[00:00:51] *** saintlou has joined #chromium-os
[00:00:51] *** ChanServ sets mode: +v saintlou
[00:00:56] *** sadrul has quit IRC
[00:02:11] *** magn3ts has joined #chromium-os
[00:06:28] *** jennb has joined #chromium-os
[00:06:28] *** ChanServ sets mode: +v jennb
[00:18:49] *** saintlou has quit IRC
[00:29:45] <gsam> is there a workaround for cros_image_to_target failures?
[00:32:18] <zbehan> gsam: I'm hoping more like a fix soon
[00:32:20] *** saintlou has joined #chromium-os
[00:32:20] *** ChanServ sets mode: +v saintlou
[00:33:24] <darinski> gsam: it should be fixed already
[00:33:48] * gsam goes to sync his tree
[00:34:07] <zbehan> oh, i guess i missed the fun then
[00:34:41] <darinski> commit 613c123d1e2e9a71a257780bb674af31dca8e4a8 in platform/dev
[00:36:07] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "arm generic PFQ" from fe0e85ba64e15a23c3d0e2fc125c1b336992dcf7: mukesh agrawal <quiche at chromium dot org>)'
[00:36:38] <crosbot> tree became 'Tree is closed (sosa lookin)'
[00:37:16] <sosa> quiiiiche!
[00:39:37] *** kinaba1 has joined #chromium-os
[00:39:38] *** ChanServ sets mode: +v kinaba1
[00:39:42] <quiche> d'oh!
[00:39:43] <quiche> on it!
[00:46:37] <crosbot> tree became 'Tree is open (revert in for test ebuild change)'
[01:05:19] <ellyjones> hi guys
[01:05:28] <ellyjones> I'm going to land a change that has previously broken the tree now
[01:05:42] <ellyjones> specifically http://gerrit.chromium.org/gerrit/#change,6085
[01:05:48] <ellyjones> I believe everything is fine now but we'll see
[01:05:57] <gsam> when's your flight home?
[01:06:06] <ellyjones> friday
[01:06:09] <gsam> :)
[01:06:12] <ellyjones> I should've committed from the airport :P
[01:07:34] <ellyjones> anyway, it's in
[01:07:53] <ellyjones> it's basically like a flag flip for lots of new code :P
[01:08:47] <gauravsh> don't worry, those things never break things.
[01:12:29] <ellyjones> yeah
[01:12:47] <ellyjones> the last time I did this change it was four interlocking CLs that all had to land atomically :)
[01:12:49] <ojn> submit and leave, i'd say. it'll be alright, i'm sure.
[01:13:30] <gauravsh> the PFQ will rev for both host and board, right?
[01:13:49] <ellyjones> I hope so!
[01:13:57] <ellyjones> if not there will be much excitement :)
[01:16:33] <cmasone> kuan: I believe that http://src.chromium.org/viewvc/chrome?view=rev&revision=97925 broke the ToT chromium build for cros
[01:16:46] <cmasone> http://build.chromium.org/p/chromiumos/builders/x86%20generic%20TOT%20chrome%20PFQ/builds/503/steps/BuildTarget/logs/stdio
[01:18:36] <cmasone> kuan: this shouldn't close our tree, but it does need to get addressed either chrome side, or by us figuring out that we need to set some GYP_DEFINE in our ebuild or something
[01:30:48] *** saintlou is now known as saintlou|away
[01:33:18] *** srao has quit IRC
[01:33:39] * ellyjones anxiously watches http://build.chromium.org/p/chromiumos/builders/x86%20generic%20PFQ/builds/509/steps/cbuildbot/logs/stdio
[01:35:12] *** saintlou|away is now known as saintlou
[01:36:20] <dhendrix> dumb q about the updated "Building Chromium OS" instructions: Where should cros_sdk be run from? It's located in /path/to/chromiumos/chromite/bin/, but the doc seems to imply I should run it from /path/to/chromiumos/ .
[01:36:34] <ellyjones> it can be run from anywhere
[01:36:39] <ellyjones> update your depot_tools
[01:36:54] <dhendrix> ellyjones:  ah! thanks, I'll give that a shot.
[01:42:41] <dhendrix> (it worked)
[01:45:21] *** Sarten-X2 has quit IRC
[01:46:24] <sosa> i think it can run from anywhere in your src tree)
[01:46:39] <sosa> rather than anywhere anywhere
[01:48:58] <ellyjones> build 509 passed! :D
[01:50:58] *** hashimoto has joined #chromium-os
[01:51:11] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_kaen-binary" from 5d5cc65756ad8f193205a8a03ae1e8225733a8ce: tony at chromium dot org <tony at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[01:51:57] <crosbot> tree became 'Tree is closed (sosa looking at build_image failure)'
[01:52:54] <sosa> verity: /usr/lib/gcc/x86_64-pc-linux-gnu/4.3.4/libstdc++.so.6: version `GLIBCXX_3.4.11' not found (required by verity)
[01:53:21] <ellyjones> woah oO
[01:53:28] *** Sarten-X has joined #chromium-os
[01:54:26] *** JasonO has joined #chromium-os
[01:54:26] *** JasonO has joined #chromium-os
[01:55:17] <sosa> could be raymes change that hasn't been tested by the pfq yet (guessing)
[02:00:25] <kuan> cmasone: i've pinged pawel and then discussed with sosa, sosa will help fix it.  thanks sosa!
[02:00:32] *** kuan has left #chromium-os
[02:01:55] *** wfrichar has quit IRC
[02:03:54] <frankf> Has anyone seen this problem: repeated calls to build_image causes a leak in loop devices, eventually build_image spits out: "losetup: could not find any free loop device"
[02:04:27] <ellyjones> I own that bug
[02:04:34] <ellyjones> if you are interested in it, feel free to take it
[02:04:40] <ellyjones> (I don't know how to repro it)
[02:05:53] *** lipsinV2 has joined #chromium-os
[02:13:08] *** achuith has joined #chromium-os
[02:13:08] *** ChanServ sets mode: +v achuith
[02:15:28] <crosbot> tree became 'Tree is open (prebuilts race condition where binary build picked up full builds prebuilts)'
[02:15:42] <sosa> grawg
[02:16:35] <ellyjones> RAWR
[02:20:07] *** magn3ts has quit IRC
[02:20:56] *** magn3ts has joined #chromium-os
[02:22:38] <stevenjb_> sosa, kuan, wyck, clchiou - Heads up, this change may have broken LOCAL chrome on chromeos builds: http://codereview.chromium.org/7655046
[02:23:04] <sosa> it also break the TOT chrome pfq
[02:23:07] <sosa> broke*
[02:23:12] <sosa> notice it's redness :)
[02:23:20] <stevenjb_> OK, cool, I'll assume you're on it then :)
[02:23:40] <sosa> yes, we have to add a Dep on the kerberos headers to chromeos-chrome DEPENDS'
[02:23:49] <sosa> just looking to figure out where that is
[02:24:23] <stevenjb_> What waterfall has the tot chrome pfq btw?
[02:25:15] *** stalled has quit IRC
[02:25:20] <sosa> chromiumos
[02:25:38] <sosa> also the chromium FYI waterfall
[02:25:48] <stevenjb_> nothing red here: http://build.chromium.org/p/chromiumos/waterfall
[02:25:59] <sosa> http://build.chromium.org/p/chromiumos/waterfall?builder=x86%20generic%20TOT%20chrome%20PFQ
[02:26:15] <sosa> depending on the size of your screen you may want to scroll right.
