[00:02:33] <dennisjeffrey__> reopen?  Seems to be crosbug.com/20323 again
[00:03:51] <crosbot> tree became 'Tree is open (seems to be crosbug.com/20323 again :-\)'
[00:07:37] *** Solet has quit IRC
[00:07:50] <kliegs> dennisjeffrey__: can you put some pressure back on Zel?  this bug has been showing up more and more lately
[00:14:24] <dennisjeffrey__> kliegs: i think this issue is currently being investigated by xiyuan and cmasone, but there isn't an explanation yet
[00:15:04] <dennisjeffrey__> kliegs: recommendations for effective ways to put pressure? :-)
[00:15:08] <kliegs> dennisjeffrey__: ok.  just thinking it might need more attention as I was seeing it last week as sheriff and is still there
[00:15:20] <kliegs> dennisjeffrey__: walk over and talk to him. i'm in the wrong building :)
[00:16:19] <kliegs> getting a relevant PM, TL or TLM involved can also help as they can better assess priorities).  and probably worth mentioning how many times the trees been closed for this bug (so make sure to bump the bug with "closed trree xxx times today" at the end of each day
[00:17:02] <kliegs> in the end they may already be working hard - I haven't been tracking this - so it may not be needed.  just good to make sure that they realize the impact its having
[00:17:57] <dennisjeffrey__> kliegs: good ideas!  Yes, there is already a lot of recent work on this issue by xiyuan, but it's good to keep the pressure on
[00:18:48] <kliegs> excellent. glad to ehar
[00:18:52] <kliegs> *hear
[00:19:12] <marcheu> kliegs: I'm going to try the whole thing now with leaving the old ebuilds around, that should work for people who ended up using those
[00:19:41] <kliegs> marcheu: there should really only ever be a single set of stable ebuilds checked in
[00:19:54] <kliegs> marcheu: if the new ones you checked in aren't intended to be used yet, they should all be marked as unstable
[00:19:57] <marcheu> why so? how do I transition otherwise
[00:20:09] <kliegs> marcheu: they should all be unstable until its time to make them all stable
[00:20:12] <marcheu> you do know we use the gentoo overlay right
[00:20:19] <marcheu> so there are a lot of unmasked ebuilds
[00:20:32] <kliegs> marcheu: the highest stable ebuild is what we should build from
[00:20:43] <marcheu> I don't see why not, it would allow smooth transition
[00:20:48] <kliegs> if the highest one isn't intended to be used yet (like xlsclients-1.2) it should be marked untable
[00:21:12] <kliegs> marcheu: because having an ebuild marked as stable which won't build due to a missing dependency forces the build to backtrack to find the correct version
[00:21:17] <kliegs> this has two problems
[00:21:20] <kliegs> 1 - it makes the build take longer
[00:21:21] <marcheu> well given that you guys are already using a different version than us, I think the problem is already there
[00:21:26] <marcheu> so at least lets try to solve it smoothly
[00:21:39] <marcheu> and leaving the old ebuilds around is a solution
[00:22:01] <kliegs> 2 - there is a limit to the backtracking done.  depending on the overlays used, its possible to exceed it. which is why relying on backtracking isn't good
[00:22:13] <marcheu> well try proposing a solution then
[00:22:14] <kliegs> who is "you guys" in this case?  as far as I know we're building the same version as everyone else
[00:22:20] <marcheu> I have a solution here
[00:22:36] <kliegs> my proposal is that the ebuilds remain unstable until they are ready
[00:22:51] <marcheu> then we can't update the WAT people who are using a higher xlsclients
[00:22:53] <kliegs> should I have been building from xlsclients-1.1 or 1.0
[00:22:58] <marcheu> 1.0
[00:23:05] <kliegs> we are trying to use 1.0
[00:23:06] <marcheu> because that means we downgrade them, which doesn't work
[00:23:27] <kliegs> we're using the same version everywhere. and if you look at the people above it wasn't only waterloo experiencing the problem
[00:23:44] <marcheu> I don't know, I have yet to see a single log
[00:23:48] <kliegs> ellyjones isn't in waterloo and saw it. quiche is somewhere else .
[00:23:58] <marcheu> it's like shots in the dark for me
[00:24:09] *** Solet has joined #chromium-os
[00:24:16] <kliegs> the problem we were reporting isn't that we wanted to use xlsclients-1.1.  Its that portage was forcing us to because you marked it as stable
[00:24:27] <kliegs> an ebuild shouldn't be checked in as stable unless its intended that people use it
[00:24:38] <kliegs> check the ebuilds in as unstable and unmask them for whatever testing or profiles you need
[00:24:48] *** thieule has joined #chromium-os
[00:24:48] *** ChanServ sets mode: +v thieule
[00:25:26] <marcheu> well that was the case for 1.1.0 ebuild, so now lets leave it around until everyone is on 1.1.1
[00:26:01] <marcheu> otherwise I don't see a path out of this
[00:26:20] <marcheu> (in other news, there are many ebuilds which are on different versions on ARM and x86, which wouldn't be possible if we had the policy you describe)
[00:26:22] <kliegs> i'm not sure what you're actually trying to accomplish
[00:26:39] <marcheu> eventually, upgrade X
[00:26:43] <kliegs> ok. but right now?
[00:26:56] <marcheu> upgrade X's dependencies that live in portage-stable
[00:27:03] <kliegs> as for the policy it does hold.  the highest version of each ebuild should be marked stable for the appropriate platform
[00:27:19] <marcheu> looking at the portage spreadsheet that's not the case
[00:27:20] <kliegs> in cases where that's not possible through the ebuild the portage package mask files are used
[00:27:32] <marcheu> it sounds like that policy isn't enforced, or doesn't exists
[00:27:35] <kliegs> I'm sure there are spots its not done properly.
[00:28:00] <kliegs> but that doesn't mean we should continue to check in things which can cause problems.
