[00:00:22] *** teo__ has joined #chromium-os
[00:02:36] <teo__> hi, does anyone can suggest me some notes on how to compile a simple command line application to be run from terminal in dev mode?
[00:03:48] <sergiu> teo__, run it from the terminal on Chrome OS ?
[00:04:16] <teo__> chromium os
[00:04:29] *** behdad has joined #chromium-os
[00:05:05] <sergiu> ok, then making sure you're compiling it on 32bits should be enough (using -m32)
[00:05:43] <sergiu> not "on" but "for 32 bits" *
[00:06:22] <teo__> but I have to compile from inside cros_sdk?
[00:06:34] <sergiu> no, it works from outside as well
[00:07:05] <sergiu> it really depends on what you're trying to do, but I've tested it with a simple app and it works
[00:07:31] <teo__> so somethng like:   gcc test.c -o test -m32
[00:07:43] <sergiu> yep, that should work
[00:08:29] <teo__> and to configure includes to look at kernels ones?
[00:08:39] *** unreal has quit IRC
[00:09:26] <sergiu> umm, I'm talking about user space programs here
[00:09:46] *** yoshiki_ has quit IRC
[00:09:51] <sergiu> anyone looking at the tree?
[00:10:02] <teo__> yes, me too. But I need to read a driver include
[00:10:11] *** yoshiki_ has joined #chromium-os
[00:11:58] <chocobo__> i'm looking at the new failure
[00:12:00] <teo__> ok sergiu.
[00:12:19] <jhorwich> chocobo: thank you. I saw errors in VMTest on that canary but haven't yet deciphered
[00:12:37] *** behdad has quit IRC
[00:13:04] <chocobo__> looks like a whole slew of builds are currently broken: http://chromeos-botmaster.mtv.corp.google.com:8026/waterfall
[00:13:15] <davidjames> chocobo__: mario canary has Chrome stack trace, it's uploading now to https://sandbox.google.com/storage/?arg=chromeos-image-archive/x86-mario-release/0.16.1043.0-a1-b1009
[00:13:54] <davidjames> chocobo__: It's at test_harness/testUpdateKeepStateful/2_verify/suite_Smoke/desktopui_WindowManagerFocusNewWindows/sysinfo/var/spool/crash/chrome.20110915.143805.4436.dmp.txt
[00:14:09] <teo__> sergiu, sorry but to compile from inside thecros_sdk the same command as before is good? and more which is a good place for a test application?
[00:15:32] <chocobo__> looks like this build has been broken for a while now: http://chromeos-botmaster.mtv.corp.google.com:8026/builders/x86-zgb_he%20canary
[00:16:04] <davidjames> chocobo__: Most of that is ellyjones' breakage
[00:16:19] *** Inumedia has quit IRC
[00:16:35] <davidjames> But VMTest failure is different
[00:17:03] <davidjames> chocobo__: SIGABRT @ chrome!SpellCheckProfile::~SpellCheckProfile [spellcheck_profile.cc : 23 + 0x0]
[00:17:04] *** Inumedia has joined #chromium-os
[00:17:45] <sergiu> teo__, I'm not sure what you're trying to do exactly but the command should work from inside as well afaik
[00:19:49] <teo__> I' just tring to communicate to a device by sysfs (so user space) an I need to compile the app (command line) maybe find a good place for it
[00:21:15] <sergiu> well, then you would probably need just to read/write to sysfs, I think you can put it in /home/chronos to test it on the machine/vm image, but
[00:21:33] <sergiu> but I'm not 100% sure
[00:22:13] <teo__> ok thank you very much  for the info. very kind. ciao.
[00:22:46] *** Solet has quit IRC
[00:23:21] <sergiu> sure, no problem
[00:23:28] *** teo__ has left #chromium-os
[00:25:49] <crosbot> tree became 'Tree is closed (chocobo investigating x86-zgb_he canary failure)'
[00:28:02] *** unreal has joined #chromium-os
[00:33:56] <chocobo__> looks like it's timing issue and related to this bug: http://code.google.com/p/chromium-os/issues/detail?id=19005
[00:34:31] <davidjames> chocobo__: You saw that mario canary has a stack trace in there ?
[00:35:38] <davidjames> chocobo__: Even though build succeeded we still have a stack trace, which shows Chrome is crashing on shut down... that's a problem
[00:39:28] <chocobo__> how is the mario canary build related to this zgb canary failure?
[00:40:14] <chocobo__> the zgc canary build didn't crash on chrome shutdown. the mario build looks fine now, right?
[00:40:38] <davidjames> chocobo__: Mario build is not fine because Chrome crashed... we are planning on changing the test suite to fail on shutdown crashes
[00:41:25] <davidjames> chocobo__: So any shutdown crashes are worth filing bugs for if we have a stack trace
[00:42:01] <chocobo__> ok
[00:42:53] <jhorwich> nothing I saw in 19005 looks like a crash in the same smoke test as the run in question
[00:44:31] *** behdad has joined #chromium-os
[00:45:07] <chocobo__> 19005 has both the desktopui_WindowManagerFocusNewWindows and the login_CryptohomeMounted failure
[00:45:16] <jhorwich> (only crash I saw in anything linkedfrom crosbug 19005 was in suite_Smoke/desktopui_Screenlocker on x86-zgb-canary, which is no longer happening)
[00:45:32] <jhorwich> chocobo: I saw login_CryptohomeMounted failures, but not of the crashing type
[00:46:10] <davidjames> jhorwich: stumpy canary both failed and has a stack trace
[00:47:20] <chocobo__> jhorwich: same 2 test failed. stack trace a bit different maybe due to it being a timing issue?
[00:49:31] <jhorwich> chocobo: possible - I was just pointing out that I'm not convinced this is another instance of 19005
[00:51:58] <chocobo__> jhorwich: yeah, not 100% sure. seems like it to me.
[00:52:22] <chocobo__> davidjames: looks like suite_Smoke/login_LoginSuccess.default crashes. is that the SpellCheckProfile crash?
[00:52:33] <davidjames> chocobo__: Nope, different one
[00:52:48] <davidjames> chocobo__: http://pastebin.com/cpJ4wsC5
[00:58:41] <chocobo__> hmm, investigating
[00:59:38] <crosbot> tree became 'Tree is closed (chocobo investigating x86-zgb_he canary failure) (stumpy/lumpy failure)'
[01:04:18] *** Solet has joined #chromium-os
[01:09:22] <cmasone> jesus.  The openssl thing is still causing problems
[01:10:30] <jhorwich> cmasone: argh, I just noticed the full builders are still failing
[01:12:30] *** unreal has quit IRC
[01:14:20] <gauravsh> is that the update_engine openssl failure I am seeing?
[01:14:52] <cmasone> gauravsh: 'tis
[01:14:58] <jhorwich> gauravsh: looks like it.
[01:15:04] *** lipsinV2 has joined #chromium-os
[01:16:05] <jhorwich> building the new revision (-r8) of openssl seemed to complete in 8m2.0s
[01:16:24] <cmasone> did the host like...not get that new version?
[01:16:51] <cmasone> no, it did, you can see that in BuildBoard
[01:17:00] <cmasone> did the configuration file not get overwritten?
[01:17:05] <cmasone> davidjames: ^^^
[01:21:14] <cwolfe> sheriff: I broke the asymptote canary. Have a patch that will fix it, may I check it in? (It doesn't affect the external builders)
[01:23:43] <jhorwich> cwolfe: please hold on a moment, we are chasing two disjoint tree closers ATM
[01:23:47] <grundler> crosbot, sheriffs?
[01:23:50] <cwolfe> jhorwich: alright, thanks
[01:24:37] <davidjames> cmasone: I guess the chromiumos-sdk bot pushed an sdk with 0.9.8r-r4 of openssl, is that a busted version of openssl?
[01:24:41] <chocobo__> I'm talking to oshima about chrome crashes in stumpy canary
[01:24:54] <jhorwich> davidjames: any insight into full builder failures?
[01:25:26] <gauravsh> davidjames: how do you parse this line in the output
[01:25:27] <gauravsh> [binary     U  ] dev-libs/openssl-0.9.8r-r8 [0.9.8r-r4] USE="(sse2) zlib -bindist -gmp -kerberos -pkcs11% -test"
[01:25:45] <crosbot> tree became 'Tree is closed (chocobo investigating x86-zgb_he canary failure) (edison/newton failure)'
[01:25:47] <gauravsh> is the version enclosed in [ ] the one that is actually being used?