[02:26:28] <sosa> :)
[02:27:55] <gauravsh> why doesn't it show it in the list on the top?
[02:29:44] <ellyjones> frankf: ok, I'm annoyed enough by that to try tracking it down now
[02:30:15] *** patcito has quit IRC
[02:30:18] *** patcito_ has joined #chromium-os
[02:31:51] *** patcito_ is now known as patcito
[02:31:51] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl-binary" from f3e814d7ad65525f61580c7525e8824c2b0137ca: _third_party_ at chromium dot org, tony at chromium dot org <tony at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[02:33:53] <crosbot> tree became 'Tree is closed (Clobbering all binary builders)'
[02:34:48] <ellyjones> frankf: are you using --noenable_rootfs_verification?
[02:36:50] <sosa> gauravsh: it's not a TreeCloser
[02:38:28] <gauravsh> sosa: are the ones listed under the "important" line meant to be tree closers? or is that everything in that summary matrix on the top (important, x86, arm and chrome)
[02:39:06] <sosa> i only included treeclosers in the lists
[02:40:46] <gauravsh> i think this was answered a few emails ago, but what is special about the "important" builders then?
[02:45:24] <sosa> easy way to see status of closers
[02:45:27] <sosa> in a list
[02:45:37] <sosa> and clicking on the important link only gives you a view of just them
[02:45:41] <sosa> to make it easier to sheriff?
[02:45:49] <ellyjones> while trying to reproduce the loop_dev failure: 847G  804G   11M 100%
[02:46:56] <crosbot> tree became 'Tree is open (If binary builder fails, clobber it)'
[02:47:10] <gauravsh> sosa: oh ok. so the important ones are also duplicated in the sections below (x86, arm, chrome). Didn't realize that.
[02:50:49] *** Keybuk has quit IRC
[02:53:21] *** stalled has joined #chromium-os
[02:54:33] *** JasonO has quit IRC
[02:57:51] <stevenjb_> sosa: is someone working on fixing or reverting http://codereview.chromium.org/7655046 in ToT?
[02:58:15] <sosa> adding virtual/krb5 as a build dep of Chrome
[02:58:35] *** JasonO has joined #chromium-os
[02:58:35] *** JasonO has joined #chromium-os
[02:59:12] <stevenjb_> cool. is there a quick workaround?
[02:59:56] <sosa> sudo emerge -1 virtual/krb5
[03:01:37] <sosa> (i'm testing it myself)
[03:02:39] <ellyjones>   --download            Download and download a prebuilt SDK
[03:02:43] <ellyjones> thanks, cros_sdk!
[03:03:11] <cmasone> sosa: if I land a change to masters/master.chromiumos/master.cfg, I'll need to do a master restart to see the effects, right?
[03:03:20] <sosa> no
[03:03:24] <sosa> make reconfig
[03:03:25] <cmasone> orly?
[03:03:33] <sosa> doesn't require all bots to restart
[03:03:43] <sosa> yeah
[03:10:55] <cmasone> sosa: uh...make reconfig does not do the thing
[03:11:01] <cmasone> do I have to ssh to the master...?
[03:11:09] <sosa> uh yah
[03:11:29] <cmasone> sosa: that makes sense
[03:12:39] <stevenjb_> sosa: that didn't seem t do the trick :(
[03:13:58] *** yosafbri` has quit IRC
[03:15:07] *** aaronp_ has quit IRC
[03:15:42] *** yosafbridge has joined #chromium-os
[03:18:36] *** aaronp has quit IRC
[03:21:47] *** stevenjb_ has quit IRC
[03:23:47] <crosbot> tree became 'Tree is open (sheriffs off-duty - If binary builder fails, clobber it)'
[03:24:36] *** jrbarnette has quit IRC
[03:28:24] <ellyjones> did something bad happen to system_api?
[03:28:33] <ellyjones> I'm getting file collisions from it now for dbus/service_constants.h
[03:28:42] <cmasone> ellyjones: i burned that motherfucker down
[03:28:58] <ellyjones> aha
[03:29:04] <sosa> stevenjb: works for me?  perhaps you didn't do emerge rather emerge-board?
[03:29:26] <sosa> or maybe you need the latter
[03:29:35] <sosa> i added it as a DEPEND so i'm not sure what that ends up being
[03:29:46] <sosa> i originally thought the former, but now I'm wondering if it's the latter
[03:30:55] *** stevenjb_ has joined #chromium-os
[03:32:04] *** saintlou has quit IRC
[04:05:50] *** sadrul has joined #chromium-os
[04:05:50] *** ChanServ sets mode: +v sadrul
[04:08:10] *** dennisjeffrey has quit IRC
[04:11:59] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "arm generic full" from fe0e85ba64e15a23c3d0e2fc125c1b336992dcf7: Thieu Le <thieule at chromium dot org>, mukesh agrawal <quiche at chromium dot org>)'
[04:17:52] <crosbot> tree became 'Tree is open (git hang)'
[04:23:17] *** stevenjb_ has quit IRC
[04:32:06] *** stevenjb_ has joined #chromium-os
[04:37:03] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb_he canary" from None: )'
[04:59:24] *** BladeFreak has quit IRC
[05:06:11] <crosbot> tree became 'Tree is open (sheriffs off-duty - x86-zgb_he canary disk full)'
[05:25:32] *** aaronp has joined #chromium-os
[05:26:40] *** aaronp_ has joined #chromium-os
[05:32:15] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 6402e501258c9f64373dcd05271c6c7dd75e20bc: Chris Masone <cmasone at chromium dot org>)'
[05:33:29] <sosa> cmasone...unittests?
[05:37:52] <crosbot> tree became 'Tree is open (reverted bad change)'
[05:38:28] <sosa> i see :)
[06:03:59] *** JasonO has quit IRC
[06:07:42] *** e_t_ has quit IRC
[06:21:58] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl-binary" from 6402e501258c9f64373dcd05271c6c7dd75e20bc: Chris Masone <cmasone at chromium dot org>, abodenha at chromium dot org <abodenha at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[06:24:33] <crosbot> tree became 'Tree is open (sheriffs off-duty - If binary builder fails, clobber it)'
[06:26:10] *** stalled has quit IRC
[06:33:50] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_arthur-binary" from 313bbf5ef2c0a1c428e3d92847c1b4b665b71706: abodenha at chromium dot org <abodenha at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[06:34:51] <crosbot> tree became 'Tree is open (missed clobber)'
[06:39:29] *** cooled_ has joined #chromium-os
[06:39:56] *** stalled has joined #chromium-os
[06:40:20] *** cooled has quit IRC
[06:42:58] *** sergiu has quit IRC
[06:45:52] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from None: )'
[06:49:26] <crosbot> tree became 'Tree is open (Didn't block new Chrome correctly, fix pushed)'
[06:57:42] *** stalled has quit IRC
[07:02:18] *** gfrog has quit IRC
[07:09:33] *** gfrog has joined #chromium-os
[07:18:32] *** stalled has joined #chromium-os
[07:23:57] *** nirnimesh_ has joined #chromium-os
[07:38:43] *** magn3ts has quit IRC
[07:58:40] *** aaronp has quit IRC
[08:15:31] *** jochen__ has quit IRC
[08:15:36] *** jochen__ has joined #chromium-os
[08:15:36] *** ChanServ sets mode: +v jochen__
[08:24:46] *** Styx has joined #chromium-os
[08:35:56] *** vmil86 has joined #chromium-os
[08:58:01] <nirnimesh_> I'm seeing "gssapi.h: No such file or directory" when building chrome. Anyone seen this lately?
