 <jonmasters_> I like cov's suggestion btw
15:50:25 <pbrobinson> jonmasters_: I'm sure you can take this offline
15:50:32 <jonmasters_> sure thing
15:50:37 <pbrobinson> anyone else with anything to cover?
15:50:42 <pwhalen> or to #fedora-arm
15:50:50 <jonmasters_> to the bat cave!
15:51:01 <pwhalen> hehe :)
15:51:04 <pbrobinson> yes, that's what I meant, #fedora-arm or anywhere not the meeting :)
15:51:50 <pwhalen> many thanks to all for joining today! Please download Fedora 24 for Arm and AArch64!
15:52:02 <pwhalen> #endmeeting