[00:28:17] *** Solet has quit IRC
[00:28:22] <marcheu> how do I upgrade those dependencies then?
[00:28:22] <marcheu> if I can't get you out of it
[00:28:31] <marcheu> right now, you provide no log, no solution
[00:28:46] <kliegs> i understand eventually upgrade X - but the CL you actually checked in - what was its intent?  was it to make all the builds use the new version?
[00:28:54] <marcheu> no
[00:29:12] <marcheu> it's so that I can upgrade things in chromiumos-overlay next
[00:31:05] <kliegs> is there a reason they can't be checked in as unstable?
[00:31:22] <marcheu> when I check things into chromiumos-overlay, I'll need them
[00:31:35] <kliegs> or would that break your next step?  and what things do you mean?  i'm trying to understand what you're trying to accomplish overall to help lead to a solution
[00:32:03] <kliegs> are the pieces from chromiumos-overlay ready?  do you expect to check that in today?  next week? next month?
[00:32:07] <marcheu> next step is basically bring xcb and friends up to speed
[00:32:14] <marcheu> yeah in a while, packaging takes time
[00:32:19] <marcheu> maybe week-ish
[00:33:02] <kliegs> I'd recommend checking them in as unstable then. and for your testing enabling them (either through a package.mask or just modifying the ebuilds locally).  and get all the ebuilds staged as unstable
[00:33:13] <kliegs> once you've got the whole thing ready, then set them all to stable so everything gets upgraded together
[00:33:33] <marcheu> that sounds like a doomsday scenario
[00:33:41] *** saintlou has quit IRC
[00:33:47] <kliegs> portage will not always downgrade to a lower version if a dependency is missing - and that failure to do so can be different on different environments
[00:33:56] <kliegs> its the same effect as what you're doing now isn't it? just more controlled?
[00:33:58] <marcheu> I was trying to upgrade things separately, so as to avoir breaking everything at once
[00:34:04] <kliegs> right now you're rolling out a lot of new ebuilds that no one should be using
[00:34:12] <marcheu> no, right now some of the ebuilds I added were used, some others weren't
[00:34:17] <kliegs> and then plan on upgrading one last package that makes them all turn on
[00:34:23] <kliegs> in that case you should probably split your CL
[00:34:32] <kliegs> check in the ebuilds as stable that should be used
[00:34:37] <kliegs> then a second CL to checkin the unstable ones
[00:35:00] <marcheu> ah well, I've already spent a lot of time on this today
[00:35:04] <marcheu> I guess that can wait a couple of days
[00:35:53] <kliegs> yah.  we ended up spending a bit of time as well
[00:36:16] <kliegs> I'd also recommend adding a TEST=clause to your CL - was surprised Mike +2'd it without one
[00:36:32] <kliegs> I've checked in bad CL's before and others seeing what I did to test it saw quickly why it broke in one situation but not from my own tests
[00:38:28] <marcheu> well, I own about 100 chrome OS packages. I guess if I do ~1 CL a day, we'll have an X update in january
[00:38:57] <marcheu> it's just not possible
[00:39:52] <kliegs> what isn't possible? and why do you need to update 1 at a time?
[00:40:05] <marcheu> if you start breaking the update in small chunks
[00:40:43] *** saintlou has joined #chromium-os
[00:40:53] <kliegs> It sounds like you had a reasonable split - packages everyone should be using the new version of and packages they shouldn't
[00:41:01] <kliegs> that's two chunks
[00:41:04] *** ChanServ sets mode: +v saintlou
[00:41:15] <marcheu> that CL is only one of the CLs that I'm going to need
[00:41:21] <marcheu> we're still pretty far from a solution
[00:41:23] <marcheu> but, ah well
[00:42:30] <kliegs> creating multiple CL's isn't that hard.  I've also done it in the past (mainly when i had to split repos but occasionally within a repo) and would batch the testing
[00:42:54] <kliegs> I'm pretty sure that's accepted (at least the reviewers let me do it and I've done the same for others) so the overhead isn't that high.
[00:43:16] <kliegs> I'd make a bug to link all the changes together so there was a single point you could see everything
[00:43:53] <kliegs> so a few minutes to make the separate CL's.  a single test run to test them as a unit. and then the reviewers would review them as a group.  it wasn't that much extra work for me
[00:47:11] *** wfrichar has quit IRC
[00:49:16] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 619c91c1b4e13d416304947e100041770ab4968a: Noah Richards <noahric at chromium dot org>)'
[00:51:03] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 619c91c1b4e13d416304947e100041770ab4968a: Noah Richards <noahric at chromium dot org> --> sheriffs investigating)'
[00:54:58] <dennisjeffrey__> suite_Smoke/desktopui_WindowManagerFocusNewWindows again
[00:56:57] <crosbot> tree became 'Tree is open (seems to be crosbug.com/20323 again :-\ again :-\)'
[01:02:56] <kliegs> anyways I need to take off. 7 PM here.  Probably won't be active on irc or email as much tomorrow but feel free to get me by chat/phone if you want to discuss more.
[01:05:24] *** mkunze has quit IRC
[01:22:35] *** tbroch has quit IRC
[01:24:19] *** vmil86 has quit IRC
[01:30:43] *** rbyers_ has joined #chromium-os
[01:37:03] *** tbroch has joined #chromium-os
[01:37:03] *** ChanServ sets mode: +v tbroch
[01:46:43] *** Sergiu has quit IRC
[01:46:55] *** Sergiu has joined #chromium-os
[01:52:30] *** Sergiu has quit IRC
[02:04:19] *** rbyers_ has quit IRC
[02:08:42] *** lipsinV2 has joined #chromium-os
[02:24:27] *** rbyers_ has joined #chromium-os
[02:24:56] *** dalecurtis has quit IRC
[02:38:45] <Keybuk> davidjames: we don't use sys-kernel/linux-headers except in the host though, right?