[01:26:20] *** ojn has quit IRC
[01:26:54] *** gspencer has joined #chromium-os
[01:26:54] *** ChanServ sets mode: +v gspencer
[01:26:56] *** achuith has joined #chromium-os
[01:26:56] *** ChanServ sets mode: +v achuith
[01:27:09] <davidjames> gauravsh: No, the one on the left is the one that's being used
[01:27:58] <davidjames> jhorwich: http://gerrit.chromium.org/gerrit/7829 should revert the commit that broke the tree
[01:28:34] <cmasone> gauravsh: that means that it's Updating to openssl-0.9.8r-r8 (using a binary package) from -r4
[01:29:06] <cmasone> gauravsh: and the USE flag stuff is kinda voodoo.  - means that flag is not on, but the parens and the % confuse me
[01:29:07] *** ojn has joined #chromium-os
[01:29:20] <crosbot> tree became 'Tree is closed (chocobo investigating x86-zgb_he canary failure) (lumpy-binary, lumpy-canary failure)'
[01:29:30] <jhorwich> davidjames: looking, thank you
[01:30:28] <gauravsh> is it me, or is this stuff rather arcane and confusing?
[01:30:30] <davidjames> jhorwich: Hold on, just updating that CL
[01:31:04] <cmasone> gauravsh: all computer stuff is arcane and confusing.
[01:31:58] <jhorwich> looks like I picked the wrong week to quit sniffing glue
[01:32:21] <bleung> apparently i'm a popcorn engineer
[01:32:49] <davidjames> jhorwich: Ok, pushed :)
[01:33:10] <davidjames> gauravsh: So the issue was that ellyjones pushed his fix, but then bot built a new sdk without his fix and deployed it
[01:33:12] <jhorwich> davidjames: thank you!
[01:33:40] <davidjames> gauravsh: Ironically, the bot that built the sdk without his fix was kicked off *by* his fix, but because preflight had not run yet, it did not include his fix.
[01:34:46] <davidjames> gauravsh: Further, when people try to install the fix, it is not included because of http://crosbug.com/13987
[01:35:35] <davidjames> ^: s/his/her/g
[01:37:19] <gauravsh> davidjames: thanks for the explanation. yes, I kind of understand all of this. it's just that an hour earlier, I didn't even know about this cros_sdk interaction.
[01:37:19] *** gspencer has quit IRC
[01:37:50] <gauravsh> I guess my main concern is that without you being there, I doubt any of the sheriffs or us would be able to diagnose these issues and have a fix out.
[01:38:18] <davidjames> gauravsh: So next week we are migrating to commit queue and getting rid of cros-workon, so when a fix is pushed it goes live everywhere
[01:38:40] <chocobo__> gauravsh: i vote for making davidjames permanent sheriff
[01:38:56] <gauravsh> and I would like to understand all of this stuff, but I don't know where to start. I thought after sosa's presentation, i knew all i needed to know about the tree and the build process. but i guess not. :)
[01:39:11] <cmasone> chocobo__: and what do you do when he burns out in a quarter, or facebook hires him away?
[01:39:14] <davidjames> gauravsh: Maybe we should have more presentations? :)
[01:39:15] <chocobo__> lol
[01:39:41] <cmasone> gauravsh: we're actively trying to improve things, which means that changes happen :-/
[01:39:46] *** ojn has quit IRC
[01:39:55] <chocobo__> oshima claims that the chrome crash is due to a race condition. i filed a bug (http://code.google.com/p/chromium-os/issues/detail?id=20504) and they will look into it
[01:40:06] <jhorwich> I suppose I should stop and restart the (doomed?) full builders
[01:40:07] <jhorwich> right?
[01:40:47] <gauravsh> davidjames: i think a talk by you about portage and how all of this interacts would be a great idea (once we switch over to the commit queue).
[01:41:12] <davidjames> gauravsh: Yeah, once the commit queue is in the talk will be a lot easier because there is less to explain :)
[01:41:15] <cmasone> gauravsh: also, failures at this level seem more like "trooper" stuff, which means that the thing to do is help sheriffs realize this, and then train up a couple more troopers so taht david isn't the single point of failure
[01:41:37] <crosbot> tree became 'Tree is closed (still fixing openssl failure, flakey lumpy canary failure will be looked into by nkostylev and oshima)'
[01:41:49] <quiche> +1 for the talk
[01:43:19] <davidjames> quiche: Good idea, I'll schedule one!
[01:43:21] <gauravsh> cmasone: agreed.
[01:43:37] <gauravsh> my trybot run ended, so I will get back to actual work. :)
[01:44:09] <davidjames> In general failures like this aren't supposed to happen :) Just bugs like http://crosbug.com/13987 make everything really hard to debug and understand
[01:50:57] *** unreal has joined #chromium-os
[01:53:32] *** ojn has joined #chromium-os
[01:58:23] <davidjames> chocobo__: FYI, all builders should be green now
[01:58:56] <davidjames> chocobo__: chromiumos-sdk actually pushed working SDK at 4pm, at 4:30pm I pushed another working SDK (older, but still working)
[01:59:47] <davidjames> The broken SDK was actually pushed at ~1pm, so it was broken from 1pm-4pm
[01:59:52] <chocobo__> ok, but do you think it's safe to open the tree up?
[02:00:20] <davidjames> chocobo__: If your only concern is the BuildTarget breakage, yes
[02:01:31] *** rbyers_ has joined #chromium-os
[02:11:37] <crosbot> tree became 'Tree is open (openssl failure fixed, should cycle green)'
[02:16:25] *** stevenjb_ has quit IRC
[02:20:39] *** semenzato_ has joined #chromium-os
[02:22:46] *** stevenjb_ has joined #chromium-os
[02:22:47] *** ChanServ sets mode: +v stevenjb_
[02:24:57] *** m1k3l_ has joined #chromium-os
[02:27:59] *** stevenjb_ has quit IRC
[02:28:23] *** semenzato_ has quit IRC
[02:28:33] *** semenzato_ has joined #chromium-os
[02:30:36] *** jep has joined #chromium-os
[02:35:03] *** grundler has quit IRC
[02:39:05] *** behdad has quit IRC
[02:39:51] *** stevenjb_ has joined #chromium-os
[02:50:17] *** ChanServ sets mode: +v ojn
[02:56:16] <cmasone> Wow, Chrome died SIGILL -- Illegal Instruction -- on x86-generic VM testing
[02:58:07] <adlr> illin'
[02:58:40] <davidjames> cmasone: Woah, I wonder if we used a instruction that the VM doesn't support
[02:58:41] <jhorwich> licensed to ill
[02:59:20] *** thieule has quit IRC
[03:00:03] <marcheu> davidjames: doesn't necessarily have to be, could also be jumping to random memory
[03:00:11] <marcheu> I guess we need a core to say for sure
[03:02:32] *** unreal has quit IRC
[03:03:35] *** unreal has joined #chromium-os
[03:05:35] *** sadrul has joined #chromium-os
[03:05:35] *** ChanServ sets mode: +v sadrul
[03:12:30] *** unreal has quit IRC
[03:12:39] *** behdad has joined #chromium-os
[03:16:56] *** rbyers_ has quit IRC
[03:17:04] *** rbyers_ has joined #chromium-os
[03:17:05] *** ChanServ sets mode: +v rbyers_
[03:17:41] <davidjames> marcheu: Yeah that might explain why so many of our stack traces have corrupted stacks
[03:19:26] *** zork has joined #chromium-os
[03:26:36] <chocobo__> i'm heading off soon. zork will be taking over for sheriffing
[03:29:48] <jhorwich> I'm also about to depart after a fun day of sheriffing
[03:31:58] *** stevenjb_ has quit IRC
[03:37:36] *** jhorwich has quit IRC
[03:40:47] *** semenzato_ has quit IRC
[03:58:03] *** yusukes has quit IRC
[04:07:44] *** m1k3l_ has quit IRC
[04:19:06] *** achuith_ has joined #chromium-os
[04:19:06] *** ChanServ sets mode: +v achuith_
[04:19:19] *** achuith_ has quit IRC
[04:21:39] *** achuith has quit IRC
[04:22:10] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from 27687933a0a53a4b62ed1b0992cb8dbf00abda29: mseaborn at chromium dot org <mseaborn at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[04:40:40] *** saintlou has quit IRC
[04:50:36] *** zmedico has quit IRC
[04:51:42] *** rbyers_ has quit IRC
[05:03:15] <crosbot> tree became 'Tree is open (Fix for crosbug.com/20034 landed, TOT preflight queue cycled green)'
[05:04:22] *** cmasone_ has joined #chromium-os
[05:05:02] <crosbot> tree became 'Tree is open (Canaries failing due to Chrome flake -> sheriffs)'
[05:06:16] *** Solet has quit IRC
[05:06:18] *** [S]olet has joined #chromium-os
[05:10:59] *** cmasone_ has quit IRC
[05:16:14] *** behdad has quit IRC
[05:16:39] *** behdad has joined #chromium-os
[05:16:39] *** ChanServ sets mode: +v behdad
[05:20:47] *** Solet has joined #chromium-os
[05:21:01] *** [S]olet has quit IRC
[05:24:15] *** dennisjeffrey has quit IRC
[05:33:43] *** zmedico has joined #chromium-os
[06:50:29] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 generic full" from 3ddd73a078519767e0d89e06660e9c2a9c044351: Andrew de los Reyes <adlr at chromium dot org>)'
[07:10:16] *** semenzato_ has joined #chromium-os
[07:30:31] *** semenzato_ has quit IRC
[07:34:47] *** vmil86 has joined #chromium-os
[07:36:48] <crosbot> tree became 'Tree is closed (zork investigating)'
[07:37:20] *** zork_ has joined #chromium-os
[07:38:05] <zork_> x86 generic full seems to be failing the Report phase regularly
[07:38:12] <zork_> Does anyone know about this?
