[00:04:59] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from 717cc2eb8be26d1553e141aa246e7d23be439e87: Ryan Cui <rcui at chromium dot org>)'
[00:05:14] *** belak has joined #chromium-os
[00:05:20] *** achuith has quit IRC
[00:05:29] <belak> Is there a recommended way of trying chromium-os?
[00:05:54] <ellyjones> belak: you can use hexxeh's vanilla builds
[00:06:46] *** achuith has joined #chromium-os
[00:06:46] *** ChanServ sets mode: +v achuith
[00:08:47] <belak> ellyjones: thanks
[00:11:36] <crosbot> tree became 'Michael Krebs thinks this will not reproduce; he's looking into what he believes is periodic flakiness.'
[00:12:14] <cmasone> um...was the tree supposed to get opened?
[00:12:47] <ellyjones> dunno, tlambert set the status
[00:12:53] <crosbot> tree became 'Tree is open(Michael Krebs thinks this will not reproduce; he's looking into what he believes is periodic flakiness.)'
[00:15:32] <belak> What distro is Chromium OS built on?
[00:15:57] <ellyjones> gentoo
[00:16:12] <belak> Ah, ok
[00:16:21] *** nirnimesh has joined #chromium-os
[00:16:22] *** ChanServ sets mode: +v nirnimesh
[00:16:32] *** tty234 has quit IRC
[00:16:48] *** tty234 has joined #chromium-os
[00:45:26] *** cooled has quit IRC
[00:45:34] *** sbyer has quit IRC
[00:49:55] <kliegs> Is anyone seeing issues with openssh-server not starting on images?
[00:50:15] <Hexxeh> i saw that on a build that's like two days old? not sure if it's the same thing because my builds from the past two days are failing sorry
[00:50:18] <cmasone> kliegs: is this the thing that dale emailed about earlier today, that sam is working on?
[00:50:28] <kliegs> I don't know.  i haven't read all my email today
[00:50:42] <Hexxeh> terminated with status 3
[00:50:44] <cmasone> kliegs: the problem that's made it so we can't run any autotests on images, and haven't been able to all day
[00:50:55] <Hexxeh> that error from openssh-server?
[00:51:19] <Hexxeh> also, running on ARM, shell isn't available under crosh, presumably because i'm running unofficial hardware, what's the most sane way to force-enable dev mode?
[00:51:48] <Hexxeh> crossystem returns a boatload of errors, mostly relating to "unable to open FDT property"
[00:53:30] 
[00:53:50] <Hexxeh> sadly no idea what kind of battery life it's getting because i haven't written the battery driver yet :V
[00:56:22] *** Mazon has joined #chromium-os
[00:56:58] <Mazon> any @ awake? got a ping from freenode staff, that they'd like channels that log, to put a notice in topic (fx: IRC Logs:  http://echelog.com/?chromium-os)
[00:58:07] <adlr> echelog-1: help
[00:58:52] <adlr> does anyone know if we're logging this channel?
[00:59:04] <crosbot> tree became 'Tree is open(http://code.google.com/p/chromium-os/issues/detail?id=14663)'
[00:59:06] <cmasone> adlr: someone is.  Not Google
[00:59:27] <Mazon> has been for a long time. was talked about it in the group some time ago afaik
[00:59:36] <adlr> cmasone: link?
[00:59:45] <adlr> Mazon: if i can find the link, i'll get it in the topic
[01:00:01] <Mazon> http://echelog.com/?chromium-os <-- is the link ;)
[01:00:17] <Mazon> or you mean the dicussion ?
[01:00:21] <Mazon> discussion
[01:01:13] *** ChanServ sets mode: +o adlr
[01:01:47] *** sbyer has joined #chromium-os
[01:01:47] *** ChanServ sets mode: +v sbyer
[01:02:38] *** adlr changes topic to "Official Chromium OS channel | No official binaries are available at this time | http://bit.ly/chromeos-faq | http://www.chromium.org/chromium-os | Want crosbot spam? #chromium-os-changes | This is the dev channel. User chat? #chromium-os-users | IRC Logs:  http://echelog.com/?chromium-os"
[01:02:57] *** ChanServ sets mode: -o adlr
[01:03:54] <Mazon> this channel started as #chromium-dev?
[01:04:54] <cmasone> Mazon: no.  that's a different channel
[01:05:07] <Mazon> wasn't there somekind of split at some point ?
[01:05:25] <cmasone> Mazon: no
[01:05:39] <Mazon> hmm, ignore me then
[01:06:20] <Mazon> the only reference I could find for the logging is for #chromium & #chromium-dev
[01:06:44] <Mazon> neither of which is -os or -support :)
[01:07:43] <Mazon> so in that case - this would be a good time as any to complain if the logging should be stopped and removed :)
[01:09:40] <adlr> Mazon: I'm not sure what you're getting at. Is there something you want us to do?
[01:10:01] <Mazon> 2 things:
[01:10:11] <Mazon> add something in the topic - just done
[01:10:35] <Mazon> tell me now if logging should be removed
[01:10:48] <Mazon> because, from the records I can find
[01:11:08] <Mazon> I only see 'evmar' giving the go ahead for  #chromium & #chromium-dev
[01:11:40] <adlr> i can't speak for everyone, but i think we're generally happy w/ things here
[01:12:15] <Mazon> aye
[01:12:22] <Mazon> I just need to make sure staff is happy
[01:12:29] <Mazon> so adding to topic should suffice
[01:12:33] <adlr> great
[01:12:54] <Mazon> again, let me know if there are any issues - or change of heart
[01:13:03] <adlr> okay
[01:14:43] *** cooled has joined #chromium-os
[01:14:53] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from bb5003ac2afa32d3ac1f17140b62d6c559e115d9: evan at chromium dot org <evan at chromium dot org@78cadc50-ecff-11dd-a971-7dbc132099af>)'
[01:19:56] *** e_t_ has quit IRC
[01:20:31] <cmasone> I'll look at the x86-generic full failure once the results get archived
[01:21:12] <sosa> do you want to keep the tree closed?
[01:21:17] <sosa> it's still another instance of the same Chrome crasher
[01:21:26] <cmasone> sosa: that's the PFQ
[01:21:31] <cmasone> the full failure is a different thing
[01:22:17] <cmasone> sosa: but...we might have crash reports for the PFQ failure!
[01:22:46] <sosa> ah ok
[01:22:46] <cmasone> thieule: hey, if we did get crash reports for the most recent PFQ failure, where would they be in the test_results.tgz archive?
[01:23:10] <thieule> under the test folder and iteration_*
[01:23:31] <thieule> got a test result i can look at?
[01:23:46] <cmasone> https://sandbox.google.com/storage/chromeos-image-archive/x86-generic-pre-flight-queue/0.15.869.0-a1-b124/_index.html
[01:23:54] <cmasone> that one might have the goods
[01:23:57] <thieule> looking
[01:23:57] *** JasonO has joined #chromium-os
[01:23:58] *** JasonO has joined #chromium-os
[01:24:13] <cmasone> thieule: thx
[01:24:26] *** petermayo has quit IRC
[01:25:02] *** tlambert has joined #chromium-os
[01:25:02] *** ChanServ sets mode: +v tlambert
[01:25:40] *** sadrul has quit IRC
[01:29:09] <crosbot> tree became 'Tree is closed (tlambert looking now... Automatic: "cbuildbot_master" on "x86 generic PFQ" from bb5003ac2afa32d3ac1f17140b62d6c559e115d9: evan at chromium dot org <evan at chromium dot org@78cadc50-ecff-11dd-a971-7dbc132099af>)'
[01:35:22] *** tlambert has quit IRC
[01:38:27] *** tlambert has joined #chromium-os
[01:38:27] *** ChanServ sets mode: +v tlambert
[01:39:27] <cmasone> once x86-generic full finished archiving its results, I'll investigate that.
[01:40:04] <thieule> so close...
[01:40:38] <cmasone> thieule: per kmixter, it sounds like we can grab the chrome debug symbols from a full builder and use those, as long as they're the same chrome version it should work
[01:40:54] <thieule> nice
[01:54:55] <crosbot> tree became 'Tree is open (same transient test failure, it succeeded on the re-test, thieule and cmasone are on it)'
[01:56:57] <sergiu> which test is failing?
