19:30:00 <pwhalen> #startmeeting Fedora ARM weekly status meeting
19:30:00 <zodbot> Meeting started Wed Aug 14 19:30:00 2013 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:30:00 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
19:30:00 <pwhalen> #chair pwhalen jonmasters bconoboy ctyler pbrobinson dgilmore dmarlin masta handsome_pirate msalter ahs3 agreene jcapik ddd_
19:30:00 <zodbot> Current chairs: agreene ahs3 bconoboy ctyler ddd_ dgilmore dmarlin handsome_pirate jcapik jonmasters masta msalter pbrobinson pwhalen
19:30:08 <bconoboy> .fas blc@
19:30:09 <zodbot> bconoboy: blc '' <blc@redhat.com>
19:30:10 <pwhalen> afternoon everyone
19:30:15 <pwhalen> .fas pwhalen
19:30:16 <zodbot> pwhalen: pwhalen 'Paul Whalen' <pwhalen@redhat.com>
19:31:48 <pwhalen> #topic 0) Status of ACTION items from our previous meeting
19:31:48 <pwhalen> #link http://meetbot.fedoraproject.org/fedora-meeting-1/2013-08-07/fedora-meeting-1.2013-08-07-19.29.html
19:32:08 <pwhalen> did we want to talk about the viable boards discussed at flock now?
19:32:45 <bconoboy> is that elsewhere on the agenda?
19:32:48 <bconoboy> if not, now is good
19:32:52 <jcapik> .fas jcapik
19:32:53 <zodbot> jcapik: jcapik 'Jaromír Cápík' <jcapik@redhat.com>
19:33:24 <pwhalen> could be covered here or in the kernel portion i guess, kylem__ ?
19:34:04 <pwhalen> I pinged Peter, I havent seen him on today
19:34:10 <bconoboy> peter is traveling
19:34:17 <bconoboy> let's cover it during kernel, hopefully kyle will be back by then
19:34:18 <kylem__> sure.
19:34:24 <bconoboy> there he is! get him!
19:34:28 <pwhalen> yea, thought that may be the case
19:34:48 <kylem__> i *think* we said that wandboard, utilite and beaglebone black
19:34:50 <Guest80927> kylem__ you don't hide very well
19:35:01 <bconoboy> #link Peter's State of the ARM Union lists new boards: http://nullr0ute.com/wp-content/uploads/2013/08/Flock2013-ARM-StateOfTheUnion.pdf
19:35:50 <bconoboy> #info Likely new devices in F20: Beaglebone, Wandboard, Utilite
19:36:24 <kylem__> i've got a utilite now, but i need to bug them for their source. ships with 3.0 on it.
19:36:38 <bconoboy> #info Other possibilities include AC100, Tegra3/4 in general, Chromebook, Arndale, Odroid
19:36:57 <bconoboy> kylem__: ugh- does it have an embedded dtb?
19:37:39 <kylem__> bconoboy, looks it
19:38:02 <kylem__> let me turn it back on
19:38:23 <dgilmore> pwhalen: peter is on his way back to the UK
19:38:24 <pwhalen> big list of targets to test
19:38:51 <pwhalen> dgilmore, ok, thought that may be the case
19:39:19 <bconoboy> we're going to do a decision point on which devices are in and out
19:39:47 <pwhalen> sounds good, do we know at which point we'll do that?
19:39:58 <bconoboy> So, what is the standard by which we are willing to say "We will support $BOARD in F20?"  And when do we need to decide?
19:39:59 <kylem__> i think there's a question of what do we want to /support/ and what do we just want to enable
19:40:38 <dgilmore> kylem__: right, some stuff we will enable, it may or may not work
19:40:52 <dgilmore> but some boards we want to be sure they work
19:40:59 <kylem__> agreed.
19:41:00 <bconoboy> support implies we have the capacity to do QE
19:41:59 <pwhalen> so when do we revisit and make that decision?
19:42:14 <bconoboy> perhaps the right first step is discussing on the mailing list
19:42:28 <bconoboy> I'm not sure all the stake holders are here now
19:43:23 <bconoboy> Okay, so I propose this:
19:43:49 <bconoboy> Let's discuss minimal criteria for support on arm@lists.fp.o, possibly including devel@lists.fp.o as well
19:43:58 <bconoboy> Then we can discuss here again next week
19:44:04 <bconoboy> if needed
19:45:04 <bconoboy> can I get a +1? :-)
19:45:06 <pwhalen> sounds good, who would like to email the list?
19:45:06 * masta is here late
19:45:10 <pwhalen> +1
19:45:13 <bconoboy> I'll kick off the discussion
19:45:28 <pwhalen> thanks
19:46:13 <pwhalen> #action bconoboy to email the arm/devel mailing lists to begin supportable hw in f20 discussion
19:46:29 <pwhalen> #topic 1) Problem packages
19:48:03 <pwhalen> anyone able to call out some problem packages right now?
19:48:29 <bconoboy> Is glibc still a problem?
19:48:36 <bconoboy> (neon instructions)
19:48:58 <pwhalen> it appears so , dgilmore mentioned it should be resolved this week?
19:49:29 <bconoboy> #link glibc has neon instructions https://bugzilla.redhat.com/show_bug.cgi?id=985342
19:49:36 <bconoboy> I don't see an update to the BZ since July 24
19:49:39 <bconoboy> dgilmore?
19:51:10 <bconoboy> That's the only big one I'm aware of
19:51:32 <kylem__> that's next on my list unless carlos gets free cycles
19:52:16 <bconoboy> okay, cool
19:52:23 <pwhalen> thanks kylem__ , he may have looked today as well
19:52:31 <pwhalen> #topic 2) Kernel Status Update
19:52:48 <kylem__> no change on my end due to flock, but at least i have a flyswatter now
19:53:19 <pwhalen> awesome
19:53:47 <kylem__> poking at openocd now, hopefully should have some progress after the meeting or tomorrow morning
19:54:04 <bconoboy> #info No updated since last week (SIGFLOCK)
19:54:16 <masta> kylem__: what boards are you puting the flyswatter first?