[02:40:41] <davidjames> Keybuk: Yup
[02:40:50] <davidjames> But you're right, linux-headers is fine
[02:43:13] <davidjames> Just concerned because our snapshot has a lot of references to dev-manager
[02:44:31] <Keybuk> well, with my update installed, if I do ...
[02:44:47] <Keybuk> $ emerge -pqv -v virtual/dev-manager
[02:44:49] <Keybuk> I get nothing
[02:45:04] <Keybuk> err
[02:45:06] <Keybuk> $ emerge -pqv -c virtual/dev-manager
[02:45:08] <Keybuk> obviously
[02:48:10] <Keybuk> I'm going on the theory if we pull anything new in from our snapshot that requires it, it'll fail, and at least force them to grab something newer from cros/gentoo into portage-stable first
[02:50:02] *** Keybuk has quit IRC
[02:56:30] *** achuith has joined #chromium-os
[02:56:30] *** ChanServ sets mode: +v achuith
[02:58:54] *** Sarten-X has quit IRC
[03:02:13] <dennisjeffrey__> MTV sheriffs going off duty for the day; back at 8am tomorrow
[03:06:56] *** dennisjeffrey has quit IRC
[03:07:11] *** Sarten-X has joined #chromium-os
[03:08:15] *** sadrul has joined #chromium-os
[03:08:15] *** ChanServ sets mode: +v sadrul
[03:19:28] <grundler> ./scripts/Lindent: line 3: indent: command not found
[03:19:47] * grundler sighs...drawback of working in a chroot
[03:30:10] <dhendrix> grundler:  sudo emerge dev-util/indent?
[03:31:05] *** saintlou has quit IRC
[03:32:44] *** saintlou has joined #chromium-os
[03:32:45] *** ChanServ sets mode: +v saintlou
[03:32:46] *** Sarten-X has quit IRC
[03:39:37] *** Sarten-X has joined #chromium-os
[03:50:46] *** saintlou has quit IRC
[04:11:36] *** rbyers_ has quit IRC
[04:11:44] *** rbyers_ has joined #chromium-os
[04:11:44] *** ChanServ sets mode: +v rbyers_
[04:17:58] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" from 960ae9887ec3611678e658c6923046aaaef134d1: Chris Sosa <sosa at chromium dot org>)'
[04:19:32] <grundler> dhendrix: oh...doh. I keep forgetting it's really gentoo :)
[04:36:28] <crosbot> tree became 'Tree is open (crosbug.com/19198 - Fix for specific instance submitted)'
[04:50:02] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb_he canary" from None: )'
[05:00:01] <crosbot> tree became 'Tree is open (crosbug.com/19005)'
[05:05:08] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 10af6d3cb034f5c6f1f8d9e709df57ed1c637df7: Chris Sosa <sosa at chromium dot org>)'
[05:16:38] *** sque1 has quit IRC
[05:19:43] <crosbot> tree became 'Tree is open (Fix already in.)'
[05:25:20] <crosbot> tree became 'Tree is open (Fix already in for getupdate failure in cros_run_vm_update)'
[05:25:35] <crosbot> tree became 'Tree is open (Fix already in for getopt failure in cros_run_vm_update)'
[05:29:10] *** rbyers_ has quit IRC
[05:38:39] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex-binary" from 25bccb73925acbd4c32d99ceb8f74e907429b857: Chris Sosa <sosa at chromium dot org>)'
[05:49:26] *** sadrul has quit IRC
[06:01:54] <crosbot> tree became 'Tree is open (Fix already in for getopt failure in cros_run_vm_update)'
[06:09:27] *** FusionX has quit IRC
[06:14:05] *** FusionX has joined #chromium-os
[06:15:33] *** TW1920 has joined #chromium-os
[06:31:03] *** FusionX has quit IRC
[06:32:51] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy-binary" from 0dce056181b3675d2b266232808284e512fba09d: Chris Sosa <sosa at chromium dot org>)'
[06:35:36] *** FusionX has joined #chromium-os
[06:39:50] *** FusionX has quit IRC
[06:40:31] <crosbot> tree became 'Tree is open (Fix already in for getopt failure in cros_run_vm_update)'
[06:42:06] *** FusionX has joined #chromium-os
[07:10:45] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" from 0dce056181b3675d2b266232808284e512fba09d: Chris Sosa <sosa at chromium dot org>)'
[07:11:47] <crosbot> tree became 'Tree is open (Fix already in for getopt failure in cros_run_vm_update)'
[07:21:41] *** magn3ts has joined #chromium-os
[07:21:54] *** unreal has quit IRC
[07:31:54] *** aaronp has quit IRC
[07:58:35] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 generic full" from cdffbd104d66ad2319364ae2e841389610116c2d: Chris Sosa <sosa at chromium dot org>)'
[08:03:58] <crosbot> tree became 'Tree is open (x86 generic full -> 20323 )'
[08:08:02] *** Honoome has joined #chromium-os
[08:24:43] *** unreal has joined #chromium-os
[08:37:59] *** Honoome has quit IRC
[08:38:21] *** Honoome has joined #chromium-os
[09:13:39] *** ttuttle|work has quit IRC
[09:22:16] *** silverroots has joined #chromium-os
[09:37:55] *** patcito has quit IRC
[10:05:13] *** shankar_ has quit IRC
[10:05:41] *** shankar_ has joined #chromium-os
[10:20:01] *** rosso_ has joined #chromium-os
[10:36:55] *** silverroots has quit IRC
[11:34:44] *** Sergiu has joined #chromium-os
[11:42:25] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from ccb5e36475dfbdbfdc7bfeba2ca7b8f5dc77796f: Luigi Semenzato <semenzato at chromium dot org>)'
[11:49:50] <crosbot> tree became 'Tree is closed (x86 pineview full -> 20323)'
[11:50:05] <crosbot> tree became 'Tree is open (x86 pineview full -> 20323)'
[11:55:13] *** Honoome has quit IRC
[12:11:30] *** Adys has quit IRC
[12:15:06] *** Adys has joined #chromium-os
[12:22:30] *** Inumedia has quit IRC
[12:29:08] *** Inumedia has joined #chromium-os
[12:35:18] *** Inumedia has quit IRC
[12:35:56] *** Inumedia has joined #chromium-os
[14:38:26] *** jochen__ has quit IRC
[14:38:32] *** jochen__ has joined #chromium-os
[14:38:32] *** ChanServ sets mode: +v jochen__
[14:46:09] *** rbyers_ has joined #chromium-os
[14:46:28] *** rbyers_ has joined #chromium-os
[14:50:15] *** rharrison_chrome has joined #chromium-os
[14:50:15] *** ChanServ sets mode: +v rharrison_chrome
[14:55:47] *** mkunze has joined #chromium-os
[15:31:41] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard canary" from None: )'
[15:35:30] *** Adys has quit IRC
[15:45:54] *** shankar_ has quit IRC
[15:51:12] *** Adys has joined #chromium-os
[16:14:47] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_seaboard canary" from None: -> sheriff investigating)'
[16:17:01] <rjkroege> is anybody investigating the seaboard canary fail?