[02:03:13] *** jennb has quit IRC
[02:03:43] *** BThompsonGR has joined #chromium-os
[02:05:36] *** Keybuk has quit IRC
[02:06:49] *** Keybuk has joined #chromium-os
[02:06:49] *** ChanServ sets mode: +v Keybuk
[02:09:26] *** Keybuk has quit IRC
[02:10:50] *** Keybuk has joined #chromium-os
[02:10:50] *** ChanServ sets mode: +v Keybuk
[02:13:03] *** rbyers_ has joined #chromium-os
[02:20:49] <Hexxeh> what's the correct way to enable touchui when building chromiumos, and not just the browser seperately?
[02:21:08] <Hexxeh> s/seperately/separately/
[02:26:07] *** BladeFreak has quit IRC
[02:27:09] *** lipsinV2 has joined #chromium-os
[02:30:25] *** sadrul has joined #chromium-os
[02:30:25] *** ChanServ sets mode: +v sadrul
[02:31:06] *** bpontes has quit IRC
[02:32:08] <cmasone> tlambert: ah!  my change to address the DNS-timing-out thing during an autotest had not yet been marked as stable
[02:32:43] <cmasone> tlambert: so, the full build was not using it. This time, it should.  So, hopefully, it that "Timed out waiting to revert DNS" failure is actually fixed
[02:34:13] *** mazda has quit IRC
[02:34:49] *** chocobo__ has quit IRC
[02:35:29] <tlambert> that would be cool.
[02:35:45] *** wfrichar has quit IRC
[02:41:56] *** BladeFreak has joined #chromium-os
[02:46:45] *** LarXa has joined #chromium-os
[02:48:35] *** JasonO has quit IRC
[02:50:18] *** LaraX has quit IRC
[02:50:45] *** ajsdlk has joined #chromium-os
[02:55:00] *** LarXa is now known as LaraX
[02:56:46] *** ajsdlk has quit IRC
[03:01:48] *** TW1920 has quit IRC
[03:05:23] *** mazda has joined #chromium-os
[03:06:23] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from d48cfcf9445109c2af8c93acb7727317479c06d3: eae at chromium dot org <eae at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>)'
[03:06:32] *** mazda has left #chromium-os
[03:07:52] *** JasonO has joined #chromium-os
[03:07:53] *** JasonO has joined #chromium-os
[03:16:36] <crosbot> tree became 'Tree is open (same transient test failure, thieule and cmasone are on it)'
[03:17:22] <crosbot> tree became 'Tree is closed (restarting sdk builder to pick up important change)'
[03:17:52] <crosbot> tree became 'Tree is open (same transient test failure, thieule and cmasone are on it)'
[03:18:38] <crosbot> tree became 'Tree is open (crosbug.com/17899 -- but we've got a dump, and will symbolize tomorrow)'
[03:23:29] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from d5b4038d4b751c5f961d903d340c5f67f414310a: Chris Sosa <sosa at chromium dot org>)'
[03:26:02] <crosbot> tree became 'Tree is open (same problem: crosbug.cm/17899)'
[03:26:29] <sergiu> crosbug.com *
[03:26:42] <sergiu> any idea why I can't see that issue, it gives me a 403?
[03:27:05] <sergiu> I don't think it affects me, more out of curiosity :)
[03:27:27] <davidjames> sergiu: That issue probably has some private information so permissions are restricted
[03:27:50] <sergiu> ah, that makes sense, I thought it was something like that
[03:29:17] *** josephsih has joined #chromium-os
[03:32:50] <davidjames> sergiu: Opened up http://crosbug.com/17899 since there's nothing secret there
[03:33:10] <davidjames> sergiu: Just the name of our build machine which isn't terribly secret :)
[03:39:04] *** Keybuk has left #chromium-os
[03:42:45] *** rbyers_ has quit IRC
[03:46:47] <josephsih> hello, I am sheriff today in Taipei office.
[04:02:49] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from a3081e08a372dcf29a82fd0f33491252697c1851: _third_party_ at chromium dot org)'
[04:11:34] <josephsih> djkurtz / adlr are working on the bug that caused the tree closed.
[04:12:32] *** jrbarnette has quit IRC
[04:15:56] <adlr> fix has been LGTMed. should go in any second now
[04:19:29] <adlr> pushed
[04:22:15] <crosbot> tree became 'Tree is open (Fix is in for crosbug.com/18887, arm builders should cycle green)'
[04:26:01] <sergiu> davidjames, thanks, now I see that it was definitely not related to what I was doing, but I like to keep an eye for the tree to get a feel of how things work (+ it sometimes is related) :)
[04:26:50] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb canary" from None: )'
[04:27:01] <MostAwesomeDude> Evening. Reporting back on my issue; I don't know whether my Chromebook is on R14. It says that it's got Chrome 13.0, but I don't know if that's an indicator of the entire system version.
[04:27:27] <MostAwesomeDude> Also, chroot'd xev + all system libs still fails to receive input.
[04:32:47] *** bbck has quit IRC
[04:34:29] <crosbot> tree became 'Tree is open (same problem: crossbug.com/18788)'
[04:35:31] <crosbot> tree became 'Tree is open (same problem: crosbug.com/18788)'
[04:40:38] <marcheu> MostAwesomeDude: yeah chrome 13 == R13
[04:47:33] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-mario canary" from None: )'
[04:48:02] <MostAwesomeDude> marcheu: Okay. So I should figure out how to force an update while in dev mode. I'm not super-sure on how to do this and I can't find docs; maybe I should save my stuff, switch off dev mode, and autoupdate?
[04:48:25] <marcheu> I don't know
[04:48:53] <MostAwesomeDude> Alright. I'll figure it out. I'm good with computer.
[04:49:53] <marcheu> yeah sorry, I have no idea about the update stuff, I obviously don't use it
[04:59:47] <MostAwesomeDude> No worries. I'm trying to achieve things which people can do without having to spin an entire image, so I'm trying very hard to avoid touching the build system.
[05:00:32] *** bbck has joined #chromium-os
[05:00:50] *** BladeFreak has quit IRC
[05:09:46] *** mao has joined #chromium-os
[05:10:09] <mao> hello
[05:10:38] <mao> who can give me a chrome-os download link
[05:10:44] <mao> thanks
[05:13:34] <crosbot> tree became 'Tree is open (same problem: crosbug.com/17899)'
[05:13:49] <sergiu> mao, you can try Hexxeh's builds: http://chromeos.hexxeh.net/ but you should know that (as the topic says) there are no official builds right now
[05:15:16] *** BThompsonGR has quit IRC
[05:16:10] <mao> sergiu: thank you a lot......i'm a rookie
[05:17:04] *** Inumedia has joined #chromium-os
[05:19:21] *** Inumedia_ has quit IRC
[05:25:48] *** Paul_C|away has joined #chromium-os
[05:25:50] *** Paul_C|away is now known as Paul_C
[05:26:24] *** Paul_C has left #chromium-os
[05:31:12] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from e0d13d42c9428d2da55fe5b2f32c146441de9e8b: dhollowa at chromium dot org <dhollowa at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, varunjain at chromium dot org <varunjain at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[05:40:15] *** D|sT has quit IRC
[05:49:05] <crosbot> tree became 'Tree is open (same problem: crosbug.com/17899)'
[05:55:08] *** D|sT has joined #chromium-os
[06:04:47] *** JasonO has quit IRC
[06:05:12] *** sergiu has quit IRC
[06:20:28] <cmasone> MostAwesomeDude: you may have to make sure that xev can get at the appropriate .Xauthority file.  It's not in a standard place
[06:35:25] *** shaggymane has joined #chromium-os
[06:45:10] *** Xavierdarkness has left #chromium-os
[06:48:19] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from ecd11ad7e3dcfef640f951e5036bcd82cd86f3bb: Benson Leung <bleung at chromium dot org>, Chris Sosa <sosa at chromium dot org>, Simon Que <sque at chromium dot org>, _third_party_ at chromium dot org)'
[06:51:09] <yoshiki> same 'sig 11' problem?
[07:03:53] <crosbot> tree became 'Tree is open (same problem: crosbug.com/17899)'
[07:14:57] *** stalled has quit IRC
[07:17:32] *** mao has quit IRC
[07:24:42] *** Xavierdarkness has joined #chromium-os
[07:25:23] <MostAwesomeDude> cmasone: That's already handled just fine. sgt-puzzles works.