[09:01:18] <zbehan> nirnimesh: in the chroot or outside?
[09:02:47] <zbehan> nirnimesh: hmm, in, i guess, otherwise you'd be asking at #chromium :) try running equery-$board belongs gssapi.h
[09:07:48] *** ellyjones has quit IRC
[09:25:00] *** stevenjb_ has quit IRC
[09:27:03] *** vmil86 has quit IRC
[09:39:15] *** nirnimesh_ has quit IRC
[09:41:43] *** vmil86 has joined #chromium-os
[09:53:25] *** hajovonta has joined #chromium-os
[10:10:12] *** hajovonta has left #chromium-os
[10:34:48] *** patcito has quit IRC
[10:50:11] *** MX80 has quit IRC
[10:50:35] *** snickersnack has joined #chromium-os
[10:51:40] *** snickersnack has quit IRC
[10:52:15] *** snickersnack has joined #chromium-os
[10:53:17] *** arun_ has quit IRC
[11:07:45] *** arun_ has joined #chromium-os
[11:29:07] *** arun_ has quit IRC
[11:38:12] *** silverroots has joined #chromium-os
[11:43:15] *** arun_ has joined #chromium-os
[12:19:39] *** MX80 has joined #chromium-os
[14:11:06] *** jochen___ has joined #chromium-os
[14:11:06] *** ChanServ sets mode: +v jochen___
[14:12:39] *** jujugre has joined #chromium-os
[14:13:01] *** jochen__ has quit IRC
[14:13:01] *** jochen___ is now known as jochen__
[14:57:50] <crosbot> tree became '?'
[14:58:06] <crosbot> tree became 'Tree is open (Didn't block new Chrome correctly, fix pushed)'
[15:03:26] *** MX80 has quit IRC
[15:19:52] *** rharrison_chrome has joined #chromium-os
[15:20:07] *** ChanServ sets mode: +v rharrison_chrome
[15:20:15] <rharrison_chrome> East Coast Sheriff checking in
[15:28:28] *** MX80 has joined #chromium-os
[15:33:41] *** silverroots has quit IRC
[15:40:12] *** petermayo has joined #chromium-os
[15:40:12] *** ChanServ sets mode: +v petermayo
[15:45:04] *** rbyers_ has joined #chromium-os
[15:45:20] *** rbyers_ has quit IRC
[15:53:50] *** borios has joined #chromium-os
[15:54:17] *** oc80z has quit IRC
[15:54:30] <borios> Chrome OS has a wiki page ?
[15:55:12] *** oc80z has joined #chromium-os
[15:56:34] *** varunjain has quit IRC
[15:56:41] *** varunjain has joined #chromium-os
[15:56:41] *** ChanServ sets mode: +v varunjain
[16:07:11] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy canary" from None: )'
[16:17:26] *** mnissler has quit IRC
[16:17:40] *** mnissler has joined #chromium-os
[16:17:40] *** ChanServ sets mode: +v mnissler
[16:18:57] *** penghuang has joined #chromium-os
[16:18:57] *** ChanServ sets mode: +v penghuang
[16:21:30] <crosbot> tree became 'Tree is closed (Investigating -> rharrison)'
[16:21:33] *** borios has quit IRC
[16:21:55] *** borios has joined #chromium-os
[16:39:27] <crosbot> tree became 'Tree is open (Potential flake on stumpy_canary, hoping build goes green -> rharrison)'
[16:49:59] *** cros_ has joined #chromium-os
[17:08:22] *** borios has quit IRC
[17:08:24] *** borios_ has joined #chromium-os
[17:09:57] *** borios_ has quit IRC
[17:12:45] *** jrbarnette has joined #chromium-os
[17:15:08] *** Styx has quit IRC
[17:16:43] *** Styx has joined #chromium-os
[17:29:59] *** greboid3 has joined #chromium-os
[17:42:06] *** wfrichar has joined #chromium-os
[17:42:07] *** ChanServ sets mode: +v wfrichar
[17:50:50] *** Greboid has quit IRC
[17:52:31] *** Flameeyes has quit IRC
[17:52:41] *** greboid3 is now known as Greboid
[17:54:42] *** Flameeyes has joined #chromium-os
[18:01:05] *** JasonO has joined #chromium-os
[18:03:36] *** lipsinV2 has quit IRC
[18:10:06] <ojn> large branch push going on right now, hopefully infrastructure will handle it. :)
[18:10:49] *** D|sT has quit IRC
[18:12:09] *** stevenjb has quit IRC
[18:14:10] * gsam_ looks for davidjames
[18:14:25] <gsam_> i am still stuck on the build_payload breakage
[18:14:35] <gsam_> did repo sync
[18:14:40] <gsam_> build_packages etc
[18:14:47] <gsam_> but cros_image_to_target continues to fail
[18:14:57] <gsam_> anyone know of another step required to fix the chroot?
[18:15:10] <davidjames> gsam_: emerge cros-devutils ?
[18:15:16] <gsam_> k
[18:15:43] <davidjames> gsam_: Also could you run cros_workon --host list and unworkon any packages you are done with there?
[18:15:58] <gsam_> nothing listed
[18:16:09] <gsam_> davidjames, do i need to build_packages again after emerge?
[18:16:31] <davidjames> gsam_: Nope just emerge cros-devutils is enough. But build_packages should have done that for you so I am puzzled
[18:16:39] <gsam_> still fails
[18:16:47] <gsam_> mount: can't find /tmp/state.o80wV4 in /etc/fstab or /etc/mtab
[18:16:47] <gsam_> umount: /tmp/state.o80wV4: not mounted
[18:16:47] <gsam_> ERROR: generate_payload failed
[18:17:00] *** dennisjeffrey has joined #chromium-os
[18:17:00] *** ChanServ sets mode: +v dennisjeffrey
[18:17:12] <gsam_> >>> Emerging (1 of 1) chromeos-base/cros-devutils-0.0.1-r118 from chromiumos
[18:18:03] <davidjames> gsam_: ( cd ~/trunk/src/platform/dev && git rev-parse HEAD )
[18:18:21] <gsam_> 613c123d1e2e9a71a257780bb674af31dca8e4a8
[18:19:03] <davidjames> gsam_: Ah I see the problem, your chromiumos-overlay didn't update right during repo sync?
[18:19:18] <davidjames> gsam_: Maybe you have local changes there and need to rebase your branch?
[18:19:30] <gsam_> had local changes but did rebase, will check
[18:19:31] <gsam_> thansk
[18:20:57] <gsam_> git rev-parse is unchanged
[18:21:18] <gsam_> what am i looking for in git log of chromiumos-overlay?
[18:21:28] <gsam_> is that where the reverrt cl landed?
[18:21:31] <davidjames> gsam_: Actually I was wrong, your chromiumos-overlay is up to date, you've got r118
[18:21:53] <davidjames> gsam_: How about diff -u ~/trunk/src/platform/dev/host/cros_generate_update_payload /usr/bin/cros_generate_update_payload
[18:22:20] <gsam_> no differences
[18:22:54] <davidjames> gsam_: Ok that's really puzzling.
[18:23:06] <davidjames> gsam_: Will followup over chat
[18:23:10] <gsam_> k
[18:23:12] <gsam_> thank you
[18:39:52] *** saintlou has joined #chromium-os
[18:39:52] *** ChanServ sets mode: +v saintlou
[18:41:38] *** jujugre has left #chromium-os
[18:42:26] *** vbendeb has joined #chromium-os
[18:42:26] *** ChanServ sets mode: +v vbendeb
[18:44:09] <vbendeb> guys, anyone knows what's the deal with x86 generic chrome PFQ and friends?
