20110707

June 7th, 2011, 15:00 UTC in #ubuntu-meeting.

Agenda

Action Items from last meeting

Standing Items

Meeting Outcome

[15:00] <ogra_> #startmeeting
[15:00] <MootBot> Meeting started at 10:00. The chair is ogra_.
[15:00] <MootBot> Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE]
[15:00] <ogra_> agenda is at https://wiki.ubuntu.com/MobileTeam/Meeting/2011/20110707
[15:00] <ogra_> is everyone here ?
[15:00]  * ogra_ pokes the arm team
[15:01] <GrueMaster> zzzz
[15:01] <ogra_> infinity, around ?
[15:01] <ppisati> o/
[15:01]  * ogra_ sees janimo and pokes to see if he's alive 
[15:02] <janimo> o hai
[15:02] <ogra_> ah, good
[15:02] <ogra_> so only infinity is missing
[15:02] <ogra_> likely asleep
[15:02] <ogra_> lets get started then
[15:02] <ogra_> [topic] standing items
[15:02] <MootBot> New Topic:  standing items
[15:02] <ogra_> http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel.html
[15:03] <MootBot> LINK received:  http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel.html
[15:03] <ogra_> http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel-oneiric-alpha-2.html
[15:03] <MootBot> LINK received:  http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel-oneiric-alpha-2.html
[15:03] <ogra_> i moved most items to A3 that were not DONE
[15:04] <ogra_> if you have any items that actually are DONE and just miss the paperwork, please correct :)
[15:04] <infinity> o/
[15:04] <ogra_> infinity needs to move his remaining three items too :)
[15:04] <GrueMaster> I can move the usbboot wi to A3.  Won't get more done until next week.
[15:05] <ogra_> rsalveti, are you here ? seems there is an A2 item on the workitem tracker for you
[15:05] <GrueMaster> netinstall testcase documentation will be added today or tomorrow.
[15:05] <rsalveti> ogra_: yup, for kernel I believe, let me check
[15:05] <ogra_> can you move, drop, close it (whatever is appropriate)
[15:06] <rsalveti> sure, will move it for now
[15:06]  * ogra_ notes doko and lag also have items on the arm tracker
[15:06] <rsalveti> yep
[15:06] <ogra_> anyway, lets see that we have that cleaned up by tomorrow
[15:06] <lag> ogra_: Link
[15:06] <ogra_> moving on
[15:07] <ogra_> lag, above :)
[15:07] <ogra_> the alpha-2 tracker
[15:07] <ogra_> [topic] arm server status (NCommander, Daviey )
[15:07] <MootBot> New Topic:  arm server status (NCommander, Daviey )
[15:07] <ogra_> hmm, i guess both are on vacation
[15:08] <ogra_> we have server images thanks to tireless testing work from GrueMaster, a d-i upload from infinity to fix netboot and apw for a linux-omap4 upload !
[15:08] <ogra_> \o/
[15:08] <ogra_> that was awesome (cross)team work !
[15:09] <ogra_> sadly it didnt fix desktop or omap3
[15:09] <ogra_> anything else for server to add ?
[15:09] <infinity> Nope.
[15:09] <ppisati> ogra_: did we miss the usb fix patch for omap3?
[15:09] <ogra_> ppisati, nope, but it segfaults on boot
[15:10] <ppisati> ogra_: DOH!
[15:10] <ogra_> patch seems to be in
[15:10] <GrueMaster> ppisati: Not sure.  It has other more serious issues.
[15:10] <ppisati> GrueMaster: ah, you mean the oneiric kernel? ok
[15:10] <ogra_> [topic] kernel status ppisati, cooloney)
[15:10] <MootBot> New Topic:  kernel status ppisati, cooloney)
[15:10] <ppisati> Kernel development: usual CVE fixes across different branches.
[15:10] <ppisati> Oneiric/ti-omap4: we reached an agreement with agreen/linaro to get the TI BSP on top of a vanilla kernel, as a result a new oneiric/master-next-based kernel is in the work - see git://kernel.ubuntu.com/ppisati/ubuntu-oneiric.git ti-omap4-next. Still a WIP, kernel panics on boot ATM, do not use.
[15:10] <GrueMaster> ppisati: See bug 806113
[15:10] <ubottu> Launchpad bug 806113 in linux (Ubuntu) "Series of segfaults early in the kernel boot process on omap." [Undecided,New] https://launchpad.net/bugs/806113
[15:10] <ogra_> ppisati, yes, oneiric omap
[15:10] <ppisati> ok, i'll assign it to me
[15:10] <ogra_> (not omap4)
[15:11] <ppisati> for tegra: ask to orga&c :)