[07:25:36] <MostAwesomeDude> And, I think, xev works with R14. Not sure yet, but I think that's fine.
[07:25:52] *** shaggymane has quit IRC
[07:28:58] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 generic full" from d5b4038d4b751c5f961d903d340c5f67f414310a: Chris Sosa <sosa at chromium dot org>, _third_party_ at chromium dot org)'
[08:06:16] *** oc80z has quit IRC
[08:06:45] *** oc80z has joined #chromium-os
[08:08:33] <crosbot> tree became 'Tree is open (same problem: crosbug.com/17899)'
[08:14:09] *** Styx has joined #chromium-os
[08:15:42] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "arm generic full" from f26699024f21ebc7f13436fbcf1c28627febc6a6: Daniel Kurtz <djkurtz at chromium dot org>)'
[08:22:11] <MostAwesomeDude> cmasone: The .Xauthority did move in R14, but I found it and updated my scripts accordingly. xev works just fine now. The fullscreen apps (vbam, zsnes) are still not grabbing keyboard, though. :c
[08:58:04] <crosbot> tree became 'Tree is open (yoshiki reverting)'
[09:00:38] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 seaboard full" from a546540419e04a6b3c1e3eaad12976d308016de6: Daniel Kurtz <djkurtz at chromium dot org>, tkent at chromium dot org <tkent at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>)'
[09:03:35] <yoshiki> This seems caused by where I fixed. Opening.
[09:03:57] <crosbot> tree became 'Tree is open (yoshiki reverting)'
[09:08:32] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl canary" from None: )'
[09:09:33] <crosbot> tree became 'Tree is closed (same problem: yoshiki reverted)'
[09:09:48] <crosbot> tree became 'Tree is open (same problem: yoshiki reverted)'
[09:19:15] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_arthur canary" from None: )'
[09:21:58] *** gfrog has quit IRC
[09:23:50] <crosbot> tree became 'Tree is open (same problem: yoshiki reverted)'
[09:25:58] *** gfrog has joined #chromium-os
[09:30:45] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 full" from a546540419e04a6b3c1e3eaad12976d308016de6: tkent at chromium dot org <tkent at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>)'
[09:39:13] *** stalled has joined #chromium-os
[09:40:44] <crosbot> tree became 'Tree is open (yoshiki reverted)'
[09:45:15] *** Styx has quit IRC
[09:48:50] *** Styx has joined #chromium-os
[10:20:04] *** vmil86 has joined #chromium-os
[10:29:20] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_kaen canary" from None: )'
[10:31:47] *** msb___ has joined #chromium-os
[10:31:47] *** ChanServ sets mode: +v msb___
[10:32:20] *** micahc_ has joined #chromium-os
[10:32:20] *** ChanServ sets mode: +v micahc_
[10:32:21] *** xiyuan_ has joined #chromium-os
[10:32:21] *** ChanServ sets mode: +v xiyuan_
[10:32:24] *** gauravsh_ has joined #chromium-os
[10:32:24] *** ChanServ sets mode: +v gauravsh_
[10:32:25] *** cmasone_ has joined #chromium-os
[10:32:25] *** ChanServ sets mode: +v cmasone_
[10:32:28] *** cowbud has joined #chromium-os
[10:32:28] *** ChanServ sets mode: +v cowbud
[10:32:58] *** achuith_ has joined #chromium-os
[10:32:58] *** ChanServ sets mode: +v achuith_
[10:33:09] *** achuith has quit IRC
[10:33:09] *** achuith_ is now known as achuith
[10:33:43] *** pstew has quit IRC
[10:34:17] *** katier1 has joined #chromium-os
[10:34:17] *** ChanServ sets mode: +v katier1
[10:35:08] *** msb__ has quit IRC
[10:35:10] *** katier has quit IRC
[10:35:27] *** gauravsh has quit IRC
[10:35:28] *** vlaviano has quit IRC
[10:35:30] *** _scottz has quit IRC
[10:35:34] *** cmasone has quit IRC
[10:35:51] *** micahc has quit IRC
[10:36:06] *** xiyuan has quit IRC
[10:43:41] *** patcito has quit IRC
[10:44:37] *** msb___ has quit IRC
[10:46:18] *** katier1 has quit IRC
[10:46:52] *** micahc_ has quit IRC
[10:47:01] *** cowbud has quit IRC
[10:47:03] *** gauravsh_ has quit IRC
[10:47:12] *** cmasone_ has quit IRC
[10:47:21] *** xiyuan_ has quit IRC
[10:48:23] *** msb__ has joined #chromium-os
[10:48:23] *** ChanServ sets mode: +v msb__
[10:48:43] *** micahc has joined #chromium-os
[10:48:43] *** ChanServ sets mode: +v micahc
[10:48:51] *** gauravsh has joined #chromium-os
[10:48:55] *** cmasone has joined #chromium-os
[10:49:01] *** ChanServ sets mode: +v gauravsh
[10:49:01] *** ChanServ sets mode: +v cmasone
[10:49:04] *** xiyuan has joined #chromium-os
[10:49:04] *** ChanServ sets mode: +v xiyuan
[10:59:45] *** katier has joined #chromium-os
[10:59:45] *** ChanServ sets mode: +v katier
[11:00:06] *** cowbud has joined #chromium-os
[11:00:06] *** ChanServ sets mode: +v cowbud
[11:11:43] *** oc80z has quit IRC
[11:11:43] *** oc80z has joined #chromium-os
[11:51:04] <crosbot> tree became 'Tree is open (yoshiki reverted &amp; bots restarted)'
[11:56:19] *** gfrog is now known as GNUfrog
[11:58:34] <crosbot> tree became '?'
[11:58:49] <crosbot> tree became 'Tree is open (yoshiki reverted &amp; bots restarted)'
[12:06:36] *** GNUfrog is now known as gfrog
[12:08:01] *** jochen__ has quit IRC
[12:08:06] *** jochen__ has joined #chromium-os
[12:08:06] *** ChanServ sets mode: +v jochen__
[12:13:24] *** lipsinV2 has quit IRC
[12:15:52] *** LaraX has quit IRC
[12:23:01] <crosbot> tree became '?'
[12:27:52] <crosbot> tree became 'Tree is open (yoshiki reverted &amp; bots restarted)'
[12:37:28] <crosbot> tree became '?'
[12:37:44] <crosbot> tree became 'Tree is open (yoshiki reverted &amp; bots restarted)'
[12:39:53] *** quannnum has quit IRC
[12:40:47] *** quannnum has joined #chromium-os
[12:40:57] *** pstew has joined #chromium-os
[12:40:57] *** ChanServ sets mode: +v pstew
[12:42:59] *** vlaviano has joined #chromium-os
[12:43:00] *** ChanServ sets mode: +v vlaviano
[12:52:24] *** lipsinV2 has joined #chromium-os
[13:03:21] *** lipsinV2 has quit IRC
[13:14:48] *** gfrog has quit IRC
[13:26:16] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 57471c4efc263ece81583ac1ee0aa3f8cb164a3f: Hung-Te Lin <hungte at chromium dot org>)'
[13:29:50] <crosbot> tree became 'Tree is closed (hungte checking -> login_CryptohomeMounted)'
[13:30:07] *** stalled has quit IRC
[13:30:21] <crosbot> tree became 'Tree is open (same problem: crosbug.com/17899)'
[13:35:32] *** gfrog has joined #chromium-os
[13:36:03] <crosbot> tree became '?'
