20:00:31 <pwhalen> #startmeeting Fedora ARM weekly status meeting
20:00:31 <zodbot> Meeting started Wed Jul 24 20:00:31 2013 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:31 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
20:00:31 <pwhalen> #chair pwhalen jonmasters bconoboy ctyler pbrobinson dgilmore dmarlin masta handsome_pirate msalter ahs3 agreene jcapik ddd_
20:00:31 <zodbot> Current chairs: agreene ahs3 bconoboy ctyler ddd_ dgilmore dmarlin handsome_pirate jcapik jonmasters masta msalter pbrobinson pwhalen
20:00:36 * masta is here
20:00:39 <dmarlin> .fas dmarlin
20:00:39 <pwhalen> good afternoon everyone
20:00:39 <zodbot> dmarlin: dmarlin 'David A. Marlin' <dmarlin@redhat.com>
20:00:44 <pwhalen> .fas pwhalen
20:00:44 <zodbot> pwhalen: pwhalen 'Paul Whalen' <pwhalen@redhat.com>
20:00:50 * dgilmore is here
20:00:51 * pbrobinson is here
20:00:59 * msalter is here
20:01:21 <agreene_> .fas agreene
20:01:21 <zodbot> agreene_: agreene 'Andrew Greene' <andrew.greene@senecacollege.ca> - tag4fedora 'Tim Greene' <tagreene@flowserve.com>
20:01:36 <oatley_> .fas oatley
20:01:37 <zodbot> oatley_: oatley 'Andrew Oatley-Willis' <andrew.oatley-willis@senecacollege.ca>
20:01:49 <masta> hey everyone
20:02:05 <bconoboy> .fas blc@
20:02:06 <zodbot> bconoboy: blc '' <blc@redhat.com>
20:02:11 <pwhalen> #topic 0) Status of ACTION items from our previous meeting
20:02:12 <pwhalen> #link http://meetbot.fedoraproject.org/fedora-meeting-1/2013-07-17/fedora-meeting-1.2013-07-17-20.02.html
20:03:57 <pwhalen> #info ACTION: dgilmore to setup koji hub&db for aarch64
20:04:09 <dgilmore> still not had a chance to
20:04:45 <pwhalen> #info ACTION: handsome_pirate Is talking to ghc upstream
20:04:54 <pbrobinson> focus has been on push to primary so I suggest we push this back to post mass rebuild
20:05:02 <pbrobinson> that's for koji btw
20:05:50 <pwhalen> #info ACTION: msalter is working on openjdk for aarch64
20:06:03 <pwhalen> #info complete, built!
20:06:07 <msalter> yay
20:06:15 <pbrobinson> woo!
20:06:15 <pwhalen> #info ACTION: kylem to create a list of new supported platforms in 3.11
20:06:40 <kylem> the list includes, and is limited to, what i have in my cube. ;-P
20:06:51 <pbrobinson> boo ;-)
20:06:55 <bconoboy> So... if we store stuff in your cube...
20:06:56 <jcapik1> .fas jcapik
20:06:57 <zodbot> jcapik1: jcapik 'Jaromír Cápík' <jcapik@redhat.com>
20:08:08 <bconoboy> kylem: carry over for next week?
20:08:24 <pbrobinson> might be worth putting it out to the mailing list
20:08:31 <bconoboy> definitely
20:08:51 <kylem> bconoboy, yeah
20:09:02 <kylem> it's too early to tell if it's going to be different from f19
20:09:46 <pwhalen> #topic 1) Problem packages
20:10:02 <pbrobinson> java (being worked on) python3 (being worked on)
20:10:18 <pbrobinson> redhat-lsb (being worked on I believe)
20:10:25 <bconoboy> BZs?
20:10:33 * pbrobinson looking already
20:10:53 <pbrobinson> java https://bugzilla.redhat.com/show_bug.cgi?id=985811
20:11:10 <pwhalen> #info java (being worked on)
20:11:15 <bconoboy> #link java: https://bugzilla.redhat.com/show_bug.cgi?id=985811
20:11:56 <pbrobinson> #info python3 https://bugzilla.redhat.com/show_bug.cgi?id=986713
20:12:26 <pbrobinson> can't find lsb but the maintainer has a device he's testing on in beaker
20:12:57 <pbrobinson> ghc was blocking but was cleared this morning and catching up on ghc* and friends backlog pretty quickly
20:13:18 <bconoboy> #info redhat-lsb has issues, but is being worked on by its maintainer
20:13:18 <pbrobinson> and that's it I think
20:13:24 <bconoboy> I have one other problem package
20:13:32 <bconoboy> glibc is being built with neon instructions
20:13:41 <bconoboy> this works *great* on highbank
20:13:47 <pbrobinson> :-(
20:13:50 <bconoboy> but means f20 can't be used on tegra2
20:13:54 <pbrobinson> or OLPC
20:13:59 <kylem> we're working on it
20:14:05 <bconoboy> kylem: have the bz# handy?
