15:00:13 <pwhalen> #startmeeting Fedora ARM & AArch64 Status Meeting
15:00:13 <zodbot> Meeting started Tue Feb 24 15:00:13 2015 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:13 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:18 <pwhalen> good morning folks
15:00:22 <pwhalen> .fas pwhalen
15:00:23 <zodbot> pwhalen: pwhalen 'Paul Whalen' <pwhalen@redhat.com>
15:01:08 <pwhalen> good morning kylem
15:01:10 * pbrobinson is here
15:01:12 <dmarlin_> .fas dmarlin
15:01:13 * kylem waves.
15:01:13 <zodbot> dmarlin_: dmarlin 'David A. Marlin' <dmarlin@redhat.com>
15:01:19 <kylem> .fas kyle
15:01:20 <zodbot> kylem: kingdomkeeper 'Kyle Keevill' <tgxkingdomkeeper@gmail.com> - escapist 'Kyle King' <krk126@live.com> - kylechen1990 'kyle chen' <rhce.kyle@gmail.com> - indemnity83 'Kyle Klaus' <kklaus@indemnity83.com> - bluelightning32 'Kyle Stemen' <jks6csmm9y@snkmail.com> - kylezk '张奎' <kylezk@smail.hust.edu.cn> - caymon 'Kyle Munley' <komunley@gmail.com> - kymully 'Kyle Mullan' <kylemullan@gmail.com> - kylezk00 (12 more messages)
15:01:38 <kylem> hmm, i'm none of those jerks. ;-)
15:02:05 <pbrobinson> hehe
15:02:20 <dmarlin_> The Kyle Identity   :)
15:02:26 <kylem> .fasinfo kyle
15:02:28 <zodbot> kylem: User: kyle, Name: Kyle McMartin, email: kmcmartin@redhat.com, Creation: 2007-11-07, IRC Nick: kyle / kylem, Timezone: America/New_York, Locale: en, GPG key ID: 191fcd8a, Status: active
15:02:31 <zodbot> kylem: Approved Groups: fedorabugs cla_fedora cla_done packager cla_redhat provenpackager @gitkernel-arm64
15:03:34 <bconoboy> .fasinfo blc@
15:03:35 <zodbot> bconoboy: User "blc@" doesn't exist
15:03:39 <bconoboy> .fasinfo blc
15:03:42 <oatley> .fas oatley
15:03:42 <zodbot> bconoboy: User: blc, Name: None, email: blc@redhat.com, Creation: 2008-06-20, IRC Nick: None, Timezone: None, Locale: None, GPG key ID: None, Status: active
15:03:45 <zodbot> bconoboy: Approved Groups: cla_fedora cla_done cla_fpca gitarm +aarch64 +gitarm-boot-config packager fedorabugs @gitkernel-arm64
15:03:48 <zodbot> oatley: oatley 'Andrew Oatley-Willis' <andrew.oatley-willis@senecacollege.ca>
15:04:17 <pwhalen> i didnt see many agenda requests for today, lets go through some obvious topics first
15:04:47 <hrw> .fas mjuszkiewicz
15:04:48 <pwhalen> #topic 1) ==== Fedora 22 Testing ====
15:04:48 <zodbot> hrw: hrw 'Marcin Juszkiewicz' <mjuszkiewicz@redhat.com>
15:05:17 <pbrobinson> I've tested in the last week on Wandboard, CT, BBB
15:05:24 <pbrobinson> doing some more testing on some other platforms
15:06:19 <pwhalen> we have a TC4, not announced due to issues with the images
15:07:04 <pwhalen> repo also has some issues as well, TC5 should be request soon
15:07:15 <pbrobinson> I've also been working on improving some kernel <-> userspace bits as well as seeing what bits of accelerated X of varying types we can improve
15:07:37 <pwhalen> accelerated x woudl be awesome
15:08:52 <pwhalen> #info F22 TC4 posted but with issues, TC5 should be requested shortly. watch for an announcement
15:10:29 <pwhalen> #topic 2) ==== Kernel Testing ====