[18:45:38] *** saintlou has quit IRC
[18:45:43] *** saintlou has joined #chromium-os
[18:45:43] *** ChanServ sets mode: +v saintlou
[18:46:19] <davidjames> vbendeb: Looks to me like a release was cut that doesn't build, but tip of tree builds fine
[18:46:38] <davidjames> vbendeb: So x86 generic chrome pfq won't go green until chrome cuts another release (that builds)
[18:47:56] <vbendeb> davidjames: thanks! Do you know if the new crom release is in the works - are they aware of the problem?
[18:51:31] *** jrbarnette_ has joined #chromium-os
[18:51:31] *** ChanServ sets mode: +v jrbarnette_
[18:51:42] <davidjames> FYI, gsam_ and I figured out the issue. gsam_ had two chroots, one named chroot and one named chroot-arm. cros_image_to_target.py has a bug and was using the wrong chroot
[18:52:03] *** sque has joined #chromium-os
[18:52:09] *** benchan has joined #chromium-os
[18:52:09] *** ChanServ sets mode: +v benchan
[18:55:06] *** jrbarnette has quit IRC
[18:55:06] *** jrbarnette_ is now known as jrbarnette
[19:02:06] <gsam_> http://code.google.com/p/chromium-os/issues/detail?id=19596
[19:03:11] *** chocobo__ has joined #chromium-os
[19:03:11] *** ChanServ sets mode: +v chocobo__
[19:07:50] *** sergiu has joined #chromium-os
[19:07:50] *** sergiu has joined #chromium-os
[19:11:02] <msb__> win 18
[19:12:59] <ojn> that's a horrible password. :)
[19:15:26] <davidjames> vbendeb: Don't know, could you get in touch with Chrome sheriffs and ask?
[19:15:49] <davidjames> vbendeb: Since they fixed the issue there's a good chance they're aware. Releases are usually cut nightly.
[19:15:55] *** sadrul has quit IRC
[19:16:13] <vbendeb> davidjames: will do - I was not sure if they were aware of the problem. BTW,
[19:16:23] <vbendeb> looking at one of the failure logs: http://build.chromium.org/p/chromiumos/builders/x86%20generic%20chrome%20PFQ/builds/85/steps/cbuildbot_chrome/logs/stdio/text
[19:16:48] <vbendeb> it is not obvious that it is failing due to a Chrome build errors - how do you tell?
[19:17:06] <davidjames> vbendeb: chromeos-chrome-15.0.861.0_rc-r1: ./net/http/http_auth_gssapi_posix.h:9:20: error: gssapi.h: No such file or directory
[19:17:37] <davidjames> vbendeb: Also look at http://build.chromium.org/p/chromiumos/builders/x86%20generic%20chrome%20PFQ/builds/85 -- it shows BuildTarget phase failed
[19:17:56] <davidjames> vbendeb: Then look at BuildTarget phase, e.g. http://build.chromium.org/p/chromiumos/builders/x86%20generic%20chrome%20PFQ/builds/85/steps/BuildTarget/logs/stdio and error is shown at bottom
[19:18:36] <vbendeb> davidjames: got you, thanks!
[19:19:07] <crosbot> tree became 'Tree is open'
[19:19:43] <davidjames> vbendeb: Looking at the error though, I am wondering why it is that gssapi.h is missing. Hopefully it is a Chrome issue that is fixed in top of tree (since top of tree builder is working)
[19:22:31] <vbendeb> davidjames: is PFQ using a different Chrome version? Or should it be just restarted to see if the problem cleared itself?
[19:24:56] <davidjames> vbendeb: Yes PFQ uses a different Chrome version, it uses latest release candidate
[19:25:00] *** sque has quit IRC
[19:25:01] *** JasonO has quit IRC
[19:25:29] <davidjames> vbendeb: Would be worthwhile to run it again, but I think you'll see same problem since Chrome version hasn't changed
[19:25:58] <vbendeb> davidjames: is there a Chrome IRC channel? how do I find out who's their sheriff today?
[19:27:57] <gauravsh> vbendeb: #chromium
[19:27:59] <davidjames> vbendeb: http://build.chromium.org/p/chromium/console lists sheriffs
[19:28:16] <nirnimesh> I was seeing the gssapi.h problem last night as well. chrome tree at the cutoff point looks broken
[19:28:27] <nirnimesh> tot seems fine now though
[19:32:55] <vbendeb> davidjames: do they have a separate IRC channel too?
[19:32:59] *** sque has joined #chromium-os
[19:35:35] <grundler> <gauravsh> vbendeb: #chromium
[19:36:19] <vbendeb> that's quite intuitive :)
[19:38:25] *** rbyers_ has joined #chromium-os
[19:39:26] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from be44c2b278ccbb8602ff254806e160e606661c3b: dpolukhin at chromium dot org <dpolukhin at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, reed at google dot com <reed at google dot com@0039d316-1c4b-4281-b951-d872f2087c98>, ttuttle <ttuttle at chromium dot org>)'
[19:40:24] *** ellyjones has joined #chromium-os
[19:40:24] *** ChanServ sets mode: +v ellyjones
[19:41:22] <vbendeb> there wAS a revert eralier today to fix the missing gssapi.h problem
[19:41:25] <davidjames> vbendeb: zgb failure looks like flake, file a bug against me for the parallel_emerge traceback
[19:41:37] <vbendeb> how do we pick it up?
[19:41:52] <vbendeb> ok, will do
[19:42:28] <davidjames> vbendeb: chrome revert will get picked up automatically tonight, it's possible to make it happen earlier but probably not necessary to do that
[19:42:43] <nirnimesh> I'm also curious. if there was a real build breakage, how was the chromiumos bots green?
[19:42:59] <vbendeb> davidjames: it's just a pain to see PFQs red though
[19:42:59] <davidjames> nirnimesh: Chrome PFQ is doing its job and preventing broken Chrome from breaking rest of Chrome OS
[19:43:27] <nirnimesh> davidjames: chrome hasn't been uprev'd in days. I'd strongly recommend doing it
[19:43:46] <nirnimesh> davidjames: several P0/S0 fixes are waiting to show up
[19:44:32] <davidjames> nirnimesh: Hmm OK, can you ask Chrome team to cut a new release that builds? Old release didn't work either
[19:45:13] <davidjames> nirnimesh: As soon as Chrome team cuts a new release candidate Chrome OS picks it up automatically
[19:45:26] <davidjames> (Well, the builder starts building and testing the new release )
[19:45:50] <crosbot> tree became 'Tree is open (flaky build, filing bug to davidjames for parallel_emerge traceback)'
[19:46:13] <nirnimesh> Don't know who to ask for it. Maybe sosa knows.
[19:46:37] <davidjames> nirnimesh, vbendeb: So sosa@ committed a chang last night to roll back Chrome to 15.0.859, was the issue that broke Chrome OS fixed yet?
[19:47:41] <davidjames> Note that even if Chrome cuts a new release, we are sticking with 859 because Chrome >=860 is marked as unstable in package.keywords
[19:48:25] <davidjames> nirnimesh, vbendeb: See http://gerrit.chromium.org/gerrit/6504
[19:48:29] <nirnimesh> I think yes. http://code.google.com/p/chromium-os/issues/detail?id=19551 has had fixes applied
[19:48:53] <davidjames> nirnimesh, vbendeb: Er, http://gerrit.chromium.org/gerrit/6505 :)
[19:50:03] <nirnimesh> What was the reason for http://gerrit.chromium.org/gerrit/6505?