[13:36:11] *** lipsinV2 has joined #chromium-os
[13:36:19] <crosbot> tree became 'Tree is open (same problem: crosbug.com/17899)'
[13:41:19] *** stalled has joined #chromium-os
[13:46:57] *** lipsinV2 has quit IRC
[13:47:48] *** cowbud has quit IRC
[13:48:22] *** cowbud has joined #chromium-os
[13:48:22] *** ChanServ sets mode: +v cowbud
[14:01:49] *** lipsinV2 has joined #chromium-os
[14:21:06] *** McMAGIC--Copy has quit IRC
[14:21:40] *** McMAGIC--Copy has joined #chromium-os
[14:22:07] *** SquiRoFL has joined #chromium-os
[14:22:14] <SquiRoFL> hi
[14:28:23] *** lipsinV2 has quit IRC
[14:28:23] <SquiRoFL> wow, all these users but everyone is AFK
[14:28:26] <SquiRoFL> heh
[14:32:49] *** SquiRoFL has quit IRC
[14:36:42] *** lipsinV2 has joined #chromium-os
[14:46:28] *** lipsinV2 has quit IRC
[14:46:47] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard canary" from None: )'
[14:46:54] *** seanpaul has quit IRC
[14:54:01] *** lipsinV2 has joined #chromium-os
[14:54:05] *** lipsinV2 has joined #chromium-os
[15:21:37] *** echelog-1 has joined #chromium-os
[15:24:25] <crosbot> tree became '?'
[15:24:41] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard canary" from None: )'
[15:25:00] *** lipsinV2 has left #chromium-os
[15:45:14] *** SoulRaven has joined #chromium-os
[15:46:38] *** rharrison_chrome has joined #chromium-os
[15:46:39] *** ChanServ sets mode: +v rharrison_chrome
[16:06:25] *** rbyers_ has joined #chromium-os
[16:10:03] *** ScriptRipper has quit IRC
[16:16:05] *** ScriptRipper has joined #chromium-os
[16:27:11] <ellyjones> crosbot: sheriffs?
[16:27:11] <crosbot> ellyjones: sheriffs: tlambert, singh, yoshiki, josephsih
[16:28:52] <ellyjones> it looks like someone deleted a python file that ibus needs from chromiumos-overlay between those two runs
[16:29:28] <ellyjones> er, that ibus-xkb-layots
[16:29:31] <ellyjones> layots
[16:29:33] <ellyjones> layouts
[16:30:53] *** josephsih has quit IRC
[16:31:24] *** saintlou has joined #chromium-os
[16:31:24] *** ChanServ sets mode: +v saintlou
[16:31:59] <ellyjones> this is certainly the guilty commit: chromiumos-overlay:b6665fa7f53a26bd187faa85ac46b02ee0b56f39
[16:32:31] <ellyjones> I wonder why a change from a week ago broke the canary though
[16:35:41] <yoshiki_> the commit breaking ibus-xkb-layots was already revetred.
[16:35:46] <ellyjones> O_o
[16:35:54] <ellyjones> but the canary doesn't have the revert?
[16:36:26] <ellyjones> ah
[16:36:29] <ellyjones> yeah, it doesn't
[16:36:33] <ellyjones> the revert is eb410877e8966af57b9b7dbeb1ea4894d66aa512
[16:36:38] <ellyjones> ok, reopening :)
[16:36:56] *** Mazon has left #chromium-os
[16:37:39] <crosbot> tree became 'Tree is open (ellyjones: tegra2_seaboard-canary will pick up revert for cros-overlay:eb410877)'
[16:54:52] *** sadrul has quit IRC
[17:02:17] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from None: )'
[17:03:17] <ellyjones> hmm
[17:03:36] <ellyjones> chrome sig 11
[17:03:36] <ellyjones>   suite_Smoke/login_CryptohomeMounted
[17:03:47] <ellyjones> screw that bug, I am assigning it to rtc
[17:04:37] <ellyjones> oh, crosbug is readonly
[17:04:56] *** rbyers_ has quit IRC
[17:08:11] *** rbyers_ has joined #chromium-os
[17:10:05] <cmasone> ellyjones: that won't help
[17:10:28] <cmasone> ellyjones: we have a stack trace now, but that's not going to help all that much either unfortunately :-/
[17:10:58] <cmasone> ellyjones: someone needs to go back and figure out at least a rough window of what chrome version introduced the failure
[17:11:01] *** simon___ has joined #chromium-os
[17:11:16] *** rbyers_ has quit IRC
[17:11:59] <simon___> I'm looking to switch to chromium os on my current linux/xbmc media center. Is there any media player that supports hardware acceleration etc. for chromium?
[17:12:16] <simon___> like NVIDIA VDPAU
[17:12:28] <ellyjones> cmasone: ok; I will do that shortly
[17:13:20] *** rbyers_ has joined #chromium-os
[17:18:51] *** rbyers_ has quit IRC
[17:21:47] *** rbyers_ has joined #chromium-os
[17:22:54] *** rbyers_ has quit IRC
[17:23:19] *** saintlou has quit IRC
[17:23:27] *** saintlou has joined #chromium-os
[17:29:12] *** cros_ has joined #chromium-os
[17:29:52] <cros_> Hi. Is there anyone that knows about the window manager and/or open GL ES integration?
[17:31:15] *** vmil86 has quit IRC
[17:32:46] <ellyjones> cros_: I bet reveman does
[17:33:32] <cros_> ellyjones: Thanks, I'll try to ping him.
[17:33:38] <ellyjones> good luck :)
[17:33:52] <reveman> cros_: I might. what about it?
[17:35:54] <cros_> reveman: If my graphics driver directly accesses the X server for display purposes are there any synchronization issues that could arise?
[17:39:27] <reveman> cros_: what do you mean by "directly accessing the X server"? Are you using the X11 protocol to talk to the X server in the graphics driver?
[17:40:02] <cros_> reveman: I mean making a call to XPutImage().
[17:42:07] <reveman> cros_: ok, does all the output produced by your driver get to the server through XPutImage requests?
[17:42:45] <cros_> reveman: I'm not sure what you mean. Are you asking if the requests make it to the XServer or if that is the only kind of request my driver makes?
[17:44:13] <reveman> cros_: are there other resources shared by the driver and the X server? Does the driver do both direct rendering and XPutImage calls?
[17:46:00] <reveman> cros_: if everything displayed by the driver is getting to the X server and the display device through XPutImage calls, then you don't really have to worry much about synchronization issues
[17:46:14] <ellyjones> does anyone know why the build numbers on the bots were reset? x86-gen-full is at build #52, which is clearly wrong
[17:46:19] <ellyjones> (and build #1 is dated august 4)
[17:46:31] <ellyjones> I need history from the old builds to figure out where the chrome sig 11 appeared
[17:46:56] <rochberg> Did they reset when the masters were reset to reflect the new waterfall ordering?
[17:47:27] *** sadrul has joined #chromium-os
[17:47:27] *** ChanServ sets mode: +v sadrul
[17:47:30] <ellyjones> who would one even ask
[17:47:34] <ellyjones> is sosa the right person?
[17:48:34] <reveman> cros_: if the driver is both using direct rendering to buffers shared by the X server and XPutImage calls, then you need to know that the sequence of DRI calls and XPutImage calls might not be the same as the order they are displayed
[17:49:44] <reveman> cros_: unless you do XSync after each XPutImage call
[17:52:36] <cros_> reveman: Give me a second, I'm trying to get answers, not my area of expertise.
[17:53:53] *** jrbarnette has joined #chromium-os
[18:09:57] *** jrbarnette_ has joined #chromium-os
[18:12:09] *** jrbarnette has quit IRC
[18:12:09] *** jrbarnette_ is now known as jrbarnette
[18:15:06] *** seanpaul has joined #chromium-os
[18:15:06] *** ChanServ sets mode: +v seanpaul
[18:20:51] <simon___> Is it possible to playback locally stored video using chromium, also - are there any streaming services that can be installed on my file server to serve my chromium computer with content (HD-content)?
[18:21:19] <simon___> linux file-server ofc
[18:23:30] <redpig> I've used the media player off of Ctrl+M to playback videos (Chrome OS/Chromium OS) though codecs may vary by build type (Chrome v Chromium)
[18:23:49] <redpig> streaming-wise, you can check out what <video> options there are. I'm not sure about direct file hosting (not my area)
[18:24:17] <redpig> You can also write your own html5-based video player as a Chrome App if you want diff features (it's a fun exercise :))
[18:26:47] *** wfrichar has joined #chromium-os
[18:26:48] *** ChanServ sets mode: +v wfrichar
[18:32:03] <simon___> redpig: yeah, I could :p thanks!
[18:32:51] *** mascondante has quit IRC
[18:42:33] *** quiche has quit IRC
[18:42:44] <ellyjones> crosbot: sheriffs?
[18:42:45] <crosbot> ellyjones: sheriffs: tlambert, singh, yoshiki, josephsih
[18:49:26] <crosbot> tree became 'Tree is closed (Internal PFQ is out of space.)'
