[00:02:53] *** m1k3l has quit IRC
[00:04:51] *** achuith has quit IRC
[00:05:11] *** achuith has joined #chromium-os
[00:05:11] *** ChanServ sets mode: +v achuith
[00:07:12] *** m1k3l has joined #chromium-os
[00:12:27] *** unreal has joined #chromium-os
[00:18:27] *** petermayo has left #chromium-os
[00:18:59] *** davemoore has quit IRC
[00:19:23] *** davemoore has joined #chromium-os
[00:27:00] *** shawn186|i7 has joined #chromium-os
[00:27:04] <shawn186|i7> Howdy
[00:27:18] <shawn186|i7> will Chromium os work on a compaq presario?
[00:27:29] *** davemoore has quit IRC
[00:29:18] *** m1k3l has quit IRC
[00:30:27] *** lipsinV2 has joined #chromium-os
[00:30:31] *** lipsinV2 has joined #chromium-os
[00:30:41] *** davemoore has joined #chromium-os
[00:32:14] <ellyjones> maybe
[00:35:35] <shawn186|i7> ah
[00:35:42] <shawn186|i7> *maybe*
[00:36:20] <ellyjones> yes :P
[00:39:33] *** GodoPPL has quit IRC
[00:46:12] *** bleung has joined #chromium-os
[00:46:12] *** ChanServ sets mode: +v bleung
[00:48:19] *** kliegs has quit IRC
[00:49:51] <cmasone> davemoore: x86 pineview full is going to fail with 19005
[00:50:08] <cmasone> please append the logs there once it does and I'll look at it in an hour
[00:54:33] *** gspencer has joined #chromium-os
[00:54:34] *** ChanServ sets mode: +v gspencer
[00:58:45] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from 5d68e0f78a184b2890202880ea4289cda41b1c01: Mike Frysinger <vapier at chromium dot org>, Paul Stewart <pstew at chromium dot org>)'
[01:02:06] *** mkunze has quit IRC
[01:08:08] <thutt> Most recent failure is timed out waiting to revert DNS.  19005?
[01:15:11] <crosbot> tree became 'Tree is open (Probable 19005; failure waiting for revert of DNS; watching to see if turns green)'
[01:40:08] *** wfrichar has quit IRC
[01:52:35] *** seventh has joined #chromium-os
[01:55:47] *** rosso_ has quit IRC
[02:04:33] <ellyjones> hm
[02:04:43] <ellyjones> I'm pleasantly surprised that I haven't broken the tree yet today
[02:05:15] <gauravsh> the flakes are masking your tree breaking check-ins
[02:05:59] <ellyjones> :D
[02:09:57] <kees> ellyjones: I reviewed your code! :)
[02:10:00] <kees> (lgtm)
[02:10:16] <ellyjones> which?
[02:10:23] <ellyjones> oh, awesome
[02:10:29] <ellyjones> it is too bad two +1s do not make a +2 :P
[02:10:43] <ellyjones> hey, gauravsh!
[02:10:55] <ellyjones> want to review http://gerrit.chromium.org/gerrit/8144?
[02:10:59] <gauravsh> but 2 +1s and a -1 do make a -1
[02:11:41] * ellyjones makes a pass through her bugs to close some
[02:11:53] <gauravsh> ellyjones: i totally would review it but I am just not familiar enough with minijail to +2 it.
[02:12:20] *** magn3ts has quit IRC
[02:14:15] <ellyjones> it's a small and ugly refactor :P
[02:14:36] <gauravsh> famous last words. I am not falling for that again. :)
[02:14:46] <ellyjones> darn it!
[02:14:53] <ellyjones> it worked on the split-home change the first... six? times :P
[02:15:08] <gauravsh> i'd hate to trample on redpig's authoritah
[02:15:17] <gauravsh> haha, as they say, the 7th time is a charm
[02:15:17] <ellyjones> you respect it?
[02:16:50] <gauravsh> sure if and where he is around.
[02:17:56] <ellyjones> he's totally afk right now
[02:27:47] *** rbyers_ has joined #chromium-os
[02:28:29] *** magn3ts has joined #chromium-os
[02:31:21] <bleung> Anyone notice there's some autotest login problems?
[02:31:39] <quiche> anyone else having trouble with xlsclients-1.1.2? it seems to depend on libxcb-1.7. but i don't see a libxcb-1.7 anywhere.
[02:31:48] <marcheu> quiche: \o/
[02:32:00] <marcheu> okay I guess I should fix that
[02:32:09] <quiche> yes, please. :)
[02:32:22] <marcheu> I'll add you as reviewer for punishment
[02:32:26] <marcheu> :D
[02:32:28] <quiche> doh!
[02:33:11] <marcheu> quiche: funnily enought, I still get xlsclients 1.0.2 here
[02:33:16] <marcheu> enough
[02:33:45] <quiche> hmm. is there something i might have misconfigured?
[02:34:49] <marcheu> I'm not sure
[02:34:54] <marcheu> I want to update libxcb anyway
[02:34:57] <quiche> ok
[02:35:20] <marcheu> couldn't do it with the previous commit, because it needs to be in chromiumos-overlay
[02:35:50] <quiche> i thought http://gerrit.chromium.org/gerrit/#change,3775 was in chromiumos-overlay, no?
[02:36:07] <marcheu> it never went in
[02:36:32] <marcheu> all the other mess I just did will help me upgrade libxcb actually
[02:36:37] <marcheu> and also the Xserver
[02:36:39] <quiche> ok
[02:41:07] <shawn186|i7> yayy
[02:41:11] <shawn186|i7> got all my pcs back here :D
[02:41:17] <shawn186|i7> gonna try somethin
[02:41:25] * shawn186|i7 evil grin
[02:41:31] <shawn186|i7> oops
[02:41:37] <shawn186|i7> wrong channel
[02:56:48] <ellyjones> jglasgow just committed and ran :P if it breaks he said just revert everything
[02:57:21] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from 8f0c4dc3e8e3887e7a5820c63955c8c53b7a22d0: Andrew de los Reyes <adlr at chromium dot org>)'
[03:00:08] <thutt> eating dinner. back in a bit
[03:05:07] *** davemoore has quit IRC
[03:06:49] *** davemoore has joined #chromium-os
[03:09:30] <davemoore> I just got back, and my mac crashed for the third time today (reboot is *way* slower than cros) what's the status on the closure?
[03:15:32] <kees> ellyjones: yeah, wandering through all the open CLs, that one had enough context that I felt like I could review it. Unfortunately it seems like I need to read the style guide a little more closely (re: gauravsh's comment).
[03:15:38] *** benchan has quit IRC
[03:18:31] <ellyjones> I don't think the style guide mentions any of gaurav's comments
[03:19:03] <kees> ah, well. okay :)
[03:25:39] *** Sergiu has quit IRC
[03:31:01] <crosbot> tree became 'Tree is open (davemoore -> vmtest failures are flaky, may be crosbug.com/19005)'
[03:31:47] *** saintlou has quit IRC
[03:34:33] <ellyjones> ugh
[03:34:34] *** Keybuk has quit IRC
[03:34:47] <ellyjones> my productivity has gone to basically zero
[03:34:51] * ellyjones ->
[03:35:46] <thutt> Why has your productivity plummeted, elly?
[03:45:22] *** aaronp has quit IRC
[03:46:48] <ellyjones> because I have been at work for 12 hours so far
[03:47:07] <ellyjones> my fiancee is taking classes at MIT, which results in my staying at work until 2230 to go home with her
[03:53:47] <thutt> That's a bummer.
[03:55:25] *** thutt has quit IRC
[03:56:11] <ellyjones> it's not that bad, really, but it's screwing up my work/life balance
[03:56:26] <ellyjones> because, well, work is across the street from MIT, so I just keep staying at work...