[19:50:15] <davidjames> nirnimesh: Ah ok, so then once release is cut that includes the fix we can revert sosa's CL
[19:51:20] <nirnimesh> I'm not sure how chrome revision is vetted. if we revert sosa's CL, will chromeos pick up tot chrome?
[19:51:22] <vbendeb> davidjames: ok, bug filed (19603 in case you have not yet seen it :)
[19:51:52] <davidjames> vbendeb: Thanks!
[19:53:53] *** rbyers_ has quit IRC
[19:55:36] *** jrbarnette_ has joined #chromium-os
[19:57:08] *** Styx has quit IRC
[19:59:48] *** jrbarnette has quit IRC
[20:01:43] *** jrbarnette_ has quit IRC
[20:03:04] <vbendeb> from the #chrome channel:  The reverts for the problem are in the trunk of Chrome - any revision past 98029 should be OK
[20:04:29] <benchan> Has anyone seen this message from build_image this morning?   "verity: /usr/lib/gcc/x86_64-pc-linux-gnu/4.3.4/libstdc++.so.6: version `GLIBCXX_3.4.11' not found (required by verity)"
[20:05:29] <davidjames> benchan: Sounds like your toolchain was upgraded but verity was not recompiled with new toolchain
[20:07:29] <davidjames> vbendeb: Right, yes, our TOT builders that build from tip of chrome work
[20:07:49] <davidjames> vbendeb: Just not the ones that build from release candidates (release candidates are cut by the Chrome team nightly)
[20:08:07] <vbendeb> davidjames: I know, just in case it is not clear which version to grab for the PFQ
[20:09:25] <davidjames> vbendeb: It would be possible to mark one of our tip of tree builds as stable
[20:09:57] <benchan> davidjames: it looks like the opposite happened.  verity was compiled with newer c++ libs, but chroot has older c++ libs?
[20:10:03] <davidjames> vbendeb: So if you edit package.keywords you could mark a tip of tree build as stable, just note that it didn't get testing on arm or tegra etc
[20:10:28] <davidjames> benchan: Any chance you have multiple toolchains in your chroot?
[20:10:44] <davidjames> benchan: E.g. have you been testing gcc 4.6
[20:11:43] <crosbot> tree became 'Tree is open (parallel_emerge traceback -> crosbug.com/19603)'
[20:13:36] <benchan> davidjames: yes, I tested gcc 4.6 earlier this month but then reverted to 4.3.4.  this lib match problem started to appear this morning after i repo sync and use cros_sdk to enter chroot. so perhaps it picks up the gcc 4.6 toolchain again
[20:14:35] <davidjames> benchan: I don't think it's cros_sdk, that's just a wrapper around enter_chroot
[20:14:48] <benchan> davidjames: thanks for the hint. let me look into the gcc version being picked up
[20:15:28] <davidjames> benchan: re-emerging whatever pkg installs verity would probably work around the problem, just curious what lead to you having a broken verity and if there's a bug
[20:19:16] <benchan> davidjames
[20:20:08] <benchan> davidjames: emerged verity, and verity still has the GLIBCXX_3.4.11 tag
[20:20:53] <davidjames> benchan: Hmm, maybe your x86_64 toolchain is broken
[20:21:55] <benchan> davidjames: that may be the case, but I didn't know what triggered that though
[20:23:37] <benchan> davidjames: i did a repo sync as usual this morning and didn't notice any error.  is there some way to rebuild the toolchains beside blowing away the chroot?
[20:25:13] <davidjames> benchan: Recreating the chroot should be pretty fast with cros_sdk
[20:25:32] <benchan> davidjames: let me give it a try. thanks
[20:27:34] *** quiche has quit IRC
[20:30:29] <davidjames> benchan: Also note toolchain is included with fresh cros_sdk, so no need to update toolchain once you get the sdk
[20:32:36] <benchan> davidjames: great, rebuilding chroot, will see how it goes
[20:32:43] <sosa> yeah
[20:32:55] <sosa> if you scroll up from last night's message all the binary builders ran into this issue
[20:33:16] <sosa> i had to clobber them up
[20:34:52] <darinski> 401174
[20:35:14] <darinski> oops, wrong window for my otp :/
[20:35:16] <davidjames> sosa: Do you know what introduced it?
[20:35:32] *** seumas has quit IRC
[20:35:41] *** quiche has joined #chromium-os
[20:35:41] *** ChanServ sets mode: +v quiche
[20:36:16] *** saintlou has quit IRC
[20:36:44] *** patcito has joined #chromium-os
[20:37:10] *** saintlou has joined #chromium-os
[20:37:10] *** ChanServ sets mode: +v saintlou
[20:40:33] <sosa> i had thought the issue was from verity getting an old prebuilt from the full prebuilt binhost after re-building it from a compile dep ... but having the full prebuilt override it with an old dep
[20:40:59] <sosa> all the PFQ's the actually revved verity themselves didn't run into the issue
[20:41:15] *** saintlou has quit IRC
[20:41:44] <sosa> it can also probably be fixed by just re-emerge verity locally
[20:41:48] <sosa> re-emerging*
[20:42:17] *** Keybuk has joined #chromium-os
[20:42:18] *** ChanServ sets mode: +v Keybuk
[20:42:49] <benchan> sosa: re-emerging verity locally didn't seem to fix my local chroot
[20:43:01] <sosa> with --nousepkg?
[20:43:21] <sosa> it's probably a bad binary
[20:43:36] <sosa> from the full builders you're just installing
[20:43:45] <benchan> sosa: i see
[20:43:52] <sosa> (i'm only conjecturing though)
[20:44:49] *** seumas has joined #chromium-os
[20:44:49] *** ChanServ sets mode: +v seumas
[20:56:31] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 034b942065301a66c1f5fe2bb08a46b0711a5a34: ttuttle <ttuttle at chromium dot org>)'
[20:57:09] * ttuttle|work curses.
[20:57:55] <sque> i think it might have been this: http://gerrit.chromium.org/gerrit/#change,5503
[20:58:27] <davidjames> sque: Yup
[20:58:42] <vbendeb> sque: yes, I have already reverted it (thanks to davidjames: head up!) - please reopnen the tree
[20:58:49] <ttuttle|work> how do I see what failure actually caused it to close the tree?
[20:59:44] <davidjames> ttuttle|work: Listed here: http://build.chromium.org/p/chromiumos/builders/x86%20generic%20PFQ/builds/525
[20:59:55] <davidjames> ttuttle|work: Ignore the list at the bottom, look at list under LKGMCandidateSync
[21:00:16] *** rbyers_ has joined #chromium-os
[21:02:11] *** sque has quit IRC
[21:04:27] <crosbot> tree became '?'
[21:04:42] <crosbot> tree became 'Tree is open (Reverted http://gerrit.chromium.org/gerrit/#change,5503)'
[21:05:57] *** zhengda has joined #chromium-os
[21:05:58] *** ChanServ sets mode: +v zhengda
[21:06:21] *** sadrul has joined #chromium-os
[21:06:21] *** ChanServ sets mode: +v sadrul
[21:09:08] *** jrbarnette has joined #chromium-os
[21:09:08] *** ChanServ sets mode: +v jrbarnette
[21:10:22] <gsam_> has lkgm been reset to a working version yet?
[21:13:59] <cmasone> have we managed to roll chrome versions yet?
