[00:12:07] *** bleung has left #chromium-os
[00:20:12] <jamescook> tegra2_seaboard-tangent-binary failed in chromeos-bootimage-0.0.2-r11, near a log line saying "Flash map entries overlap by 256 bytes".  Does that sound familiar?  dhendrix?
[00:20:40] <dhendrix> jamescook:  ping Stepan
[00:21:11] <crosbot> tree became 'Tree is open (jamescook investigating tegra2_seaboard-tangent-binary chromeos-bootimage )'
[00:22:11] *** sadrul has quit IRC
[00:22:40] <dhendrix> jamescook:  vadim bendebury might also have an idea
[00:24:36] <Stepan> huh
[00:27:57] <Stepan> I checked in a change to the x86-alex device tree/fmap in uboot
[00:28:20] <Stepan> if that causes trouble with seaboard something is severely wrong ;)
[00:31:49] *** McMAGIC--Copy has quit IRC
[00:40:19] <crosbot> tree became 'Tree is open (gabeblack investigating tegra2_seaboard-tangent-binary chromeos-bootimage -> crosbug.com/19306 )'
[00:46:57] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 seaboard full" from e2fb3eb255ddc91624bb8fbbe2756b4abe353544: Ryan Cui <rcui at chromium dot org>, Todd Broch <tbroch at chromium dot org>, _third_party_ at chromium dot org, aa at chromium dot org <aa at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>, tony at chromium dot org <tony at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>)'
[00:47:39] <jamescook> gabeblack: tegra2 seaboard full is the same issue
[00:48:41] *** McMAGIC--Copy has joined #chromium-os
[00:50:31] <crosbot> tree became 'Tree is closed (sheriffs investigating "tegra2 seaboard full" from e2fb3eb255ddc91624bb8fbbe2756b4abe353544: Ryan Cui <rcui at chromium dot org>, Todd Broch <tbroch at chromium dot org>, _third_party_ at chromium dot org, aa at chromium dot org <aa at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>, tony at chromium dot org <tony at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>)'
[00:54:12] *** vmil86 has quit IRC
[00:54:19] <jamescook> Could this change be related? http://gerrit.chromium.org/gerrit/#change,6111
[00:58:22] <gabeblack> jamescook: I'm really starting to suspect it is
[00:58:34] <gabeblack> I'm attempting to build to get a better idea
[00:58:40] <gabeblack> I'll keep you posted
[01:01:55] <gabeblack> building chromeos-bootimage worked for me
[01:02:03] <gabeblack> I'll sync and try again
[01:03:17] <jamescook> davidjames: looks like tegra2_seaboard-tangent-binary is stuck in sync.  Should I kill the build?
[01:04:27] <jamescook> gabeblack: we could experimentally revert it
[01:04:45] <gabeblack> I'm trying to reproduce the problem so I can do exactly that
[01:05:02] <jamescook> I'm going to stop the build on tegra2_seaboard-tanget-binary, which is stuck in sync
[01:05:41] <davidjames> jamescook: Thanks :)
[01:05:49] <jamescook> done
[01:14:59] <jamescook> davidjames: tegra2_seaboard-tangent-binary claims to have 4 pending, but doesn't seem to be picking up automatically?  do I need to force a build?
[01:17:13] <gabeblack> I just built chromeos-bootimage again, and it still worked
[01:17:22] <gabeblack> I got a couple of warnings
[01:17:31] <gabeblack> Warning: Flash map has a gap of 61184 bytes: ro-frid: 00101000-00101100, gbb: 00110000-00190000
[01:17:31] <gabeblack> Warning: Flash map has a gap of 61184 bytes: ro-frid: 00101000-00101100, gbb: 00110000-00190000
[01:17:38] <gabeblack> but not the one about an overlap
[01:17:45] <gabeblack> so maybe that was a transient problem?
[01:18:16] <jamescook> What platform did you build it for?
[01:18:30] <gabeblack> tegra2_asymptote
[01:18:49] <jamescook> is that the same as tegra2_seaboard ?
[01:18:58] <gabeblack> sort of
[01:19:18] <gabeblack> I'll try seaboard too
[01:19:24] <davidjames> jamescook: tangent is a shared builder shared with tangent canary
[01:19:36] <davidjames> so it won't start until tangent canary finishes
[01:19:46] <jamescook> ah, ok
[01:20:18] <gabeblack> oh, did I build for the wrong platform :-0
[01:20:45] <gabeblack> Doing to many things at once, I think
[01:23:31] *** thieule has joined #chromium-os
[01:23:32] *** ChanServ sets mode: +v thieule
[01:30:03] <gabeblack> ah, there we go
[01:30:13] <gabeblack> I got it to fail for me
[01:37:38] <jamescook> gabeblack: other platforms seem ok, should we reopen while you investigate further?