[04:13:57] *** dennisjeffrey has quit IRC
[04:33:48] *** yusukes has joined #chromium-os
[04:34:01] *** chrelad_ has joined #chromium-os
[04:56:29] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from fcc8d6e8161f5221211a282026f90fe4d9aff662: Andrew de los Reyes <adlr at chromium dot org>, Darren Krahn <dkrahn at chromium dot org>, Elly Jones <ellyjones at chromium dot org>, Simon Que <sque at chromium dot org>)'
[05:04:41] <crosbot> tree became 'Tree is closed (mkrebs looking)'
[05:11:21] <crosbot> tree became 'Tree is open (mkrebs reverted enabling unittest)'
[05:14:10] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb_he canary" from None: )'
[05:20:04] <crosbot> tree became 'Tree is open (crosbug.com/20323 and mkrebs revert in)'
[05:25:36] *** rbyers_ has quit IRC
[05:28:14] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex canary" from None: )'
[05:31:35] <crosbot> tree became 'Tree is open (20323, 18622 revert in)'
[05:34:51] *** yusukes has quit IRC
[05:43:25] *** sosa has quit IRC
[06:13:00] *** chrelad_ has quit IRC
[06:15:11] *** davemoore has quit IRC
[06:16:52] *** TW1920 has quit IRC
[06:16:54] *** TW1920_ has joined #chromium-os
[06:24:35] *** hashimoto has joined #chromium-os
[06:37:30] *** sadrul has quit IRC
[06:41:45] *** HW__ has joined #chromium-os
[06:43:38] *** HW_ has quit IRC
[06:45:21] *** nona has left #chromium-os
[06:47:35] *** sadrul has joined #chromium-os
[06:47:35] *** ChanServ sets mode: +v sadrul
[07:18:06] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex-binary" from 0654881988d41beef0a4ea26192b82ee2e05e493: Chris Sosa <sosa at chromium dot org>)'
[07:38:42] <crosbot> tree became 'Tree is open (20323 again)'
[08:06:36] *** ferringb has quit IRC
[08:06:49] *** ferringb has joined #chromium-os
[08:06:49] *** ChanServ sets mode: +v ferringb
[08:17:18] *** ferringb has quit IRC
[08:19:15] *** ferringb has joined #chromium-os
[08:19:15] *** ChanServ sets mode: +v ferringb
[08:26:53] *** ferringb has quit IRC
[08:28:35] *** ferringb has joined #chromium-os
[08:28:35] *** ChanServ sets mode: +v ferringb
[08:37:17] *** ferringb has quit IRC
[08:39:00] *** ferringb has joined #chromium-os
[08:39:00] *** ChanServ sets mode: +v ferringb
[08:46:37] *** patcito has quit IRC
[08:47:34] *** ferringb_ has joined #chromium-os
[08:47:35] *** ChanServ sets mode: +v ferringb_
[08:47:35] *** ferringb has quit IRC
[08:52:10] *** ferringb_ is now known as ferringb
[08:57:08] *** ferringb has quit IRC
[08:58:46] *** ferringb has joined #chromium-os
[08:58:46] *** ChanServ sets mode: +v ferringb
[09:17:20] *** ferringb has quit IRC
[09:18:29] *** ferringb has joined #chromium-os
[09:18:29] *** ChanServ sets mode: +v ferringb
[09:22:52] *** ferringb has quit IRC
[09:41:06] *** shawn186|i7 has quit IRC
[10:00:08] *** akrpic77 has joined #chromium-os
[10:00:15] <akrpic77> hello.
[10:02:42] *** hashimoto has quit IRC
[10:03:13] *** rosso_ has joined #chromium-os
[10:06:39] <akrpic77> NPAPI extensions are disabled. as in, can't be installed by users. but if I put my extension into /opt/google/chrome/plugins it will still be picked up? according to comments in http://codereview.chromium.org/3159040/patch/12001/5003 yes
[10:06:58] <akrpic77> can someone confirm this?
[10:08:18] *** hashimoto has joined #chromium-os
[10:24:32] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "arm generic PFQ" from 74802fe9263f88995da4964b8e927b82281a0a37: Jun Mukai <mukai at chromium dot org>)'
[10:53:04] <nkostylev> akrpic77: I'm not sure but even if they might get installed, NPAPI plugins just won't work http://code.google.com/p/chromium-os/issues/detail?id=6084
[10:56:47] *** nexusz99 has quit IRC
[10:56:51] <crosbot> tree became 'Tree is opened ("arm generic PFQ" -> error when cloning repo, mukai has clobbered buuildbot)'
[10:59:02] *** nexusz99 has joined #chromium-os
[11:05:32] *** Malmis has quit IRC
[11:06:13] *** Malmis has joined #chromium-os
[11:15:17] *** seventh has quit IRC
[11:30:02] *** marcheu has quit IRC
[11:48:02] *** akrpic77 has quit IRC
[11:55:06] *** akrpic77 has joined #chromium-os
[12:02:46] <shankar_> am building chromium under cros_sdk and build fails.
[12:04:04] <shankar_> used : USE="-build_tests" FEATURES="-usersandbox" CHROME_ORIGIN=LOCAL_SOURCE emerge-tegra2_seaboard chromeos-chrome
[12:04:21] <shankar_> do we need to define any other gyp variables ?
[12:04:31] <shankar_> any help please ?
[12:16:22] *** Inumedia has quit IRC
[12:29:51] *** mkunze has joined #chromium-os
[12:34:54] *** vmil86 has joined #chromium-os
[12:41:37] *** Inumedia has joined #chromium-os
[12:51:05] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from 74802fe9263f88995da4964b8e927b82281a0a37: Jun Mukai <mukai at chromium dot org>)'
[13:08:31] <crosbot> tree became 'Tree is opened ("x86 pineview full" > 19005, "x86 generic full" > ?, watching waterfall)'
[13:18:59] *** akrpic77 has quit IRC
[14:38:10] *** jochen__ has quit IRC
[14:38:16] *** jochen__ has joined #chromium-os
[14:38:16] *** ChanServ sets mode: +v jochen__
[15:12:33] *** rbyers_ has joined #chromium-os
[15:14:29] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_asymptote canary" from None: )'
[15:14:43] *** sadrul has quit IRC
[15:34:35] *** rbyers_ has quit IRC
[15:37:02] *** rbyers_ has joined #chromium-os
[15:37:03] *** ChanServ sets mode: +v rbyers_
[15:50:07] *** rharrison_chrome has joined #chromium-os
[15:50:08] *** ChanServ sets mode: +v rharrison_chrome
[15:53:46] <vapier> Pending 35, Ready 0, Running 7, Retrying 0, Total 443 [Time 28m48.9s Load 33.55 24.84 23.0]
[15:53:49] <vapier> is that load normal ?
[16:00:02] *** rbyers_ has quit IRC
[16:04:03] *** rbyers_ has joined #chromium-os
[16:04:26] *** aaronp has joined #chromium-os
[16:09:00] *** kliegs has joined #chromium-os
[16:09:01] *** ChanServ sets mode: +v kliegs
[16:09:16] <kliegs> crosbot: sheriffs?
[16:09:16] <crosbot> kliegs: sheriffs: dennisjeffrey, tbroch, rjkroege
[16:09:40] *** wfrichar has joined #chromium-os
[16:09:41] *** ChanServ sets mode: +v wfrichar
[16:11:37] *** Flameeyes has quit IRC
[16:11:52] <kliegs> dennisjeffrey__, rjkroege: ping
[16:15:58] *** Flameeyes has joined #chromium-os
[16:16:16] *** Flameeyes has joined #chromium-os
[16:17:25] <rjkroege> am here
[16:17:31] <rjkroege> kliegs: ^^
[16:17:50] <kliegs> rjkroege: i'm looking into asymptote canary failure
[16:20:49] <rjkroege> kliegs: excellent
[16:22:29] <vapier> the load is crazy high
[16:22:41] <kliegs> vapier: the load on what?
[16:22:56] <vapier> Pending 35, Ready 0, Running 7, Retrying 0, Total 443 [Time 28m48.9s Load 33.55 24.84 23.0]
[16:23:05] <kliegs> which bot?
[16:23:27] <vapier> that was x86 generic full, but it seems it came down
[16:23:33] <vapier> so maybe hitting a load of 40 is normal ?
[16:23:49] <vapier> Pending 212, Ready 48, Running 16, Retrying 0, Total 443 [Time 15m36.4s Load 50.33 38.89 23.53]
[16:23:50] <kliegs> vapier: believe so
[16:24:27] <kliegs> vapier: portage will launch N packages in parallel where possible.  but each package can spawn N processes with make -j N
[16:24:32] <kliegs> where N is numcores I believe.
[16:25:02] *** rbyers_ has quit IRC
[16:26:57] <kliegs> vapier: i think the asymptote link errors aren't errors - its the ebuild trying to link locally on host and having the target library directory in the linker path. but skipping those libraries and picking up the host ones anyways
[16:27:12] <kliegs> ibus-mozc-chewing-1.1.773.102-r3: ./unix/ibus/key_translator.h:38:33: error: session/commands.pb.h: No such file or directory
[16:27:21] <kliegs> I think that's the error.  but I don't know much about ibus or what it does
[16:27:30] <kliegs> and I can't reproduce locally
[16:27:37] <rjkroege> ibus is part of the ime sub-system
[16:27:57] *** rbyers_ has joined #chromium-os
[16:27:57] *** ChanServ sets mode: +v rbyers_
[16:28:17] <rjkroege> penghuang: bryueng: ^^ ibus issues -- any ideas?