[21:14:00] *** stevenjb has joined #chromium-os
[21:14:00] *** ChanServ sets mode: +v stevenjb
[21:15:56] *** saintlou has joined #chromium-os
[21:15:56] *** ChanServ sets mode: +v saintlou
[21:16:21] <stevenjb> sosa/davidjames or anyone - remind me which chrome builder(s) produce the binary used for chromeos?
[21:16:51] <cmasone> stevenjb: actually publish the binary pkg, or declare a version of chrome acceptable for inclusion in cros?
[21:17:16] <cmasone> stevenjb: the latter is x86-generic chrome PFQ
[21:17:24] <cmasone> stevenjb: the former, I'm not sure
[21:18:03] <stevenjb> The former I think is what I want? (The version that anyone not building chrome from source would get)
[21:19:18] <cmasone> stevenjb: let's step back for a sec...why do you want to know?  is there a corrupt binary you're getting, or are you looking for what bot decides which version that everyone not building chrome from source gets?
[21:19:38] <stevenjb> The latter.
[21:19:51] <cmasone> stevenjb: then the answer is x86 generic chrome PFQ
[21:20:31] <stevenjb> There were a couple of chrome bugs that passed the bots but prevent anyone developing networking code on ChromeOS.
[21:21:01] <cmasone> stevenjb: yes, that sucked :-/
[21:21:10] <stevenjb> Yes, yes it did.
[21:21:15] <cmasone> stevenjb: are you wondering why we didn't catch that?
[21:21:22] <stevenjb> I was looking to track when a binary might be ready. It looks like the PFQ is not building atm?
[21:21:29] <stevenjb> No, I know exactly why we didn't catch that.
[21:21:52] <stevenjb> I am trying to answer sam with "build from source" or "there will be a good binary soon"
[21:21:55] <cmasone> stevenjb: Ah!  Yah...I don't know why they're not building.  As far as I know, sosa added the needed package to the chrome ebuild yesterday afternoon
[21:21:59] <stevenjb> It looks like the answer is the former.
[21:22:15] <cmasone> stevenjb: but people have kicked the bot a couple times to make it build, and still no dice
[21:22:27] <cmasone> stevenjb: afaik no one is looking at what's going on to try to fix it.
[21:22:50] <stevenjb> It looks like it succeeded at 10:50, #524
[21:22:59] <stevenjb> Then broke again at 11:30.
[21:23:29] <cmasone> stevenjb: you're looking at the ToT chrome PFQ
[21:23:34] <cmasone> http://build.chromium.org/p/chromiumos/builders/x86%20generic%20chrome%20PFQ
[21:24:24] <stevenjb> I'm looking at http://build.chromium.org/p/chromiumos/builders/x86%20generic%20PFQ
[21:24:24] <cmasone> stevenjb: ^^^ that's the one that vets versions of chrome.  I beleive this is discussed in http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium-os/pre-flight-queue-faq
[21:25:00] <stevenjb> This succeeded at 11:20: http://build.chromium.org/p/chromiumos/builders/x86%20generic%20PFQ/builds/524
[21:25:17] <cmasone> stevenjb: yes, that is not the bot you need to be looking at
[21:25:31] <cmasone> stevenjb: x86-generic _chrome_ PFQ
[21:25:37] * stevenjb slams his head into the desk
[21:26:14] <cmasone> stevenjb: perhaps we could rename it to "CHROME the x86-generic CHROME bot that CHROME builds CHROME CHROME CHROME"
[21:26:44] *** rbyers_ has quit IRC
[21:27:13] <stevenjb> or name the otherone x86-generic-PFQ-with-old-chrome-binary
[21:28:11] <stevenjb> And I'm not sure I want to ask the difference between x86-generic-CHROME-pfq and x86-generic-TOT-CHROME-pfq...
[21:28:28] <stevenjb> The latter is, I guess, obvious...
[21:28:30] <cmasone> stevenjb: it's described in the FAQ about PFQs that I linked above
[21:28:31] *** benchan has quit IRC
[21:28:38] *** benchan has joined #chromium-os
[21:28:38] *** ChanServ sets mode: +v benchan
[21:29:42] <stevenjb> "The various Chrome PFQs are similar conceptually to the other PFQs, but function rather differently and serve a different purpose."
[21:29:46] <stevenjb> cool. :)
[21:30:05] <cmasone> stevenjb: :-P
[21:30:57] *** Sarten-X has quit IRC
[21:31:27] <stevenjb> Actually, I do have a constructive suggestion. It sounds like x86-generic-chrome-pfq could be named x86-generic-chrome-nightly-pfq
[21:33:29] <cmasone> stevenjb: SGTM. This is a waterfall config change.  I think if you file a cros issue in Area-Build, someone might take care of it in th enext couple days
[21:34:08] <cmasone> stevenjb: it's basically a string change that you could do ypourself, but you have to checkout the buildbot code and yadda yadda yadda
[21:36:31] *** dgarrett has joined #chromium-os
[21:36:31] *** ChanServ sets mode: +v dgarrett
[21:39:12] *** Sarten-X has joined #chromium-os
[21:40:12] <stevenjb> cmasone: Filed 19616.
[21:41:53] *** benchan has quit IRC
[21:41:58] *** unreal has quit IRC
[21:43:15] <davidjames> vbendeb: FYI pfq failed, might want to close the tree
[21:43:21] <davidjames> vbendeb: http://build.chromium.org/p/chromiumos/builders/x86%20generic%20PFQ/builds/527/steps/BuildTarget/logs/stdio
[21:43:55] <davidjames> (Tree only autocloses if bot *newly* fails)
[21:45:03] <stevenjb> davidjames: Do you have any idea why the *chrome* pfq build isn't building? I was seeing the same error its last build showed yesterday, but not when I built around 9am.
[21:45:50] <davidjames> stevenjb: It's using yesterday's build
[21:45:55] *** jhorwich has joined #chromium-os
[21:45:55] *** ChanServ sets mode: +v jhorwich
[21:46:04] <davidjames> stevenjb: It won't upgrade to new build until someone cuts a new Chrome release
[21:46:12] <stevenjb> Right. right. ouch, my head.
[21:46:22] <stevenjb> OK, is it possible to modify the rev it uses?
[21:46:26] <cmasone> davidjames: that doesn't explain why it's still failing on the kerberos stuff
[21:46:35] <cmasone> davidjames: there was no chrome-side change needed to fix that
[21:46:41] <cmasone> davidjames: as far as I know
[21:50:02] <stevenjb> cmasone: Are you aware of any open issues to improve the vetting of the nightly pfq builds? If not I'd like to open one. (i.e. they should pass some sort of auto test run on a device)
[21:51:04] <davidjames> stevenjb: You could also push the TOT chrome to be stable
[21:51:24] <davidjames> stevenjb: Or you could cut a new release, then it would go through all the bots the normal way
[21:51:47] <cmasone> stevenjb: there is alredy an issue for getting that pfq to run tests on a live device
[21:51:54] <crosbot> tree became 'Tree is closed (pfq failure -> sheriffs)'
[21:52:22] <stevenjb> davidjames: cool. Can you point me to a doc for the first suggestion? That sounds like the quickest way to egt a stable build to developers, since TOT chrome pfq is green.