15:10:59 <pbrobinson> so 4.0rc1 is out (aka the kernel formerly known as 3.20rc0)
15:11:01 <pwhalen> woke up this morning to see - kernel-4.0.0-0.rc1.git0.1.fc22
15:11:27 <pwhalen> #link armhfp - http://koji.fedoraproject.org/koji/buildinfo?buildID=614361
15:11:45 <pbrobinson> there's a known issue on Tegra platforms due to shitting IOMMU drivers from other SoCs. I'm reviewing the proposed upstream patches with the view of pulling them in
15:11:46 <pwhalen> #link scratch aarch64 build - http://arm.koji.fedoraproject.org/koji/taskinfo?taskID=2894832
15:12:11 <pwhalen> scratch so we didnt need to wait for koji-shadow to pick it up
15:12:42 <hrw> I just booted to 3.20rc
15:12:57 <pwhalen> #info known issue on Tegra platforms due to IOMMU drivers from other SoCs. Peter Robinson is reviewing the proposed upstream patches with the view of pulling them in
15:13:03 <hrw> 3.20.0-0.rc0.git7.3.fc23.aarch64
15:14:45 <hrw> kernel oops when NM gets ethernet up
15:14:46 <pwhalen> any other kernel news, testing?
15:14:54 <hrw> [   19.017346] Call trace:
15:14:54 <hrw> [   19.019784] [<fffffe000060b10c>] xgene_enet_rd_mcx_mac.isra.11+0x5c/0xe0
15:14:58 <hrw> [   19.026457] [<fffffe000060b6e8>] xgene_gmac_tx_enable+0x28/0x50
15:16:19 <kylem> what firmware version, acpi or not?
15:16:24 <pbrobinson> hrw: backtraces are not really for a general ARM meeting
15:16:25 <kylem> (what's the bugzilla #?)
15:16:48 <kylem> toss all that stuff into #info and i'll look at it afterwards.
15:16:58 <hrw> kylem: 0.15, default (so dtb)
15:17:02 <hrw> kylem: ok
15:17:18 <pwhalen> #info kernel-4 : https://lwn.net/Articles/634346/
15:17:32 <pwhalen> #link https://lh5.googleusercontent.com/-W_apgVmij20/VOAvvgpevvI/AAAAAAAAO3A/fPG5gFS2Rg8/w953-h536-no/mate_2015.02.15.051456.utc-smplayer-2009-terminator-salvation-mp4_16-9.jpg
15:17:46 <pwhalen> yea, i googled that this morning
15:18:25 <pwhalen> #topic 3) ==== Package Status & Issues  ====
15:19:07 <pbrobinson> no major issues for aarch64 other than ghc
15:19:22 <pbrobinson> gcc5 issues are slowly starting to settle down
15:20:22 <pbrobinson> would be interested in people's opinion on https://fedoraproject.org/wiki/Changes/Harden_all_packages_with_position-independent_code and what impact it might have on aarch64/ARMv7
15:20:47 <kylem> there's very little impact on either.
15:21:11 <pbrobinson> kylem: good to know
15:21:12 <dgilmore> it does seem the arch most impacted is i686
15:21:20 <kylem> the only reason x86 cares is it's register starved
15:21:48 <kylem> so throwing one register away for the GOT increases register pressure in busy functions badly.
15:23:29 <pwhalen> #topic 4) ==== F22 Documentation ====
15:23:29 * ctyler pops in with his coffee, sits at back of room
15:24:48 <pbrobinson> jsmith: you're up
15:24:55 <pwhalen> just pinged, not in channel
15:25:47 <jsmith> Sorry I'm late
15:25:50 <jsmith> Got side-tracked
15:26:01 <pwhalen> np., thanks for joining
15:26:16 <jsmith> Anyhoo -- I thought it would be worth discussing what we want in terms of documentation for the F22 release.