[01:38:12] <gabeblack> yes, I think so
[01:38:19] <jamescook> I'll open
[01:38:58] <crosbot> tree became 'Tree is open (gabeblack investigating "tegra2 seaboard full" flash map issue -> crosbug.com/19306 )'
[01:42:07] *** P1ersson has quit IRC
[01:42:18] *** P1ersson has joined #chromium-os
[01:44:04] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from e2fb3eb255ddc91624bb8fbbe2756b4abe353544: Ryan Cui <rcui at chromium dot org>, _third_party_ at chromium dot org)'
[01:44:34] <cmasone> x86-pineview full is about to go red because of http://code.google.com/p/chromium-os/issues/detail?id=19005
[01:44:37] <cmasone> oh
[01:44:39] <cmasone> it already happened
[01:44:50] <cmasone> append to that bug, and I'll look at it, and repopen
[01:45:10] <jamescook> will do
[01:46:27] *** gsam_ has joined #chromium-os
[01:46:28] *** ChanServ sets mode: +v gsam_
[01:46:33] <gsam_> oops, tree was closed
[01:47:39] <crosbot> tree became 'Tree is closed (gabeblack investigating "tegra2 seaboard full" flash map issue -> crosbug.com/19306, x86 pineview full -> crosbug.com/19005 )'
[01:47:54] <crosbot> tree became 'Tree is open (gabeblack investigating "tegra2 seaboard full" flash map issue -> crosbug.com/19306, x86 pineview full -> crosbug.com/19005 )'
[01:49:30] *** ambro718 has quit IRC
[01:49:45] <gabeblack> yes, I think it was 6111
[01:49:55] <jamescook> revert!
[01:50:10] <gabeblack> I think Simon missed a file
[01:50:18] <gabeblack> I'm trying to verify that a revert fixes it
[01:50:25] <gabeblack> I tried that a minute ago and it didn't for some reason
[01:50:27] <gabeblack> I'm trying again
[01:50:33] <gabeblack> just to make sure I don't fat finger git or something
[01:50:53] <cmasone> gabeblack: you can have gerrit revert for you
[01:51:09] <gabeblack> oh, yeah!
[01:51:11] <gabeblack> I forgot about that
[01:51:44] <gabeblack> and it just successfully built with the revert too
[01:52:11] <gabeblack> did that actually revert it, or just make a change to revert it?
[01:52:23] <srao> gabeblack: it just makes a change to revert it
[01:52:25] <jamescook> made the change
[01:53:10] <jamescook> er, just makes the change
[01:53:18] <gabeblack> it's up for review
[01:53:53] <gabeblack> http://gerrit.chromium.org/gerrit/#change,6194
[01:54:17] <cmasone> gabeblack: pushed
[01:54:36] <jamescook> he beat me to it :-)
[01:54:37] <gabeblack> ok, so seaboard should be good now
[01:54:40] <gabeblack> heh :-)
[01:57:05] <crosbot> tree became 'Tree is open (x86 pineview full -> crosbug.com/19005 )'
[02:00:53] <jamescook> gabeblack: how do we let Simon know his change was reverted?
[02:01:23] <cmasone> I assume gerrit emailed him.  If not...gchat will work
[02:01:38] <gabeblack> he's not at his desk
[02:02:01] <jamescook> I tried gchat, he's not on
[02:02:05] <gabeblack> I've been watching for him to let him know
[02:02:07] <cmasone> is he WFH today?  or did he land and run?
[02:02:13] <gabeblack> he's been in
[02:02:14] <cmasone> email will probably work too :)
[02:02:23] <gabeblack> yeah, lets do that
[02:02:38] <gabeblack> I'll also try to tell him in person in case he misses the email, it gets filtered, etc.
[02:02:56] <cmasone> gabeblack: tell him with a smack upside the head :-)
[02:03:11] <gabeblack> two smacks
[02:03:17] <gabeblack> one for making my hand hurt with the first one
[02:04:10] <jamescook> gabeblack: I'm going to head home, try to beat the worst of the traffic.  I'll be back online in about an hour and check in.
[02:04:19] <gabeblack> ok
[02:05:43] *** jamescook has left #chromium-os
[02:05:44] <gabeblack> email sent
[02:06:37] *** sadrul has joined #chromium-os
[02:06:39] *** ChanServ sets mode: +v sadrul
[02:08:44] *** lipsinV2 has joined #chromium-os
[02:09:35] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy canary" from None: )'
[02:13:19] <gabeblack> I'm looking at the stumpy canary problem
[02:13:40] <gabeblack> was that expected?
[02:13:58] <gabeblack> It's complaining about setlocale: LC_ALL
[02:14:44] <gabeblack> and then a call with uname in it had some problems
[02:14:45] <gabeblack> Exception: Call to 'uname -m | sed -e "s/i.86/ia32/;s/x86_64/x64/;s/amd64/x64/;s/arm.*/arm/;s/i86pc/ia32/"' returned exit status 0. while trying to load src/build/all.gyp
[02:14:48] <cmasone> davidjames: vapier ^^^
[02:14:57] <cmasone> I think we fixed that hours ago...when did that build start?
[02:15:45] <gabeblack> cmasone: around 1:30, it looks like
[02:15:59] <gabeblack> so that's probably a non-issue
[02:16:59] <crosbot> tree became 'Tree is open (Stumpy canary failure was fixed after the build started)'
[02:18:12] <davidjames> cmasone, gabeblack: Thanks! :)
[02:35:23] *** GTAXL has quit IRC
[02:35:38] *** kerz has quit IRC
[02:52:55] *** wfrichar has quit IRC
[03:41:27] *** grundler has quit IRC
[03:42:56] *** Keybuk has quit IRC
[04:01:35] *** jrbarnette has quit IRC
[04:11:40] <crosbot> tree became '?'