[16:19:22] <rjkroege> tbroch: what did you discover about the build bot?
[16:19:26] <rjkroege> it looks green now?
[16:20:34] *** sadrulhc has joined #chromium-os
[16:20:34] *** ChanServ sets mode: +v sadrulhc
[16:21:06] <ellyjones> holy crap, crosbot :P
[16:29:04] <rjkroege> dennisjeffrey: tbroch: i asked petermayo: about the build failure on seaboard and he mumbled something I don't understand about version changes and needing a trooper
[16:29:27] <rjkroege> and that scottz: would know better
[16:37:30] <crosbot> tree became 'Tree is open (tbroch: looks like archive checking issue filing bug)'
[16:41:23] *** rosso_ has quit IRC
[16:41:51] <crosbot> tree became 'Tree is open (tbroch: looks like archive checking issue crosbug.com/20944)'
[16:46:39] *** rbyers_ has quit IRC
[16:48:21] *** ChanServ sets mode: +v vapier
[16:53:10] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-mario canary" from None: )'
[16:54:18] *** rbyers_ has joined #chromium-os
[16:54:52] *** aaronp has joined #chromium-os
[16:55:26] *** Sergiu has quit IRC
[16:57:11] <rjkroege> build failure is the same as the previous failure. opening tree
[16:58:18] <crosbot> tree became 'Tree is open (crosbug.com/20944)'
[17:05:55] <dennisjeffrey__> MTV sheriff reporting for duty
[17:13:45] *** cooled has quit IRC
[17:15:13] *** cooled has joined #chromium-os
[17:23:24] <ellyjones> I wish the recovery ui worked properly on kaen :P
[17:23:29] * ellyjones can't tell if her test is succeeding or not
[17:32:22] <tbroch> any troopers out there?
[17:32:51] <tbroch> please have a look at crosbug.com/20944
[17:35:45] *** dkrahn has left #chromium-os
[17:51:19] *** rbyers_ has quit IRC
[18:00:55] *** dennisjeffrey has joined #chromium-os
[18:00:59] *** ChanServ sets mode: +v dennisjeffrey
[18:18:07] *** jochen__ has quit IRC
[18:19:12] *** gman1 has joined #chromium-os
[18:23:12] *** jochen__ has joined #chromium-os
[18:23:12] *** ChanServ sets mode: +v jochen__
[18:23:51] *** silverroots has joined #chromium-os
[18:25:48] *** lipsinV2 has quit IRC
[18:28:06] <ellyjones> my change http://gerrit.chromium.org/gerrit/#change,8288 does not seem to be entering the commit queue
[18:30:23] <ellyjones> 2011/09/28 08:59:36 - validation_pool.py - DEBUG   : Some changes could not be applied cleanly.
[18:30:27] <ellyjones> 2011/09/28 08:59:36 - validation_pool.py - INFO    : Change ellyjones:8288 did not apply cleanly.
[18:30:30] <ellyjones> hmm.
[18:31:39] <ellyjones> INFO: RunCommand: ['git', 'rebase', '--strategy', 'resolve', '-X', 'trivial', '--onto', 'patch_branch', u'cros/master', 'FETCH_HEAD'] in /b/cbuild/src/scripts
[18:31:42] <ellyjones> First, rewinding head to replay your work on top of it...
[18:31:45] <ellyjones> error: Merge requires file-level merging
[18:31:48] * ellyjones rebase
[18:35:44] <grundler> The authenticity of host '[gerrit.chromium.org]:29418 ([74.125.248.80]:29418)' can't be established.
[18:35:48] <grundler> RSA key fingerprint is d0:a8:99:57:f8:bc:1c:ec:c5:57:d8:57:6c:58:2d:b5.
[18:35:52] <grundler> uhm...how do I verify that?