[15:11] <ogra_> heh
[15:11] <ogra_> well, i plan to move to .38 soon
[15:11] <ogra_> with the a100 package
[15:11] <ogra_> once i got that it will also go to oneiric
[15:11] <ppisati> ogra_: btw, i tested the ac100, and it seems the wireless tends to loose pkts
[15:11] <ppisati> did you notice it too?
[15:12] <ogra_> some people do
[15:12] <ogra_> seems hw specific
[15:12] <ppisati> k
[15:12] <ogra_> i have one device with crappy wlan and one where it is rock solid
[15:12] <ogra_> same model
[15:12] <ogra_> same parts
[15:12] <ppisati> same revision? 10z etcetc
[15:12] <ogra_> yep
[15:12] <ppisati> crap
[15:12] <ogra_> 10V
[15:12] <ogra_> anything else for kernel ?
[15:13] <ppisati> nope from me
[15:13] <ogra_> .
[15:13] <ogra_> ..
[15:13] <ogra_> ...
[15:13] <ogra_> moving on then
[15:13] <GrueMaster> ppisati: Can you look at bug 779410 for possible race condition in the kernel?
[15:13] <ubottu> Launchpad bug 779410 in flash-kernel (Ubuntu) "package initramfs-tools 0.98.8ubuntu3 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/779410
[15:13] <ogra_> [topic] arm porting/ ftbfs status (NCommander, janimo)
[15:13] <MootBot> New Topic:  arm porting/ ftbfs status (NCommander, janimo)
[15:13] <ogra_> GrueMaster, i doubt it is the kernel
[15:14]  * ogra_ got pinged today that json-glib has a fix in debian
[15:14] <ogra_> thats holding up a lot other stuff, we need to get it synced
[15:14] <infinity> Can do.
[15:15] <ogra_> not sure how important the other stuff on the ftbfs list is
[15:15] <ogra_> (who uses emacs23 anyway)
[15:15] <infinity> Oh, it was synched 10 minutes ago. :)
[15:15] <ogra_> apaort from that http://qa.ubuntuwire.org/ftbfs/ looks pretty good for that state of the cycle
[15:15]  * janimo is curious what the fix was, so goes take a look
[15:16] <ogra_> universe could need some love though
[15:16] <GrueMaster> There is no love ein the universe?
[15:17] <ogra_> i wonder how we could get the community more active ... PPAs would be good :P
[15:17] <ogra_> GrueMaster, there is, but apparently not in the places i look at :/
[15:17] <persia> No, not PPAs: that just gets more stuff fixed *somewhere* that isn't useful.
[15:18] <ogra_> for testbuilds that helps a lot
[15:18] <persia> That's true, although it's sadly not used that way usually (from what I see)
[15:18] <ogra_> its a lot less effort to just enable a PPA than to set up a local virtualization env or some weird cross env
[15:19] <ogra_> well, thats a matter of communication i guess
[15:19] <ogra_> anyway ...
[15:19] <ogra_> anything else for ftbfs ?
[15:19] <ogra_> .
[15:19] <ogra_> ..
[15:19] <ogra_> ...
[15:19] <ogra_> moving on then :)
[15:19] <ogra_> [topic] arm image status (NCommander, ogra)
[15:19] <MootBot> New Topic:  arm image status (NCommander, ogra)
[15:20] <ogra_> well, desktop seems badly broken
[15:20] <infinity> Just a tad.
[15:20] <ogra_> we get a non resize friendly filesystem since the switch to live-builder ... and jasper additionally has bugs
[15:20] <GrueMaster> I plan on building a desktop image from a server install today so that I can at least test the other bits.
[15:20] <ogra_> i'll look into jasper asap
[15:21] <GrueMaster> I'm not convinced that jasper has a big issue.
[15:21] <GrueMaster> Since server works on the same SD card.
[15:21] <ogra_> must not be big
[15:21] <ogra_> but apparently it has issues
[15:21] <ogra_> how long does the resize on desktop take ?
[15:22] <ogra_> server took 15min for me
[15:22] <ogra_> (class 4 card though)
[15:22] <GrueMaster> desktop took 30 minutes.
[15:22] <GrueMaster> I think
[15:22] <ogra_> phew
[15:22] <GrueMaster> (I have multiple systems and don't watch that closely).
[15:23] <ogra_> well, high prio to fix that :)
[15:23] <ogra_> anything else about images ?
[15:23] <ogra_> .
[15:23] <ogra_> ..
[15:23] <ogra_> ...
[15:23] <ogra_> moving on
[15:23] <GrueMaster> I'm thinking there is a regression in the kernel.  I'll test with a maverick kernel later today or tomorrow.
[15:23] <janimo> GrueMaster, happens on all SD cards tried?