[04:11:55] <crosbot> tree became 'Tree is open (Stumpy canary failure was fixed after the build started)'
[04:25:12] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl-binary" from a9d1a103348d92bdbb072c19e71610a401e48885: Ryan Cui <rcui at google dot com>, bbudge at chromium dot org <bbudge at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, sail at chromium dot org <sail at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[04:51:16] *** petermayo has joined #chromium-os
[04:51:18] *** ChanServ sets mode: +v petermayo
[04:56:24] *** saintlou has quit IRC
[05:05:36] *** petermayo has left #chromium-os
[05:06:22] *** Hexxeh has quit IRC
[05:06:57] *** Hexxeh has joined #chromium-os
[05:15:44] *** deshantm is now known as deshantm_away
[05:16:56] *** GTAXL has joined #chromium-os
[05:17:04] *** kinaba_ has joined #chromium-os
[05:17:21] *** kinaba_ has quit IRC
[05:17:27] *** kinaba_ has joined #chromium-os
[05:17:51] *** kinaba_ has quit IRC
[05:19:30] *** satorux_ has quit IRC
[05:27:38] *** satorux_ has joined #chromium-os
[05:27:38] *** ChanServ sets mode: +v satorux_
[05:36:45] *** oc80z has quit IRC
[05:36:45] *** oc80z has joined #chromium-os
[06:18:50] *** Daxvex has quit IRC
[06:30:06] *** gfrog has quit IRC
[06:47:01] *** gfrog has joined #chromium-os
[07:00:37] *** vmil86 has joined #chromium-os
[07:19:28] *** gfrog has quit IRC
[07:30:11] *** arun has quit IRC
[07:32:43] *** gfrog has joined #chromium-os
[07:57:08] *** sergiu has quit IRC
[08:12:46] *** Sergiu has joined #chromium-os
[08:12:46] *** Sergiu has joined #chromium-os
[08:15:24] *** Sergiu has quit IRC
[08:21:46] *** jochen__ has quit IRC
[08:21:52] *** jochen__ has joined #chromium-os
[08:21:52] *** ChanServ sets mode: +v jochen__
[08:30:13] *** Sergiu has joined #chromium-os
[08:30:13] *** Sergiu has joined #chromium-os
[08:35:27] *** cjb` has joined #chromium-os
[08:35:35] *** cjb has quit IRC
[08:36:03] *** Styx has joined #chromium-os
[08:36:15] *** Mkools has joined #chromium-os
[08:38:30] <Mkools> Hi.guys.good.to.see.u.after.6.months.hope.project.is.going.great:)
[09:10:51] *** GTAXL has quit IRC
[09:20:49] <aaronp> Wow, the new new tab page is lame.
[09:44:42] *** Sergiu has quit IRC
[09:46:07] <Mkools> exit
[09:46:14] *** Mkools has left #chromium-os
[10:04:12] *** patcito has quit IRC
[10:25:05] *** rosso has joined #chromium-os
[10:32:33] *** green_ has joined #chromium-os
[11:54:28] *** arun__ has joined #chromium-os
[11:56:26] *** arun__ has quit IRC
[11:56:26] *** arun__ has joined #chromium-os
[11:56:46] *** arun__ is now known as arun
[12:45:03] *** kochi has quit IRC
[12:54:11] *** sadrul has quit IRC
[13:00:28] *** green_ has quit IRC
[13:14:32] *** andybleaden has joined #chromium-os
[13:19:17] *** sadrul has joined #chromium-os
[13:49:04] *** petermayo has joined #chromium-os
[13:49:05] *** ChanServ sets mode: +v petermayo
[14:35:51] *** rbyers_ has joined #chromium-os
[14:53:05] *** deshantm_away is now known as deshantm
[15:13:00] *** sakoman has quit IRC
[15:14:55] *** sakoman has joined #chromium-os
[15:19:49] *** cros_ has joined #chromium-os
[15:29:03] *** rharrison_chrome has joined #chromium-os
[15:29:04] *** ChanServ sets mode: +v rharrison_chrome
[15:37:53] *** rbyers_ has quit IRC
[15:41:11] *** sakoman has quit IRC
[15:41:57] <reveman> x86-zgb-binary failure looks like crosbug.com/19298
[15:42:57] *** sakoman has joined #chromium-os
[15:56:42] *** djkurtz has quit IRC
[15:59:22] *** Styx has quit IRC
[16:00:50] *** zmedico has quit IRC
[16:05:16] <reveman> and x86-generic-chrome-PFQ failure looks like crosbug.com/16591
[16:09:18] *** zmedico has joined #chromium-os
[16:12:06] *** Styx has joined #chromium-os
[16:20:09] *** yusukes has quit IRC
[16:34:02] <rosso> has anyone an idea how to remount the / read-write on chromium os?
[16:34:20] <rosso> The obvius: mount -o remount,rw / would fail
[16:38:10] <ellyjones> you can use make_dev_ssd to disable rootfs verification
[16:39:35] <rosso> ellyjones: in the build you mean?
[16:39:58] <ellyjones> if you are in dev mode and have a root shell, you can just run that - it's in /usr/share/vboot/bin
[16:42:32] 
[16:42:42] <ellyjones> 'somehow fails'?
[16:42:44] <ellyjones> what does it say?
[16:43:14] <rosso> "you are not using developer firmware" [...]
[16:43:24] <rosso> obviously i am not
[16:43:39] <ellyjones> did you run make_dev_ssd --remove_rootfs_verification?
[16:43:51] <ellyjones> yeah, it is probably pretty sad on random hardware
[16:44:58] <rosso> everything of the above fails even with --force
[16:45:05] <rosso> damn i did this before ...