[18:36:40] <ellyjones> did your kerberos tickets expire? that happens to me when ssh can't read my known_hosts
[18:37:26] *** shankar has joined #chromium-os
[18:38:40] <ellyjones> I have d0:a8:99:57:f8:bc:1c:ec:c5:57:d8:57:6c:58:2d:b5 for gerrit.chromium.org:29418, according to `ssh-keygen -l -f ~/.ssh/known_hosts`
[18:39:08] *** silverroots has quit IRC
[18:43:55] *** BladeFreak has joined #chromium-os
[18:48:34] *** patcito has joined #chromium-os
[18:49:23] *** danakj has quit IRC
[18:53:36] *** jennb has quit IRC
[18:54:51] *** gman1 has quit IRC
[18:57:01] *** rharrison_chrome has quit IRC
[18:58:56] *** saintlou has joined #chromium-os
[18:58:56] *** ChanServ sets mode: +v saintlou
[19:01:20] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy-binary" from af26631333984d187d7e365fd94fc21461555912: Taylor Hutt <thutt at chromium dot org>)'
[19:01:52] <dennisjeffrey__> failed: suite_Smoke/login_BadAuthentication
[19:02:05] *** jennb has joined #chromium-os
[19:02:06] *** ChanServ sets mode: +v jennb
[19:03:34] <dennisjeffrey__> "Timed out waiting to revert DNS"; I think it's this issue again crosbug.com/19005
[19:03:42] <dennisjeffrey__> will reopen tree
[19:04:59] <crosbot> tree became 'Tree is open ("Timed out waiting to revert DNS" --> crosbug.com/19005)'
[19:06:16] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from beb5466ec8e87ba5a87534ffa106e02ab1f2d81c: Paul Taysom <taysom at google dot com>)'
[19:07:36] <dennisjeffrey__> 2 smoke suite failures
[19:08:47] *** rbyers_ has joined #chromium-os
[19:08:48] *** ChanServ sets mode: +v rbyers_
[19:08:52] <dennisjeffrey__> (1) suite_Smoke/login_CryptohomeUnmounted --> i think crosbug.com/20323 again
[19:09:46] <dennisjeffrey__> (2) suite_Smoke/security_ProfilePermissions.login --> i think crosbug.com/20887 again
[19:09:50] <dennisjeffrey__> will reopen tree
[19:10:53] <crosbot> tree became 'Tree is open (crosbug.com/20323, crosbug.com/20887)'
[19:13:39] *** rbyers_ has quit IRC
[19:20:12] <davidjames> kvm 0.14.1 is being rolled out to all buildbots now, this should fix bug crosbug.com/20305
[19:20:50] <davidjames> note that one side effect of the rollout is a one-time reboot, that'll happen after each bot on external tree finishes its work, and may cause the bot to go purple for a bit
[19:20:57] <davidjames> crosbot: sheriffs?
[19:20:57] <crosbot> davidjames: sheriffs: dennisjeffrey, tbroch, rjkroege
[19:22:18] <tbroch> thanks for headsup david
[19:23:25] <davidjames> the x86-generic-full and commit queue purpleness interrupts were due to the above rollout, more purple should be coming soon
[19:25:51] *** rbyers_ has joined #chromium-os
[19:26:06] *** ChanServ sets mode: +v rbyers_
[19:34:10] *** rbyers_ has quit IRC
[19:37:00] *** rbyers_ has joined #chromium-os
[19:43:16] *** danakj has joined #chromium-os
[19:43:16] *** ChanServ sets mode: +v danakj
[19:47:52] *** Keybuk has joined #chromium-os
[19:47:52] *** ChanServ sets mode: +v Keybuk
[19:48:50] *** rbyers_ has quit IRC
[19:50:44] *** ttuttle|work has joined #chromium-os
[19:51:49] *** sakoman__ has joined #chromium-os
[19:52:05] *** sakoman_ has quit IRC
[19:53:50] *** saintlou has quit IRC
[19:54:39] *** jennb has quit IRC
[19:59:06] *** saintlou has joined #chromium-os
[19:59:07] *** ChanServ sets mode: +v saintlou
[20:01:02] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy-binary" from eb300c59634e8504d3e28fce7b9992fe12c058e7: Elly Jones <ellyjones at chromium dot org>)'
[20:01:12] <ellyjones> hmm
[20:01:41] *** saintlou has quit IRC
[20:01:42] <ellyjones> Timed out waiting to revert dns
[20:02:04] <dennisjeffrey__> crosbug.com/19005 again
[20:02:12] *** rbyers_ has joined #chromium-os
[20:02:58] <crosbot> tree became 'Tree is open ("Timed out waiting to revert DNS" --> crosbug.com/19005)'
[20:08:22] *** sakoman has joined #chromium-os
[20:08:34] *** sakoman__ has quit IRC
[20:12:07] *** rharrison_chrome has joined #chromium-os
[20:12:07] *** ChanServ sets mode: +v rharrison_chrome
[20:13:38] <ttuttle|work> hey guys, how do I find whether a certain commit of mine made it in to a particular (shipped) version of cros?
[20:14:56] <ellyjones> ttuttle|work: look at the manifest for that version
[20:14:57] <tbroch> ttuttle: can try http://chromeos-images.corp.google.com/diff/
[20:14:57] <davidjames> marcheu: Prepare for Tree Closure: http://build.chromium.org/p/chromiumos/builders/x86%20generic%20PFQ/builds/1058/steps/BuildTarget/logs/stdio
[20:15:40] 
[20:16:28] <ellyjones> !!! The ebuild selected to satisfy "media-sound/adhd" for /build/tegra2_kaen/ has unmet requirements.
[20:16:31] <ellyjones> that's novel
[20:16:33] <ellyjones> (local build)
[20:16:35] <davidjames> tbroch: transfer closed with 1905131 bytes remaining to read, and then could not be resume
[20:16:40] <ttuttle|work> davidjames: sounds familiar
[20:17:35] <ttuttle|work> ellyjones: where do I get those?
[20:17:49] <dennisjeffrey__> maybe this? http://code.google.com/p/chromium-os/issues/detail?id=20828
[20:18:22] <davidjames> dennisjeffrey__: Yes, but it will fail again I expect because we don't have any verification that download was correct
[20:19:11] <dennisjeffrey__> marcheu: should we revert http://gerrit.chromium.org/gerrit/#change,8404 to get the tree green while we figure things out?
[20:19:22] <dennisjeffrey__> (assuming that's will make the tree green...)
[20:19:26] <ellyjones> ttuttle|work: hm, I actually can't remember
[20:19:29] <ellyjones> ask trond
[20:19:31] <marcheu> dennisjeffrey__: hmm you should fix the infrastructure instead
[20:19:39] <marcheu> it's not my fault
[20:20:16] <dennisjeffrey__> marcheu: let's figure out who to get on it and how long it will take
[20:20:19] <dennisjeffrey__> troopers?