[16:30:45] <kliegs> nkostylev: ping
[16:30:57] <penghuang> rjkroege, I think key_translator.h is not belong to ibus. I don't know where dose it come from. :(
[16:32:05] <rjkroege> am investigating the vm test failure in x86_generic_full
[16:32:56] *** dennisjeffrey has joined #chromium-os
[16:32:56] *** ChanServ sets mode: +v dennisjeffrey
[16:33:31] <kliegs> penghuang: any idea where it comes from?
[16:34:36] <penghuang> kliegs, I am asking the tokyo team. he said the issue should be fixed already.
[16:34:49] <penghuang> kliegs, probably you need repo sync again
[16:35:11] <kliegs> penghuang: its the full builder I"m trying to debug - my local build works with the same version number as the builder that fails
[16:35:17] <kliegs> penghuang: can you see if they can join in irc?
[16:35:38] <kliegs> penghuang: what's confusing is two different arm builders using the same version with builds at the same time had different results
[16:35:45] <penghuang> http://git.chromium.org/gitweb/?p=chromiumos/overlays/chromiumos-overlay.git;a=tree;f=app-i18n/ibus-mozc-chewing;hb=HEAD
[16:35:51] <kliegs> one succeeded and one failed and their configurations are almost identical
[16:35:53] <penghuang> here is the fix
[16:36:05] <penghuang> probably sdk version is different?
[16:36:36] <rjkroege> 457 was the last succeeding build
[16:37:04] <kliegs> rjkroege: 457 of which builder?
[16:37:35] <rjkroege> sorry. am looking at x86_generic_full as a starting point
[16:38:13] *** petermayo has joined #chromium-os
[16:38:13] *** ChanServ sets mode: +v petermayo
[16:38:45] <nkostylev> kliegs: yes
[16:39:05] <kliegs> penghuang: ok. that looks like it fixes this error.. and explains randomness on it
[16:39:37] <kliegs> rjkroege: recommend you clear the asymptote builder failure witha reference to crosbug.com/20668 as being already resoolved
[16:39:40] <kliegs> penghuang: thanks
[16:40:01] <kliegs> nkostylev: you reopened the tree earlier.  can you talk to rjkroege so he doesn't reexamine the same things you did?
[16:40:36] <crosbot> tree became 'Tree is open ("tegra2_asymptote canary" ->crosbug.com/20668, fix is submitted)'
[16:41:18] <nkostylev> rjkroege: 2 reasons were documented in tree status, x86 generic full - I was not sure at that point
[16:41:28] <petermayo> nkostylev:  opened the tree for you ... knew the issue since I was CC'd on it and recognized it.
[16:41:43] <ellyjones> hrm
[16:41:51] <petermayo> hrm?
[16:41:52] <ellyjones> are we ever likely to care about non-glibc environments?
[16:41:56] <rjkroege> so what was the issue?
[16:42:06] <nkostylev> "x86 generic full" > ?
[16:42:13] <rjkroege> yes.
[16:42:13] <nkostylev> let me dig logs again
[16:42:53] <nkostylev> oh, it's a known issue too
[16:42:54] <nkostylev> desktopui_UrlFetch.not-live
[16:43:00] <nkostylev> http://crosbug.com/20429
[16:44:11] <kliegs> petermayo: is x86 generic full failure known?
[16:44:18] <kliegs> ok
[16:44:33] <kliegs> petermayo: can you forward things that closed the tree to the list? so incoming sheriffs have more info?
[16:47:41] <petermayo> kliegs: haven't looked at the x86 generic,   I looked at the closer on asymptote.
[16:48:38] <petermayo> kliegs: which list?  chromium-os-dev?
[16:50:08] *** Sergiu has joined #chromium-os
[16:50:08] *** Sergiu has joined #chromium-os
[16:53:50] *** rbyers_ has quit IRC
[16:59:22] *** rbyers_ has joined #chromium-os
[17:01:43] <dennisjeffrey__> in MTV, reporting for first-time sheriff duty
[17:03:13] *** rbyers_ has quit IRC
[17:13:34] <kliegs> dennisjeffrey__: good luck!  if you need help feel free to ask
[17:14:09] <kliegs> dennisjeffrey__, rjkroege: I strongly recommend you two start a shared doc to track issues.  its great for keeping a history of what happened and really helps the next set of sheriffs figure out what's going on when they start
[17:14:31] <dennisjeffrey__> kliegs: thank you!
[17:20:56] *** tbroch has joined #chromium-os
[17:21:22] *** ChanServ sets mode: +v tbroch
[17:33:00] *** rbyers_ has joined #chromium-os
[17:33:10] *** rosso_ has quit IRC
[17:35:28] *** yoshiki_ has quit IRC
[17:37:43] *** aaronp_ has joined #chromium-os
[17:39:43] *** varunjain has quit IRC
[17:39:49] *** yoshiki has joined #chromium-os
[17:39:50] *** varunjain has joined #chromium-os
[17:39:50] *** ChanServ sets mode: +v varunjain
[17:41:36] <dennisjeffrey__> kliegs: great idea about the shared doc; we've started one up
[17:42:00] <kliegs> dennisjeffrey__: excellent.  it helped me a lot when I was sheriffing
[17:42:22] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard-tangent-binary" from ef0f110bd7f0fae89dfbd2e7b11f2d5238767921: Mike Frysinger <vapier at chromium dot org>)'
[17:43:40] <grundler> crosbot: sheriffs?
[17:43:41] <crosbot> grundler: sheriffs: dennisjeffrey, tbroch, rjkroege
[17:43:57] <vapier> hmm, nacl should have worked
[17:44:11] <tbroch> good morning ... looking now
[17:44:20] <vapier> err wait, there's already a fix for that
[17:44:48] <vapier> http://gerrit.chromium.org/gerrit/8343
[17:44:54] <vapier> that should fix that error
[17:45:20] <vapier> the build started before that was merged, so that makes sense
[17:45:30] *** aaronp has quit IRC
[17:45:33] <kliegs> dennisjeffrey__, rjkroege: i'll have a fix in shortly.  however it will take a day to propagate unless we force a build rev of chrome
[17:45:53] *** aaronp has joined #chromium-os
[17:46:34] <rjkroege> kliegs: so, should we?
[17:47:06] *** lipsinV2 has quit IRC
[17:47:42] <kliegs> rjkroege: probably. the asymptote builders will start failing shortly
[17:47:51] <grundler> kliegs: unless the process is smoother, that will take nearly all day too...almost not worth it.
[17:48:09] <kliegs> grundler: last time it took ~4 hours?
[17:48:15] <kliegs> so should be clear a bit after noon maybe
[17:48:25] <kliegs> grundler: we do need to clean up this process
[17:48:27] <grundler> 4h to get a new chrome and another 3-4h to "clear the pipeline"
[17:48:31] <kliegs> grundler: true
[17:48:39] <kliegs> grundler: i just don't like leaving builders red all day
[17:48:57] <dennisjeffrey__> grundler: would the tree be closed for that long?
[17:49:16] <kliegs> dennisjeffrey__: tree would be opened, just a few builders would remain red all day
[17:49:20] <kliegs> there's no point closing the tree for this
[17:49:22] <grundler> kliegs: yeah, I guess the answer is to repeat the fire drill so it's obvious we need a faster process
[17:49:53] <grundler> dennisjeffrey__: no, that's the downside of being sheriff when chrome is b0rked...get to nurse the tree along all day. :(
[17:50:08] <kliegs> grundler: I don't think its a huge burden on people at least.
[17:50:26] <grundler> kliegs: true, it wasn't the only problem we had last week.
[17:50:46] <tbroch> so sounds like we're opening the tree?