[16:46:22] <ellyjones> build the image with --noenable_rootfs_verification
[16:46:45] <rosso> ellyjones: yeah, will have to rebuild :(
[16:46:48] <ellyjones> oh well
[16:47:01] <rosso> thanx for your support
[16:51:18] *** andybleaden has quit IRC
[16:54:32] *** wfrichar has joined #chromium-os
[16:54:33] *** ChanServ sets mode: +v wfrichar
[17:04:39] <reveman> cmasone: the TOT-pfq failure looks like crosbug.com/16591 again, right?
[17:12:43] <cmasone> reveman: nope
[17:13:00] <cmasone> logging_CrashSender failed
[17:13:09] <cmasone> Command <chown chronos:chronos "/home/chronos/Consent To Send Stats"> failed, rc=1, Command returned non-zero exit status
[17:13:16] <cmasone> it does look transient, tho
[17:13:26] <cmasone> file an issue and reopen, if it doesn't happen aain
[17:13:30] <cmasone> in the next cycle
[17:18:07] *** Styx has quit IRC
[17:19:06] *** McMAGIC--Copy has quit IRC
[17:19:31] *** McMAGIC--Copy has joined #chromium-os
[17:20:23] *** Styx has joined #chromium-os
[17:24:46] *** SoulRaven has joined #chromium-os
[17:26:25] *** rosso has quit IRC
[17:26:35] *** BladeFreak has joined #chromium-os
[17:26:41] *** seanpaul has joined #chromium-os
[17:26:41] *** ChanServ sets mode: +v seanpaul
[17:30:22] *** grundler has joined #chromium-os
[17:30:23] *** ChanServ sets mode: +v grundler
[17:34:37] *** JasonO has joined #chromium-os
[17:34:38] *** JasonO has joined #chromium-os
[17:48:28] *** sadrul has quit IRC
[17:49:32] <grundler> Can someone explain why "find" isn't listing some files in /sys ?
[17:49:35] <grundler> localhost sys # find | fgrep temp
[17:49:35] <grundler> ./devices/platform/tegra-i2c.1/i2c-2/2-000b/power_supply/battery/temp
[17:49:35] <grundler> ./devices/platform/tegra-i2c.1/i2c-2/2-0020/temp_deactivated
[17:49:35] <grundler> ./devices/platform/tegra-i2c.3/i2c-4/4-004c/temperature
[17:49:35] <grundler> ./devices/platform/tegra-i2c.3/i2c-4/4-004c/ext_temperature
[17:49:36] <grundler> localhost sys # find -name temp
[17:49:38] <grundler> ./devices/platform/tegra-i2c.1/i2c-2/2-000b/power_supply/battery/temp
[17:50:01] <grundler> oh...I forgot the \*
[17:50:26] <grundler> hrm. Now it find them.
[17:50:53] <grundler> Oh...I know what I did wrong before...symlinks weren't being followed and I was in /sys/bus/i2c
[17:51:18] <grundler> sorry for the noise
[18:00:06] *** |simo- has quit IRC
[18:01:53] *** saintlou has joined #chromium-os
[18:01:53] *** ChanServ sets mode: +v saintlou
[18:04:12] <saintlou> Your friendly West Coast sheriff is reporting for duty. Unfortunately I notice I am the only one on West Coast today.
[18:05:29] <saintlou> varunjain, reveman  it seems that East Coast was blessed with 2 sheriffs, please ping me if there is anything that I need to know.
[18:07:49] <varunjain> saintlou, reveman has already investigated the x86-generic-chrome-PFQ failure... except for that everything seems fine
[18:08:04] <saintlou> thanks varunjain
[18:08:38] <saintlou> Let me know when you guys take off for lunch. I have a meeting from 11.30 to 1pm Pacific, and will appreciate if you can cover for me since I am alone on the West Coast
[18:10:06] <reveman> saintlou: sure, no prob
[18:10:17] <saintlou> thanks!
[18:10:21] <reveman> saintlou: x86-zgb-binary failure -> crosbug.com/19298 and x86-generic-chrome-pfq failure -> crosbug.com/16591
[18:11:06] <reveman> that's the issues I've seen since this morning
[18:11:28] <cmasone> reveman: no, the pfq was not 16591
[18:12:00] <cmasone> reveman: if you look at the logs, tehre was an actual test failure...I responded to your IRC ping when I got in
[18:16:06] <reveman> cmasone: hm that was the recent tot-pfq failure. I saw your response and filed a bug for that one as you told me. the x86-generic-chrome-pfq failure was a different one but I might be wrong about that one too
[18:16:26] <cmasone> reveman: oohhhhhhh, my bad
[18:16:35] *** jrbarnette has joined #chromium-os
[18:18:02] <saintlou> cmasone: do you know what is happening, are you investigating, and if so should we put this in the Tree Status?
[18:18:17] <cmasone> saintlou: I am not looking at anything
[18:18:57] *** jujugre has left #chromium-os
[18:19:07] *** achuith has joined #chromium-os
[18:19:07] *** ChanServ sets mode: +v achuith
[18:20:33] *** jrbarnette_ has joined #chromium-os
[18:20:51] <saintlou> cmasone: no pb, thanks
[18:21:00] <saintlou> cmasone: no pb, thanks (need more coffee)
[18:21:47] *** oc80z has quit IRC
[18:22:34] *** oc80z has joined #chromium-os
[18:23:04] *** jrbarnette has quit IRC
[18:23:05] *** jrbarnette_ is now known as jrbarnette
[18:23:49] *** D|sT has quit IRC
[18:24:24] *** cjb` is now known as cjb
[18:24:40] <reveman> saintlou: I don't understand why tegra2_aebl-binary closed the tree, looks like the latest build was successful
[18:25:05] *** |simo| has joined #chromium-os
[18:25:31] *** |simo| is now known as Guest57323
[18:26:05] <saintlou> reveman: let me see
[18:26:41] <cmasone> reveman: hm.  There was a weird issue yesterday where this happened with kaen-binary
[18:27:09] <cmasone> reveman: it had to do with there being multiple slaves running a build for the same config at the same time, and some kind of out-of-order completion going on
[18:27:32] <cmasone> reveman: saintlou sounds like something to escalate to the troopers?