[07:39:07] *** Styx has joined #chromium-os
[07:42:17] <crosbot> tree became 'Tree is open (Report failure doesn't seem to be a real error)'
[07:57:34] *** hashimoto has quit IRC
[08:16:12] *** hashimoto has joined #chromium-os
[08:24:50] *** behdad1 has joined #chromium-os
[08:26:40] *** behdad has quit IRC
[08:30:52] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex-binary" from b75a63ad7a1c0d2846bc71cab94e66c5d86d2d03: prasadt at chromium dot org <prasadt at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, zmo at google dot com <zmo at google dot com@0039d316-1c4b-4281-b951-d872f2087c98>)'
[08:34:16] *** behdad1 has quit IRC
[08:40:21] *** sadrul has quit IRC
[08:40:39] *** jochen__ has quit IRC
[08:40:53] *** jochen__ has joined #chromium-os
[08:40:53] *** ChanServ sets mode: +v jochen__
[08:49:43] *** stalled has quit IRC
[09:03:10] <crosbot> tree became 'Tree is open (VMTest timeout on DNS)'
[09:06:37] *** stalled has joined #chromium-os
[09:06:42] *** jsalz has quit IRC
[09:08:59] <petermayo> zork_: The report phase is telling you VMTests failed  -- on x86 generic notice the chrome sig 11 in VMTests,
[09:09:26] <petermayo> cbuildbot is the wrapper in which both VMTests and Report run.
[09:12:03] <zork_> Ok, so it's just the DNS error we've been seeing, then?
[09:20:31] *** jochen___ has joined #chromium-os
[09:20:32] *** ChanServ sets mode: +v jochen___
[09:21:52] *** jochen__ has quit IRC
[09:21:54] *** jochen___ is now known as jochen__
[09:29:06] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from aa48237c7b263ae44f94e3631ac9c9c18b0467df: Chris Masone <cmasone at chromium dot org>)'
[09:30:57] <petermayo> 1 was a chrome crash, the other was a timeout waiting for a DNS revert, let's see about this one.
[09:33:18] <petermayo> this one, rsync died in a weird way.
[09:33:18] <zork_> It seems to be another timeout
[09:33:53] <crosbot> tree became 'Tree is closed (zork invenstigating)'
[09:34:02] <petermayo> 10s for a banner exchange, seems short.
[09:34:29] <zork_> I'm not entirely sure what that is
[09:35:45] <petermayo> Suggest: grab the test results, attach them to a bug, copy in the VMTest output that looks different.
[09:36:00] <zork_> I'll do that.
[09:36:51] <petermayo> I don't think that other changes are going to confound the exploration of these things.  But I am veery weak in the ways of debugging VMTests.
[09:37:26] <zork_> A timeout on 127.0.0.1 seems like it could be caused by a DNS error.
[09:37:34] <zork_> So, I'm happy with the bug idea.
[09:43:54] <crosbot> tree became 'Tree is open (see bug http://crosbug.com/20514)'
[09:56:12] *** nexusz99_ has quit IRC
[10:00:25] *** SoulRaven has joined #chromium-os
[10:02:57] *** yogeshrt has joined #chromium-os
[10:03:20] <yogeshrt> how to disable audio runtime on chromium os ?
[10:05:02] <yogeshrt> I would like to know what is userspace manager for Audio management in Chromium OS , I would like to disable/stop
[10:15:52] *** patcito has quit IRC
[10:25:02] *** nexusz99 has joined #chromium-os
[10:27:26] *** Inumedia has quit IRC
[10:33:12] *** petermayo has quit IRC
[10:48:23] *** yusukes has joined #chromium-os
[10:49:07] *** chocobo__ has quit IRC
[10:49:15] *** chocobo__ has joined #chromium-os
[10:49:15] *** ChanServ sets mode: +v chocobo__
[10:49:33] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb_he canary" from None: )'
[10:56:09] *** Styx has quit IRC
[10:57:56] *** nexusz99 has quit IRC
[11:00:00] *** yogeshrt has quit IRC
[11:07:19] *** hashimoto has quit IRC
[11:09:33] *** zork_ has quit IRC
[11:10:11] *** hashimoto has joined #chromium-os
[11:41:06] *** Inumedia has joined #chromium-os
[12:35:03] *** unreal has joined #chromium-os
[12:50:56] *** yoshiki_ has quit IRC
[12:51:21] *** yoshiki_ has joined #chromium-os
[13:05:25] *** SoulRaven has quit IRC
[13:12:14] *** Styx has joined #chromium-os
[13:44:04] *** Malmis has joined #chromium-os
[13:44:11] *** stalled has quit IRC
[13:45:12] *** Malmis_ has quit IRC
[13:52:12] *** stalled has joined #chromium-os
[14:11:49] *** Styx has quit IRC
[14:14:55] *** Styx has joined #chromium-os
[14:52:36] <ellyjones> zgb canary has a chrome crash
[14:53:29] <crosbot> tree became 'Tree is open (ellyjones: x86-zgb-he-canary chrome crashed)'
[15:53:42] *** rharrison_chrome has joined #chromium-os
[15:53:43] *** ChanServ sets mode: +v rharrison_chrome
[16:01:03] *** Hexxeh has quit IRC
[16:01:38] *** Hexxeh has joined #chromium-os
[16:05:46] *** Solet has quit IRC
[16:16:42] *** petermayo has joined #chromium-os
[16:16:43] *** ChanServ sets mode: +v petermayo
[16:18:02] *** Solet has joined #chromium-os
[16:19:39] *** yoshiki_ has quit IRC
[16:20:05] *** yoshiki_ has joined #chromium-os
[16:22:26] *** Solet has quit IRC
[16:22:39] *** Solet has joined #chromium-os
[16:22:51] <crosbot> tree became '?'
[16:23:06] <crosbot> tree became 'Tree is open (ellyjones: x86-zgb-he-canary chrome crashed)'
[16:31:45] <ellyjones> I love the smell of TEST=None in the morning
[16:38:18] <redpig> hehe
[16:46:15] <petermayo> ellyjones: I love having to upload a new patch to change the TESTED= status of the previous patch.   :-)
[16:46:40] <ellyjones> petermayo: it's best when that invalidates all your +2s
[16:47:24] <ellyjones> due to a deficiency in the way my trybot wrappers work, they push the trybot logs with different uuids each run even if the CL is unchanged, so the trybot logs are almost gauranteed to be stale :P maybe I should have them do a fixed thing instead
[17:07:35] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 PFQ" from efba40e7ad292aa9adcc741dfaf828ca1a1e8488: Taylor Hutt <thutt at chromium dot org>)'
[17:12:44] *** grundler has joined #chromium-os
[17:12:44] *** ChanServ sets mode: +v grundler
[17:14:10] <ellyjones> hmm
[17:14:32] <ellyjones> adhd missing board-tegra2.h
[17:14:39] <ellyjones> blame thutt
[17:14:49] <ellyjones> crosbot: sheriffs?
[17:14:50] <crosbot> ellyjones: sheriffs: chocobo, jhorwich, zork
[17:16:15] * ellyjones mails thutt, waits 15m
[17:16:49] <grundler> thutt is looking at it now
[17:17:03] <ellyjones> great
[17:17:31] <grundler> looks like it's missing board-x86-generic.h for x86-generic build too.
[17:17:38] <cmasone> grundler: can he update teh tree status to indicate that, or respond on the buildbot thread, or post in IRC?
[17:17:49] <ellyjones> or do all three of those things :)
[17:17:53] <grundler> yeah...I'll remind :)
[17:19:13] <grundler> taylor says he'll have a fix ready in a minute or two.
[17:22:01] <ellyjones> perhaps I should gently remind people at the next team meeting that people in remote offices can't tell what's going on in meatspace
[17:23:01] <grundler> yeah, I'll point that out again to taylor...IRC is just part of working in a distributed team.
[17:25:07] <ellyjones> wait, is it just me or are there no EST sheriffs?
[17:25:18] <cmasone> ellyjones: zork is TOK
[17:25:22] <ellyjones> ah
[17:25:23] <cmasone> ellyjones: not sure about jhorwich
[17:25:29] <ellyjones> jhorwich is listed as 1950
[17:25:41] <cmasone> ellyjones: in that case, I have bad news...