[17:51:18] <grundler> dennisjeffrey__, rjkroege : https://sites.google.com/a/google.com/chromeos/resources/tree-closure-postmortems for previous summaries of sheriffing (last week)
[17:51:19] <tbroch> what is actionable w/ respect to uprev of chrome ... anything
[17:51:32] <kliegs> its fine to reopen the tree for this
[17:51:49] <kliegs> mark it as "kliegs working on patch, new chrome rev needed to resolve"
[17:52:11] <kliegs> eblake is the person to release a new buildspec for chrome. so one of the sheriffs should contact him
[17:52:34] <dennisjeffrey__> i can update the status
[17:52:40] <rjkroege> thanks
[17:53:01] <grundler> kliegs: looking at notes from last week, we didn't get a new buildspec until almost 1pm...so we have a 4h headstart today.
[17:53:08] <crosbot> tree became 'Tree is open (kliegs working on patch, new chrome rev needed to resolve)'
[17:53:27] <tbroch> thanks for info kliegs/grundler
[17:53:27] <dennisjeffrey__> tbroch: beat me to it :-)
[17:53:34] <kliegs> grundler: yah.  if its not a huge amount of work we should start the process.
[17:53:46] <grundler> https://sites.google.com/a/google.com/chromeos/resources/tree-closure-postmortems/20110922
[17:53:48] * grundler nods
[17:53:49] <tbroch> gonna read post-mortem gits
[17:53:57] <tbroch> gits-bits
[17:54:32] *** aaronp_ has quit IRC
[17:54:50] <grundler> kliegs: and it was 3h from buildspec to new chrome bits available for chromeOS builds.
[17:54:57] <grundler> (not 4h)
[17:55:52] * grundler reminds dennisjeffrey__ , tbroch, rjkroege that a "journal" of the day is helpful for reconstructing previous, repeated issues.
[17:56:19] <rjkroege> is started
[17:56:32] <rjkroege> fwiw
[17:57:12] *** Sergiu_ has joined #chromium-os
[17:57:30] *** Sergiu has quit IRC
[18:01:01] <kliegs> tbroch, rjkroege, dennisjeffrey__: crosbug.com/20761 while not the cause of the bug,  should help fix it
[18:02:05] <kliegs> and is someone reaching out to eblake?
[18:03:25] <dennisjeffrey__> kliegs: I'll reach out to eblake
[18:04:30] <kliegs> dennisjeffrey__: thanks.  working to get the updated patch file generated.  should be <30 minutes
[18:04:42] *** saintlou has joined #chromium-os
[18:04:42] *** ChanServ sets mode: +v saintlou
[18:06:20] *** patcito has joined #chromium-os
[18:07:55] <dennisjeffrey__> kliegs: eblake offline and not at his desk right now.  will keep an eye out for when he gets in
[18:08:06] <kliegs> dennisjeffrey__: great, thanks
[18:11:18] *** rbyers_ has quit IRC
[18:16:45] <ellyjones> manifest change landed
[18:24:20] *** rginda_home has quit IRC
[18:25:39] *** dkrahn has joined #chromium-os
[18:25:42] *** benchan has joined #chromium-os
[18:26:11] *** ChanServ sets mode: +v dkrahn
[18:26:11] *** ChanServ sets mode: +v benchan
[18:34:25] *** saintlou has quit IRC
[18:36:12] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" from ef0f110bd7f0fae89dfbd2e7b11f2d5238767921: Mike Frysinger <vapier at chromium dot org>)'
[18:37:15] <ellyjones> that is unlikely to be vapier's fault
[18:37:25] <ellyjones> power_manager unittest failed
[18:37:26] <dennisjeffrey__> investigating...
[18:38:41] *** unreal has quit IRC
[18:38:52] *** unreal has joined #chromium-os
[18:40:20] <kees> morning!
[18:41:39] *** saintlou has joined #chromium-os
[18:41:39] *** ChanServ sets mode: +v saintlou
[18:45:53] *** wfrichar has quit IRC
[18:46:45] <dennisjeffrey__> possibly related?  http://gerrit.chromium.org/gerrit/#change,7231
[18:49:48] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" --> sheriffs investigating>)'
[18:51:34] <dennisjeffrey__> Simon's recent CL is reinstating the xidle unit tests, and one of those tests failed.
[18:51:36] <dennisjeffrey__> http://gerrit.chromium.org/gerrit/#change,7231
[18:51:41] <dennisjeffrey__> he's not in yet, so will revert for now
[18:52:53] *** wfrichar has joined #chromium-os
[18:52:54] *** ChanServ sets mode: +v wfrichar
[18:54:30] <dennisjeffrey__> i have a CL to revert: http://gerrit.chromium.org/gerrit/#change,8353
[18:55:01] <dennisjeffrey__> when i try to publish and submit, it says "server error: submit no permitted"
[18:55:24] <dennisjeffrey__> any ideas what I might be doing wrong?
[18:57:33] <kliegs> dennisjeffrey__: it needs a +2 bit set
[18:57:34] *** saintlou has quit IRC
[18:57:51] *** saintlou has joined #chromium-os
[18:57:52] *** ChanServ sets mode: +v saintlou
[18:58:03] <dennisjeffrey__> kliegs: thanks, it worked
[18:58:20] *** marcheu has joined #chromium-os
[18:59:55] <dennisjeffrey__> ok to reopen tree?
[19:01:44] <dennisjeffrey__> (http://gerrit.chromium.org/gerrit/#change,8353 has been merged)
[19:03:31] <dennisjeffrey__> will reopen the tree...
[19:04:10] *** rbyers_ has joined #chromium-os
[19:04:10] <crosbot> tree became 'Tree is open (revert in for http://gerrit.chromium.org/gerrit/#change,7231)'
[19:04:11] *** ChanServ sets mode: +v rbyers_
[19:08:06] <grundler> anyone know what happened to r16 autotest results? I'm trying to look at crosbug.com/20676 and URLs in that bug are getting a 404 :(
[19:09:05] * grundler goes trolling on the other side of the building...
[19:11:25] *** saintlou has quit IRC
[19:12:22] <dennisjeffrey__> grundler: it's likely due to the recent move of the autotest lab to MTV; the machines are being set up now
[19:12:34] <grundler> dennisjeffrey: hrm. Ok.
[19:13:12] <grundler> "As far as the dashboard, results emails, and LabTest are concerned the transition should be invisible."
[19:13:12] *** davidjames has joined #chromium-os
[19:13:13] *** ChanServ sets mode: +v davidjames
[19:13:44] *** saintlou has joined #chromium-os
[19:13:44] *** ChanServ sets mode: +v saintlou
[19:15:45] <dennisjeffrey__> grundler: Yes, I remember reading that.  If you go to the autotest results dashboard now, you'll see the data presented in the old format, and the data itself is also old
[19:16:21] <grundler> dennisjeffrey__: ok...I've just sent and email to dale curtis and cc'd autotest team
[19:16:58] <grundler> I'll try to dig up the logs I'm looking for directly on the cautotest server
[19:18:11] <dennisjeffrey__> question: on the chromiumOS sheriff details page, it says that to revert a broken commit, we should "push without an LGTM"
[19:18:21] *** Sergiu_ has quit IRC
[19:18:36] <dennisjeffrey__> but I found in practice that I needed to approve it (+2) to commit it.  Should I update the documentation on the sheriff page?
[19:18:55] *** Sergiu has joined #chromium-os
[19:18:55] *** Sergiu has joined #chromium-os
[19:19:23] <grundler> dennisjeffrey__: sounds like a good idea. It's possible to original text meant to "don't wait for someone else to +2 - DIY"
[19:19:55] <dennisjeffrey__> grundler: aha, that makes sense.  I'll clarify the documentation a bit on the webpage.  Thanks
[19:20:19] <ellyjones> LGTM basically by definition comes from someone else :) a change you put up implicitly LGTY
[19:21:47] <dennisjeffrey__> ellyjones: haha, it's so confusing when two people are talking about "me" and "you"; easy to get them mixed up!  But that does make sense ;-)
[19:31:36] *** rbyers_ has quit IRC
[19:32:30] <kliegs> Us and Them
[19:32:38] <kliegs> And after all, we're all just ordinary men
[19:32:45] *** saintlou has quit IRC
[19:33:53] *** rbyers_ has joined #chromium-os
[19:33:53] *** ChanServ sets mode: +v rbyers_
[19:34:50] *** dalecurtis has joined #chromium-os
[19:36:05] <dennisjeffrey__> kliegs: I contacted eblake, who recommended we contact chrome-pmo@ regarding the chrome buildspec (but he said he could do it if they don't)
[19:37:06] *** saintlou has joined #chromium-os
[19:37:06] *** ChanServ sets mode: +v saintlou
[19:37:25] <kliegs> dennisjeffrey__: ok. sounds good. finishing test of patch file now
[19:46:54] <seanpaul> marcheu: so emerge will automatically mask xlsclients-1.1.2 if the dependencies aren't satisfied?