[18:28:31] <saintlou> cmasone: I think you are right this does not look correct all the aebl builds are fine, there is some redness but elsewhere
[18:31:12] *** saintlou_ has joined #chromium-os
[18:31:12] *** ChanServ sets mode: +v saintlou_
[18:31:48] *** Guest57323 has quit IRC
[18:32:12] <reveman> cmasone: ok, do you know if a bug report exists for this out-of-order issue?
[18:34:11] <mononoke> does anybody know who "scottzcr" is and why he uses a gmail.com address in the chromium-os bug database?
[18:34:33] *** saintlou has quit IRC
[18:35:53] <saintlou_> reveman, varunjain: sent an email to troopers
[18:37:16] *** benchan has joined #chromium-os
[18:37:16] *** ChanServ sets mode: +v benchan
[18:37:36] <reveman> saintlou_: thanks
[18:37:59] <reveman> saintlou_: should we reopen the tree?
[18:39:12] <saintlou_> reveman: yes I was going to suggest that
[18:39:33] <saintlou_> please go ahead (put something like pending trooper investigation on aebl in the status)
[18:40:21] <reveman> done
[18:42:57] <benchan> ellyjones: I tried minijail0 and libminijailpreload. seccomp is too restrictive but caps seems to work in my case.
[18:46:11] <redpig> benchan: a new seccomp is on the way ;)
[18:46:25] <redpig> awesome that caps worked!
[18:46:27] *** |simo- has joined #chromium-os
[18:46:42] <benchan> redpig: with individual syscall filtering?
[18:46:53] *** lipsinV2 has quit IRC
[18:46:59] <redpig> benchan: yeah
[18:47:05] <redpig> needed to fix it up for arm before i enabled it
[18:47:30] <redpig> I had added unneeded dependencies :/
[18:47:31] <benchan> redpig: great, i'll wait in line to try it out
[18:47:58] <redpig> hehe me too :p
[18:50:30] <benchan> redpig: I'm still looking for a more systematic way, besides strace, to see what syscalls could be used by a program
[18:52:15] <redpig> really that's one fo the best -- with this change, you'll be told which syscall was blocked too
[18:52:33] <redpig> so at least there's a way to figure out what happened - but it's nicer to figure it out first :p
[18:52:46] *** sadrul has joined #chromium-os
[18:52:47] *** ChanServ sets mode: +v sadrul
[18:56:35] <benchan> redpig: strace requires the program being ptrace-able though. but if the new secommp can log what syscalls are blocked, I can use an autotest to monitor that
[18:58:30] <redpig> yup - that'll give us a good feedback option
[19:03:35] <ellyjones> benchan: seccomp mode 1 is too restrictive for almost anything
[19:04:10] <ellyjones> caps should be an improvement though
[19:04:19] <ellyjones> you might also try pid namespace and vfs namespace
[19:04:41] <ellyjones> pid namespace has the pleasant property that minijail_kill() atomically kills everything inside the namespace
[19:04:44] <sosa> PSA:  I turned off parallelism in the test harness during the TestStage to hopefully remove any flakiness in the test harness from parallelization and tests that fail without reason.  Let me know if it happens again after the change makes its way through
[19:05:03] <ellyjones> sosa: \o/
[19:05:54] <sosa> since the errors are quite inexplicable i'm not sure it'll help but it'll def eliminate a possiblity of why it's happening if it still happens
[19:07:38] <cmasone> sosa: how much slower did the tests get?
[19:07:48] <reveman> sosa: ok, will let you know if it happens again
[19:08:39] <reveman> saintlou_, varunjain: are you around for the next 30 min? I need to grab some lunch
[19:08:49] <benchan> ellyjones: I ran into some problem with pid/vfs namespace, but haven't got a chance to debug it
[19:08:50] <saintlou_> reveman: go ahead!
[19:08:58] <varunjain> reveman, sure
[19:09:39] <ellyjones> benchan: ah, ok
[19:09:47] <ellyjones> benchan: pid namespace breaks strace
[19:10:02] <ellyjones> (in particular, trying to strace minijail0 -p will cause strace to hang forever)
[19:10:07] <ellyjones> er, strace -f
[19:17:28] *** benchan has quit IRC
[19:17:34] *** benchan has joined #chromium-os
[19:17:34] *** ChanServ sets mode: +v benchan
[19:17:57] <benchan> ellyjones: yes, i need to find another way to debug it
[19:18:25] *** chocobo__ has joined #chromium-os
[19:18:25] *** ChanServ sets mode: +v chocobo__
[19:20:19] *** saintlou_ is now known as saintlou
[19:22:11] <davidjames> cmasone: The out-of-order completion is confusing, but I'm not sure it's a bug in buildbot -- it's more the display on the waterfall that's buggy (that it doesn't show you very well that you have multiple builds running in the same column on multiple builders)
[19:22:35] <cmasone> davidjames: sure.  Why would we ever have that happen, though?
[19:22:37] <zbehan> davidjames: just curious though, why would two slaves build the same board?