[15:24] <ogra_> (or not :) )
[15:24] <janimo> I wonder if any of my partitioning changes could affect resize
[15:24] <GrueMaster> I get mixed results.
[15:24] <ogra_> janimo, unlikely
[15:24] <janimo> except in the way intended of course
[15:24] <GrueMaster> Larger cards are more troublesome.
[15:24] <ogra_> janimo, it could possibly affect the repartitioning
[15:24] <ogra_> but the slowness is caused at filesystem creation
[15:25] <ogra_> the inode size is to small
[15:25] <ogra_> so resizing has to move way more then necessary
[15:26] <ogra_> we used to use 8k for that in livecd-rootfs
[15:26] <ogra_> anyway, we'll fix it :)
[15:26] <ogra_> moving on for real now
[15:27] <ogra_> [topic] Qa? (GrueMaster)
[15:27] <MootBot> New Topic:  Qa? (GrueMaster)
[15:27] <GrueMaster> Testing this release has been very problematic (as noted earlier).
[15:27] <ogra_> yeah
[15:28] <ogra_> the date of the sprint week wasnt helping with that either
[15:28] <GrueMaster> I'm running into issues getting benchmarks going on btrfs.  Seems natty image doesn't care for conversion.
[15:28] <ogra_> well, just format an SD and copy a configured FS over
[15:28] <GrueMaster> system fails to mount rootfs using that method.
[15:29] <ogra_> hmm, btrfs-tools installed ?
[15:29] <GrueMaster> I have tried several different things (including nuke & pave).
[15:29] <ogra_> it puts some bits into initrd
[15:29] <GrueMaster> yes
[15:29] <ogra_> weird
[15:29] <ogra_> then it should work
[15:29] <GrueMaster> It is failing in initrd
[15:30] <GrueMaster> At any rate, it was lower priority this week vs release testing.
[15:30] <ogra_> yeah, file a bug and one of us should look into it
[15:30] <GrueMaster> netboot install works for the most part.  Some tweeking needed for the sd boot image size (10M is way too small).
[15:31] <ogra_> ac100 lives with 8M :P
[15:31] <GrueMaster> Easily worked around though, and the rest of the install works fine.
[15:31] <ogra_> our initrds are way to large nowadays imho
[15:31] <GrueMaster> ac100 doesn't have x-loader & u-boot in the boot partition.
[15:31] <ogra_> but hey, who wouldnt want the intel drm drivers :)
[15:31] <GrueMaster> Need to be ~30M I would think.
[15:31] <ogra_> or nouveau ...
[15:32] <persia> Those need to be there with the current model.
[15:32] <GrueMaster> They have nothing to do with boot partition size.
[15:32] <persia> The way that libdrm works, it needs to have *all* implementations available.
=== Ursinha is now known as Ursinha-lunch
[15:32] <ogra_> MLO and u-boot are 250k together
[15:33] <GrueMaster> At any rate, I tested both SD & usbboot methods for netinstall on to sata USB drives.
[15:33] <GrueMaster> ogra_: The problem is the filesystem runs out of space for whatever reason at 10M.
[15:33] <ogra_> yeah
[15:33] <GrueMaster> So make it bigger.
[15:33] <ogra_> sure
[15:34] <ogra_> i'm just ranting ... :)
[15:34] <GrueMaster> Server install works cleanly, which iis odd given the issues with desktop.
[15:34] <ogra_> having 5M initrds is to big too :)
[15:35] <ogra_> yeah, jasper should do exactly the same on both
[15:35] <GrueMaster> In between testing, I have been struggling to get jenkins running.  I am able to get it to remotely log into the pandas and start a slave instance.  Now to get it to do something usefull.
[15:35] <ogra_> modulo the PPA stuff
[15:36] <GrueMaster> Not much else to report here.
[15:36] <ogra_> awesome progress !
[15:36] <ogra_> anything else ?
[15:36] <ogra_> .
[15:36] <ogra_> ..
[15:36] <ogra_> ...
[15:36] <ogra_> moving on
[15:36] <ogra_> [topic] any other business ...
[15:36] <MootBot> New Topic:  any other business ...
[15:37] <ogra_> anything ... anyone ... ?
[15:37] <ogra_> doesnt look like
[15:37] <ogra_> going once
[15:38] <ogra_> going twice
[15:38] <ogra_> ending here ...
[15:38] <ogra_> #ednmeeting
[15:38] <ogra_> bah
[15:38] <ogra_> #endmeeting
[15:38] <MootBot> Meeting finished at 10:38.

Action Items

ARM/Meeting/2011/20110707 (last edited 2011-07-28 17:58:23 by davidm)