[17:25:57] <ellyjones> :<
[17:26:54] *** thutt has joined #chromium-os
[17:26:54] *** ChanServ sets mode: +v thutt
[17:27:22] <thutt> I broke the build.  I have checked in a fix, but how do I make the broken target restart?
[17:27:40] <ellyjones> it restarts on its own, it is a pfq
[17:28:07] <cmasone> ellyjones: at least, it's supposed to; it was missing the signal yesterday, but that is a bug that is filed somewhere
[17:28:19] <ellyjones> hrm
[17:28:26] <cmasone> I don't see the fix on the waterfall
[17:28:30] <ellyjones> well, if the commit shows up and it doesn't start, we can kick it
[17:29:03] <thutt> You don't see *my* fix, or the fix for the waterfall issue?
[17:29:06] <ellyjones> thutt: are we talking about http://gerrit.chromium.org/gerrit/#change,7855 ?
[17:29:12] <ellyjones> I don't see your fix specifically
[17:29:21] <cmasone> there it is now
[17:29:25] <ellyjones> there we go
[17:29:29] <thutt> Yes, 7855
[17:29:32] <cmasone> it took a bit to get picked up I suppose
[17:29:34] *** semenzato_ has joined #chromium-os
[17:29:36] <ellyjones> tegra's going to fail again though
[17:29:49] <ellyjones> see http://build.chromium.org/p/chromiumos/builders/tegra2%20PFQ/builds/894/steps/BuildTarget/logs/stdio
[17:30:23] *** Sarten-X has quit IRC
[17:30:56] *** thutt has quit IRC
[17:31:12] *** thutt has joined #chromium-os
[17:31:12] *** ChanServ sets mode: +v thutt
[17:31:18] <thutt> On it.
[17:35:01] <ellyjones> thutt: do you have an approximate ETA? if it's not gonna be fixed within the next 10m or so we should just revert and reopen
[17:37:47] *** rbyers_ has joined #chromium-os
[17:37:47] *** ChanServ sets mode: +v rbyers_
[17:40:28] *** Sarten-X has joined #chromium-os
[17:40:28] <thutt> I've checked in a fix for tegra2.
[17:40:56] <ellyjones> now we wait
[17:41:12] <thutt> I'm trying to look at tegra2_chrome, x86-gcc-generic-46 and amd64-generic-full, but they don't seem to have anything in their columns.
[17:41:16] <thutt> How do I look at them.
[17:41:24] <ellyjones> they are probably fine to ignore
[17:41:32] <ellyjones> you can click the bot names
[17:42:14] <ellyjones> x86-generic-gcc_46 is the gcc 4.6 test bot and it never passes
[17:42:23] *** wfrichar has joined #chromium-os
[17:42:26] *** ChanServ sets mode: +v wfrichar
[17:42:40] <ellyjones> amd64-generic-full is targeting amd64 and never passes either
[17:42:44] *** rbyers_ has quit IRC
[17:42:48] <ellyjones> tegra2_chrome I have no idea, probably another chrome pfq
[17:43:41] *** rbyers_ has joined #chromium-os
[17:44:29] <ellyjones> pity there isn't a doc listing which bots are what and whether they matter
[17:46:06] *** patcito has joined #chromium-os
[17:48:09] <ellyjones> ibus-1.3.99.20110817-r1: /usr/lib64/portage/bin/misc-functions.sh: line 729: 24882 Segmentation fault      rm -Rf "${root}"/${no_inst} &>/dev/null
[17:48:13] <ellyjones> (my local build)
[17:52:00] <cmasone> The link at the top of the waterfall (called "important") brings you to a view of only the bots that matter
[17:52:10] <ellyjones> rockin'
[17:52:35] <ellyjones> the tegra2 pfq doesn't seem to have kicked off
[17:53:10] <cmasone> ellyjones: the PFQs all kick off together
[17:53:21] <cmasone> ellyjones: so, the x86 onre needs to go first
[17:53:24] <ellyjones> ah
[17:53:24] <cmasone> and I don't know why it's not
[17:53:36] <ellyjones> it says 'building' but has no build
[17:53:47] * ellyjones kicks it
[17:54:01] <cmasone> ellyjones: that's the bug david mentioned yesterday
[17:54:09] <ellyjones> oh cool
[17:54:12] <ellyjones> well, I kicked it
[17:56:27] *** rharrison_chrome has quit IRC
[17:57:19] <crosbot> tree became 'Tree is open (ellyjones: thutt's fixes for t2 pfq and x86-gen pfq are in, pfqs going green)'
[17:57:23] *** rbyers_ has quit IRC
[17:57:30] <ellyjones> and now, lunch
[18:04:27] *** saintlou has joined #chromium-os
[18:04:27] *** ChanServ sets mode: +v saintlou
[18:06:22] *** behdad has joined #chromium-os
[18:09:55] *** dennisjeffrey has joined #chromium-os
[18:09:55] *** ChanServ sets mode: +v dennisjeffrey
[18:10:49] <gauravsh> crosbot: sheriffs?
[18:10:49] <crosbot> gauravsh: sheriffs: chocobo, jhorwich, zork
[18:11:33] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex-binary" from f1c3e3dd6bedf51b70170d2cf3fce9fc9c6bf426: agl at chromium dot org <agl at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, backer at chromium dot org <backer at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[18:11:35] <gauravsh> pushing my opencryptoki cros_workonify ebuild change. do not expect fallout, but you never know.
[18:11:39] <gauravsh> ahhhh damn it
[18:11:54] <cmasone> gauravsh: your timing, it is amazing
[18:12:25] <gauravsh> 09/16 08:56:37 ERROR|logging_ma:0560| [2261:2261:0916/085637:138281257:ERROR:proxy_launcher.cc(108)] WaitForInitialLoads failed.
[18:12:29] <gauravsh> 09/16 08:56:37 ERROR|logging_ma:0560| [2261:2261:0916/085637:138281373:ERROR:proxy_launcher.cc(557)] Failed to ConnectToRunningBrowser
[18:12:46] <gauravsh> is that one of the flaky failures related to the race condition?
[18:13:08] <gauravsh> suite_Smoke/desktopui_UrlFetch.not-live failed
[18:13:25] <gauravsh> cmasone: did you change to clean up VMTest output not go in? or is it only live on some builders?
[18:13:38] <cmasone> gauravsh it failed at login, though.  Or at least I think so.
[18:13:42] <cmasone> gauravsh: I was looking at that
[18:13:50] <cmasone> you'll see the BuildException is gone
[18:14:10] <cmasone> but now there's some OTHER stupid stack trace that tells no one anything :-P
[18:14:27] <gauravsh> crosbug.com/20323
[18:14:36] <gauravsh> looks similar to the failure we are seeing
[18:14:54] <cmasone> gauravsh: mebbe
[18:15:40] <gauravsh> it's the exact same error but in a different test - "Unhandled JSONInterfaceError: Automation call {'username': 'performancetestaccount at gmail dot com', 'password': 'perfsmurf', 'command': 'Login'} received empty response.  Perhaps the browser crashed.
[18:15:41] <cmasone> gauravsh: this NEW trace comes from ctest, which is not used on the external bots...I think
[18:16:16] *** semenzato_ has quit IRC
[18:16:24] *** rbyers_ has joined #chromium-os
[18:19:24] <gauravsh> did zgb-binary just fail too with the exact same test failure?
[18:19:36] <gauravsh> or was it already failing before alex-binary died?
[18:20:23] <gauravsh> never mind, it failed on the previous run. but same error - that is suspicious.
[18:26:35] <gauravsh> logs should be displayed in the reverse order.
[18:27:45] *** patcito has quit IRC
[18:28:43] *** patcito has joined #chromium-os
[18:31:34] *** rharrison_chrome has joined #chromium-os
[18:31:35] *** ChanServ sets mode: +v rharrison_chrome
[18:33:16] *** zmedico has quit IRC
[18:38:05] <ellyjones> is anyone looking at the x86-alex-bin failure?
[18:38:30] <cmasone> ellyjones: I think garuav was, above
[18:38:43] <ellyjones> gauravsh: reopening or what?
[18:41:04] *** jhorwich has joined #chromium-os
[18:41:04] *** ChanServ sets mode: +v jhorwich
[18:41:27] <ellyjones> alright, I'm reopening it
[18:41:55] *** dennisjeffrey has quit IRC
[18:42:32] <crosbot> tree became 'Tree is open (ellyjones: x86-alex-bin is crosbug.com/20323)'
[18:42:57] <jhorwich> jhorwich starting sheriff duty
[18:43:04] <ellyjones> hi
[18:43:12] <jhorwich> goot morning/afternoon
[18:43:20] <jhorwich> oops - s/goot/good
[18:47:25] <cmasone> jhorwich: you went to Dartmouth?  I'm an '02!