[20:20:49] <davidjames> dennisjeffrey: Just rm'd the file
[20:21:12] <davidjames> dennisjeffrey: So it will download again hopefully well this time
[20:21:51] <dennisjeffrey__> davidjames: thanks - so should we reopen and see if it's green on the next cycle?
[20:22:09] <davidjames> dennisjeffrey__: Yes, if PFQ is only reason to close the tree
[20:22:42] <davidjames> dennisjeffrey__: Canaries look pretty red, did you check those are all being worked on?
[20:24:21] <dennisjeffrey__> davidjames: I believe canaries are red due to this: http://code.google.com/p/chromium-os/issues/detail?id=20944
[20:24:27] <dennisjeffrey__> tbroch: correct me if i'm wrong
[20:24:41] <tbroch> dennisjeffery__: you're correct
[20:24:50] <dennisjeffrey__> will reopen the tree
[20:25:20] <davidjames> Thanks :)
[20:26:24] <crosbot> tree became 'Tree is open (download problem --> crosbug.com/20828, should cycle green)'
[20:44:18] *** dianders has joined #chromium-os
[20:44:19] *** ChanServ sets mode: +v dianders
[20:47:05] <dennisjeffrey__> tree is going red soon
[20:47:19] <dennisjeffrey__> bunch of VMTest failures: "DBusException: org.freedesktop.DBus.Error.ServiceUnknown: The name org.chromium.SessionManager was not provided by any .service files"
[20:48:02] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from f48633a62658ba003ca5a554545c9685a881f37f: Scott Zawalski <scottz at google dot com>)'
[20:48:48] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from f48633a62658ba003ca5a554545c9685a881f37f: Scott Zawalski <scottz at google dot com> --> sheriffs investigating)'
[20:51:21] *** rbyers_ has quit IRC
[20:53:16] <ttuttle|work> okay, it looks like my change is *not* in 0.15.1011.30.
[20:53:20] *** rbyers_ has joined #chromium-os
[20:53:20] *** ChanServ sets mode: +v rbyers_
[20:53:55] <dennisjeffrey__> ttuttle|work: is your comment related to the current tree closure?
[20:54:01] <ttuttle|work> absolutely not
[20:54:11] <dennisjeffrey__> ttuttle|work: ok
[20:54:14] <ttuttle|work> it is related to some recent kernel crashes, though
[20:59:33] <ttuttle|work> alright, how do I figure out what branch in git gets merged into a given series of releases?
[21:00:57] *** rbyers_ has quit IRC
[21:03:38] <tbroch> @closure, seems like crosbug.com/20576
[21:04:49] <crosbot> tree became 'Tree is open (crosbug.com/20576)'
[21:11:15] <davidjames> tbroch: pfq failed
[21:12:48] <dennisjeffrey__> davidjames: i'm investigating, seems to be compile failure
[21:13:36] *** gman1 has joined #chromium-os
[21:13:36] *** ChanServ sets mode: +v gman1
[21:22:18] <cmasone> dennisjeffrey__: tbroch we should probably get the tree closed, and y'all should talk to keybuk about it :-)
[21:23:12] <dennisjeffrey__> cmasone: ok, i'll close tree
[21:24:09] <crosbot> tree became 'Tree is closed (BuildTarget failure in x86 generic PFQ --> sheriffs investigating)'
[21:24:14] <Keybuk> I has a fix
[21:25:03] <dennisjeffrey__> Keybuk: awesome!
[21:25:19] <Keybuk> http://gerrit.chromium.org/gerrit/8432
[21:26:27] <cmasone> I think the thingy is to get dennis to push...?  as he is sheriff, he's allowed?
[21:26:36] <Keybuk> that's my understanding
[21:26:50] <dennisjeffrey__> cmasone: trying to push...
[21:27:02] <redpig> cmasone: is that enforced now or just current best practice?
[21:27:10] *** jennb has joined #chromium-os
[21:27:10] *** ChanServ sets mode: +v jennb
[21:27:17] <cmasone> redpig: I don't have a submit button when I look at the CL
[21:27:23] <redpig> cmasone: interesting :)
[21:27:25] <cmasone> redpig: so I assume that it's enforced
[21:27:35] <dennisjeffrey__> ok, it's been merged
[21:28:02] <dennisjeffrey__> ok to reopen tree?
[21:28:04] <cmasone> dennisjeffrey__: I think now we wait for the waterfall poller to notice and kick the bot
[21:28:21] <cmasone> dennisjeffrey__: wait for the build to get past BuildTarget, I think?
[21:28:53] <dennisjeffrey__> cmasone: ok, i'll watch it until it gets past that step
[21:31:06] <cmasone> damned slow poller
[21:31:21] <crosbot> tree became 'Tree is closed (fix in, waiting to ensure BuildTarget goes green)'
[21:31:40] <Keybuk> also see http://gerrit.chromium.org/gerrit/8433, but that can wait until tree goes green
[21:34:56] <Keybuk> just running to lunch, since it'll take about that time to cycle through again
[21:48:35] *** rbyers_ has joined #chromium-os
[21:51:25] <crosbot> tree became 'Tree is open (BuildTarget cycled green on x86 generic PFQ)'
[21:53:41] *** rbyers_ has quit IRC
[21:53:51] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from None: )'
[21:55:02] <cmasone> shit
[21:55:07] <cmasone> http://code.google.com/p/chromium-os/issues/detail?id=20964
[21:55:18] <dennisjeffrey__> cmasone: yes
[21:56:08] <cmasone> and the PFQ too
[21:56:48] <dennisjeffrey__> cmasone:  for x86 pineview full --> http://code.google.com/p/chromium-os/issues/detail?id=20576
[21:56:55] <dennisjeffrey__> i think it's that issue
[21:57:09] <dennisjeffrey__> same on PFQ
[21:57:11] <cmasone> dennisjeffrey__: no, it's the one I pointed to above
[21:57:16] <cmasone> I just diagnosed that issue
[21:57:20] <cmasone> and filed a new one
[21:57:23] <cmasone> because it's different
[21:57:27] <dennisjeffrey__> cmasone: oh, ok
[21:57:35] <cmasone> (I own the session manager, which is why I looked at)
[21:57:46] <cmasone> *at it
[21:58:03] <cmasone> davidjames: I wonder if this might have to do with the new kvm...?