20:14:16 <kylem> 985342
20:14:37 <pbrobinson> can we please ensure we add all ARM bugs to the ARMTracker so they're easy to follow
20:14:43 <bconoboy> #link glibc is being built with neon, breaking tegra2/olpc: https://bugzilla.redhat.com/show_bug.cgi?id=985342
20:14:50 <bconoboy> Adding now
20:15:01 <bconoboy> Actually, it's there
20:15:13 <bconoboy> that's it for me
20:15:17 <pbrobinson> cool, wonder why I didn't see that
20:15:32 <pwhalen> anyone have any others?
20:15:59 <dgilmore> statistics: {'older': 230, 'local_only': 0, 'remote_only': 248, 'same': 13513, 'newer': 0, 'total_missing_builds': 309}
20:16:01 <pbrobinson> no major blockers, I'm slowly working through the rest of the list and filing bugs
20:16:09 <dgilmore> thats where we are right now
20:16:29 <bconoboy> that's with pending ghc packages still to be built?
20:16:32 <pbrobinson> OK, so we're about 230 behind where f19 was at release
20:16:52 <dgilmore> bconoboy: thats with whats built of ghc
20:16:52 <pbrobinson> there's probably 100 odd of those by guestimaiton and a number blocked by java
20:17:06 <dgilmore> we were just over 300 before ghc was fixed
20:17:25 <dgilmore> and over half that was because of ghc
20:17:30 <pbrobinson> there's still quite a bit of ghc to build
20:17:42 <dgilmore> right
20:17:57 <dgilmore> we should be just over 100 when ghc is done
20:18:40 <bconoboy> so between java and ghc we'll be in good shape
20:18:43 <pbrobinson> plus remote only
20:18:50 <pbrobinson> yes, pretty much I think
20:19:05 <dgilmore> which many of the remote_only exclude arm
20:19:18 <pbrobinson> yep
20:19:42 <pbrobinson> and are things like ada/pascal that need a bring up
20:19:55 <pbrobinson> but aren't high prio
20:20:40 <pwhalen> #topic 2) Kernel Status Update
20:21:01 * ahs3 apologizes for being late, stands in the corner with a dunce hat for a bit
20:21:39 <pwhalen> this is usually pbrobinson's topic, but i guess is now kylem's
20:22:14 <pwhalen> kylem, did you want to update on 3.11? or either of you ?
20:22:16 <kylem> right, this week i've been beating tegra back into working in rawhide
20:22:38 <kylem> next up is figuring out why omap isn't working after decompressing the kernel
20:22:46 <pbrobinson> I've noticed a usb patch, should that fix the fun ssd issues?
20:23:07 <bconoboy> #info In 3.11 there is considerable breakage.  Tegra is being worked on now.  Next is OMAP4.
20:23:13 <pbrobinson> and kylem do you have a flyswatter?
20:23:20 <kylem> no, but i'm about to order one
20:23:27 * handsome_pirate waves from the late seat
20:23:32 <pbrobinson> kylem: ping me offline before you do
20:23:58 <kylem> ok. trimslice will be in booting shape with 3.11 once i rebuild dracut (patch is upstream now)
20:24:01 <pbrobinson> ahs3: pass the dunce hat to handsome_pirate ;-)
20:24:12 <kylem> i poked the linux-tegra guys to get them to submit the fix to 3.11-rc2
20:24:15 <kylem> s/rc2/rc3/
20:24:29 <pbrobinson> cool
20:25:05 * handsome_pirate dons the dunce hat
20:25:23 <dgilmore> kylem: im starting a thread on the u-boot list to work on standardising features to make things simpler for us to support. one thing i want us to do is to use extlinux.conf everwhere with raw kernels and initrds
20:25:36 <dgilmore> hopefully that should simplify things for us greatly
20:25:48 <handsome_pirate> dgilmore:  Could you cc me to that?