[18:47:31] <cmasone> jhorwich: (stalked you on who/)
[18:49:01] <jhorwich> cmasone: indeed. get off my lawn! I'm a '92
[18:49:19] * jhorwich feels old
[18:50:04] <ellyjones> yikes :P
[18:50:23] * kliegs feels young for once
[18:50:33] <cmasone> jhorwich: you should.  I was just back there for a wedding last weekend, and the were having DOC trips for the '15s.  Those kids were born in what...1993?
[18:50:47] <cmasone> jhorwich: that made me feel terrrrrrrribble
[18:50:51] * ellyjones CMU '10 :P
[18:51:08] <cmasone> jhorwich: so, you can feel terrrrrrrribble + 10
[18:51:20] <jhorwich> what a way to start my morning ;)
[18:51:50] <cmasone> ellyjones: CMU 'gofuckyourself
[18:52:19] <kliegs> Go Tartans!
[18:52:28] <ellyjones> cmasone: you're just jealous
[18:52:39] <ellyjones> :P
[18:52:52] <cmasone> kliegs: I would make fun of you, if my own school's mascot was not simply the color green.
[18:53:10] <ellyjones> a tartan is a pretty terrible mascot
[18:53:19] <kliegs> cmasone: It's proof we're more complex.  Our colors are a pattern - Tartan!
[18:53:19] <ellyjones> I like the CS department's dragon
[18:53:32] <cmasone> ellyjones: is it in a dungeon?
[18:53:39] <ellyjones> it used to be
[18:53:49] <ellyjones> but we moved out of that building and gave it to the math and physics departments
[18:53:49] <cmasone> ellyjones: or your mother's basement?
[18:54:01] <ellyjones> haha :P
[18:54:16] <kliegs> ellyjones: Our mascot is the Scotty Dog!  Kids these days, they don't learn anything in school
[18:54:25] <ellyjones> I did a lot of my work during undergrad in a windowless room in the bowels of this building: http://upload.wikimedia.org/wikipedia/commons/c/c0/Wean_hall.jpg
[18:54:42] <ellyjones> that's as close at it gets to dungeon
[18:54:53] <ellyjones> (the sorry world is not for the architecture, either)
[18:54:58] <cmasone> http://www.ists.dartmouth.edu/images/Sudikoff_Lab.JPG
[18:55:14] <cmasone> the windows are tall and skinny because it used to be a sanitarium
[18:55:25] <ellyjones> that seems fitting
[18:55:27] <cmasone> also, Mr. Sudikoff went to federal prison for embezzling
[18:55:38] <ellyjones> that also seems fitting
[18:56:45] <jhorwich> cmasone: back in my day, we were in the "shower towers" like Gerry
[18:57:06] <quiche> "shower tower"?
[18:57:39] *** rbyers_ has quit IRC
[18:57:40] <jhorwich> their exterior architecture consisted of large color rectangles
[18:57:49] <jhorwich> made them look like a shower, i guess
[18:57:54] <quiche> ellyjones: i got you beat -- CMU '11 :P
[18:58:13] <kliegs> CMU '97.  You young kids don't know what Skibo really was
[18:58:38] <ellyjones> quiche: excellent!
[18:58:41] <quiche> i take it you're not talking about skibo gym.
[18:58:42] <ojn> Bah. None of you kids have had to drag your sorry asses to class in -40, I bet.
[18:58:51] <cmasone> ojn: nah, only -20
[18:58:54] <ellyjones> quiche: when's your birthday? :P
[18:58:58] <quiche> ojn: just once, in ann arbor.
[18:59:04] <jhorwich> uphill, both ways
[18:59:15] <kliegs> ojn: You didn't see the winter in Pittsburgh in '95.  It was pretty close to -40.  And the heat was out in the lecture halls
[18:59:25] <ellyjones> kliegs: gross
[18:59:27] <jhorwich> picture of the former "shower towers" ... http://i391.photobucket.com/albums/oo351/ccg654/showertwr06041320102.jpg
[18:59:41] <ellyjones> jhorwich: what the heck is that
[18:59:43] <cmasone> jhorwich: those were the worst.  The new math building is much less heinous
[18:59:50] <cmasone> ellyjones: our old math buildings
[18:59:54] <ellyjones> nasty
[19:00:10] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 full" from 0bd4feab7779e0fc8281322ffc8bde030d4aaae7: Taylor Hutt <thutt at chromium dot org>)'
[19:00:15] <quiche> kliegs: i think we had the same in ann arbor in '95.
[19:00:22] <ojn> kliegs, only if you consider -22 the same as -40, right? http://www.pittsburghcitypaper.ws/gyrobase/Content?oid=oid%3A27570
[19:00:32] <kliegs> which scale are we using?
[19:00:33] <quiche> one of the few times umich actually cancelled classes.
[19:00:40] <kliegs> and it sure felt like -40.  was there wind or something?
[19:00:41] <ellyjones> CMU has obvious architectural phases - a lot of dignified old buildings from around 1900, then a brutalist CS building from the 1960s (?), and now a tr\'es moderne CS building in the style of the Stata Center at MIT
[19:00:45] <ellyjones> er, hm, tree
[19:01:09] <ellyjones> tegra2 full does not have thutt's tegra2 fix from like 298364832 hours ago
[19:01:14] <quiche> jhorwich: the name is quite apt.
[19:01:34] <cmasone> quiche: what, "shower towers"?  Yeah
[19:01:41] <quiche> cmasone: yeah
[19:02:02] <jhorwich> looking at that tree thing
[19:02:11] <ellyjones> just kick the bot :P
[19:02:28] <cmasone> quiche: yeah, what gets me is that a committee of people had to look at a bunch of options and collectively say "yes, that.  Let's do that one there, with the shower tiling"
[19:02:50] <jhorwich> missing board-tegra2.h. lovely
[19:03:04] <quiche> cmasone: well, a committee decision would certainly explain it. :)
[19:03:38] <ellyjones> cmasone: all truly ugly things come from committees
[19:03:44] *** lipsinV2 has quit IRC
[19:03:49] <wfrichar> Pfft. Clemson, class of 85. http://goo.gl/rpWra
[19:03:54] <quiche> could be worse... IIRC, after they built the new CS building at UCSD, they realized they didn't plan enough cooling for the machine rooms.
[19:04:08] <ellyjones> wfrichar: what even is that
[19:04:15] <wfrichar> EE lab.
[19:04:22] <ellyjones> quiche: haha, the machine rooms in the old CS building would routinely flood
[19:04:24] <quiche> is  that from frankenstien or something?
[19:04:30] <wfrichar> yes. ;-)
[19:04:34] <ellyjones> this was found out the hard way
[19:04:35] <wfrichar> But it looked about like that.
[19:04:55] <cmasone> http://dartmouth.org/classes/66/webcam/ live stream of college :-)
[19:04:58] <quiche> ellyjones: actually i think they've had some flooding in the new one too, no?
[19:04:59] <ellyjones> because one day people were like "is that... a river...?" and they pulled up the raised floor and the cables were submarine cables
[19:05:12] <ellyjones> quiche: that one was an AC unit sucking, though
[19:05:15] <gauravsh> i may be in the wrong channel, and this may be off-topic but...
[19:05:16] <wfrichar> We had 220v 3-phase connectors that were 1/2" metal-plated holes in the wall, unguarded, about shoulder height.
[19:05:18] <crosbot> tree became 'Tree is open (thutt: board-tegra2.h has been committed.)'
[19:05:19] <ellyjones> not an actual design failure
[19:05:21] <gauravsh> what's the deal with the tree?
[19:05:23] <quiche> ellyjones: ah, ok.
[19:05:29] <wfrichar> Do NOT place your hand across the terminals.
[19:05:31] <ellyjones> gauravsh: it's green
[19:05:50] <ellyjones> wfrichar: haha, oh wow
[19:05:51] <gauravsh> the alex-binary failure, was that because of thutt's change?
[19:06:01] <ellyjones> alex-binary was crosbug.com/20323, I believe
[19:06:08] <wfrichar> cmasone: I didn't know they had webcams in 1966.
[19:06:22] <ellyjones> wfrichar: I have taken 110V through bits of my hands more often than I'd like
[19:07:09] <cmasone> wfrichar: they didn't, until we developed time travel in 1985
[19:07:10] <wfrichar> same here. Scary, but not as painful as an electric fence.
[19:07:12] <gauravsh> crosbot: sheriffs?
[19:07:12] <crosbot> gauravsh: sheriffs: chocobo, jhorwich, zork
[19:07:15] <gauravsh> landing ebuild change
[19:07:18] <vpalatin> ellyjones: sometimes it's useful to live in an inefficient/energy-wasting country with 110V instead of 240V
[19:07:21] <wfrichar> cmasone: most excellent!