[19:23:30] <davidjames> cmasone: Well, tegra2_aebl is configured to run on multiple slaves, so if it has multiple requests to build, it tries to fulfill as many of the requests as possible using the machine resources available
[19:24:22] <cmasone> davidjames: either way, this is a sheriffability challenge :-)  Can someone own making it make sense?
[19:28:39] <cmasone> that failure looks like 16591 or whatever
[19:28:43] <cmasone> is anyone looking?
[19:29:00] <cmasone> ellyjones: where's crosbot?
[19:29:03] <saintlou> cmasone: I am looking at pineview
[19:29:14] <saintlou> is that what you aked?
[19:29:34] <cmasone> saintlou: can you update the status to indicate that someone's looking?  The new sheriffing FAQ asks for this, so that people don't start duplicating effort
[19:30:18] <ellyjones> crosbot: tree?
[19:30:18] <crosbot> ellyjones: tree: Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl-binary" from a9d1a103348d92bdbb072c19e71610a401e48885: Ryan Cui <rcui at google dot com>, bbudge at chromium dot org <bbudge at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, sail at chromium dot org <sail at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)
[19:30:23] <saintlou> cmasone: I was looking for your new FAQ BTW can you send me the link :-) I read it before but need a refresher :-)
[19:30:38] <ellyjones> but that message is stale...
[19:30:39] <ellyjones> odd
[19:30:45] <cmasone> saintlou: it's the same URL as before: http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium-os
[19:31:07] *** crosbot has quit IRC
[19:31:14] *** crosbot has joined #chromium-os
[19:31:18] <ellyjones> crosbot: announce +tree
[19:31:18] <crosbot> ellyjones: announcing tree
[19:31:24] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from a19d1e2ae9301fdc5d0610475b4693e28f46f5d2: Anne Kenyon <annemkenyon at google dot com>, halyavin at google dot com <halyavin at google dot com@fcba33aa-ac0c-11dd-b9e7-8d5594d729c2>)'
[19:31:26] <ellyjones> I need to figure out why it does that
[19:31:55] <crosbot> tree became 'Tree is closed - sheriff looking into it (Automatic: "cbuildbot" on "x86 pineview full" from a19d1e2ae9301fdc5d0610475b4693e28f46f5d2: Anne Kenyon <annemkenyon at google dot com>, halyavin at google dot com <halyavin at google dot com@fcba33aa-ac0c-11dd-b9e7-8d5594d729c2>)'
[19:33:25] <cmasone> http://code.google.com/p/chromium-os/issues/detail?id=16591
[19:33:29] <cmasone> it's that bug, again
[19:34:08] <saintlou> cmasone: by that you mean the git/gerrit sync issue is what causing the pineview failure?
[19:34:51] <cmasone> saintlou: nope.  That's not what the bug is about.  And that builder failed in the test phase...how could that have anything to do with gerrit?
[19:35:43] <saintlou> cmasone: Sorry just looking at the output of the test, I got confused.
[19:35:57] *** grundler has quit IRC
[19:37:38] <saintlou> cmasone: looking at the update_engline.log I see the flakyness
[19:37:50] <saintlou> " it's that bug, again" threw me off
[19:38:34] <davidjames> cmasone: (was just off talking with eblake about this)
[19:38:50] <crosbot> tree became 'Tree is open (flakyness from http://code.google.com/p/chromium-os/issues/detail?id=16591 )'
[19:39:10] <davidjames> cmasone: eblake offered to fix the sherrifability challenge, i'll file a bug for him
[19:41:26] *** micahc has quit IRC
[19:42:24] *** sergiu has joined #chromium-os
[19:42:24] *** sergiu has joined #chromium-os
[19:43:26] *** grundler has joined #chromium-os
[19:43:26] *** ChanServ sets mode: +v grundler
[19:44:49] *** Styx has quit IRC
[19:47:58] <davidjames> cmasone: Filed http://crosbug.com/19324
[19:49:37] <ellyjones> mononoke: tpm bGlobalLock for x86 recovery: ?
[19:51:16] <mononoke> Hi
[19:51:36] <ellyjones> I have not done a successful recovery on x86 in some time because bGlobalLock cannot be taken
[19:51:40] <ellyjones> is there a bug?
[19:51:45] <gauravsh> it should be fixed now
[19:51:55] <mononoke> It's fixed
[19:52:03] <ellyjones> just need to sync my tree? excellent
[19:52:08] <mononoke> yes
[19:52:12] <mononoke> funny bug too
[19:52:29] <ellyjones> oh?
[19:52:35] <mononoke> the script contained exec &> /dev/tty2
[19:52:44] <mononoke> we switched to a new busybox
[19:52:47] <gauravsh> surprisingly, it was anush's busybox update that caused the breakage
[19:52:50] <ellyjones> haha
[19:52:52] <mononoke> with a more primitive bash
[19:52:58] <ellyjones> the only thing in the blamelist!
[19:52:58] <gauravsh> i will let mononke continue... :)
[19:53:05] <mononoke> so &> became & >
[19:53:05] <ellyjones> I wrote that off as not being possible
[19:53:08] <ellyjones> oops!