[18:52:20] *** quiche has joined #chromium-os
[18:52:21] *** ChanServ sets mode: +v quiche
[18:58:11] *** DoDzy has quit IRC
[18:59:26] *** sjg has quit IRC
[19:01:32] *** DoDzy has joined #chromium-os
[19:03:33] <cros_> reveman: What do you mean by direct rendering? Do you mean direct rendering to the framebuffer?
[19:03:40] *** chocobo__ has joined #chromium-os
[19:03:41] *** ChanServ sets mode: +v chocobo__
[19:03:53] *** rginda has joined #chromium-os
[19:03:53] *** ChanServ sets mode: +v rginda
[19:13:19] *** Keybuk has joined #chromium-os
[19:13:20] *** ChanServ sets mode: +v Keybuk
[19:15:39] *** JasonO has joined #chromium-os
[19:15:39] *** JasonO has joined #chromium-os
[19:18:24] <crosbot> tree became 'Tree is closed (sosa is cleaning up space on internal pfq.)'
[19:22:13] *** grundler has joined #chromium-os
[19:22:14] *** ChanServ sets mode: +v grundler
[19:30:51] <sosa> Looks like the chroot/tmp was at 380GB on the builder
[19:31:12] <sosa> because we never wipe chroots and tmp in the chroot doesn't have the standard tmp characteristics of being wiped on reboot like /tmp
[19:31:19] <ellyjones> holy crap, that is huge
[19:31:22] <sosa> scottz is filing a bug to fix that
[19:31:25] <ellyjones> cool
[19:31:44] <sosa> it's chroot/tmp never being wiped in probably a year on that builder
[19:31:54] *** sergiu has joined #chromium-os
[19:31:54] *** sergiu has joined #chromium-os
[19:32:04] <sosa> and other processes treating it as normal /tmp in the chroot
[19:32:10] <ellyjones> yeah
[19:32:12] <reveman> cros_: yes
[19:32:53] <ellyjones> sosa: when you get a bug number, please stick it in the tree status? :)
[19:33:18] <crosbot> tree became 'Tree is closed (sosa is cleaning up space on internal pfq. crosbug.com/18909)'
[19:33:45] <ellyjones> sosa: danke
[19:35:08] *** rbyers_ has joined #chromium-os
[19:35:19] <sosa> scott is finishing clean up cause i gottta run to a meeting
[19:35:21] <crosbot> tree became 'Tree is closed (scottz is cleaning up space on internal pfq. crosbug.com/18909)'
[19:35:59] <cros_> reveman: The graphics driver does NOT do direct rendering to the framebuffer.
[19:37:12] <cros_> reveman: Backing up a bit, in the UI composition does the window manager have any specific requirements on the Open GL ES implementation such as timing, etc.?
[19:38:18] *** stevenjb has joined #chromium-os
[19:38:19] *** ChanServ sets mode: +v stevenjb
[19:40:59] <reveman> cros_: not really
[19:41:56] *** BladeFreak has joined #chromium-os
[19:42:24] <reveman> cros_: so what is the driver doing? client side rendering and XPutImage calls?
[19:42:35] <cros_> reveman: I don't see any usage of DRI in the UI composition.
[19:42:55] <jrbarnette> gee... should we change enter_chroot to mount /tmp as tmpfs?
[19:42:56] <Hexxeh> Got Tegra drivers running, but ChromiumOS performance is still pitifully slow, is there some flag required to enable accel or something? Xorg log reports that everything is fine, driver loaded without issue
[19:43:07] <marcheu> reveman: the best way to do compositing is software rendering + XPutImage, everyone knows that!
[19:43:42] *** saintlou has quit IRC
[19:43:57] <jrbarnette> also, maybe we should clean various scripts to leave less junk in /tmp
[19:44:15] <reveman> marcheu: heh
[19:44:48] <marcheu> btw fixed those two bugs yesterday, now to the antialiasing one, it seems a bit more tedious
[19:44:55] <reveman> cros_: with "UI composition", do you mean chromeos-wm compositor code?
[19:45:27] <cros_> reveman: The GPU will output to the render target and use XPutImage to show on the screen.
[19:45:51] <marcheu> cros_: no, really, no.
[19:46:49] <cros_> reveman: Yes, that is what I mean by UI composition.
[19:49:31] <cros_> marcheu: So no one uses the GPU?
[19:50:00] <marcheu> everyone uses it. I'm saying that "The GPU will output to the render target and use XPutImage to show on the screen." is a completely wrong statement
[19:50:50] <cros_> marcheu: How so? Can you explain?
[19:51:38] <marcheu> compositing uses GL or GLES depending on your platform. look at the source code, it's all available
[19:52:17] <cros_> marcheu: That is what we are trying to debug. We are using GLES hardware solution but are running into problems.
[19:52:51] <marcheu> what GPU?
[19:53:30] *** LaraX has joined #chromium-os
[19:53:32] <crosbot> tree became 'Tree is open (internal PFQ should cycle green)'
[19:53:59] *** seumas has left #chromium-os
[19:55:36] <ellyjones> chrome has been crashing 100% of the time lately on x86-alex-bin
[19:56:21] <cros_> reveman: The minimum timeout in the gl_compositor.cc is 16 ms. Where does that number come from?
[19:57:04] <Hexxeh> Is OpenMAX required to have things not run like molasses on Tegra?
[19:57:54] <reveman> cros_: 60 fps, 1000 / 60 = 16ms
[19:58:01] *** phuang has quit IRC
[19:58:09] <cros_> reveman: The problem we're facing is for our graphics driver we run an independent graphics composition benchmark (glcompbnch) and the driver works fine. However, when we integrate the graphics driver with the chromium window manager we see 4 login screens show up side by side instead of 1.
[19:58:40] *** phuang has joined #chromium-os
[19:58:40] *** ChanServ sets mode: +v phuang
[19:58:54] <marcheu> cros_: so what GPU is that?
[20:01:32] <aaronp> Hexxeh, what are you trying to run, exactly?
[20:01:41] <aaronp> OpenMAX is definitely required for video to not run like molasses.
[20:01:50] <Hexxeh> Asus Transformer with ChromiumOS tegra2
[20:02:15] <Hexxeh> Added the LDK alpha files to the image, Xorg log says the driver loaded fine, but everything runs very slowly
[20:02:25] <Hexxeh> HTML5 video and Flash are slideshows
[20:02:32] <Hexxeh> Most WebGL demos won't load, etc
[20:02:32] *** jrbarnette_ has joined #chromium-os
[20:02:38] <aaronp> Is the clock scaling governor raising the clocks?
[20:02:55] <aaronp> Until very recently, the timing for the 'ondemand' governor was way out of whack and it took forever to raise the clocks.
[20:02:59] <Hexxeh> How would I check that?
[20:03:16] <aaronp> Try setting it to 'performance' and see if it gets better.
[20:03:55] *** JasonO has quit IRC
[20:04:17] *** dots has joined #chromium-os
[20:04:22] *** JasonO has joined #chromium-os
[20:04:22] *** JasonO has joined #chromium-os
[20:04:46] <Hexxeh> I presume this is somewhere under /sys, but I'm not sure where?
[20:04:53] *** seumas has joined #chromium-os
[20:04:53] *** ChanServ sets mode: +v seumas
[20:04:54] <aaronp> I don't remember.
[20:05:18] <aaronp> (stupid developer mode screen taking 30 seconds ... grr)
[20:05:21] *** LaraX has left #chromium-os
[20:05:29] <Hexxeh> Things like es2gears don't even load, too, which I guess is related
[20:05:34] <cmasone> aaronp: ctrl-D
[20:05:35] <adlr> aaronp: hit Ctrl-D to bypass
[20:05:37] <ellyjones> aaronp: you can hit C-d
[20:05:45] <adlr> jinx!
[20:05:47] <aaronp> Not on Springbank, I can't.
[20:05:49] <cmasone> I wonder what happens if you hit ctrl-D...
[20:05:49] <ellyjones> dammit :P
[20:06:00] <eggy> lol..
[20:06:01] *** rginda has quit IRC
[20:06:01] <sergiu> :))
[20:06:09] <aaronp> It has a USB keyboard and the dev-mode switch is hard-wired to on.  :(
[20:06:12] *** jrbarnette has quit IRC
[20:06:15] <Hexxeh> Apparently that's due to an EGL bug in the LDK alpha stuff?