[19:08:15] <ellyjones> vpalatin: yes :P
[19:08:18] <jhorwich> gauravsh: anything we need to do specifically, or just keep our eye open?
[19:08:20] <wfrichar> vpalatin: I much prefer our dainty plugs, too.
[19:10:10] <gauravsh> jhorwich: just keep eyes open. i will be monitoring the tree too. it's similar to elly's change from yesterday but for a different package. But it shouldn't fail in the same way if at all. i think.
[19:10:37] <jhorwich> gauravsh: thank you for the heads-up
[19:10:44] <ellyjones> if anyone didn't see it, the postmortem for yesterday: https://docs.google.com/a/google.com/document/d/1GWqfzHDbGhUDVRZ0AC5q0cIW5VLZQ2M4Pq_JyQFyJ6o/edit?hl=en_US
[19:11:21] <ellyjones> props go to davidjames for helping me fix the fallout
[19:11:50] <gauravsh> ellyjones: that doesn't open without a @google.com account
[19:12:19] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 generic full" from 0bd4feab7779e0fc8281322ffc8bde030d4aaae7: Taylor Hutt <thutt at chromium dot org>)'
[19:12:32] <jhorwich> ellyjones: should we add some more to the timeline (the part where we got tripped up later by the ebuild not getting promoted to stable until davidjames re-triggered PFQs)?
[19:13:11] <davidjames> jhorwich: Plan is for changes to only be pushed by commit queue, which also uprevs ebuilds at same time
[19:13:19] <vpalatin> wfrichar: sure, nowadays they are kind of unique in the world. I cannot remember having seen other plug type where you can see the metal pins while it's still powered. This adds a chromy touch that remembers the fire engines.
[19:13:31] <davidjames> jhorwich: So that means, when it's pushed, it's stable, there's no difference
[19:13:36] <ellyjones> gauravsh: yeah, I know
[19:13:38] <davidjames> Thus much easier to debug
[19:13:49] <jhorwich> tree issue caused by missing board-x86-generic.h
[19:14:19] <ellyjones> we have discovered a new and fun way to inflict misery on the sheriffs
[19:14:25] <davidjames> thutt, jhorwich: Same failure again, suggest reopening and also checking whether any other bots are running with versions before thutt fixed the issue
[19:14:29] <wfrichar> :-)
[19:14:36] <ellyjones> craft a change that passes the pfq but not the full bots
[19:14:39] <jhorwich> davidjames: agreed
[19:15:13] <crosbot> tree became 'Tree is open (thutt: board-tegra2.h has been committed.)'
[19:15:59] <crosbot> tree became 'Tree is open (board-x86-generic.h had already been committed)'
[19:21:53] *** Sarten-X has quit IRC
[19:22:05] *** cros_ has joined #chromium-os
[19:22:31] *** raymes has joined #chromium-os
[19:22:31] *** ChanServ sets mode: +v raymes
[19:22:57] *** Guest64910 has quit IRC
[19:22:57] *** Guest64910 has joined #chromium-os
[19:22:57] *** adams.freenode.net sets mode: +v Guest64910
[19:23:02] *** Guest64910 is now known as ttuttle|work
[19:23:12] <cros_> Has anyone built the chromium browser from the source included when you pull down the chromium OS source? There is no makefile in the browser source directory so it fails when I try to emerge.
[19:24:06] <kliegs> cros_: To build with that source you need to set CHROME_ORIGIN=GERRIT_SOURCE
[19:24:14] <kliegs> or probably call cros_workon start chromeos-chrome
[19:24:29] <kliegs> and then the ebuild will build it
[19:24:36] <kliegs> although that workflow is still in beta I believe.
[19:26:52] <cros_> kliegs: Will it use that exact source then if i say CHROME_ORIGIN=GERRIT_SOURCE? I really want to build the source bundled with chromium OS because otherwise I run into version mismatch issues between the browser and libcros, etc.
[19:27:37] <kliegs> cros_: I believe it would but not 100% positive.  Are you looking to make changes to the code? Or just want a clean tree to build from?
[19:28:51] <cros_> kliegs: For now, I just want to build with different options like proprietary_codecs, maybe source changes later. Anyways, looking at the chrome ebuild it looks like GERRIT_SOURCE may be what I want. I was trying using LOCAL_SOURCE.
[19:29:06] <kliegs> cros_: GERRIT_SOURCE is probably what you want once its finished
[19:29:18] <kliegs> cros_: For right now LOCAL_SOURCE should be fine. I build with it a lot and so do others here
[19:29:27] *** Sarten-X has joined #chromium-os
[19:29:29] <kliegs> cros_: You might want to do a cros_workon start libcros
[19:29:52] <cros_> kliegs: The problem with LOCAL_SOURCE is that the chromium browser source included with chromium OS won't build since it has no Makefile.
[19:29:55] <kliegs> cros_: And then if you keep your source directory at tip of tree and keep libcros synced they should always be there
[19:30:11] <kliegs> cros_: Do a separate checkout using a gclient file for local source
[19:30:34] <cros_> kliegs: That is what I have done before but I don't want to do that. I don't want to have to update my chromium OS source too.
[19:30:40] <gauravsh> crosbot: sheriffs?
[19:30:40] <crosbot> gauravsh: sheriffs: chocobo, jhorwich, zork
[19:30:48] <cros_> kliegs: Why include the chromium browser source if you can't build it?
[19:31:06] *** Solet has quit IRC
[19:31:11] <gauravsh> jhorwich: the x86 PFQ is about to blow up with - "Timed out waiting to revert DNS." on one of the test
[19:31:16] <kliegs> cros_: Its included for the GERRIT_SOURCE workflow that's still being developed.  I don't know the exact state of it but it wasn't working the last time I tried
[19:31:18] *** Solet has joined #chromium-os
[19:31:25] <kliegs> cros_: However it is being actively developed so might work better for you now
[19:31:36] <kliegs> try it by doing:  cros_workon start chromeos-chrome
[19:31:39] <cros_> kliegs: Alright, that is a bummer. Appreciate the help.
[19:31:41] <kliegs> Aad see if that works
[19:31:55] <kliegs> cros_: It was pretty close when I did it.  Mostly just some dependencies missing and such. So it might work
[19:31:57] <jhorwich> gauravsh: thank you for the warning.
[19:32:01] <gauravsh> jhorwich: looks unrelated to my change. i think we have seen a few of these flaky failures recently.
[19:32:15] <kliegs> cros_: I can probably give some assistance if you get errors but I won't be here that late today
[19:32:19] <cros_> kliegs: I still don't see how it will build since the ebuild calles emake and there is no makefile but I'll try.
[19:32:19] <jhorwich> ok
[19:32:41] <cmasone> cros_: in src_prepare, it runs a program that generates makefuiles
[19:32:46] <ellyjones> gauravsh: better land my changes fast then ;)
[19:33:25] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 257e7f750dd576dcc357a4ad1f983ef546b8f4f0: Gaurav Shah <gauravsh at chromium dot org>)'
[19:33:33] <gauravsh> ellyjones: the lesson i learnt from the post-mortem was - never let elly submit a change. :)
[19:33:41] *** stevenjb has quit IRC
[19:34:00] <kliegs> gauravsh: +1
[19:34:14] <cros_> cmasone: I don't see that. It just seems to do a gclient runhooks --force. There is no .gclient file there either by default.
[19:34:40] <cmasone> cros_: guess what gclient runhooks does?
[19:34:54] <davidjames> gauravsh: Bad news, looks like buildbots flaked out on tests while running your change. Not your fault, though
[19:35:04] <ellyjones> gauravsh: mwahahaha.
[19:35:06] <cros_> cmasone: Ok, gclient runhooks --help didn't seem to indicate that.
[19:35:24] <kliegs> cros_, cmasone: Using GERRIT_SOURCE it doesn't call runhooks
[19:35:40] <kliegs> cros_, cmasone: It calls chrome_set_ver.py instead I think
[19:35:46] <kliegs> (I could be wrong - this is my read of the ebuild)
[19:35:56] <kliegs> and I believe chrome_set_ver does what gclient runhooks did
[19:36:06] <cros_> kliegs, cmasone: You are right I missed that so maybe it works :). I best try it before I speak further ;-)
[19:36:26] <kliegs> cros_: I can confirm that it does create makefiles.  I've had it start building on me before but hit other errors
[19:36:42] * davidjames notes that desktopui_WindowManagerFocusNewWindows and desktopui_UrlFetch.not-live seem to tickle Chrome flakiness more than other tests
[19:36:43] <kliegs> but I know there's been active work fixing the errors, I'm just not sure what the exact status of it is
[19:37:03] <cros_> kliegs: No problem, thanks for the help, I'll let you know how it goes.
[19:37:13] <crosbot> tree became 'Tree is open ("cbuildbot_master" on "x86 generic PFQ" probable flake)'
[19:37:43] <kliegs> cros_: sounds good and good luck.
