15:00:33 <pwhalen> #startmeeting Fedora ARM & AArch64 Status Meeting
15:00:33 <zodbot> Meeting started Tue Jun 16 15:00:33 2015 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:33 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:33 <pwhalen> #chair pwhalen bconoboy pbrobinson dgilmore hrw jsmith kyle dmarlin
15:00:33 <zodbot> Current chairs: bconoboy dgilmore dmarlin hrw jsmith kyle pbrobinson pwhalen
15:00:46 <pwhalen> good morning folks, whos here today?
15:00:49 <jsmith> Greetings, all
15:01:51 <pwhalen> lets give it a couple of minutes
15:03:05 * pbrobinson is here
15:03:14 <hrw> hi
15:03:52 <pwhalen> looks like we have a quorum, thanks for coming guys.
15:04:04 <pwhalen> #topic 1) ==== Package Status & Issues  ====
15:04:15 <pwhalen> how does aarch64 look?
15:04:20 <pbrobinson> it's looking reasonable
15:04:50 <pwhalen> you mentioned qemu?
15:05:00 <pbrobinson> I'm trying to work out which bit is broken that is causing the qemu build to fail, which I believe is blocking a bunch of other packages
15:05:01 <hrw> pbrobinson: lot of perl stuff to build in queue?
15:05:17 <pbrobinson> hrw: getting there, let me type!
15:05:23 <hrw> mkey
15:05:55 <pwhalen> #info Qemu appears to be blocking Koji Shadow builds
15:05:59 <pbrobinson> I believe it's due to relro/hardening but I'm trying to work out which package isn't hardended causing the failure bit of a rabbit hole
15:06:36 <pbrobinson> but to rule out a bunch of packages I've had to fix them on primary which has cost me time today but I think I'm almost there
15:06:50 <pbrobinson> from there it should with luck unblock the rest of the perl issues
15:07:11 <pbrobinson> I think I fixed most of the perl issues that affect most package builds yesterday
15:07:21 <hrw> pbrobinson: you are awesome
15:08:00 <pbrobinson> in other news the gold linker is now fixed in rawhide and I've now reverted the work around in systemd for this and it's had the side effect of fixing ghc-*
15:08:20 <pbrobinson> the later should be cleaned up nicely in rawhide with the mass rebuild
15:09:04 <pbrobinson> other than the above two issues I'm not aware of any other aarch64 issues ATM but we're currently about 2000 builds behind primary
15:09:05 <pwhalen> #info Gold linker is now fixed in rawhide, reverted the work around in systemd. Should be cleaned up nicely in rawhide with the mass rebuild
15:09:30 <pwhalen> #info Rawhide aarch64 roughly 2000 builds behind primary
15:09:56 <pwhalen> thanks pbrobinson.
15:10:10 <pwhalen> anyone have anything else?
15:10:50 <pbrobinson> not from me
15:10:55 <hrw> me neither
15:11:03 <pwhalen> ok
15:11:09 <pwhalen> #topic 2) ==== Kernel Status ====
15:11:28 <hrw> pcie delay issue identified upstream and fix got created
15:11:36 <pwhalen> #info latest - kernel-4.1.0-0.rc8.git0.1.fc23
15:12:01 <pbrobinson> hrw: was that fixed in rc8 or is it landed in something later?
15:12:01 <hrw> my mustang went from 34 minutes to 24 seconds. but it is not in mainline yet (last checked) so not sent for fedora
15:12:18 <hrw> pbrobinson: have to check rc8. was on linux-pci ML
15:12:20 <pbrobinson> hrw: got a link to the upstream thread?
15:12:45 <hrw> will search
15:12:53 <pbrobinson> hrw: thx
15:13:28 <pwhalen> #info AArch64 pcie delay issue identified upstream, should be patched soon
15:13:33 <pbrobinson> from my PoV it's fairly flat at the moment, 4.1 is mostly done now so please test on devices and provide feedback
15:13:58 <pbrobinson> as far as I'm concerned 4.0.x in the stable releases is fairly boring
15:15:30 <pwhalen> no major breakage so far in testing 4.1
15:15:55 <pwhalen> anything else?