[19:47:31] <marcheu> not mask, but it won't pick it up
[19:47:46] <kliegs> marcheu: we shouldn't have ebuilds marked as stable that aren't expected to compile
[19:48:11] <kliegs> the more of those we have, eventually portage will fail - it has a limit on how much backtracking it does so introducing too many of these cases and eventually one of them will break
[19:48:13] <marcheu> they are, if you have libxcb 1.7
[19:48:18] <marcheu> we have tons of ebuilds like that
[19:48:46] <kliegs> marcheu: we shouldn't though.  I've gone through in the past and marked some of them as unstable to prevent this problem
[19:49:22] *** rcui has joined #chromium-os
[19:49:22] *** ChanServ sets mode: +v rcui
[19:49:31] <marcheu> I'd rather move forward and add libxcb 1.7 now
[19:49:37] <seanpaul> marcheu: FYI, x11-apps/xwininfo-1.1.2 requires libxcb-1.6, which we also don't have
[19:49:48] <seanpaul> marcheu: seems like revving libxcb would fix my world
[19:50:06] <kliegs> marcheu: is libxcb safe to rev now and deploy?
[19:50:09] <marcheu> yeah, the issue is that I coudln't add it to the same overlay because we need to modify it
[19:50:22] <seanpaul> I'm still confused about why I don't fall back to xlsclients-1.0.2
[19:51:07] *** rbyers_ has quit IRC
[19:51:54] <kees> ellyjones: hey, so, was looking at your CL again (looks fine), and I noticed that libminijail.c mixes positive and negative return values (return -EINVAL vs return errno). is there a reason for this, or should I submit a CL that fixes all the "return errno" into "return -errno" ?
[19:52:01] <kliegs> vapier, davidjames: ^^?
[19:52:37] <vapier> what's the thread we're looking at ?  libxcb ?
[19:52:44] <kliegs> marcheu: I can confirm that emerge-tegra2_seaboard -pv xlsclients fails for me
[19:52:52] <kliegs> vapier: yes - one some systems its not falling back to 1.02 properly
[19:53:00] <kliegs> and instead failing due to missing dependencies
[19:53:06] <ellyjones> kees: ah... yeah, that may be a good thing to clean up
[19:53:09] <seanpaul> vapier: I'm getting dependency issues when building (no thread)
[19:53:32] <vapier> what board ?
[19:53:46] <kliegs> vapier: tegra2_seaboard
[19:55:26] *** Solet has quit IRC
[19:58:07] <vapier> it's emerging x11-libs/libxcb-1.5-r1 and x11-apps/xwininfo-1.0.5 for me on a clean system
[19:59:04] <seanpaul> vapier: hrmm. You're falling back to xlsclients-1.0.2?
[19:59:42] <vapier> i'm getting x11-apps/xlsclients-1.1.1
[19:59:59] <seanpaul> ah
[20:00:15] <seanpaul> ok, well I can try cleaning up my chroot, see if that helps
[20:01:56] <kees> ellyjones: are the Area- and Sev- things documented somewhere? I haven't managed to get through all the public docs yet.
[20:02:01] <vapier> actually i lied, i'm not getting 1.0.2 :)
[20:02:23] <ellyjones> kees: if they are, I don't know where it is
[20:03:23] <kees> ellyjones: oh, heh, quiche just showed that there are ... drop-downs. :)
[20:03:33] <ellyjones> yes :P
[20:03:34] <seanpaul> vapier: same result :) I'm deleting my chroot now, it's probably overdue anyways
[20:03:58] <vapier> seanpaul: the 1.1.2 automatically gets ignored due to: emerge: there are no ebuilds to satisfy ">=x11-libs/libxcb-1.7" for /build/tegra2_seaboard/.
[20:04:28] <seanpaul> vapier: yeah, mine gets that error, but instead of falling back it just failed
[20:04:45] <seanpaul> same thing happened to kliegs when he tried to emerge it
[20:05:26] <vapier> there might be a bug in downgrading an installed pkg vs installing a new one
[20:05:54] <quiche> FWIW, i was getting failure with buildpackages --nousepkg, but not when emerging xlsclients directly.
[20:06:18] <vapier> you could prob force the issue with `emerge-tegra2_seaboard -C xlsclients`
[20:07:11] <seanpaul> vapier: could be a bug, I can't be sure about which version was previously installed
[20:08:18] <vapier> portage doesn't seem to save a log when using a diff sysroot
[20:12:04] <redpig> kees: Area- is best guess
[20:12:27] <redpig> kees: Sev- is not used by many people anymore. it was an experiement but I think largely only Zel and Area-UI still  useit
[20:12:35] <redpig> kees: I'll often still set it reflexively
[20:12:39] *** benchan has quit IRC
[20:12:56] *** rbyers_ has joined #chromium-os
[20:13:29] <kliegs> vapier: I'm getting this error in a cleancheckout. it has no version of xlsclients installed
[20:13:46] <kliegs> maybe if you have the old version it ignores the fact you can't build the new
[20:14:16] <vapier> kliegs: i ran ./setup_board; ./build_packages
[20:14:21] <vapier> i don't have anything installed for this board :)
[20:14:24] <kliegs> with success?
[20:14:28] <kliegs> which board?
[20:14:35] <vapier> tegra2_seaboard
[20:14:43] <kliegs> internal or external repo?
[20:14:45] <vapier> ext
[20:14:52] <kliegs> could be a bug with internal repo
[20:15:05] <vapier> perhaps ... i dont have int configured anywhere to test out
[20:15:07] <kliegs> possibly related to the metadayout/layout.xml i updated
[20:15:37] <kliegs> i'm somewhat tempted to force a clobber build on an internal bot. but kinda afraid
[20:16:03] <vapier> the ext repo def has a bug that needs to be resolved ... either removing xlsclients-1.1.2 or adding libxcb-1.7
[20:16:10] <vapier> i think marcheu wants to add xcb 1.7
[20:16:28] <marcheu> yeah I'm working on that
[20:16:44] <marcheu> I want to do this no matter what, it's on the dependency path of a lot of things
[20:16:53] <vapier> you know best wrt X
[20:17:01] <seanpaul> vapier: new chroot, same old problem :)
[20:17:02] <marcheu> I wish I didn't
[20:17:05] <kliegs> lets do one of the other please.
[20:17:22] <kliegs> as this is blocking builds.  i'll trust your judgement on which but one needs to happen soon
[20:20:33] <kees> redpig: okay, noted. I figure if I can find a matching Area in the src/ tree names, I'll use it (i.e. platform)
[20:21:43] <marcheu> http://gerrit.chromium.org/gerrit/#change,3775 let me know how that flies, I'm going to lunch
[20:22:48] <ellyjones> redpig: wait, Sev is not used? I always set it :)
[20:26:00] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from 916e0fcd1a0b33ee4c8c99a127697ab2be930b9c: Mike Frysinger <vapier at chromium dot org>)'
[20:27:05] <vapier> not it :) FAIL VMTest
[20:27:11] *** Solet has joined #chromium-os
[20:27:48] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from 916e0fcd1a0b33ee4c8c99a127697ab2be930b9c: Mike Frysinger <vapier at chromium dot org> --> dennisjeffrey investigating)'
[20:28:41] <dennisjeffrey__> investigating...
[20:31:29] <dennisjeffrey__> seems to be this failure: suite_Smoke/security_ProfilePermissions.login
[20:31:36] *** rginda_home has joined #chromium-os
[20:34:15] <kliegs> Building tip of tree Chrome on arm I'm getting /usr/x86_64-pc-linux-gnu/armv7a-cros-linux-gnueabi/binutils-bin/2.21/ld.gold.real: error: cannot find -lGLESv2
[20:34:52] <kliegs> i saw a post on the mailing list about this and how they copied those libraries from host into the chroot. so confused
[20:40:01] <ellyjones> libxcb?
[20:40:07] <ellyjones> what're we doing about that, if anything?
[20:41:09] <ellyjones> crosbot: sheriffs?
[20:41:10] <crosbot> ellyjones: sheriffs: dennisjeffrey, tbroch, rjkroege
[20:41:56] <ellyjones> marcheu?
[20:46:06] <quiche> marcheu sent over http://gerrit.chromium.org/gerrit/#change,3775, and then went to lunch.