[21:58:18] *** mmocny_ has joined #chromium-os
[21:58:19] *** ChanServ sets mode: +v mmocny_
[21:58:47] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from None: --> sheriffs, cmasone investigating)'
[21:58:58] <cmasone> no, OI
[21:59:08] <cmasone> I'm not investigating...it's 20964
[21:59:36] <cmasone> which is filed on derat
[21:59:38] <dennisjeffrey__> cmasone: sorry, updating message
[21:59:40] <cmasone> and he's looking at it
[22:00:19] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from None: --> crosbug.com/20964, derat investigating)'
[22:00:31] *** behdad has joined #chromium-os
[22:00:32] *** ChanServ sets mode: +v behdad
[22:10:26] <dennisjeffrey__> cmasone: derat updated the bug
[22:11:28] <cmasone> dennisjeffrey__: I'm glad he's looking at it :-)
[22:12:25] <dennisjeffrey__> cmasone: he thinks there's no reason to keep the tree closed over this, but it does seem like it's happening a lot recently. What do you recommend?
[22:13:59] *** Sergiu has joined #chromium-os
[22:14:00] *** Sergiu has joined #chromium-os
[22:18:20] <crosbot> tree became 'Tree is open (crosbug.com/20964)'
[22:18:33] *** derat has joined #chromium-os
[22:18:33] *** ChanServ sets mode: +v derat
[22:18:39] <cmasone> do we have any reason to believe builds will begin passing?
[22:19:48] <davidjames> cmasone: Good point, a lot of tests are failing, so maybe it is a reproducible failure now?
[22:20:12] <cmasone> davidjames: yes, 20964 is a new failure that's happening all the time on all builds
[22:20:27] <cmasone> at least x86 builds
[22:20:52] <dennisjeffrey__> derat: what do you think?
[22:22:01] <davidjames> dennisjeffrey__: If failure happens reliably I think tree should be closed. And to me it looks like it's happening reliably
[22:22:41] <dennisjeffrey__> davidjames: agreed, and no strong reasons to keep tree open.  i'll close it
[22:22:43] *** wfrichar has joined #chromium-os
[22:22:44] *** ChanServ sets mode: +v wfrichar
[22:22:57] <davidjames> dennisjeffrey__: I suspect marcheu's upgrade to mesa broke all builders
[22:23:28] <crosbot> tree became 'Tree is closed (crosbug.com/20964 happening consistently)'
[22:23:40] <davidjames> dennisjeffrey__: Builds that started before the upgrade worked, any that started after seem to be failing consistently
[22:24:02] <davidjames> marcheu: Any chance you tried trybot with your change? :) https://sites.google.com/a/chromium.org/dev/chromium-os/local-trybot-documentation
[22:24:06] <marcheu> davidjames: it might merely have exposed a bug in our xcb I think, the crash is not in the same path
[22:24:41] <marcheu> we've have xcb oddities forever, and they seem to be race conditions, so that depending on how fast other things go they'll appear/go away
[22:24:57] <marcheu> we've had*
[22:25:10] <cmasone> can we revert something to get things green again and handle this without holding everyone up?
[22:26:03] <davidjames> Reverting http://gerrit.chromium.org/gerrit/8404 might help
[22:26:30] <marcheu> ah well, once again I'm stopped from working by irrelevant infrastructure. awesome
[22:26:58] <davidjames> marcheu: Well, the infrastructure is relevant if your change broke it :)
[22:27:07] <marcheu> it didn't break it, the bug is elsewhere
[22:27:15] <cmasone> marcheu: the integration tests are irrelevant infrastructure?
[22:27:29] <ttuttle|work> jeez, I want to transfer to the team that gets to design infrastructure solely with the purpose of tormenting innocent developers.  that sounds like a blast.
[22:28:13] <grundler> marcheu: I don't think anyone is arguing with your point.
[22:28:23] <marcheu> cmasone: it's not but the error is irrelevant to my change.
[22:29:09] <derat> marcheu: how about reverting mesa temporarily, updating xcb, and then relanding mesa?
[22:29:25] <marcheu> derat: yea, I'm not sure that'll help at all, it's just my hunch
[22:29:28] <marcheu> but, ah well
[22:29:37] *** McMAGIC--Copy has joined #chromium-os
[22:30:20] <grundler> Can someone advise me how to confirm the credit card info I gave verizon in the past (and got a "Sales Confirmation" email from them) is no longer an active account?
[22:30:22] <marcheu> derat: the problem is that we're at the point where the number of existing bugs elsewhere is preventing things from moving forward
[22:30:43] <ttuttle|work> grundler: see #chromium-os-users; this is the dev channel.
[22:30:58] <grundler> ttuttle|work: ok - I'll try there - thanks! :)
[22:31:24] <grundler> ttuttle|work: BTW, I'm here because I'm a developer...but I got your point
[22:31:30] <ttuttle|work> oh, heh
[22:31:34] <ttuttle|work> right, the voice ;)
[22:34:56] <dennisjeffrey__> it sounds like the underlying cause of the tree closure likely won't get fixed soon.