[19:53:08] <jrbarnette> hey, busybox *ain't* bash
[19:53:18] <mononoke> and there were two copies of the script running
[19:53:27] <mononoke> (I meant sh not bash)
[19:53:35] <mononoke> so the TPM was locked twice
[19:53:40] <mononoke> and the second time it failed
[19:53:46] <mononoke> blame it on the TPM as usual
[19:55:13] <davidjames> cmasone: Er, wrong link. sheriffability challenge filed as http://crosbug.com/19342
[19:55:54] *** jhorwich_ has joined #chromium-os
[19:55:55] *** ChanServ sets mode: +v jhorwich_
[19:56:38] <redpig> mononoke: oh ffs :(
[19:56:52] <redpig> nice job figuring thta out :/
[19:57:53] <mononoke> Yes well Hung-Te did.  I had to leave for the day.
[19:58:36] <mononoke> It was hard to figure out because recovery.log  was empty
[19:58:55] *** jrbarnette_ has joined #chromium-os
[19:59:02] <mononoke> In addition, when I rebuilt my tree the bug didn't reproduce.
[20:00:02] <davidjames> saintlou, varunjain, revema: Both internal and external preflight queues are hung again, gonna kill git on them to get them going again
[20:00:11] <mononoke> As if the busybox upgrade didn't get picked up
[20:00:23] <mononoke> which is totally believable given the complexity of our build system
[20:00:39] <mononoke> (I didn't start from scratch, just repo sync, build_packages, build_image etc.)
[20:00:57] <mononoke> Sorry, just venting.
[20:01:27] <ellyjones> s'ok
[20:01:36] <davidjames> mononoke: No problem, that's what IRC is for
[20:01:43] <ellyjones> yeah :P
[20:01:57] <mononoke> Ahhhh... I didn't know that.
[20:02:55] *** jrbarnette has quit IRC
[20:05:03] *** jrbarnette_ has quit IRC
[20:05:13] <davidjames> Wow that's a first, x86 generic pfq was hung on the kernel repo. Usually it's hung on chromiumos-overlay
[20:07:28] <adlr> djkurtz did just check in some stuff to that repo
[20:08:43] <ellyjones> is it gitmasters at chromium dot org or gitmaster at chromium dot org?
[20:10:17] <davidjames> ellyjones: Neither, chromium-gitmaster at googlegroups.com
[20:10:33] <ellyjones> aha
[20:14:15] <saintlou> thanks davidjames
[20:14:36] *** Keybuk has joined #chromium-os
[20:14:36] *** ChanServ sets mode: +v Keybuk
[20:14:43] <ellyjones> davidjames: chromium-gitmaster at googlegroups dot com gave me a bounce
[20:14:57] <davidjames> ellyjones: I guess they changed it, too much spam ;)
[20:15:00] * eggy bounces around
[20:15:39] <saintlou> davidjames: any clue on a permanent solution? It is my understanding from yesterday's sheriff that you were looking into it.
[20:16:12] <ellyjones> I have a fallback plan, which I call 'im scottz'
[20:17:21] <davidjames> saintlou: See http://crbug.com/90675
[20:18:31] <saintlou> thanks davidjames
[20:18:39] <davidjames> saintlou: I think GIT_HTTP_LOW_SPEED_TIME will work around it, ferringb is working on getting that into the tree
[20:21:16] <davidjames> GIT_HTTP_LOW_SPEED_TIME=30 would cause git to exit if we have more than 30 seconds where it is receiving <1 byte per second.
[20:21:46] <davidjames> We already have that enabled for curl so no reason not to enable same for git, just hope the feature actually works as it's hard to reproduce the problem locally
[20:23:26] <gauravsh> ellyjones: scottz is out until next week. there goes your fallback plan!
[20:23:41] <gauravsh> file a bug on chrome infrastructure team on crbug.com
[20:23:47] <gauravsh> they usually respond very quickly.
[20:24:16] <ellyjones> cool :)
[20:26:39] *** jrbarnette has joined #chromium-os
[20:26:40] *** ChanServ sets mode: +v jrbarnette
[20:32:48] *** rbyers_ has joined #chromium-os
[20:35:06] <davidjames> jrbarnette: Noticed some init scripts using "&>" notation. Should those switch away from that since it's not supported by busybox?
[20:35:45] <davidjames> jrbarnette: E.g. cron-lite.conf
[20:36:17] <jrbarnette> davidjames:  the scope of busybox usage is very limited
[20:36:24] <jrbarnette> basically, only initramfs right now
[20:36:42] <jrbarnette> that said, the &> notation isn't POSIX; it shouldn't be used
[20:37:04] <jrbarnette> I think dash also doesn't support that syntax; it's largely bash specific
[20:37:06] *** saintlou_ has joined #chromium-os
[20:37:07] *** ChanServ sets mode: +v saintlou_
[20:39:07] <davidjames> jrbarnette: Ooh yeah sounds like a booby-trap, if &> doesn't work, cron-lite.conf might chew a lot of CPU but no other notification
[20:39:31] <davidjames> I think init uses dash for sh
[20:40:09] *** saintlou has quit IRC
[20:42:31] *** saintlou_ is now known as saintlou
[20:45:02] <davidjames> jrbarnette: Ok, looked at it some more, looks like the code there *only* works in dash and depends on '&>' not doing what you would expect :)
[20:45:31] <mononoke> davidjames: Ouch!
[20:46:12] *** rbyers_ has quit IRC
[20:46:32] *** rbyers_ has joined #chromium-os
[20:51:46] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "TOT Pre-Flight Queue" from 8282ed6fe6c6c2965fbf99ed4373c141a9b45fd6: joaodasilva at chromium dot org <joaodasilva at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[20:53:15] <saintlou> will try to ping : joaodasilva
[20:54:39] <cmasone> saintlou: why?  There's 0 chance that the failure has anything to do with his change
[20:55:10] <saintlou> I am still trying to digest the output
[20:55:17] <saintlou> of all the logs
[20:55:18] <cmasone> saintlou: that's a chromium change, which will not appear in our builds until tomorrow at the earliest
[20:55:39] <saintlou> cmasone: Chris you are much faster than me, I am just trying to figure this thing out.