[20:46:36] <quiche> not sure who should review/commit it
[20:47:07] <dennisjeffrey__> specific error is this: Unhandled JSONInterfaceError: Automation call {'username': 'performancetestaccount at gmail dot com', 'password': 'perfsmurf', 'command': 'Login'} received empty response.  Perhaps the browser crashed.
[20:47:51] <dennisjeffrey__> nirnimesh believes this is another instance of crosbug.com/20323
[20:49:42] <dennisjeffrey__> so it is likely flakiness.  Should we reopen the tree?
[20:50:22] *** Keybuk has joined #chromium-os
[20:50:22] *** ChanServ sets mode: +v Keybuk
[20:50:37] <dennisjeffrey__> tbroch agrees; will re-open tree
[20:51:55] <crosbot> tree became 'Tree is open (likely flakiness with chrome not getting online: crosbug.com/20323)'
[21:00:47] *** sosa has joined #chromium-os
[21:01:39] *** ChanServ sets mode: +v sosa
[21:01:43] *** magn3ts has quit IRC
[21:15:03] <seanpaul> marcheu: your patch fixes the libxcb issue, but now i'm getting there are no ebuilds to satisfy ">=x11-libs/xcb-util-wm-0.3.8"
[21:15:14] <kliegs> lets just reverty it
[21:15:14] <seanpaul> required by xcb-util
[21:15:31] <kliegs> rjkroege, dennisjeffrey__ tbroch, marcheu: comments?
[21:15:53] <kliegs> We can revert, seanpaul or I or others can work with marceu to replicate and fix
[21:15:56] <kliegs> and then we can move forward
[21:16:27] <seanpaul> kliegs: +1
[21:17:03] <tbroch> just to clarify then ...revert of  http://gerrit.chromium.org/gerrit/#change,3775
[21:17:41] <dennisjeffrey__> but the status of that change says "Review in Progress"
[21:17:49] <tbroch> nevermind its not merged
[21:17:50] <quiche> i think you'll also need to revert whatever is causing the dependency on the newer xlsclients
[21:18:01] <seanpaul> tbroch: http://gerrit.chromium.org/gerrit/#change,7965
[21:18:25] <seanpaul> nvrmind, wrong one (that one was reverted)
[21:18:47] <seanpaul> tbroch: http://gerrit.chromium.org/gerrit/8284
[21:19:15] <tbroch> on it
[21:20:25] <kliegs> tbroch: right - the one seanpaul linked
[21:21:53] <tbroch> reverted 8284
[21:22:34] <seanpaul> tbroch: thanks!
[21:22:38] <kliegs> tbroch: thanks
[21:23:51] <kliegs> dennisjeffrey__, rjkroege, tbroch: http://gerrit.chromium.org/gerrit/8368 created to resolve tangent/asymptote redness
[21:24:18] <kliegs> will require a new chrome build spec to take effect.  sorry for the delays - ran into a separate bug in chromeos-chrome ebuild that I'll be fixing later
[21:30:52] <dennisjeffrey__> kliegs: thanks - once merged, I'll ask chrome-pmo@ to create a new chrome build spec
[21:30:57] <davidjames> vapier: chromiumos-sdk failure looks like you: http://build.chromium.org/p/chromiumos/builders/chromiumos%20sdk/builds/267/steps/BuildBoard/logs/stdio
[21:31:12] <vapier> let's look
[21:31:46] <vapier> davidjames: next run should fix it
[21:31:53] <vapier> that was run before x86_64-cros-linux-gnu was in the sdk config
[21:32:18] <kliegs> dennisjeffrey__: submitted
[21:32:20] <vapier> i'll post a change to the script to sanely handle non-installed gcc tuples
[21:34:07] <rjkroege> kliegs: the patch is bad
[21:34:20] <kliegs> rjkroege: it is?
[21:34:32] <kliegs> even after wjmaclean fixed it?
[21:34:44] <rjkroege> yup
[21:34:53] <kliegs> rjkroege: bad as in won't run? or just needs more stuff?
[21:35:20] <rjkroege> as in it will result in a dysfunctional build
[21:35:35] <rjkroege> let me shows you
[21:35:50] <rjkroege> or should I freshen?
[21:44:37] <dennisjeffrey__> rjkroege: you're referring to the patch that would require a new chrome build spec, right?
[21:50:28] *** srao has quit IRC
[21:50:35] <kliegs> dennisjeffrey__: he is
[21:51:34] <dennisjeffrey__> kliegs: ok - let me know when it's ok to ask chrome-pmo@ to generate a new chrome build spec; i have an e-mail ready to send to them
[21:51:44] <kliegs> dennisjeffrey__: sure thing. sorry about that
[21:52:38] <dennisjeffrey__> kliegs: np ;-)
[21:53:54] <kliegs> dennisjeffrey__: was a confusion - talked to rjkroege - its good to go ahead and push
[21:54:13] <kliegs> confusion was over reviewing a patch within gerrit which shows things as a patches - too much indirection
[21:55:26] <dennisjeffrey__> kliegs: ok, so I'll send the e-mail to chrome-pmo@
[21:55:43] *** dalecurtis has quit IRC
[21:56:39] <dennisjeffrey__> kliegs: done
[21:56:49] *** rcui has quit IRC
[21:56:54] <kliegs> dennisjeffrey__: thanks
[21:59:19] *** rbyers_ has quit IRC
[22:22:37] <ttuttle|work> the commit queue is unhappy with my commit; how do I find out why?
[22:23:36] <vapier> if it says "The Commit Queue failed to verify your change", then you have to look at the recent build log
[22:23:49] <ttuttle|work> where is that?
[22:24:02] <vapier> the recent build log ?  it's in the waterfall page
[22:24:31] <vapier> if you don't think it was your patch's fault, then remove the chrome-bot's "fail" from the reviewer list and let the bot try again
[22:25:22] <vapier> it's fairly hard atm to figure out what failure caused that message because the bot tests all the pending changes in one go ... if one of them fails, the bot kicks all the patches out
[22:25:34] <ttuttle|work> oh >.<
[22:25:38] <ttuttle|work> how do I even know which one my change was in?
[22:25:45] <vapier> you don't
[22:25:47] <vapier> :x
[22:25:51] <ttuttle|work> >.<
[22:26:01] <vapier> i think sosa is working on making this nicer
[22:27:42] <vapier> the x86-generic-commit-queue builder is the one to look at
[22:28:07] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy canary" from None: )'
[22:30:47] <dennisjeffrey__> seems to be same flakiness as the last time the tree was closed, just manifesting in a different test
[22:31:29] <dennisjeffrey__> Unhandled JSONInterfaceError: Automation call {'username': 'performancetestaccount at gmail dot com', 'password': 'perfsmurf', 'command': 'Login'} received empty response.  Perhaps the browser crashed.
[22:31:52] <dennisjeffrey__> in suite_Smoke/desktopui_WindowManagerFocusNewWindows
[22:32:13] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy canary" from None: --> sheriffs investigating )'
[22:33:01] <dennisjeffrey__> actually there's another test failure: suite_Smoke/login_CryptohomeMounted
[22:34:06] <dennisjeffrey__> this second failure seems to be an instance of 19005
[22:34:15] <dennisjeffrey__> crbug.com/19005
[22:34:44] <dennisjeffrey__> sorry, crosbug.com/19005
[22:35:21] <marcheu> kliegs: I don't get it, you can't just revert changes when both buildbots and other people can't repro
[22:35:32] <marcheu> like I don't go around reverting random stuff
[22:36:40] <tbroch> dennisjeffrey: I agree (19005)
[22:36:44] <marcheu> tbroch: ?
[22:36:54] <marcheu> tbroch: so what did that revert fix exactly
[22:37:11] <dennisjeffrey__> tbroch: ok, i will open the tree
[22:37:39] <cmasone> tbroch: dennisjeffrey__ Yeah, I'm still working on 19005 :-/  Using strace to try to suss it out
[22:37:46] <marcheu> tbroch: things work if you have a clean chroot, I tried about 4 times for both arm and x86, what am I supposed to do?
[22:37:58] <tbroch> marcheu: from seanpaul your patch fixes the libxcb issue, but now i'm getting there are no ebuilds to satisfy ">=x11-libs/xcb-util-wm-0.3.8"
[22:38:20] <marcheu> tbroch: but without my patch, things work. what are you fixing? can you point me to logs?
[22:38:37] <crosbot> tree became 'Tree is open ("Timed out waiting to revert DNS flakiness --> crosbug.com/19005)'
[22:39:08] <tbroch> marcheu: no ... kliegs, seanpaul, ellyjones?