[22:35:34] <dennisjeffrey__> if that's true, the question is whether to revert a change that causes the problem to manifest (so tree is green), or else keep tree red to force the underlying cause to get fixed
[22:36:05] <ttuttle|work> as someone who likes to build, I would vote the former, and push both changes once fixed
[22:36:24] <marcheu> the issue is that I get to shave someone else's yak before I can push anything
[22:36:37] <marcheu> it's pretty much alway the issue with packages
[22:36:42] <marcheu> but sure, go ahead
[22:39:06] <dennisjeffrey__> marcheu: thank you.  so we'll revert http://gerrit.chromium.org/gerrit/8404 then to turn the tree green
[22:39:24] <dennisjeffrey__> but we should definitely keep the pressure on getting the underlying problems fixed
[22:41:35] <Keybuk> I have a weird bug that seems to only occur with today's repo sync ... none of my VMs start, they seem to hang somewhere starting the X server
[22:41:40] <Keybuk> is this known, or should I go ahead and file it?
[22:41:52] <marcheu> Keybuk: _starting_ it?
[22:42:13] <Keybuk> I am staring at screens that just say "Starting SeaBIOS" ... "Booting from Hard Disk..."
[22:42:14] <marcheu> anyway, file a bug, I'll go take a look
[22:42:27] <marcheu> right, probably the same WM crash we're talking about then
[22:42:58] <Keybuk> yeah, "start ui" gives me lots of respawns
[22:43:05] <marcheu> yeah, that's that
[22:43:22] <marcheu> I hope it's just all xcb fail
[22:44:50] <sosa> hey sheriffs, i have a one character fix for a release blocker that's making all the canary builds ineffective right now.  mind if i commit on the closed tree?  http://gerrit.chromium.org/gerrit/#change,8441
[22:45:41] <dennisjeffrey__> sosa: ok with me
[22:45:47] <dennisjeffrey__> i think we will reopen the tree soon
[22:45:48] <tbroch> sosa: ditto
[22:47:54] <sosa> tbroch/dennisjeffrey: much  thanks
[22:48:02] <sosa> err ... many* thanks :)
[22:48:18] *** rbyers_ has joined #chromium-os
[22:48:18] *** ChanServ sets mode: +v rbyers_
[22:49:39] <crosbot> tree became 'Tree is closed (revert in for http://gerrit.chromium.org/gerrit/#change,8404, waiting for bots to cycle green)'
[22:59:20] *** frankf has joined #chromium-os
[22:59:20] *** ChanServ sets mode: +v frankf
[22:59:42] <frankf> anyone else having an issue with missing 'chromium/src/tools/grit/resource_ids' when emerging chrome?
[23:02:29] *** TW1920 has quit IRC
[23:02:41] *** dots is now known as andchat
[23:03:03] *** gman1 has quit IRC
[23:04:49] *** andchat is now known as dots
[23:08:07] *** rharrison_chrome has quit IRC
[23:08:22] *** gman1 has joined #chromium-os
[23:08:23] *** ChanServ sets mode: +v gman1
[23:13:32] *** derat has quit IRC
[23:18:51] <cmasone> I can smell the green
[23:19:15] <ellyjones> It tastes like *blue*!
[23:19:26] <cmasone> this water tastes like burning!
[23:19:35] <ellyjones> :)
[23:20:11] <crosbot> tree became 'Tree is open (bots seem to be happy again)'
[23:20:48] *** behdad has quit IRC
[23:32:22] <grundler> btw, I'm finding the "You are currently offline" message a bit annoying...when I have both ethernet and WIFI connected and working. :/
[23:32:37] * grundler looks to see how many bugs have been submitted for that so far
[23:32:58] *** jshin has joined #chromium-os
[23:32:59] *** ChanServ sets mode: +v jshin
[23:35:31] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "arm generic full" from 51c238d35606ccce7f422e265518906034b274a4: Dennis Jeffrey <dennisjeffrey at chromium dot org>)'
[23:37:40] <grundler> hrm 20014 claims to be "fixed" 5 days ago...
[23:38:01] <ellyjones> I
[23:38:15] <ellyjones> er, I am going to hack something into crosbot to automatically grep for symptoms of common bugs
[23:39:37] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "arm generic full" from 51c238d35606ccce7f422e265518906034b274a4: Dennis Jeffrey <dennisjeffrey at chromium dot org> --> sheriffs investigating)'
[23:39:44] * Keybuk raises a shy hand
[23:44:20] <Keybuk> dennisjeffrey: http://gerrit.chromium.org/gerrit/8443
[23:45:34] <dennisjeffrey__> Keybuk: would you like me to push that?
[23:45:49] <Keybuk> yup
[23:45:55] <Keybuk> well, you'd like you to push that ;)
[23:46:22] <dennisjeffrey__> ok, i was trying to figure out if it was related to the current tree closure; sounds like it is ;-)
[23:46:28] <dennisjeffrey__> will push it...
[23:46:50] <Keybuk> yeah, it was the kind of failure that was only ever going to show up on a -full
[23:46:57] <davidjames> dennisjeffrey__: Note that won't fix the tree until x86 generic pfq pushes the change
[23:47:25] <davidjames> hence ~30 minutes before anybody can build unless somebody updates stable ebuild manually
[23:48:04] <Keybuk> davidjames: I still don't understand, mind, why emerging udev works without declaring that it will break a dep on it
[23:48:20] <davidjames> Keybuk: Use --complete-graph and you'll see the problem
[23:48:24] <dennisjeffrey__> http://gerrit.chromium.org/gerrit/8443 has been merged
[23:48:29] <Keybuk> davidjames: completegraph on emerge?
[23:49:30] <Keybuk> instead of --deep?
[23:50:02] <davidjames> Keybuk: Yeah, --complete-graph does a little more checking for existing ebuilds... but actually wouldn't help because you have an existing built package which works
[23:50:18] *** rbyers_ has quit IRC
[23:59:07] <dennisjeffrey__> davidjames: what's involved with updating the stable ebuild manually?  Would it save us considerable time as compared to waiting for the bots to turn green again?
[23:59:31] <davidjames> dennisjeffrey__: Yeah it would make the bots green immediately, it would just involve copying keybuk's change over to stable ebuild





top