[20:55:47] <cmasone> saintlou: scroll to the bottom of the log for the Test stage
[20:55:55] <saintlou> getting there :-)
[20:56:25] <cmasone> saintlou: that's the right place to start, though.
[20:57:11] *** rbyers_ has quit IRC
[20:57:20] *** rbyers_ has joined #chromium-os
[20:57:20] *** ChanServ sets mode: +v rbyers_
[20:58:00] <cmasone> saintlou: how do we make it easier for sheriffs to know that the right thing to do is go to the log of the failing stage and scroll to teh end?
[20:58:15] <cmasone> saintlou: is it confusing that there are multiple red stages?
[20:59:19] <saintlou> cmasone: maybe I am just confused
[21:00:00] *** rbyers_ has quit IRC
[21:00:09] *** rbyers_ has joined #chromium-os
[21:00:10] *** ChanServ sets mode: +v rbyers_
[21:05:09] *** rjkroege has quit IRC
[21:05:16] *** rjkroege has joined #chromium-os
[21:06:48] <davidjames> saintlou: FYI the failure is "chown: cannot access `/home/chronos/Consent To Send Stats': No such file or directory"
[21:08:03] <saintlou> yes davidjames
[21:08:22] <saintlou> davidjames: it seems it was logged and cmasone has assigned the bug. thanks
[21:08:55] <cmasone> http://code.google.com/p/chromium-os/issues/detail?id=19338
[21:09:40] <crosbot> tree became 'Tree is open (http://code.google.com/p/chromium-os/issues/detail?id=19338 assigned)'
[21:16:26] *** rbyers__ has joined #chromium-os
[21:16:27] *** ChanServ sets mode: +v rbyers__
[21:16:35] <gauravsh> that looks related to 19015
[21:17:43] <cmasone> gauravsh: how do we have a p2 release blocker?
[21:18:40] <gauravsh> good question. i figure krisr just forgot to update the priority while adding the label.
[21:21:55] *** zmedico has quit IRC
[21:35:07] *** stevenjb_ has joined #chromium-os
[21:40:33] <crosbot> tree became 'Tree is open (http://code.google.com/p/chromium-os/issues/detail?id=18413 reopened)'
[21:46:08] *** patcito has joined #chromium-os
[22:01:33] *** CoYoTe has joined #chromium-os
[22:05:47] *** CoYoTe has quit IRC
[22:07:00] *** awong has joined #chromium-os
[22:07:00] *** ChanServ sets mode: +v awong
[22:07:22] <awong> hey...I need a new linker for ARM cause the chromium tree is red for the arm build.
[22:08:16] <awong> Anyone know who I can ask to get a blessed version from?  I'd rather not just build a ToT Gold or gnuLD if possible.
[22:08:51] <awong> (this is pretty urgent, caus ethe other options is to take ARM out of hte main build tree for the weekend)
[22:10:48] <grundler> awong, bhaskar I think is the name of the "toolchain" team mgr
[22:12:27] <awong> thanks!
[22:29:07] *** tlambert has quit IRC
[22:34:11] *** rbyers___ has joined #chromium-os
[22:34:12] *** ChanServ sets mode: +v rbyers___
[22:34:12] *** rbyers_ has quit IRC
[22:42:23] *** rharrison_chrome has quit IRC
[22:45:34] *** nirnimesh_ has joined #chromium-os
[22:45:34] *** ChanServ sets mode: +v nirnimesh_
[22:47:52] *** SoulRaven has quit IRC
[22:48:00] *** aaronp_ has joined #chromium-os
[22:49:18] *** aaronp has quit IRC
[22:57:44] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_kaen-binary" from 4785a128f36f207894ff70b7f229fa3ea19815f5: Paul Stewart <pstew at chromium dot org>)'
[23:00:48] <crosbot> tree became 'Tree is closed (reveman investigating "tegra2_kaen-binary" looks like crosbug.com/19342 again)'
[23:02:46] <saintlou> reveman: we can reopen the tree
[23:03:24] <reveman> saintlou: yea, that must be the out-of-order issue again, right?
[23:03:40] <saintlou> reveman: yes I confirmed with kliegs
[23:03:44] <saintlou> we should reopen
[23:04:38] <crosbot> tree became 'Tree is open ("tegra2_kaen-binary" -> crosbug.com/19342)'
[23:04:55] <kliegs> reveman: 19342 is not the bug that closed the tree.  its just the bug that caused the display to look greener than it was
[23:05:37] <kliegs> reveman, saintlou: crbug.com/90675 is what caused the error (git hanging)
[23:06:11] <reveman> kliegs: oh, my bad
[23:06:16] <saintlou> thanks kliegs
[23:07:41] <crosbot> tree became 'Tree is open ("tegra2_kaen-binary" -> crbug.com/90675)'
[23:15:41] *** D|sT has joined #chromium-os
[23:17:21] *** Styx has joined #chromium-os
[23:35:43] *** rsleevi has quit IRC
[23:37:17] *** rsleevi has joined #chromium-os
[23:52:34] *** cros_ has quit IRC
[23:56:26] *** stevenjb_ has quit IRC





top