[19:38:46] * ellyjones lands more openssl ebuild changes
[19:39:12] * kliegs is glad he has no CL's left to commit
[19:39:30] <davidjames> ellyjones: You might want to wait until Zdenek lands http://gerrit.chromium.org/gerrit/#change,7805
[19:39:39] <davidjames> (which fixes CONFIG_PROTECT)
[19:39:42] <ellyjones> davidjames: they don't alter the config at all
[19:40:09] <davidjames> ok :)
[19:40:24] *** Styx has quit IRC
[19:40:50] <ellyjones> I tried a novel technique of structuring my changes so that they do not break the tree
[19:41:15] <gauravsh> hah
[19:48:25] *** zmedico has joined #chromium-os
[19:52:04] <ellyjones> hmm
[19:52:23] <ellyjones> gauravsh: <http://gerrit.chromium.org/gerrit/#change,7713>? It passes trybots :)
[19:53:54] <gauravsh> i was waiting for your explanation of why the unittests failed last time and how this fixes that problem (since I can't diff it with the previous CL).
[19:54:17] <ellyjones> oh
[19:54:34] <ellyjones> the unit tests failed because they presume that you operate on the new home directory when you migrate
[19:54:44] <ellyjones> but instead I operate on the mountpoint
[19:54:46] <ellyjones> so like
[19:55:03] <ellyjones> before: migrate from /home/chronos/user/Downloads to /home/.shadow/lol/vault/Downloads
[19:55:21] <ellyjones> now: migrate from /home/.shadow/lol/mount/user/Downloads to /home/.shadow/lol/vault/user/Downloads
[19:55:34] <ellyjones> but the unit test was checking /home/chronos/user/Downloads still
[19:56:18] <gauravsh> nod
[19:56:40] <gauravsh> ok, your cl has my blessings
[19:56:51] <ellyjones> cool
[19:56:54] <gauravsh> but don't quote me on that
[19:56:58] <ellyjones> too late
[19:57:01] <ellyjones> it's now in my worklog
[19:59:03] <ellyjones> run_as_root delenda est!
[20:05:20] <ellyjones> do we have a chromeos code search, like the chrome one?
[20:05:33] <gauravsh> http://codesearch.google.com/codesearch#/&exact_package=chromiumos
[20:06:05] <ellyjones> awesome
[20:06:15] <ellyjones> alright
[20:06:28] <ellyjones> now as a backup: gauravsh, are we using --run_as_root anywhere that I wouldn't know about?
[20:07:19] <gauravsh> other than chromeos-install, and via factory/recovery installer - not that I know of
[20:07:32] <ellyjones> good
[20:07:43] <ellyjones> can I quote you on _that_?
[20:07:47] <gauravsh> but then, you will never find out until you push your change :P
[20:08:15] <gauravsh> sure, quote me, the operative quote being "not that *I* know of"
[20:08:29] *** [S]olet has joined #chromium-os
[20:08:30] *** Solet has quit IRC
[20:19:10] *** rbyers_ has joined #chromium-os
[20:25:08] *** ttuttle|work is now known as ttuttle|work_
[20:26:05] *** ttuttle|work has joined #chromium-os
[20:26:16] *** ttuttle|work_ has quit IRC
[20:41:54] *** behdad has quit IRC
[20:48:40] *** frhodes has quit IRC
[20:51:32] <ellyjones> the pfq is stuck again
[20:51:34] <ellyjones> davidjames: ^
[20:51:47] * ellyjones kicks it
[20:51:52] <davidjames> ellyjones: thx
[20:52:08] *** rbyers_ has quit IRC
[20:52:09] <ellyjones> why does it get stuck?
[20:52:12] <davidjames> ellyjones: They seem to often miss commits that happen during their last build
[20:52:27] <davidjames> ellyjones: So if they haven't got any new commits since they restarted they don't start
[20:52:32] <ellyjones> ew
[20:52:42] <davidjames> ellyjones: Yeah, PFQ used to not have that problem because it didn't restart
[20:53:03] <davidjames> ellyjones: But now it reboots and I guess when it reboots it forgets all its planned work
[20:53:15] <ellyjones> :(
[20:53:22] <ellyjones> is this something I can help fix?
[20:53:43] <ellyjones> gauravsh: the Cursed CL is now in again
[20:53:48] <davidjames> ellyjones: I think it's a bug in buildbot itself
[20:54:05] <davidjames> ellyjones: We're using an old version of buildbot, version 0.8
[20:54:38] <ellyjones> oh, okay
[20:55:05] <davidjames> ellyjones: crbug.com/92109
[20:57:19] <ellyjones> sheriffs: heads up, I landed a CL that has been nothing but trouble so far: http://gerrit.chromium.org/gerrit/#change,7713
[20:57:26] <ellyjones> it has passed x86 and arm trybots and unit tests
[20:58:04] <gauravsh> so you think. i am pretty sure it's going to now fail in the Archive stage. :P
[20:58:25] <ellyjones> if that happens I'm going to reassign the entire bug to sosa :)
[20:58:26] <gauravsh> think that's the only stage where versions of that CL haven't failed before.
[20:59:31] <jhorwich> ellyjones: thank you for the notice
[20:59:49] <ellyjones> jhorwich: note the profusion of 'reviewed-on:' lines on the CL :P
[21:04:00] <ellyjones> davidjames: the lkgmsync blamelist step seems... false
[21:04:07] <ellyjones> it's referring to random old copies of the same CL
[21:06:49] <jhorwich> ellyjones: maybe it's parsing those reviewed-on lines? hmmm
[21:06:58] <ellyjones> maybe, but it ought not to
[21:07:04] <ellyjones> I add those manually and they mean nothing
[21:09:43] <jhorwich> ellyjones: I also find it strange that it lists 7713 for the PFQ builds that started before 7713 got cherrypicked
[21:10:31] <jhorwich> but I could very well just be misinterpreting the point of the output
[21:10:38] <cmasone> jhorwich: this is one of the flaws of the waterfall UI.  The time the patch is marked on the waterfall does not correllate well with when it actually got merged.
[21:11:09] <cmasone> jhorwich: the inline blamelist is, in general, faaaaaaar more accurate than the blamelist that buildbot generates, or what you can judge by looking at the waterfall
[21:11:12] <ellyjones> is there some clever authoritative way other than checking the manifest?
[21:11:34] <cmasone> ellyjones: that's all it's doing to generate the LKGM thing, is checking the manifest
[21:12:06] <ellyjones> oh fuck oops
[21:12:09] *** behdad has joined #chromium-os
[21:12:31] <ellyjones> the corresponding test change is LGTMed and stuff but I can't push it with the cursed CL itself
[21:12:35] <ellyjones> http://gerrit.chromium.org/gerrit/#change,7716
[21:12:54] <ellyjones> the pfq run is going to fail without that
[21:13:00] <ellyjones> davidjames: what do?
[21:13:25] <jhorwich> cmasone: guess I'm confused. I see (for instance) build 889 x86 generic PFQ. started 11:49. change 7713 shows being cherry picked at 11:51
[21:14:33] <jhorwich> cmasone: or maybe I'm not confused - maybe this is the flaw [7713 appearing in LKGMCandidateSync even though it's not actually part of the build]
[21:15:06] *** seanpaul has quit IRC
[21:15:19] <ellyjones> 7713 is not on the waterfall at all
[21:16:29] <ellyjones> well, this is going to suck a little bit then
[21:17:00] <cmasone> jhorwich: the time that matters is when LKGMCandidateSync started
[21:17:03] <jhorwich> ellyjones: sounds like we should pull in 7716 and redo the PFQs?
[21:17:10] <ellyjones> jhorwich: I can't pull in 7716
[21:17:18] <ellyjones> because the autotest-tests repo has to go through the commit queue first
[21:17:37] <ellyjones> there is some small set of magical users that can pull it in anyway but I am not one of them
[21:17:40] <ellyjones> kliegs: ?
[21:19:15] <ellyjones> cmasone: do you know who the magic users are?
[21:20:34] <kliegs> ellyjones: eh?
[21:20:43] <ellyjones> can you push 7716 please?
[21:20:51] <ellyjones> it is in commitqueue, but its partner change 7713 wasn't
[21:21:15] <ellyjones> the x86-gen-pfq failure is to be expected without 7716
[21:21:17] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from None: )'
[21:21:19] <kliegs> this is fixing tree redness?
[21:21:21] <ellyjones> yes
[21:21:33] <kliegs> Am I a chump?