15:15:55 <hrw> same here
15:16:36 <pbrobinson> not that I can think of, 4.2 is almost baked from an ARM PoV upstream and preping for landing but not much use here atm
15:17:23 <pwhalen> #topic 3) ==== U-Boot Testing ====
15:17:41 <hrw> #link http://marc.info/?l=linux-pci&m=143415577710438&w=2 - PCI: X-Gene: Disable Configuration Request Retry Status for X-Gene v1 PCIe
15:18:04 <pbrobinson> so uboot 2015.07rc2 landed in rawhide yesterday
15:18:34 <pbrobinson> I'm spending some time preparing our patch sets to get the relevant bits upstream for easier rebasing support etc
15:18:54 <pwhalen> #info Available in Rawhide - uboot-tools-2015.07-0.1.rc2.fc23
15:18:56 <pbrobinson> so at the moment TI and i.mx6 devices don't have distro boot support
15:19:35 <pbrobinson> but AllWinner and Tegra at least should work nicely out of the box even on HDMI/USB so please test it
15:20:21 <pbrobinson> I expect to have a build adding back distro support to imx6 in the next day or so and will add the TI support back after that
15:21:02 <pwhalen> #info AllWinner and Tegra should work nicely out of the box even on HDMI/USB
15:21:37 <pwhalen> #info TI and i.mx6 devices require distro support patches
15:22:13 <pbrobinson> .... and will be fortcoming, check the changelogs ;-)
15:23:18 <pwhalen> be really nice if usb input on hdmi works on allwinner. will check that shortly
15:23:41 <pwhalen> any new devices added with?
15:23:45 <pwhalen> ..this
15:24:03 <pbrobinson> also of note is that wandboard is now a single u-boot binary for all revisions of the board
15:24:12 <hrw> cool
15:24:13 <pbrobinson> cuboxi is enabled as well
15:24:22 <pwhalen> #info Added Cuboxi support
15:24:24 <pbrobinson> also a single binary for all revisions
15:24:53 <pbrobinson> but they'll both be more interesting once the rebased distro boot support for imx6 had landed
15:25:09 <pwhalen> #info Single binary for all Wandboard revisions
15:25:18 <pbrobinson> I'll send out an email to the list once TI/imx6 has distro support back
15:25:24 <pbrobinson> with more details
15:25:24 <pwhalen> very nice. thanks.
15:27:59 <pwhalen> #topic 4) ==== F23 Planning ====
15:28:34 <pbrobinson> so from my point of view I'm planning the following for F-23
15:28:44 <pbrobinson> 1) cloud images for aarch64
15:29:22 <pbrobinson> 2) taking dmarlin's ARMv7 on aarch64 recipe and standardising it to ensure easy support for ARMv7 VMs on aarch64
15:29:56 <pbrobinson> 3) standardising of u-boot support and upstreaming it to make it easy to rebase quickly to newer upstream releases
15:30:52 <pbrobinson> 4) extending (with pwhalen) the fedora-arm-installer to make support for a number of devices, and upgrading of u-boot on existing installs
15:31:02 <pbrobinson> 5) general u-boot/kernel polish
15:31:10 <pbrobinson> 6) graphics support
15:31:13 <pbrobinson> EOF
15:31:52 <pbrobinson> what are other people looking to work on, interested in?
15:32:01 <pwhalen> are you soliciting help for any of these?
15:32:31 <pbrobinson> always!
15:32:58 <hrw> I plan to get some X stuff working on mustang
15:34:50 <pwhalen> is there anything we'd like to see for f23 that hasnt been discussed yet?
15:35:51 <pbrobinson> possibly docker support
15:36:27 <pbrobinson> I'll be working with dgilmore to enable us to more formally build ARMv7 docker images (see point 2 above)
15:36:46 <pbrobinson> and with luck if the support in golang 1.5 is complete enough we should have them for aarch64 too
15:37:41 <pwhalen> ok, i think we should keep this on the agenda as well, check in next week to see where we're at.
15:37:48 <pwhalen> anything else?
15:38:33 <pbrobinson> nope
15:38:35 <pbrobinson> ah yes!
15:38:35 <hrw> nope
15:38:42 <pbrobinson> just a sec
15:39:36 <pbrobinson> reminder that F-23 feature Changes are due in a week https://lists.fedoraproject.org/pipermail/devel-announce/2015-June/001612.html
15:40:21 <pwhalen> #info F23 Feature changes due June 23rd. Alpha release in August.
15:40:36 <pwhalen> #topic 5) == Open Floor ==
15:40:54 * jsmith doesn't have anything else to add
15:41:21 <pbrobinson> only think I have is that I pushed updates for 21/22 updates/updates-testing for aarch64 last night/this morning
15:42:36 <pwhalen> awesome, thanks
15:43:18 <pwhalen> ok, lets wrap it up. thanks for coming everyone
15:43:21 <pwhalen> #endmeeting