[21:53:25] <cmasone> stevenjb: davidjames I think that you're ignoring the actual problem.  Sosa removed the pin last night, and updated the ebuild to pull in the new build dep that is causing the breakage
[21:53:32] <cmasone> stevenjb: davidjames that bot SHOULD be green and it is not
[21:55:23] <davidjames> cmasone: Oh really, interesting
[21:55:32] <davidjames> cmasone: Looking in package.keywords on my machine it still has a pin
[21:56:00] <davidjames> cmasone: he pinned, reverted and then repinned, so maybe that is a source of confusion
[21:56:05] <cmasone> davidjames: ah.  yes, talked to sosa and the pin is in there
[21:56:08] <cmasone> davidjames: BUT
[21:56:55] <cmasone> davidjames: we still have this problem where that build of chrome (and ToT chrome) need kerberos, and that is not working for some reason on the bots
[21:57:31] <davidjames> cmasone: Interesting, and where does this kerberos problem show up? TOT builder is green -- does it not test that part?
[21:58:19] <cmasone> davidjames: the requirement for kerberos was summarily reverted while we slept, but one assumes it will land again
[21:58:59] *** zhengda has quit IRC
[21:59:13] <davidjames> cmasone: Interesting, OK. Well, one option would be to allow TOT chrome to be used (which at least passes tests, I don't know whether it really works)
[21:59:25] <crosbot> tree became 'Tree is closed (sque investigating autotest network_3GDormancyDance failure)'
[21:59:33] <sosa> ToT chrome no work with internal builds.
[21:59:42] <davidjames> Ah ok, then I guess that's not an option
[21:59:54] <cmasone> sosa: davidjames roll us a new build dep?
[22:00:02] <cmasone> buildspec
[22:02:48] <stevenjb> sosa: What is the issue with ToT chrome? I am running that fine as of around 10am.
[22:07:10] <cmasone> stevenjb: the issue is that you cannont build Google Chrome without one of these Chrome buildspecs.  Generating a buildspec takes a couple hours, as it needs to pass one of the Official Chrome build and test runs.
[22:07:44] <stevenjb> I see.
[22:08:31] <cmasone> stevenjb: and at that point...the people who need to do testing and stuff will be going home anwyay, so...what are we really gaining?  What we need is to just get it so that we can build tonight's buildspec correctly, and then the problem will have cleared out by tomorrow
[22:09:51] <stevenjb> Yeah. Sucks for devs not building chrome from source and working on networking, but it sounds like we don't have a good solution at this point.
[22:10:09] <crosbot> tree became 'Tree is closed (ttuttle working to resolve)'
[22:10:35] <stevenjb> cmasone: And there are several devs in Moscow that will be affected, so if we can make it happen sooner without a whole lot of trouble might be worth it.
[22:11:33] *** mononoke has quit IRC
[22:11:47] <cmasone> stevenjb: we have an idea, actually.  Last night's build spec is probably OK from your perspective, it just has the kerberos-related build problem
[22:11:50] <cmasone> which we need to fix anyway
[22:12:16] <cmasone> so, if we clear that up, then we can kick the bots and have last night's chrome built, tested and blessed in like an hour
[22:12:17] <stevenjb> cmasone: No, the chrome fixes didn't go in until 2:30 am
[22:12:21] <cmasone> stevenjb: shit
[22:12:46] <cmasone> stevenjb: can you just let those devs know they'll need to build from source...?
[22:13:16] <stevenjb> er, actually, scracth that, that's GMT
[22:13:23] <stevenjb> Last change we need is @ 97999
[22:13:44] <stevenjb> roughly 10pm I think.
[22:14:34] <cmasone> stevenjb: the buildspec was 98014, so I think this will work
[22:14:55] <stevenjb> cool.
[22:15:04] <cmasone> sosa: ?  we need the stuff as a hard-host dep?
[22:15:29] <stevenjb> cmasone: Anything short term I can help with?
[22:15:46] *** saintlou is now known as saintlou|away
[22:15:47] <crosbot> tree became 'Tree is open (fixed autotest syntax error)'
[22:18:13] *** Adys has quit IRC
[22:18:32] *** Adys has joined #chromium-os
[22:18:48] *** ttuttle|work is now known as ttuttle|duncecap
[22:19:17] * ttuttle|duncecap resets whiteboard: "0 days since I last broke the build"
[22:21:11] <ellyjones> it's ok, I broke the factory release signing for R13 and R14
[22:21:20] <crosbot> tree became '?'
[22:21:24] <ellyjones> crosbot: B|
[22:21:25] <crosbot> ellyjones: No such command 'B|'. Try 'help'.
[22:21:35] <crosbot> tree became 'Tree is open (fixed autotest syntax error)'
[22:34:16] *** rbyers_ has joined #chromium-os
[22:42:36] *** ttuttle|duncecap is now known as ttuttle|work
[22:45:50] <davidjames> stevenjb: FYI I just rev'd hard-host-depends, it may happen to coincidentally fix the Chrome builders
[22:46:01] <cmasone> davidjames: what'd you do?
[22:46:18] <davidjames> cmasone: Just bumped hard-host-depends up a rev so that bradnelson's changes to hard-host-depends get included
[22:46:36] <stevenjb> davidjames: That'd be great, thanks.
[22:46:38] <cmasone> davidjames: AH
[22:47:27] <davidjames> stevenjb: So anyway I just kicked the Chrome builders to see if they suddenly start working
[22:47:50] <stevenjb> fingers crossed. Thanks for letting me know.
[22:48:28] <vbendeb> guys, anyone knows what's the deal with tegra2_kaen binary? Irt's been cycling failing gclient sync:
[22:48:31] <vbendeb> http://chromeos-botmaster.mtv.corp.google.com:8026/builders/tegra2_kaen-binary/builds/1649/steps/shell/logs/stdio
[22:49:44] <davidjames> vbendeb: Hmm, can you file a bug for chrome infrastructure team for that?
[22:50:37] <davidjames> vbendeb: Go to http://crbug.com and look for 'Chrome infrastructure team' template
[22:50:39] <vbendeb> I will. Who should I ping to get this looked into? nsylvain?
[22:50:50] <davidjames> vbendeb: Probably scottz@
[22:51:46] <cmasone> vbendeb: per http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium-os "Sometimes, a tree closure is due to some kind of infrastructure problem.  In that case, quickly escalate to troopers at chromium dot org."
[22:52:02] <cmasone> vbendeb: under "What is sheriffing, and why do we do it?"
[22:52:58] <sosa> vbendb:  you should no longer be seeing this
[22:53:14] <sosa> vbendeb: current TOT shows that bot has greenified.
[22:53:33] <sosa> vbendeb: oh wait i see what you mean
[22:56:16] *** dennisjeffrey has quit IRC
[22:57:37] *** unreal has joined #chromium-os
[22:59:30] <vbendeb> davidjames: http://code.google.com/p/chromium/issues/detail?id=94146
[22:59:56] <vbendeb> davidjames: tried pinging scottz, sent email to troopes@ also
[23:02:06] *** cros_ has quit IRC
[23:03:00] *** dennisjeffrey has joined #chromium-os
[23:03:00] *** ChanServ sets mode: +v dennisjeffrey
[23:03:27] <davidjames> stevenjb: OK that failed too, it might be the case that Chrome has a board dependency too
[23:03:31] *** saintlou|away is now known as saintlou
[23:03:39] <davidjames> stevenjb: Just added virtual/krb5 to dependencies, hopefully it works now
[23:03:44] <kliegs> davidjames: can you help me track down where a USE flag comes from?
[23:03:56] <davidjames> kliegs: ferringb has a tool to do that
[23:04:13] <kliegs> ferringb: please help!
[23:04:38] <ferringb> kliegs: which profile you trying to trace the flag from?