[22:39:20] <marcheu> yeah you can't, cause that's not reproducible
[22:39:26] <ellyjones> let me sync and try again
[22:39:32] <ellyjones> my most recent is:
[22:39:33] <ellyjones> emerge: there are no ebuilds to satisfy ">=x11-libs/libxcb-1.7" for /build/tegra2_kaen/.
[22:39:51] <ellyjones> I just synced, let me try again
[22:39:52] <marcheu> ellyjones: make a clean chroot, setup_board, try again. I did that a bunch of times
[22:40:23] <ellyjones> make a clean chroot? I have a whole bunch of outstanding changes in this one D:
[22:40:25] <marcheu> it's really annoying that people can revert other's patches without any evidence whatsoever
[22:40:42] <ellyjones> is there a not-destroying-the-world way to fix this?
[22:41:00] <marcheu> I'll probably end up re-commiting the same patch
[22:41:38] <marcheu> in any case, my change passes all the testbots, works here, I'm not sure what you guys are doing but...
[22:41:49] <ellyjones> running "build_packages --oldchromebinary"
[22:41:53] <ellyjones> having just synced
[22:42:00] <marcheu> just do a clean chroot
[22:42:10] <ellyjones> as I said, I have a bunch of pending changes in this chroot
[22:42:20] <ellyjones> is there a way to fix this that does not involve destroying my old chroot?
[22:42:21] <marcheu> well then don't complain and get other people's stuff reverted...
[22:43:11] <ellyjones> last night, my build worked, and today it does not work
[22:43:26] <ellyjones> there was (and still hasn't been) a PSA sent telling people to destroy their chroots
[22:43:38] <marcheu> ok, so if I send a PSA it's ok?
[22:43:41] <ellyjones> that my chroot will continue working in the absence of such a PSA seems like an entirely reasonable idea
[22:43:46] <marcheu> I'll just push it and send a PSA
[22:43:48] <marcheu> sounds good
[22:43:57] <ellyjones> can you make it not break chroots instead?
[22:44:00] <marcheu> no
[22:44:02] <ttuttle|work> isn't there a way to make it update the chroot?
[22:44:09] <ellyjones> there's an update chroot mechanism, yeah
[22:44:12] <ellyjones> I dunno how it works
[22:44:17] <ttuttle|work> that seems like an entirely reasonable solution
[22:44:27] <ttuttle|work> mailing lists are a bad way to do change control
[22:44:28] <marcheu> well I'll send a PSA
[22:44:46] <ttuttle|work> I would look in to automating the fix :P
[22:44:52] <ellyjones> yeah
[22:45:02] <ellyjones> otherwise you will get dozens of people who didn't read the PSA complaining to you
[22:45:08] <marcheu> ah well, the prime consumer of the X update is touch anyway
[22:45:14] <ttuttle|work> like, you are pre-buffering the "sorry I broke your chroot" email
[22:46:21] <njw> src/scripts/chroot_version_hooks.d
[22:46:22] <ellyjones> I don't understand why a package update necessitates a new chroot still
[22:47:52] <marcheu> under some circumstances, it seems like emerge has that limitation that it can't sort out dependencies
[22:47:59] <marcheu> when package versions change
[22:48:02] <ttuttle|work> you mean blockers?
[22:48:03] <ellyjones> which circumstances? oO
[22:48:16] <marcheu> well downgrading version is one, but that one is something else
[22:48:25] <ttuttle|work> which one is this?
[22:48:30] <marcheu> no idea
[22:48:41] <ttuttle|work> ...but you made the change; what does it do, and what does emerge say?
[22:48:55] <marcheu> anyway, the only way to get out of this is redo the chroot I think
[22:49:02] <ttuttle|work> the whole thing?
[22:49:03] <marcheu> otherwise we'll be stuck with old packages
[22:49:30] <ellyjones> the change is portage-stable:60285ea661d9d8b936bbf7025778b3da07b03d65, which introduces a dependency on xlsclients-1.1.2, which depends on libxcb-1.7, for which there is no ebuild
[22:50:14] <marcheu> ellyjones: I do not introduce a dependency on xlsclients-1.1.2, I bring over the ebuild. it still picks up xlsclients-1.0 here
[22:50:20] <marcheu> ellyjones: that what's wrong with your setup
[22:50:44] <ttuttle|work> it's not "her setup", it's tip of tree?
[22:50:47] <marcheu> you probably need to unmerge your xlsclients before that update
[22:51:07] <marcheu> ttuttle|work: but why does it work with all the bots, and here too for x86-generic and kaen?
[22:51:13] <njw> if a developer needs to unmerge to avoid some kind of conflict, add a chroot hook to fix that for everyone.
[22:51:19] <ttuttle|work> Yeah, what njw said.
[22:51:22] <ttuttle|work> The bots don't test everything.
[22:51:37] <ttuttle|work> Someone is saying "your change breaks this by doing foo, and you need to do bar to fix it."  Please just do bar. :P
[22:52:17] <ellyjones> okay, my local environment now picks up xlsclients-1.1.1 instead, which seems to work
[22:52:26] <njw> see, for example, chroot_version_hooks.d/6_mobile_broadband_removal, which I added when I broke this.
[22:52:27] <marcheu> ttuttle|work: it is impossible to repro locally
[22:52:37] <ttuttle|work> marcheu: ...because you already fixed your chroot :P
[22:52:49] <ttuttle|work> marcheu: try syncing back before your change, blowing away your chroot, then syncing ahead of it again.
[22:52:56] <ellyjones> I didn't _change_ anything
[22:53:25] <ellyjones> that's weird
[22:54:55] <marcheu> lets do the other solution, I'll leave the old xlsclients 1.1.1 around
[22:55:21] <quiche> ellyjones: maybe something changed in prebuilds? (just a guess -- i don't know how this stuff works that well.)
[22:55:55] <ellyjones> quiche: no, this happened with no intervening sync; perhaps portage has something cached:
[22:55:58] <ellyjones> (cros-chroot) ellyjones@locus ~/trunk/src/scripts $ equery-tegra2_kaen which xlsclients
[22:56:01] <ellyjones> /home/ellyjones/trunk/src/third_party/portage-stable/x11-apps/xlsclients/xlsclients-1.1.2.ebuild
[22:56:05] <ellyjones> (cros-chroot) ellyjones@locus ~/trunk/src/scripts $ equery-tegra2_kaen which libxcb
[22:56:08] <ellyjones> /home/ellyjones/trunk/src/third_party/chromiumos-overlay/x11-libs/libxcb/libxcb-1.5-r1.ebuild
[22:56:11] <ellyjones> (cros-chroot) ellyjones@locus ~/trunk/src/scripts $ less /home/ellyjones/trunk/src/third_party/portage-stable/x11-apps/xlsclients/xlsclients-1.1.2.ebuild
[22:56:14] <ellyjones> /home/ellyjones/trunk/src/third_party/portage-stable/x11-apps/xlsclients/xlsclients-1.1.2.ebuild: No such file or directory
[22:56:16] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard-tangent canary" from None: )'
[22:56:19] <marcheu> ok let me quickly cook a change, and you guys let me know if it flies for you
[22:56:22] <ellyjones> bafflement
[22:56:32] <marcheu> I'll leave the old ebuilds around, that might allow transitioning
[22:56:35] <ellyjones> I can build packages now so I am not a good test case any more
[22:56:59] <marcheu> what so now it works
[22:57:02] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard-tangent canary" from None: -- sheriffs investigating )'
[22:57:09] <ellyjones> yeah, see the output I pasted above
[22:57:20] 
[22:57:21] <ellyjones> portage must have had cached that there was a 1.1.2 ebuild or something
[22:57:25] <ttuttle|work> ellyjones: did you rebuild your chroot to make it work?
[22:57:30] <marcheu> ellyjones: who else had the issue?
[22:57:37] <ellyjones> no, I just said 'equery-$board which xlsclients'
[22:57:53] <ellyjones> marcheu: kliegs, and I do not know who else
[22:57:56] <quiche> i was having trouble with build_packages --nousepkg
[22:58:03] <quiche> lemme see if that still fails
[22:58:03] <ellyjones> I still don't understand what changed
[22:58:40] <ellyjones> oh, perhaps I got cros_sdk'd
[22:58:52] *** stalled has quit IRC
[22:59:11] <quiche> $ ./build_packages --board=x86-generic --nousepkg
[22:59:20] <quiche> emerge: there are no ebuilds to satisfy ">=x11-libs/libxcb-1.7" for /build/x86-generic/.