[20:06:20] <Hexxeh> The es2gears, thing, that is
[20:06:27] <aaronp> It's possible, I think I saw a report of that go by.
[20:06:42] <aaronp> Oh wait, I thought the public LDK release was fine.
[20:06:48] <Hexxeh> Apparently not
[20:06:49] <aaronp> What error are you getting, exactly?
[20:06:50] *** jrbarnette_ has quit IRC
[20:07:07] <Hexxeh> EGLUT: failed to get a visual of id 0x0
[20:07:44] <aaronp> Oh, I was thinking of gles2_gears.so.
[20:08:09] <aaronp> Yeah, there's some EGL property that needs to be set.  There's definitely a bug filed, I'll see if I can find it to look up its status.
[20:08:59] <aaronp> /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
[20:09:01] <aaronp> Same for cpu1
[20:09:20] <aaronp> cat cpuinfo_cur_freq to see what the current frequency is, to make sure it goes up.
[20:09:45] <Hexxeh> ah, that might explain it, 216000 at the moment, presumably 216mhz?
[20:11:50] <aaronp> Does this sound like the right bug?  "Return correct X visual ID when EGL_NATIVE_VISUAL_ID is queried from EGLConfigs that support window rendering"
[20:12:01] *** patcito has joined #chromium-os
[20:12:07] <Hexxeh> Looking at the es2gears source, yeah that looks right
[20:12:21] <aaronp> Okay.  Bug 856226, it's assigned but not yet fixed.
[20:12:39] <cros_> reveman: Regarding the 16 ms timeout, you said it comes from 60 fps, 1000/60 = 16 ms. Where does the 1000 come from? Clock frequency?
[20:13:06] <Hexxeh> aaronp: That shouldn't cause the slowness though, I worked around that by patching chromeos-wm
[20:13:42] <aaronp> No, slowness should be unrelated.
[20:14:11] <aaronp> Do you have "perf" installed?  That's the next logical step after you verify that the clocks are being raised correctly.
[20:14:21] <Hexxeh> Yep
[20:15:07] <vpalatin> cros_: 1000  => 1000ms = 1s
[20:16:21] <cros_> cpalatin: Haha, thanks.
[20:18:59] <cros_> vpalatin: If my GPU can't reach 60 fps and say reaches 30 fps, do I need to modify that 16 ms timeout?
[20:20:39] *** jrbarnette has joined #chromium-os
[20:20:39] *** ChanServ sets mode: +v jrbarnette
[20:31:08] *** rbyers_ has quit IRC
[20:33:51] *** rbyers_ has joined #chromium-os
[20:36:56] *** rbyers_ has quit IRC
[20:43:43] <tlambert> vpalatin: what's your bounding box; time XOR frame rate?  Pick one.
[20:44:02] <Hexxeh> aha, eglGetDisplay failed with error EGL_BAD_ALLOC
[20:44:09] <Hexxeh> maybe that's why everything's so slow
[20:44:14] *** phil2 has joined #chromium-os
[20:46:13] <Hexxeh> aaronp: okay, yeah, it's related to the bug you detailed, is there any workaround listed with the bug?
[20:46:22] <Hexxeh> line 53 here http://src.chromium.org/svn/trunk/src/ui/gfx/gl/gl_surface_egl.cc
[20:49:02] <ellyjones> wfrichar: hey, do you want to review <http://gerrit.chromium.org/gerrit/#change,5573>?
[20:50:10] *** cros_ has quit IRC
[20:54:05] *** phil2 has quit IRC
[20:54:33] *** phil2 has joined #chromium-os
[20:55:16] *** phil2 has quit IRC
[20:55:20] <ellyjones> http://gerrit.chromium.org/gerrit/#change,3970 in, woo
[20:55:29] <ellyjones> (probably won't break tests :P)
[20:55:39] *** snickersnack has joined #chromium-os
[20:56:22] <cmasone> tlambert: x86-generic full failed because it got a 404 while trying to download the external Adobe Flash package.  That's not going to just go away...the tree probably should be closed til someone sorts that out
[20:57:06] <Hexxeh> Login failed: Could not mount cryptohome
[20:57:12] <Hexxeh> where should I start in debugging this?
[20:57:16] <ellyjones> Hexxeh: how new of a build?
[20:57:28] <ellyjones> cmasone: it is a pity that curl didn't say what url it wanted
[20:57:28] <Hexxeh> 3rd august
[20:57:36] <Hexxeh> everything newer is failing for me
[20:58:21] <ellyjones> hrm... check /var/log/messages for signs of distress?
[20:59:02] <ellyjones> unfortunately gerrit doesn't tag CLs with the dates they really land
[20:59:13] <Hexxeh> plenty of signs, hmm, communications failure with the TPM (there isn't one)
[20:59:39] <ellyjones> hrm
[20:59:47] <Hexxeh> Cryptohome mount failed: 19, for vault (....)
[21:00:02] <Hexxeh> keyctl_search failed: Function not implemented errno=[38]
[21:00:36] <ellyjones> that seems bad
[21:00:39] *** cros_ has joined #chromium-os
[21:00:50] <Hexxeh> running tegra2_dev-board
[21:01:25] <aaronp> Hexxeh, there's only one 24-bit visual, so just use that one.
[21:02:03] *** simon___ has quit IRC
[21:02:15] <ellyjones> Hexxeh: tot pfq 4200 on aug 2 used cryptohome:147fa3da28fbf114a3497609cd99c9b62f4bef7a; tot pfq 4300 on aug 4 was still using 147fa3da28fbf114a3497609cd99c9b62f4bef7a
[21:02:26] <ellyjones> perhaps a kernel change?
[21:02:39] <Hexxeh> yeah, not using the standard kernel, that might be why
[21:03:02] <ellyjones> does it work with the standard kernel? :)
[21:03:25] <Hexxeh> can't boot the standard kernel, not running on a dev-board, running an asus transformer with the keyboard dock
[21:03:32] <Hexxeh> is there some list of required config options?
[21:03:40] <ellyjones> dunno :\
[21:18:10] *** JasonO has quit IRC
[21:21:44] *** JasonO has joined #chromium-os
[21:21:44] *** JasonO has joined #chromium-os
[21:36:49] *** BladeFreak has quit IRC
[21:40:02] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from e2363378c8d154bf3c10ec9516de26d2573228e6: dslomov at google dot com <dslomov at google dot com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>, reed at google dot com <reed at google dot com@2bbb7eff-a529-9590-31e7-b0007b416f81>)'
[21:40:03] *** JasonO has quit IRC
[21:42:56] <ellyjones> it is plausible that that was my fault
[21:44:45] <ellyjones> is there a way to get at dmesg for the kvm image?
[21:44:55] <ellyjones> sosa? :)
[21:46:26] <ellyjones> kliegs?
[21:47:49] *** JasonO has joined #chromium-os
[21:47:49] *** JasonO has joined #chromium-os
[21:50:33] <ellyjones> ok, I am reasonably confident that it's my fault
[21:51:07] <marcheu> every time I asked, I was told that everything gets destroyed afterwards. you basically have to repro locally with a VM
[21:51:19] <ellyjones> argh
[21:51:26] <marcheu> I said that too.
[21:51:27] <ellyjones> why don't we keep the dmesg at least >.<
[21:51:32] <ellyjones> well, I landed a revert
[21:52:03] <crosbot> tree became 'Tree is open (ellyjones: my kernel change broke boot on x86 pfq, reverted)'
[21:52:20] <ellyjones> so now I need to figure out why it worked for me but not the bots
[21:52:44] *** BladeFreak has joined #chromium-os
[21:55:32] <Hexxeh> seems like cryptoloop is required
[21:56:03] *** ScriptRipper has quit IRC
[21:56:04] *** ScriptRipper has joined #chromium-os
[21:56:30] <Hexxeh> also major lol at the ascii recovery screen... :V
[21:56:34] <ellyjones> :)
[21:57:01] <Hexxeh> i patched crossystem to always say cros_debug was enabled and suddenly the recovery screen started appearing
[21:57:04] <sosa> elly: it's in the test tarball
[21:58:06] <ellyjones> sosa: oh, rock
[21:58:08] <sosa> elly:  oh it's only in the tarball for autotest tests ... we should probably add it for update test failures too
[21:58:12] <ellyjones> wait, it is? oO
[21:58:15] <ellyjones> yeah
[21:58:20] <ellyjones> I just need dmesg, pretty much
[21:58:57] <Hexxeh> what is cryptohomed supposed to do in the absence of a tpm?