15:26:27 <jsmith> Now that we have much better range of supported hardware, etc.
15:26:44 <pwhalen> i think we need to coordinate more, esp with the web team
15:27:05 <jsmith> Right -- but we need to get our expectations/thoughts in order first
15:27:13 <jsmith> and then approach the web team and the docs team...
15:27:17 <pbrobinson> jsmith: how hard would it be to do something like this in the wiki? http://archlinuxarm.org/platforms/armv7/allwinner/cubietruck
15:27:31 <pbrobinson> and http://archlinuxarm.org/platforms/armv7
15:27:43 <jsmith> I'm happy to play liason with the docs team, since I spend some time there
15:27:50 <jsmith> pbrobinson: It's a wiki -- anything's possible
15:28:01 <pbrobinson> ie an easily to create template for all the various different devices we have
15:28:09 <jsmith> Shouldn't be too hard
15:28:15 <pbrobinson> jsmith: I have no  doubt.... but how easy to create and maintain?
15:28:26 <pwhalen> i can help write things up, fancy .. up to someone else
15:28:26 <jsmith> (I'm personally not a big fan of wiki templates, and no wiki expert)
15:28:33 <pbrobinson> because it would nice to be able to do something like that with install specifics etc
15:28:43 <jsmith> But if someone can give me the data, I'll make sure it gets written up nicely somewhere
15:28:52 <jsmith> Either on the wiki or in the ARM Guide, etc.
15:28:58 <pbrobinson> jsmith: you're more of one than me or at lest know who to bribe that is ;-)
15:29:11 <jsmith> pbrobinson: Something like that :-)
15:29:31 <jsmith> I just don't know enough about a lot of these platforms to be able to document myself out of a wet paper bag
15:29:52 <pwhalen> jsmith, this is what i wrote up for f21 - https://fedoraproject.org/wiki/Architectures/ARM/F21/Installation
15:30:18 <pwhalen> not as pretty, certainly would appreciate more hands on.. reviewing it before we release
15:30:22 <pbrobinson> jsmith: it would be useful to have some means of creating them and a standard URL to be able link/find, the data for me is the easy bit, I can help with that
15:30:36 <hrw> I promise writing basic installation howto for mustang
15:30:53 <pbrobinson> hrw: could you make that generic for aarch64?
15:31:05 <ctyler> Really good templating would let you take generic installation instructions and make slightly-adapted versions for a bunch of platforms, rather than "do what it says over there [link] but change A to B and C to D and insert [this] between steps 5 and 6"
15:31:16 <hrw> pbrobinson: it will be generic
15:31:30 <hrw> pbrobinson: mustang is the only platform where I can test easily
15:31:56 <pbrobinson> ctyler: exactly
15:34:01 <pwhalen> ok. so how do we leave this?
15:34:08 <pwhalen> we need to start writing it up now
15:34:14 <pbrobinson> some action points with people assigned?
15:34:53 <pwhalen> I was going to copy and paste, sed the release and then adjust as needed. someone can then take that info and put into some fancy pants template
15:35:06 <jsmith> pwhalen: Sounds good.
15:35:16 <jsmith> I'll volunteer to figure out the template (either on the Wiki or in formal docs)
15:35:36 <pwhalen> #action pwhalen to create initial f22 installation documentation page
15:35:41 <jsmith> And we can probably coordinate the list of devices (and volunteers who know those platforms) on the mailing list
15:35:59 <pwhalen> #action jsmith to figure out the template (either on the Wiki or in formal docs)
15:36:17 <pbrobinson> sounds good
15:36:57 <kylem> need to step away for a sec.
15:37:03 <pwhalen> we'll check in next week. anything else worth mentioning?
15:37:44 <pwhalen> #topic 4) == Open Floor ==
15:37:47 <jsmith> I think that's a good start :-)
15:38:08 <hrw> how close we are to f22 install images for aarch64?