20:25:58 <kylem> ok.
20:25:58 <pbrobinson> dglimore: would that be useful for the linaro cross-distro list?
20:26:15 <dgilmore> pbrobinson: parts of it would be
20:27:02 <pwhalen> #topic 3a) Aarch64 Status Update, problem packages
20:27:05 <ahs3> +1 on linaro cross-distro list
20:27:46 <bconoboy> Biggest update for the week is that msalter got openjdk into stage4
20:28:08 <bconoboy> We have 11420/13606 packages built, 4832 of them containing aarch64 components.  The build queue is currently empty.
20:28:25 <pbrobinson> nic!
20:28:27 <pbrobinson> nice!
20:28:36 <pwhalen> perhaps worth mentioning if your running the model, disable stage3 in the mock config and yum repos
20:28:45 <bconoboy> From this point forward it's a matter of identifying packages that failed to build that will succeed if requeued, and fixing packages that failed to build.
20:28:52 <bconoboy> Ah, good point
20:29:01 <bconoboy> #info  11420/13606 packages built, 4832 of them containing aarch64 components.  The build queue is currently empty.
20:29:21 <bconoboy> #info about 2000 packages remain.  These will require requeueing or build fixes.
20:29:38 <pbrobinson> can we remove/hide the stage3 repo if it should no longer be used?
20:29:44 <bconoboy> #info Stage 3 is retired - remove it from /etc/mock/stage4-aarch64.cfg or mock will fail
20:30:18 <bconoboy> #info Stage 3 is retired - remove /etc/yum.repos.d/stage3 or yum will produce errors
20:30:41 <bconoboy> pbrobinson: It's already hidden
20:30:52 <pbrobinson> cool
20:31:00 <bconoboy> So, we have something to discuss
20:31:10 <bconoboy> Right now stage4 is almost exclusively packages from f19 GA
20:31:19 <bconoboy> No updates
20:31:27 <bconoboy> No f20 builds
20:31:51 <pbrobinson> ok, presumably f20 will be for when we get koji
20:31:53 <bconoboy> I was hoping that we could move to F20 by moving to Koji, but dgilmore has too much on his plate to get koji setup in the near future.
20:32:08 <bconoboy> So, we have a few options:
20:32:17 <bconoboy> 1: Hang out with f19 GA srpms, fixing build failures
20:32:26 <bconoboy> 2: Bring in f19 updates to stage 4
20:32:36 <bconoboy> 3: Bring in rawhide to stage 4
20:33:03 <bconoboy> Opinions?
20:33:07 <pbrobinson> if we can wait a week or two to get f19 mass rebuild and armv7 promotion out of the way I think it's probably worth it to get it working more automated and properly
20:33:26 <bconoboy> I'm leaning toward #2
20:33:32 <ddd_> bconoboy: ah, get f19 all built ... then go to sure-to-be-broken f20 pkgs ... :-/
20:33:55 <bconoboy> If we do #2 we'll be comingling f19 ga with f19 updates packages
20:34:01 <bconoboy> Which I don't personally mind
20:34:18 <bconoboy> But thought others might dislike.
20:34:21 <pwhalen> option 2 could potentially pull in some fixes
20:34:28 <pbrobinson> dgilmore: how long do you think koji will be/take with f19 mass rebuild etc?
20:35:43 * handsome_pirate is also leaning #2
20:35:44 <bconoboy> we could also have a stage4-updates repo
20:35:53 <bconoboy> that'd require making some changes to the infra
20:36:01 <handsome_pirate> However, it may not be a bad move to start thinking about building some rawhide critpath
20:37:21 <bconoboy> Okay, I'm going to go with #2 unless somebody stops me in the next 8 hours :-)
20:37:41 <pbrobinson> I think that works
20:37:50 <bconoboy> #action bconoboy to start building f19-updates in stage4
20:38:06 <ahs3> i'm inclined to complete #1, have the stage4-updates repo, then move to f20, just so there's a sense of finishing things off -- but i can go with option #2 also
20:38:31 <bconoboy> I don't think "completing" #1 is really feasible, it's just a question of when to move on
20:38:46 <pbrobinson> would be good to launch into f20 from koji clean using the f19 stage4
20:38:58 <bconoboy> y
20:39:22 <ahs3> bconoboy: agreed, which is why option #2 seems okay, too
20:39:23 <bconoboy> By the way, I'm importing the f19ga noarch rpms into stage4 later today.  Should help with some build dependencies.  We last did this in early May
20:39:34 <ddd_> bconoboy: why is #1 not feasible? ... too many bugs that updates has fixed ?