[21:59:29] <ellyjones> probably fail badly :P
[22:00:00] <Hexxeh> even on x86 non-official hw?
[22:01:30] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from ee28dc2d69e248ec95556c2635294cf6d9db84f6: _third_party_ at chromium dot org)'
[22:01:42] <Hexxeh> tscd starts and then dies it appears, too, probably related
[22:03:08] *** sadrul has quit IRC
[22:04:44] <ellyjones> sosa: how do I get a serial console from kvm?
[22:05:01] <sosa> ... i don't actually know O:-)
[22:05:05] <sosa> man kvm?
[22:05:33] * ellyjones tries -nographic, but bets cros won't boot that way
[22:06:15] <sosa> ellyjones: the bots boot with --nographics though booting with graphics is nice if you don't have serial console and no ssh :D
[22:08:00] <ellyjones> ok, then something is really badly broken, because the image I built never gets past seabios in qemu and boots fine on real hardware
[22:09:20] <ellyjones> sosa: how do I go about debugging this?
[22:09:40] *** rbyers_ has joined #chromium-os
[22:10:31] <sosa> ellyjones:  Hm, I don't know.  Maybe someone in the kernel land with more KVM knowledge might have some insight?  Ping anush?
[22:11:03] *** rbyers_ has quit IRC
[22:11:08] <Hexxeh> aaaargh cryptohomed errors, everywhere!
[22:12:39] <sosa> FYI check out the LKGMSync stage with an accurate blamelist since laast passing build
[22:14:03] <ellyjones> cool
[22:14:20] <ellyjones> hm, anush is missing
[22:14:29] <Hexxeh> wait, ARM images assume the presence of a TPM?
[22:16:31] <ellyjones> hm
[22:16:36] <ellyjones> sosa: is alex boot efi?
[22:16:59] *** powderluv has joined #chromium-os
[22:16:59] *** ChanServ sets mode: +v powderluv
[22:17:03] <ellyjones> powderluv!
[22:17:07] <powderluv> yes
[22:17:10] <sosa> ellyjones:  i dunno
[22:17:25] * powderluv missed the context
[22:17:28] <ellyjones> powderluv: is alex boot efi?
[22:17:43] <powderluv> nope not with regular bios
[22:17:56] <powderluv> unless you have some special one i think
[22:18:14] <powderluv> anyone looking at the adobe-flash failure on the external bots ?
[22:18:16] <ellyjones> (I'm trying to figure out why I have an image that boots fine on x86-alex, but does not even start in qemu)
[22:18:23] <powderluv> looks like we have to rev the ebuild to pick up a new flash player
[22:18:33] <ellyjones> (and am trying to figure out how to debug)
[22:19:07] <powderluv> for qemu you want your syslinux to be in order
[22:19:11] <powderluv> in part12
[22:19:14] <ellyjones> and for alex?
[22:19:19] <benchan> ellyjones: you may try running kvm with "-usb -monitor telnet::4444,server,nowait", and then telnet localhost 4444 to see the qemu state
[22:19:20] <powderluv> the kernel blob
[22:19:43] <powderluv> what is your qemu command line ?
[22:19:58] <ellyjones> qemu -m 1024 -vga std -sial pty -pidfile /tmp/kvm.pid -net nic,model=virtio  -net user,hostfwd=tcp::9222-:22 -hda ~/l/chromeos-x86-alex/src/build/images/x86-alex/0.15.869.2011_08_09_1342-a1/chromiumos_qemu_image.bin
[22:20:04] <powderluv> crosbot: sheriffs?
[22:20:04] <crosbot> powderluv: sheriffs: tlambert, singh, yoshiki, josephsih
[22:20:37] *** rginda has joined #chromium-os
[22:20:38] *** ChanServ sets mode: +v rginda
[22:22:02] <powderluv> does the same work with x86-generic ?
[22:22:14] <powderluv> or doesn't work with any image at all ?
[22:22:40] *** rbyers_ has joined #chromium-os
[22:22:41] <ellyjones> let me fetch an official build and try it
[22:23:27] <ellyjones> part_12 is not a valid filesystem for me
[22:23:36] <ellyjones> it actually seems to contain a bunch of C++ debug log messages
[22:23:39] <ellyjones> and not a u-boot config
[22:23:45] <ellyjones> er, and not a syslinux config
[22:24:18] <ellyjones> ~ellyjones/public/part_12 is what I have
[22:25:47] <powderluv> elly you should have a fat partition with syslinux in it
[22:26:05] <powderluv> there shouldn't be any u-boot stuff in there for x86 afaik
[22:26:20] <ellyjones> it is not a fat partition
[22:26:44] <ellyjones> either unpack_partitions.sh is messing with me or image_to_vm generated bogosity
[22:27:03] * ellyjones tries unpacking the regular image and sees what that is
[22:27:45] <ellyjones> yes
[22:27:49] <powderluv> ellyjones: no fat no go
[22:27:58] <ellyjones> unpacked from the real image, it is 'x86 boot sector, code offset 0x58, OEM-ID " mkdosfs", sectors/cluster 4, root entries 512, sectors 32768 (volumes <=32 MB) , Media descriptor 0xf8, sectors/FAT 32, heads 64, serial number 0x2711e848, label: "           ", FAT (16 bit)'
[22:29:05] <ellyjones> and from the vm image, it is 'DBase 3 index file'
[22:29:11] <ellyjones> so something's _fucked_ with image_to_vm...
[22:32:49] <ellyjones> it happens reliably
[22:32:56] <ellyjones> the partition contents seem to be just uninitialized
[22:32:57] <sosa> VM images are treated like booting from a USB image if that helps at all
[22:33:05] <adlr> ellyjones: interesting
[22:33:25] <ellyjones> yes, interesting
[22:33:36] <adlr> image_to_vm calls cros_make_image_bootable which calls e2fsck which fails
[22:34:03] <ellyjones> in fact this time it is just zeroes
[22:34:15] <adlr> hmm
[22:35:09] *** kmixter has joined #chromium-os
[22:35:21] *** ChanServ sets mode: +v kmixter
[22:35:24] <adlr> let me try to repro the image to vm thing
[22:36:22] <ellyjones> how I don't see any obvious failures
[22:36:57] <ellyjones> er, s/how//
[22:37:02] <ellyjones> there is no sign of distress in the log
[22:37:30] <ellyjones> do I need to supply --full?
[22:38:26] <adlr> ellyjones: fwiw, i'm doing: ./image_to_vm.sh --full --from ../build/images/x86-alex/latest/ --vdisk_size=6072 --statefulfs_size=3074 --board=x86-alex --test_image
[22:38:44] <ellyjones> I'm just doing: ./image_to_vm.sh
[22:39:22] <ellyjones> is --full required to produce a bootable image?
[22:39:34] <adlr> i don't know
[22:40:25] * ellyjones tries --full
[22:40:43] <adlr> this is what failed: ['./enter_chroot.sh', '--', './image_to_vm.sh', '--full', '--from=/home/chrome-bot/trunk/src/build/images/x86-zgb/0.13.587.101-a1', '--vdisk_size=6072', '--statefulfs_size=3074', '--board=x86-zgb', '--test_image']
[22:40:54] <ellyjones> that failed? hrm
[22:41:04] <crosbot> tree became 'Tree is open (where are sheriffs when you need them?  Usual flaky chrome crasher in cyrptohome test)'
[22:41:05] <adlr> ellyjones: link: http://code.google.com/p/chromium-os/issues/detail?id=18560
[22:41:17] <sosa> i passive-aggressively reopened the tree
[22:41:25] <adlr> heh
[22:41:48] <ellyjones> yeah you did :P
[22:42:13] <adlr> hrm. my image_to_vm passed
[22:42:19] <kliegs> ellyjones: sorry - been in too many meetings today.
[22:42:30] <ellyjones> adlr: even with --full, I ended up with a blank part 12
[22:42:38] <kliegs> crosbot: sheriffs?