15:38:19 <pwhalen> ah, good one
15:38:49 <pwhalen> we've discussed nightly install repos, but currently not being created (dep issues in the repo right now as well)
15:39:03 <pwhalen> pbrobinson, is this on the radar for RSN?
15:39:34 <pbrobinson> yes, hopefully this week, awaiting mainline's being signed off for the first rev
15:39:35 <pwhalen> deps were broken yesterday
15:39:40 <pwhalen> nice, thanks
15:40:35 <pwhalen> .... anything else ?
15:40:49 <bconoboy> mustang firmware?
15:41:10 <pbrobinson> what about it?
15:41:13 <hrw> bconoboy: what's with it?
15:41:14 <bconoboy> #info F22 aarch64 install images expected really soon now
15:41:14 <pwhalen> thanks!
15:41:24 <bconoboy> #info rh0.15 firmware for mustang board is now out
15:41:37 <bconoboy> that's it- seemed worth noting since rh0.14 didn't play well with f21
15:41:42 <pwhalen> #link https://mlangsdorf.fedorapeople.org/
15:41:51 <ctyler> UEFI for HiKey is just out, haven't had time to try Fedora on it yet.
15:42:05 <hrw> ctyler: good news
15:42:18 <hrw> uefi
15:42:46 <pbrobinson> ctyler: got a link for that?
15:43:03 <ctyler> sec
15:43:11 <ctyler> https://github.com/96boards/edk2
15:43:48 <pwhalen> #info UEFI for HiKey is just out
15:43:54 <pwhalen> #link https://github.com/96boards/edk2
15:43:55 <ctyler> It's first cut
15:44:33 <pbrobinson> ctyler: is there pre built binaries?
15:44:42 <hrw> https://github.com/96boards/documentation/wiki/UEFI
15:45:01 <hrw> #link https://github.com/96boards/documentation/wiki/UEFI
15:45:25 <pjones> oooh, and they've pushed their edk2 tree there.
15:45:28 * pjones ponders.
15:45:57 * pbrobinson wonders what pjones is pondering
15:46:05 <ctyler> #link https://builds.96boards.org/snapshots/hikey/uefi
15:46:13 <pwhalen> interested in hearing peoples experiences with it, please post to the list
15:46:26 <pwhalen> anything else today?
15:46:27 <ctyler> zodbot don't like me today, someone repost that link?
15:46:31 <pjones> pbrobinson: my secret project to make firmware updates work the same on arm machines as x86 machines.
15:46:50 <pwhalen> oh no, fail
15:46:52 <pwhalen> #chair pwhalen bconoboy pbrobinson dgilmore hrw jsmith kyle
15:46:52 <zodbot> Current chairs: bconoboy dgilmore hrw jsmith kyle pbrobinson pwhalen
15:46:56 <pbrobinson> pjones: now tha twould be awesome!
15:47:10 <pwhalen> we have chairs now
15:47:13 <hrw> #link https://github.com/96boards/documentation/wiki/UEFI
15:47:26 <pjones> pbrobinson: I have a half-assed version that needs some debugging on mustang already
15:47:44 <ctyler> zodbot sleeps.
15:48:01 <nirik> ctyler: it never outputs anything for #link. It's there in the notes.
15:48:11 <dgilmore> pjones: afaik the hikey doesnt have any onboard storage for the UEFI payload
15:48:18 <nirik> and actually it detects links, by http on the first part of the line. ;)
15:48:28 <dgilmore> you need to insert it into the start of the boot media
15:48:31 <ctyler> nirik: thanks!
15:48:49 <pjones> dgilmore: /probably/ shouldn't matter by the time I've got a more full project.  There's no reason the path for the firmware image can't be e.g. tftp://
15:48:59 * pjones goes back to working on that.
15:49:24 <pbrobinson> pjones: let us know if you want anything tested
15:49:30 <pjones> *nod*
15:50:20 <dgilmore> pjones: same here
15:50:59 <pwhalen> last call..
15:51:29 <pwhalen> #endmeeting