19:54:26 <kylem__> pandaboard
19:55:32 <pwhalen> #topic 3a) Aarch64 - Status Update
19:56:23 <pwhalen> how does it look bconoboy ?
19:56:28 <bconoboy> Build queue is empty
19:56:52 <bconoboy> Currently 11909/13606 packages built, including 5043 aarch64-producing packages
19:57:16 <bconoboy> We're pulling in updates every night and queueing them up
19:57:24 <bconoboy> So the good news is we have more builders than we have F19 updates
19:57:48 <bconoboy> The bad news is that we're still about 1600 packages short of the finish line
19:58:21 <pwhalen> are all of these packages that failed due to a build issue?
19:58:23 <bconoboy> #info 11909/13606 packages built
19:58:40 <bconoboy> there are 856 failed builds
19:58:49 <bconoboy> 249 failed due to dependency resolution issues
19:59:01 <bconoboy> 607 failed due to something in %build
19:59:29 <bconoboy> It's possible that if we requeued many of these they would pass upon retry, as newer versions of dependencies are always rolling in
19:59:41 <bconoboy> That may be the next logical step
20:00:26 <pwhalen> right, should we keep requeuing rather than leave builders idle
20:00:36 <bconoboy> yeah, I think so
20:01:21 <pwhalen> anything else ?
20:01:30 <bconoboy> #info Builders are idle, so we will requeue the failed 856 builds as dependencies may not be filled
20:01:41 <bconoboy> #include <stdpleaforhelp.h>
20:01:52 <bconoboy> #undo
20:01:52 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x250182d0>
20:02:00 <bconoboy> #info Builders are idle, so we will requeue the failed 856 builds as dependencies may now be filled
20:02:04 <pwhalen> :)
20:02:28 <pwhalen> #topic 3b) F19 Remix
20:03:08 <bconoboy> Just a heads up for everybody, we will be putting an F19 remix out in the next few days
20:03:26 <bconoboy> This will be a little different from the builder image we're maintaining now
20:03:44 <masta> aarch64 remix?
20:04:12 <bconoboy> It won't include the rebuild stuff- just a pure stage 4 image that people can use to test drive Fedora on AArch64
20:04:28 <masta> cool!
20:05:02 <bconoboy> We'll include a 3.11 kernel and modules suitable for use with the foundation model
20:05:36 <bconoboy> Hopefully semiconductors who are producing their own aarch64 chips will find it useful as well- bring your own kernel
20:06:35 <bconoboy> #info An F19 AArch64 remix is coming- including a Fedora 3.11 kernel/initramfs for use with foundation model.
20:06:58 <bconoboy> We'll be producing the image using live media creator
20:07:49 <bconoboy> #info livemedia-creator is capable of making these images even now
20:08:27 <pwhalen> #topic 4) Flock Wrap-up - Did you attend? Tell us about it!
20:08:58 <bconoboy> Who went? What happened? Was it awesome?
20:09:06 <dgilmore> bconoboy: ideally we use the same tools as we do for v7
20:09:12 <kylem__> i did -- it was really quite fantastic, massive thanks to ruth and spot and all the volunteers
20:09:33 <bconoboy> kylem: Any arm highlights?
20:09:38 * dgilmore was there, it was good.
20:09:52 <dgilmore> bconoboy: there was a lot of different arm things
20:09:57 <dgilmore> many at the same time
20:10:08 <dgilmore> so impossible to see all
20:10:33 * masta was there. Good times =)
20:10:38 <kylem__> yeah, my only criticism was it was hard to see everything, since there was like 8 parallel trakc.s
20:10:46 <bconoboy> that's a lot of tracks
20:10:48 <kylem__> but that's not really fixable unless i clone myself a few times.
20:11:57 <bconoboy> so... any decisions get made?  anybody on the hook for anything?
20:11:59 <dgilmore> if there was only 2 or 3 tracks chances are higher there is nothing of interest
20:12:07 <kylem__> me and probinson finally got to connect on a lot of kernel stuff too, which was good
20:12:28 <dgilmore> bconoboy: nothing really new
20:12:30 <kylem__> i also talked to npmcallum about the replacement for smolt and collecting data from arm systems
20:12:49 <bconoboy> #info Vulcan mind meld between kylem and pbrobinson for kernel maint plans
20:13:01 <dgilmore> going to look at enabling grub, though we still need cmd_pxe enabled for pxe booting
20:13:12 <bconoboy> enabling grub would be great
20:13:52 <dgilmore> kylem__: that would be nice, smolt worked at one point but people broke it for not x86 at some point
20:14:19 <dgilmore> they made it only work with dmidecode
20:14:21 <kylem__> yeah, i'm going to write some python to collect hw details
20:14:30 <kylem__> anyway, that's a tangent
20:15:29 <dgilmore> there was no huge new things as to f20 things to do
20:15:52 <bconoboy> #info Possible development pursuits include enabling grub and the new smolt replacement- these are probably f21+ activities
20:16:30 <pwhalen> #link FLOCK Talks - http://www.youtube.com/channel/UC_p--G_ebsrR4kA4zpqLSTw
20:17:19 <pwhalen> #topic 5) Open Floor
20:17:46 <pwhalen> anything else for today?
20:18:20 <bconoboy> crickets
20:18:54 <masta> =)
20:18:54 * pwhalen hears crickets too
20:19:02 <pwhalen> going once..
20:19:16 <pwhalen> #endmeeting