[22:59:20] <quiche> (dependency required by "x11-apps/xlsclients-1.1.2" [ebuild])
[22:59:20] <quiche> (dependency required by "chromeos-base/chromeos-dev-0.1.0-r46" [installed])
[22:59:20] <quiche> (dependency required by "chromeos-base/chromeos-dev" [argument])
[22:59:21] <ellyjones> I am totally baffled, but now I have a different build failure
[22:59:35] <ellyjones> quiche: equery-x86-generic which xlsclients
[22:59:43] <ellyjones> I am curious what'll happen
[22:59:49] <quiche> /home/quiche/trunk/src/third_party/portage-stable/x11-apps/xlsclients/xlsclients-1.1.2.ebuild
[23:00:02] <ellyjones> if you cat it, is it there?
[23:00:08] <quiche> yep, file exists
[23:00:17] <quiche> RDEPEND="
[23:00:17] <quiche> 	>=x11-libs/libxcb-1.7
[23:00:17] <quiche> 	>=x11-libs/xcb-util-0.3.8
[23:00:17] <quiche> "
[23:00:21] <quiche> (and some other stuff)
[23:00:25] <ellyjones> I do not have it
[23:00:29] <ellyjones> weeeird
[23:01:00] <ellyjones> ah, do you not have tbroch's be1902b6c537866ae2138fa51c3b721b59baaaad?
[23:01:17] <marcheu> yeah it seems to fly if I keep the old ebuilds around
[23:01:20] <quiche> yeah, i haven't sync-ed since he put in the revert
[23:01:21] <marcheu> I'll re-test after that meeting
[23:02:34] <ellyjones> if I undo tbroch's revert and say emerge-$board xlsclients, it looks for xlsclients 1.0.2
[23:02:58] <ellyjones> in spite of the presence of 1.1.2 in portage-stable
[23:03:02] <ellyjones> davidjames: you around?
[23:03:21] <davidjames> ellyjones: yes
[23:03:26] <crosbot> tree became 'Tree is open (webkit patch issue.  Addressed by http://gerrit.chromium.org/gerrit/#change,8368 awaiting forced chrome rev)'
[23:03:30] <ellyjones> davidjames: can you shed any light? this is freaky
[23:03:38] <ttuttle|work> argh
[23:03:42] <ttuttle|work> the autotest pages
[23:03:43] <ttuttle|work> are a maze
[23:03:45] <ttuttle|work> and I am lost
[23:04:31] <davidjames> ellyjones: Maybe libxcb-1.7 ebuild has an error, that's the message displayed when ebuild is broken
[23:04:42] <davidjames> ellyjones: Actual error is displayed above that message :)
[23:04:44] <ellyjones> davidjames: specifically, without portage-stable:be1902b6c537866ae2138fa51c3b721b59baaaad, equery-$board which xlsclients returns xlsclients-1.1.2, but emerge-$board xlsclients returns xlsclients-1.0.2
[23:04:56] <ellyjones> (or rather, tries to build xlsclients-1.0.2)
[23:05:01] *** rharrison_chrome has quit IRC
[23:05:12] <davidjames> ellyjones: Yeah it does that if xlsclients-1.0.2 has error or breaks a dependency
[23:05:28] <ellyjones> if I force building -1.1.2, I get the complaint about xlsatoms
[23:05:40] <ellyjones> er, about libxcb
[23:05:45] <davidjames> ellyjones: Hence the error that needs to be fixed :)
[23:05:59] <ellyjones> so my question is, why did it ever try to use 1.1.2 to resolve the dependency cycle, since it knows 1.1.2 is broken?
[23:06:35] <ttuttle|work> I swear there was a link here before that isn't here nw.
[23:06:37] <davidjames> ellyjones: portage has code to ensure that latest version of packages is consistent
[23:07:03] <davidjames> ellyjones: That's the autounmask feature -- it helps you by failing when you make a mistake :)
[23:07:09] <ellyjones> heh
[23:07:17] <ellyjones> great
[23:07:37] <davidjames> You can disable autounmask and then it continues to skip over buggy ebuilds happily :)
[23:07:40] <ellyjones> I wish we had a deterministic build system :P
[23:08:00] <davidjames> Well that's the reason for autounmask, to make things more deterministic, deterministically failing that is :)
[23:08:29] <ellyjones> except I didn't touch anything and portage decided suddenly that there was no longer the (reverted) 1.1.2 ebuild
[23:09:03] *** stalled has joined #chromium-os
[23:09:10] <davidjames> ellyjones: Oh, if you're talking about bots, they are probably unhappy that they were left in an inconsistent state after certain packages were upgraded and then ebuilds were pulled out from under the bot
[23:09:36] <ellyjones> I'm talking aboutmy local chroot
[23:10:01] <davidjames> Ah, and maybe you have a mix of versions thanks to upgrade / downgrade of X ?
[23:10:43] <ellyjones> I have no idea :)
[23:10:45] <davidjames> Easiest way to cleanup is setup_board --force, but if you want to investigate why Portage is confused, join #gentoo-portage and I can help you there
[23:10:59] <ellyjones> I don't care much
[23:11:00] <ttuttle|work> ...one of my tests just failed with:
[23:11:01] <ttuttle|work> 09/27 14:04:18 ERROR|logging_ma:0560| /usr/local/lib/flimflam/test/veth: 65: dhcpd: not found
[23:11:13] <ellyjones> ttuttle|work: weird
[23:11:34] * ttuttle|work installs dhcpd, goes about his day.
[23:15:13] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" from 8686a474646047331b2a43e2e30f26347c4e60e7: Mike Frysinger <vapier at chromium dot org>)'
[23:15:47] <vapier> not it: suite_Smoke/login_CryptohomeUnmounted              FAIL
[23:16:32] <ellyjones> chrome crash?
[23:16:45] *** Ruetobas has quit IRC
[23:19:33] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" from 8686a474646047331b2a43e2e30f26347c4e60e7: Mike Frysinger <vapier at chromium dot org> --> sheriffs investigating)'
[23:19:59] <tbroch> seems similar to the crash earlier this morning related to login flak
[23:20:08] <tbroch> 14:12:32 INFO | 		ERROR	login_CryptohomeUnmounted	login_CryptohomeUnmounted	timestamp=1317157953	localtime=Sep 27 14:12:33	Unhandled JSONInterfaceError: Automation call {'username': 'performancetestaccount at gmail dot com', 'password': 'perfsmurf', 'command': 'Login'} received empty response.  Perhaps the browser crashed.
[23:21:01] <tbroch> how/can we get the detailed output from autotest ... troopers?
[23:23:45] <tbroch> gonna re-open crosbug.com/20323
[23:24:39] *** Ruetobas has joined #chromium-os
[23:24:40] <crosbot> tree became 'Tree is open (crosbug.com/20323)'
[23:26:39] *** dalecurtis has joined #chromium-os
[23:28:59] *** rbyers_ has joined #chromium-os
[23:29:00] *** ChanServ sets mode: +v rbyers_
[23:29:25] <gauravsh> tborch: all the autotest logs are part of the build artifacts. which are linked at the end stage of each buildbot run.
[23:34:35] *** TW1920_ has quit IRC
[23:34:56] <tbroch> gauravsh: thanks
[23:35:10] <crosbot> tree became 'Tree is open (crosbug.com/20323, crosbug.com/20887)'
[23:42:01] <kliegs> marcheu: sorry - was in meetings
[23:42:29] <kliegs> catching up on the talk above w.r.t the x changes
[23:46:59] *** achuith has quit IRC
[23:48:13] *** rbyers_ has quit IRC
[23:48:15] *** achuith has joined #chromium-os
[23:48:15] *** ChanServ sets mode: +v achuith
[23:48:49] *** achuith has quit IRC
[23:51:53] <seanpaul> marcheu: FYI: I did a clean chroot at 11:16 and still encountered the dependency error
[23:52:01] <seanpaul> 11:16 PST
[23:56:10] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy-binary" from 8686a474646047331b2a43e2e30f26347c4e60e7: Mike Frysinger <vapier at chromium dot org>)'
[23:56:34] <vapier> suite_Smoke/desktopui_WindowManagerFocusNewWindows FAIL
[23:58:13] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy-binary" from 8686a474646047331b2a43e2e30f26347c4e60e7: Mike Frysinger <vapier at chromium dot org> --> sheriffs investigating)'





top