[23:04:40] <kliegs> cool.  i've noticed something weird with iptables builds between two different tegra variants. one is built with USE=+ipv6, one with -ipv6.  can't spot from where
[23:04:47] * ferringb hasn't gotten off his ass and released the tool yet; needs to
[23:05:23] <kliegs> ferringb: only problem is i'm using some internal profiles so haven't figured out if the external replicates the behavior
[23:05:33] <davidjames> kliegs: ipv6 is disabled in make.defaults in chromiumos-overlay
[23:05:33] <kliegs> emerge-pv shows the flag being set for one and not the other
[23:06:02] <kliegs> davidjames: from what I can see, the openssh-server init file depends on ipv6 being turned on - it calls ip6tables and fails on seaboard
[23:06:08] <kliegs> which is why I can't ssh to a seaboard
[23:06:20] <kliegs> but I'm using a different tegra variant that for some reason has it turned on
[23:06:58] <davidjames> kliegs: make.defaults:USE="${USE} -acl -cracklib -gpm -ipv6 -openmp -python -sha512"
[23:07:07] <davidjames> kliegs: package.use:net-firewall/iptables       ipv6
[23:07:15] <davidjames> kliegs: Not sure which one wins that battle :)
[23:07:22] <ferringb> depends on ordering
[23:07:28] <kliegs> davidjames: any chance its non-deterministic?
[23:07:53] <kliegs> and which directory/repo is that in?
[23:08:13] <davidjames> kliegs: ~/trunk/src/third_party/chromiumos-overlay/profiles/targets/chromeos
[23:08:59] <kliegs> ok. i see it.  is it possible that its a weird race and one turns it on and the other off?
[23:09:01] *** rharrison_chrome has quit IRC
[23:09:32] * ferringb notes it's deterministic; no race
[23:09:45] <kliegs> hmm. ok.
[23:09:45] <ferringb> it's just nonobvious stacking
[23:10:05] <ferringb> would need to see the profile tree to really answer your question
[23:10:19] <ferringb> davidjames: also, who do I bribe for a cool +v in here? ;)
[23:11:25] <ellyjones> do you have a chromium.org account?
[23:11:48] <davidjames> ellyjones: Yeah he's ferringb at chromium dot org
[23:12:37] <ellyjones> done
[23:12:43] *** ChanServ sets mode: +v ferringb
[23:12:45] <kliegs> ferringb: unfortunately I can't share the full tree, sorry.
[23:12:47] <ferringb> wheee
[23:13:16] <ferringb> kliegs: yep, figured.  watch for 'pinspect profiles'; I'll send a PSA when I finally kick the thing out the door
[23:13:38] <ferringb> still need to finish some validation on it before I'm willing to let others play with it; purpose of it is for general profile inspection, things like this
[23:14:50] <kliegs> ferringb: no problem.  will be handy to have and looking forward to it.  always wished I could add 'echo foo' stuff into the make.conf and equivalents to stack trace
[23:15:54] * ferringb nods
[23:16:59] <kliegs> Does someone have a seaboard checkout they can check something on for me?
[23:17:15] <ellyjones> seaboard specifically or any t2?
[23:17:42] <kliegs> actually any could be useful
[23:17:53] <ellyjones> ask away
[23:18:04] <kliegs> what does 'emerge-foo -pv iptables' output for the USE= bit?
[23:19:11] <ellyjones> USE="ipv6" for 1.4.8-r1
[23:19:39] <kliegs> ok.  that's the right response. my seaboard build is giving USE=-ipv6
[23:19:44] <kliegs> which is causing openssh-server to fail
[23:19:56] <ellyjones> :nfi:
[23:20:05] <davidjames> stevenjb: Bad news, krb5 doesn't compile for arm, any way you can fix that?
[23:20:31] <stevenjb> davidjames: I'm not sure I would even know where to begin.
[23:21:00] <davidjames> stevenjb: Well that's blocking PFQ from going green with new Chrome, see http://build.chromium.org/p/chromiumos/builders/arm%20generic%20chrome%20PFQ/builds/87/steps/BuildTarget/logs/stdio
[23:21:28] <davidjames> stevenjb: For now I reverted krb5 dependency but that means we have to wait for Chrome roll
[23:21:42] <davidjames> stevenjb: (I.e. until Chrome stops depending on krb5)
[23:21:52] <ellyjones> hm, that error is sensible
[23:22:15] <stevenjb> davidjames: Is that dependency actively being removed from chrome?
[23:22:40] * stevenjb looks up kbr5
[23:22:59] <cmasone> stevenjb: yeah, pavel said he took it out
[23:23:13] <stevenjb> So it's already gone in ToT?
[23:23:19] <cmasone> stevenjb: believe so
[23:23:30] <cmasone> stevenjb: but that's second hand info
[23:23:44] <stevenjb> pavel@ is the person to contact about that?
[23:24:09] <davidjames> stevenjb: phajdan
[23:25:15] <davidjames> stevenjb: Yeah he says he disabled Kerberos for Chrome OS
[23:26:22] <stevenjb> ok, cool. Do we know where and when that disable occurred?
[23:27:14] <davidjames> stevenjb: http://src.chromium.org/viewvc/chrome?view=rev&revision=98116
[23:28:12] <stevenjb> cmasone: OK, so, what would need to happen to get a Chrome buildspec >= 98116 instead of @98014?
[23:29:12] <davidjames> stevenjb: Actually, just checked, >=98028 is all you need
[23:29:32] <davidjames> stevenjb: 98116 is where the feature that required kerberos was re-added (and hopefully doesn't break chrome os, but was not tested yet)
[23:29:45] <stevenjb> which, unfortunately is still > 98014
[23:31:13] *** benchan has joined #chromium-os
[23:31:13] *** ChanServ sets mode: +v benchan
[23:31:33] <davidjames> stevenjb: full history is in http://code.google.com/p/chromium/issues/detail?id=92689
[23:32:15] <davidjames> stevenjb: I talked to cmasone / sosa and they thought it made sense just to wait for next Chrome roll
[23:32:49] <stevenjb> davidjames: Seems like it at this point. What's the approx expected eta on that?
[23:35:35] <davidjames> stevenjb: I think it happens each morning around 2:30am
[23:37:38] <stevenjb> k
[23:53:45] <ellyjones> test_build_root: Missing libcros headers.
[23:53:47] <ellyjones> "??"
[23:55:45] <stevenjb> ?
[23:55:47] *** saintlou has quit IRC
[23:56:00] <stevenjb> ellyjones: link to log?
[23:56:23] <ellyjones> that was local
[23:56:29] <ellyjones> when running build_image
[23:56:39] <stevenjb> emerge libcros?
[23:56:50] <ellyjones> I feel like I shouldn't need to, but ok :)
[23:56:55] <stevenjb> oh, I'll bet you did a gmerge of libcros?
[23:57:38] <stevenjb> Cleverly, the gmerge "cleans up" the headers it produces, so a chrome gmerge afterwards fails. You need to emerge libcros to rebuild the headers.
[23:57:48] <stevenjb> I'm pretty sure somewhere there is an issue filed for that...
[23:58:05] *** dennisjeffrey has quit IRC
[23:58:24] <cmasone> stevenjb: there is.  But I too forget which one
[23:58:32] *** saintlou has joined #chromium-os
[23:58:32] *** ChanServ sets mode: +v saintlou
[23:59:23] <stevenjb> Hm. 12788 was closed.
[23:59:28] <stevenjb> er, 12677





top