20:39:52 <bconoboy> ddd_ Hundreds of packages not yet ported.  It would take man-years.
20:40:25 <ddd_> bconoboy: and f19-updates doesnt have these hundreds of pkgs?
20:40:27 <pbrobinson> what... not ga yet :-P
20:40:54 <bconoboy> ddd_: f19-updates has hundreds of packages whose version is newer, but is otherwise already built
20:41:27 <ddd_> bconoboy: never mind.... /me confused over what's built or not built btwn f19 & f19-updates...
20:41:38 <bconoboy> Some stuff that goes into f20 ends up in f19-updates, so building f19 updates gives us a partial bridge to f20 when the time comes.
20:41:45 <bconoboy> Anyway, let's move on
20:41:51 <ddd_> bconoboy: thnx
20:42:13 <pwhalen> okay, 3b was covered (aarch64 koji)
20:42:25 <pwhalen> #topic 4) Open Floor
20:42:57 <bconoboy> I would like to mention that we're talking to Compulab about their Utilite platform.  If all goes well we will be able to support it in F20.
20:43:05 * handsome_pirate has something for open floor
20:43:06 <pbrobinson> kylem: status update on llvm?
20:43:16 <bconoboy> If things go astonishingly well we will be able to support it in F19
20:43:36 <pbrobinson> bconoboy: great news, it seems imx6 support in 3.11 is pretty good
20:43:36 <kylem> haven't had a chance to poke it yet given the tegra stuff
20:43:46 <pbrobinson> kylem: thx
20:43:57 <kylem> but, i rebuilt mesa against llvm 3.1, and gnome-shell worked, so we have a data point
20:44:11 <kylem> (i also fixed a pulseaudio bug that was hitting an assert() and crashing at start)
20:44:30 <kylem> gdm login doesn't work yet, but starting gnome-shell from xinitrc does
20:44:45 <pbrobinson> OK, that was working but I've not looked at audio for a while due to lack of time
20:44:49 <kylem> next step is to try to debug llvm 3.3 mcjit, and then trace gnome-session
20:45:01 <kylem> pbrobinson, it may only be on chromebook that it fails
20:45:12 <kylem> fixed now, at any rate
20:45:13 <bconoboy> #info Utilite support being investigated- check back in 4 weeks
20:45:23 <pbrobinson> excellent either way
20:45:28 * handsome_pirate has something
20:45:32 <bconoboy> #info mesa built with llvm 3.1 results in gnome-shell working
20:45:42 <handsome_pirate> Is anyone going to flock riding from RDU?
20:45:47 <handsome_pirate> (Raleigh)
20:45:56 <pbrobinson> nope, flying in via miami
20:46:31 <pbrobinson> handsome_pirate: there may be some RHers though
20:46:48 <pbrobinson> I would put a call out on the flock mailing list
20:47:10 <bconoboy> #action kylem now debugging llvm 3.3 mcjit, will then trace gnome-session
20:47:17 * handsome_pirate is all matter of frustrated atm
20:47:40 <kylem> (gnome session is hammering on a unix socket waiting on something, but i've not figured out what yet)
20:47:41 <pbrobinson> handsome_pirate: can we take flock off the meeting so we can close this out
20:48:08 <pbrobinson> kylem: I'm shocked!
20:48:34 <handsome_pirate> pbrobinson:  Aye
20:48:52 <bconoboy> kylem: Love, probably.
20:48:52 <pwhalen> anything else for today?
20:49:01 <kylem> heh
20:49:04 <pwhalen> :)
20:49:33 <pbrobinson> bconoboy: LMAO!
20:49:37 <pbrobinson> nope
20:49:46 <bconoboy> "Damnit /tmp/.X11-unix/X0, why don't you love me?"
20:50:10 <ahs3> you make that sound so *sultry*....
20:50:16 <bconoboy> pwhalen: I think we're done ;-)
20:50:31 <pwhalen> sound good material here though :)
20:50:36 <pwhalen> #endmeeting