[21:21:43] <kliegs> pushed
[21:21:46] <ellyjones> danke
[21:21:50] <kliegs> bitte
[21:22:03] <ellyjones> that should engreenen (?) the tree
[21:22:05] <kliegs> and sheriffs should be able to do that as well if I'm not around - cutting out early today
[21:22:08] <kliegs> oh - lunchtime for them probably
[21:22:13] <jhorwich> no I'm here
[21:22:31] <jhorwich> but was under impression I couldn't push in autotest-tests
[21:22:37] <ellyjones> yeah, I'm gonna go to the airport in 90min or so, which is why I'm anxious to get this in :)
[21:22:55] *** behdad has quit IRC
[21:23:43] <jhorwich> ellyjones: will 7716 fix the just-reported error in x86 generic PFQ log : /home/.shadow/472723a59c92b1d98f0330a4ec06f563e0c54f63/vault - Expected chronos:0710, saw root:0700
[21:23:53] <ellyjones> yes, jhorwich
[21:24:07] <ellyjones> http://gerrit.chromium.org/gerrit/#patch,sidebyside,7716,1,client/site_tests/security_ProfilePermissions/security_ProfilePermissions.py
[21:24:21] <jhorwich> ellyjones: thought so but wanted to double-check ... will re-open tree
[21:25:01] <ellyjones> cool
[21:25:12] <crosbot> tree became 'Tree is open (7716 should fix "x86 generic PFQ")'
[21:28:52] <jhorwich> folks, I'm taking a lunch break - coordinated with chocobo, he's here too
[21:32:40] <kliegs> cros_: see the email on chromium-os-dev titled "Simplified workflow for Chromium on ChromiumOS development"
[21:33:12] <ellyjones> you know, crosbot has been remarkably reliable lately
[21:35:41] <cros_> kliegs: Thanks, I'll take a look. Didn't have any luck building because it was failing on git checkout commands with access denied for some reason. Running the commands manually work.
[21:45:57] <cros_> kliegs: Isn't that mailing list no longer active? I searched but found something in the chromium developer guide but not the mailing lists.
[21:52:20] *** Malmis has quit IRC
[22:04:48] *** behdad has joined #chromium-os
[22:05:44] * jhorwich comes back with full belly
[22:06:40] <ellyjones> and the PFQ passed with the Cursed CL!
[22:09:09] <ellyjones> alright
[22:09:13] <ellyjones> leaving for the airport in 54 minutes
[22:09:16] <ellyjones> let's land more dangerous changes
[22:14:27] <ojn> better now than monday when I am sheriff!
[22:14:48] <jhorwich> monday is an excellent time to land dangerous stuff ;)
[22:15:00] <gauravsh> ojn: you underestimate the capability of the tree to remain broken through the weekend
[22:15:43] <ojn> gauravsh: Looks like my cosheriff is on vacation as well. Hopefully most people will be too busy with perf to break the tree.
[22:15:53] <ojn> Now, the tree has a tendency to break itself, so that might not be much of a help
[22:17:46] *** dhendrix has quit IRC
[22:26:32] *** dhendrix has joined #chromium-os
[22:26:33] *** ChanServ sets mode: +v dhendrix
[22:34:49] *** stevenjb has joined #chromium-os
[22:34:50] *** ChanServ sets mode: +v stevenjb
[22:37:31] *** sadrulhc has joined #chromium-os
[22:37:32] *** ChanServ sets mode: +v sadrulhc
[22:37:35] <ellyjones> hey so
[22:37:47] <ellyjones> installing when you have rootfs verification turned off has been broken for months
[22:37:53] <ellyjones> have any of you noticed, and if so, do you care?
[22:38:26] <gauravsh> broken in what way?
[22:38:41] <ellyjones> it just doesn't wor
[22:38:42] <ellyjones> work
[22:38:53] <ellyjones> we try to parse the kernel command line for dm="", it's missing, chromeos-setimage bails out
[22:38:57] <ellyjones> or does something random
[22:39:23] <redpig> seems like it broke when setimage got refactored for arm support
[22:54:44] *** rhashimoto has left #chromium-os
[22:55:20] <ellyjones> go go gadget pfq
[23:00:12] <davidjames> ellyjones: Thanks for acting as a cronjob
[23:00:32] <ellyjones> no problem
[23:00:49] <gauravsh> it's her penance for being a tree breaker :P
[23:01:37] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 PFQ" from e3693a996b06309699c5c2370ed03e65ebe320fa: Doug Anderson <dianders at chromium dot org>, Elly Jones <ellyjones at chromium dot org>, Sadrul Habib Chowdhury <sadrul at chromium dot org>, Stefan Reinauer <reinauer at chromium dot org>)'
[23:01:45] <jhorwich> ruh roh
[23:03:36] <ellyjones> ERROR: Cannot find commit id for /home/chrome-bot/trunk/src/third_party/chromiumos-overlay/sys-apps/coreboot-utils/coreboot-utils-9999.ebuild
[23:03:46] <jhorwich> yea, just saw that
[23:04:08] <ellyjones> most likely http://gerrit.chromium.org/gerrit/#change,7875
[23:04:21] <ellyjones> forgot to add a stable build with it
[23:04:31] <ellyjones> anyway, heading out for the weekend :) see you all on monday
[23:04:39] <jhorwich> ellyjones: have a good weekend
[23:04:41] <davidjames> nah, the issue is more likely that he forgot CROS_WORKON_PROJECT or somesuch :)
[23:04:58] *** cros_ has quit IRC
[23:05:13] <jhorwich> davidjames: should we revert this one, or do we have a quick fix?
[23:05:24] <davidjames> jhorwich: Would suggest revert
[23:05:48] <davidjames> jhorwich: Multiple problems with that commit, will take a few minutes to fix
[23:05:53] *** behdad has quit IRC
[23:06:46] <jhorwich> davidjames: ok
[23:08:34] <davidjames> jhorwich: reinauer is reverting that now, I'll help him get it back in after
[23:09:09] *** petermayo has quit IRC
[23:09:36] <jhorwich> davidjames: ok, I'll hold off and let reinauer revert
[23:10:23] <crosbot> tree became 'Tree is closed (reverting 7875)'
[23:11:25] <davidjames> jhorwich: Ok it's reverted
[23:12:29] *** Solet has joined #chromium-os
[23:12:36] *** [S]olet has quit IRC
[23:13:58] <jhorwich> davidjames: was the rebuild of PFQs triggered after this (and hence we should re-open)?
[23:17:08] <crosbot> tree became 'Tree is open (7882 reverted 7875, PFQ should cycle green)'
[23:22:05] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard canary" from None: )'
[23:22:59] *** [S]olet has joined #chromium-os
[23:23:16] *** Solet has quit IRC
[23:23:25] *** srao has joined #chromium-os
[23:23:25] *** ChanServ sets mode: +v srao
[23:23:51] *** cooled_ has joined #chromium-os
[23:26:24] <jhorwich> pushimage: /var/www/chromeos-official/dev-channel/tegra2-seaboard/0.16.1047.0
[23:26:27] <jhorwich> pushimage: already exists.  Unable to continue!
[23:26:32] *** cooled has quit IRC
[23:26:45] <jhorwich> I haven't seen that before
[23:27:11] *** rharrison_chrome has quit IRC
[23:27:46] <crosbot> tree became 'Tree is closed (jhorwich investigating tegra2_seaboard canary)'
[23:29:28] <gauravsh> jhorwich: this looks like something to escalate to a trooper - djmm, scottz, eblake
[23:29:52] <jhorwich> gauravsh: I'm reaching that conclusion myself
[23:30:24] *** jglasgow has left #chromium-os
[23:33:05] *** vmil86 has quit IRC
[23:36:27] *** Justasic has quit IRC
[23:37:11] *** Justasic has joined #chromium-os
[23:38:08] <crosbot> tree became 'Tree is open (djmm investigating tegra2_seaboard canary)'
[23:39:00] *** thutt has quit IRC
[23:42:43] <grundler> ojn, I've confirmed your guess by inspecting the console output:
[23:42:47] <grundler> [    0.878877] tegradc tegradc.0: fb can't be mapped
[23:42:55] <grundler> [    0.900413] Unable to handle kernel NULL pointer dereference at virtual address 00000000
[23:43:02] <grundler> [    0.900462] PC is at bitfill_aligned+0xb0/0x124
[23:43:26] <grundler> in other words, someone is trying to bitblip something onto the FB even it it's not present. Nice, robust code. :)
[23:43:55] <grundler> anyway, back to the other firmware now...
[23:44:26] <ojn> Ok. i'll do another pass looking at the backlog of skipped patches, then see what was merged into tot since I started.
[23:48:24] <nirnimesh_> Anyone know why chrome is stuck at 16.0.880.0?
[23:56:24] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex-binary" from 56f4048a7b55e60e037c06aae38c080b00c89014: _third_party_ at chromium dot org, bsalomon at google dot com <bsalomon at google dot com@2bbb7eff-a529-9590-31e7-b0007b416f81>, rsimha at chromium dot org <rsimha at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, sail at chromium dot org <sail at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'





top