[22:42:38] <crosbot> kliegs: sheriffs: tlambert, singh, yoshiki, josephsih
[22:42:41] <adlr> let me check my part 12
[22:43:38] *** kmixter has quit IRC
[22:43:52] <crosbot> tree became 'Tree is open (there are multiple crashers.  One is the signall 11 in the login test 18269; the other is 17899; these don't need fioxed to open the tree.  The other is the flash update, which does need fixed)'
[22:44:58] <rochberg> Anne and I have been looking at the adobe-flash issue
[22:45:01] <adlr> heh, that seems contradictory
[22:45:16] <tlambert> Anyone have any clue on how to do a flash update, and why when we get this failure, it doesn't happen automatically?  Neither I nor Amit have done a flash update before -- Thanks for any help.
[22:45:57] <powderluv> tlambert: I just verified flash11 compiles fine
[22:46:06] <adlr> ellyjones: ah yeah, my part12 is F'd. (meaning it's full of 00 and FF)
[22:46:25] <powderluv> you need to import https://github.com/funtoo/portage/tree/funtoo.org/www-plugins/adobe-flash/adobe-flash-11.0.1.60_beta201107131-r1.ebuild
[22:46:34] <powderluv> and files/mms.cfg
[22:46:39] <rochberg> It seems like we want 10.3.183.5 (or at least that's what one gets if one attempts a flash download).  But naively creating a new ebuild with "cp adobe-flash-10.3.181.26 adobe-flash-10.3.183.5" gets something that will build  if I ebuild merge it
[22:46:54] <powderluv> into chromiumos-overlay/www-plugins/adobe-flash/
[22:47:31] <powderluv> the 10.xx binaries are all missing
[22:47:39] <ellyjones> adlr: same
[22:47:43] <ellyjones> adlr: does your image boot in qemu?
[22:47:45] <rochberg> 10.3.183.5 is not missing
[22:48:03] <powderluv> i checked 34 and it couldn't download it
[22:48:03] <rochberg> but I can't get portage to see my new ebuild
[22:48:15] <powderluv> remove the package.unmask
[22:48:32] <powderluv> src/third_party/chromiumos-overlay/profiles/targets/chromeos/package.mask
[22:48:39] <powderluv> looks for adobe-flash
[22:48:48] <powderluv> yeah if .5 works we can go with it..
[22:49:30] <adlr> ellyjones: haven't tried
[22:49:34] * powderluv late lunch
[22:50:20] <adlr> i wish i could repro the image_to_vm command failing
[22:51:06] <ellyjones> even when it does work the resulting image is not bootable
[22:51:44] *** powderluv has quit IRC
[22:51:46] <ellyjones> (I am staring at update_bootloaders)
[22:52:59] *** rochberg____ has joined #chromium-os
[22:54:32] <adlr> ellyjones: do you know when this started failing?
[22:54:40] <adlr> is there a blamelist?
[22:55:48] <ellyjones> adlr: I do not know, because it's failing for me locally but not on the bots
[22:55:59] <adlr> oh
[22:56:11] <rochberg____> powderluv, tlambert:   Want a CL for this?
[22:56:12] <ellyjones> (which is preventing me from debugging why one of my other CLs doesn't boot on the bots :P)
[22:58:22] <ellyjones> after make_image_bootable in cros_make_image_bootable, we have an x86 boot sector
[22:58:25] *** Kernel-Panic has joined #chromium-os
[22:58:43] *** rbyers_ has quit IRC
[22:58:54] <ellyjones> oh, but it's 3GB
[22:59:05] *** Kernel-Panic has joined #chromium-os
[22:59:46] *** Kernel-Panic has joined #chromium-os
[23:04:04] <rochberg> Anne and I put up a CL for the flash issue
[23:04:39] <adlr> ellyjones: so if i understand your issue, the bots are working, but when you (and i) do image_to_vm, we get unbootable images, and part_12 is particulary suspect
[23:04:44] <ellyjones> yes
[23:05:05] <ellyjones> which is preventing me from debugging a failure of one of my changes to bo inside the vms :)
[23:05:39] <adlr> okay, i'm going to switch gears for a bit and work on another flakey AU test, then i may come back to this
[23:05:43] <adlr> but i'm curious what you learn
[23:06:17] <ellyjones> cool
[23:06:24] <ellyjones> I am adding debug output until I figure out what is going on
[23:07:41] * ellyjones unpacks an intermediate image
[23:10:26] *** rochberg____ has quit IRC
[23:10:35] <ellyjones> update_bootloaders.sh is trashing all the partitions except 1 and 3
[23:10:55] <adlr> heh
[23:20:20] <ellyjones> just before update_x86_bootloaders(), part 12 is already junk
[23:30:52] *** saintlou has joined #chromium-os
[23:31:32] *** bpontes has joined #chromium-os
[23:31:32] *** ChanServ sets mode: +v bpontes
[23:32:57] <ellyjones> before make_image_bootable(), part 12 is junk
[23:35:05] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from 1e031c316b07091ad3f5e4083d74aa7a6dc0b4bb: Chris Sosa <sosa at chromium dot org>)'
[23:35:51] <crosbot> tree became 'Tree is open (there are multiple crashers. One is the signall 11 in the login test 18269; the other is 17899; these don't need fioxed to open the tree. The other is the flash update, which does need fixed)'
[23:36:56] *** SoulRaven has quit IRC
[23:38:52] <ellyjones> adlr: I suspect these lines of image_to_vm:
[23:38:52] <ellyjones> 222 dd if="${TEMP_ESP}"    of="${TEMP_IMG}" conv=notrunc bs=512 \
[23:38:52] <ellyjones> 223   seek="${START_ESP}"
[23:39:13] <adlr> hm
[23:39:14] <ellyjones> ${TEMP_ESP} is a valid part 12 before that; I'm now seeing if unpacking ${TEMP_IMG} after that produces a valid part 12
[23:39:19] <ellyjones> but I'm sorta betting it won't
[23:39:41] <adlr> okay
[23:42:13] <ellyjones> erm
[23:42:21] <ellyjones> the offsets in unpack_partitions.sh are all wrong
[23:42:36] <jrbarnette> uh oh; i was last to touch that code :-(
[23:42:47] <ellyjones> or at least, do not match up with the offsets that are being used in image_to_vm
[23:42:58] <adlr> compare w/ cgpt?
[23:44:09] <ellyjones> gdisk -l shows part 12 starting at 6037504
[23:44:28] <ellyjones> unpack_partitions believes it starts at 3940352
[23:44:47] <ellyjones> actually all of unpack_partitions is royally screwed up; the offsets seem basically random
[23:45:33] <jrbarnette> are the offsets same as pack_partitions?
[23:45:36] *** benchan has quit IRC
[23:45:50] <ellyjones> yes
[23:45:56] <ellyjones> but not the same as what image_to_vm is using
[23:46:23] <jrbarnette> oh, that could be different
[23:46:38] <ellyjones> image_to_vm gets them from common.sh
[23:46:42] <jrbarnette> pack/unpack_partitions uses offsets specific to one image
[23:46:45] <ellyjones> unpack and pack have them hardcoded
[23:46:49] <ellyjones> wait what
[23:46:58] <jrbarnette> but if you build an image differently, it could have different offsets
[23:47:08] <ellyjones> so unpack_partitions is actually not useable on vm images?
[23:47:17] <jrbarnette> not necessarily...
[23:47:21] <ellyjones> argh
[23:47:22] <ellyjones> okay
[23:47:23] <ellyjones> adlr: ^
[23:47:26] <adlr> oh, makes sense
[23:47:34] <ellyjones> that puts me back to square one on why my image doesn't boot in the vm though
[23:47:41] <adlr> you make a new image w/ stuff in different places... then *pack_parittions won't work
[23:47:58] <adlr> i should have noticed that
[23:48:14] <ellyjones> using the right offset I do get a valid x86 boot sector in part 12
[23:51:17] <ellyjones> where does unpack_partitions.sh live?
[23:51:22] <ellyjones> I want to add a huge warning to it
[23:53:01] *** cros_ has quit IRC
[23:59:22] *** eugeni has joined #chromium-os
[23:59:59] <ellyjones> jrbarnette: ?





top