[00:01:34] <danbeam> sheriffs - just got an email about a CrOS beta buildbot failure from my 1 line (probably untested) JS change - anything I need to do or just ignore?
[00:02:55] *** rsleevi has quit IRC
[00:02:56] *** aaronp has quit IRC
[00:03:18] <jamescook> danbeam: which build failed?
[00:04:27] <danbeam> jamescook: Google Chrome Continuous on cros beta
[00:04:36] *** rsleevi has joined #chromium-os
[00:05:35] *** sosa has joined #chromium-os
[00:05:35] *** ChanServ sets mode: +v sosa
[00:05:59] <jamescook> danbeam: it's probably fine, can you forward me the email?  I'm having trouble finding the failure on the tree(s)
[00:06:02] <jamescook> jamescook@
[00:06:02] <kliegs> sosa: what does it take to kick off a tot arm build of chrome on the buildbots?  to make sure things work good?
[00:06:07] <danbeam> jamescook: yeah
[00:06:16] <kliegs> kicking off the x86 chrome pfq seemed to do a noop build if I read the bots output right
[00:06:17] <sosa> kliegs: TOT?
[00:06:29] *** aaronp has joined #chromium-os
[00:06:35] <kliegs> sosa: I wanted to test arm building with gcc-4.6 on the buildbot. but couldn't figure out how
[00:07:43] <sosa> kliegs: ah.  you can't do it with the buildbot, you'd have to use a local trybot because the Chrome PFQ builds only build when they know they'll producing something different i.e. a new chrome ebuild from a new deps file.
[00:09:15] <kliegs> sosa: hmm.  That's too bad.  I wanted to test that the problem isn't in my local configuration :(
[00:09:31] <kliegs> a bit worried that gcc-4.6 might not build with arm.  but been having trouble testing
[00:09:58] <sosa> ah and you don't want to wait till tmrw night to find out ;)?
[00:10:21] <sosa> wouldn't kicking off any build that builds using gcc-4.6 do that?  do we not have an arm bulder of that?
[00:10:28] *** danbeam has left #chromium-os
[00:10:43] <kliegs> sosa: i'm concerned it won't build chromeos-chrome, have no indication of issues with other packages
[00:10:47] *** aaronp has quit IRC
[00:11:08] <kliegs> sosa: and if i wait until it cycles tomorrow and fails, we'll lose a day playing catchup or trying to get buildspec's rev'd
[00:11:11] <sosa> sorry i don't have any context ... did gcc4-6 just become the default on chromiumos?
[00:11:33] <kliegs> sosa: I believe so. ~1 pm today (unsure which timezone)
[00:12:01] <kliegs> although i couldn't get it to use 4.6 in my chroot but someone else here had it switch on them.  i'm rebuilding my chroot now
[00:12:01] <sosa> kliegs: then the arm builders on the chromium.fyi waterfall that petermayo set up should be building chromium TOT with the chromeos-chorme ebuild with that gcc change
[00:12:13] *** radhermit has quit IRC
[00:12:16] <kliegs> sosa: oh right - I forgot about that
[00:12:35] *** ojn has quit IRC
[00:13:17] *** McMAGIC--Copy has quit IRC
[00:13:18] <kliegs> assuming its picked up 4.6 - i'm not sure why my local chroot didn't. i'll go see what its using
[00:13:26] <sosa> they seem to be doing alright, though not sure if they are using gcc4.6 yet ... they may require a clobber build to see for sure ... also any build that tries to rebuild chrome on our side ... so any arm-full builder
[00:13:44] <sosa> kliegs: it may require a fresh chroot build
[00:14:44] *** marcheu has quit IRC
[00:14:46] <kliegs> sosa: yah :(  do you have a link to that?  I can't find the chromium os builder on chrome
[00:14:52] *** marcheu has joined #chromium-os
[00:15:01] <kliegs> a full builder should catch it, just didn't want to wait that long as they have long cycle times.
[00:15:07] <kliegs> could be 6 hours to see on one of them
[00:15:52] <sosa> kliegs:  i'm not actually sure any "incremental" build that isn't clobbered would see it.  might want to try clobbering one of the builders on the chromium FYI if they aren't seeing it.
[00:16:49] <kliegs> sosa: odd. i just rebuilt my chroot and its still saying 4.4.3 for gcc
[00:17:21] <sosa> then maybe we didn't switch correctly?  i'd follow up with whoever was working on that to ask how they got it
[00:17:30] <kliegs> sosa: he said he just repo sync'd
[00:18:03] <kliegs> setup-board says gcc defaults to 4.6.0 when i do help.  but armv7a-cros-linux-gnueabi-g++ --version says 4.4.3
[00:18:10] *** behdad has quit IRC
[00:20:04] <sosa> kliegs: sorry i dunno ... talk to asharif perhaps?
[00:20:10] *** ojn has joined #chromium-os
[00:20:44] <kliegs> yah. responding on that thread..  confused
[00:23:38] *** vandebo has quit IRC
[00:23:45] *** vandebo has joined #chromium-os
[00:23:46] *** ChanServ sets mode: +v vandebo
[00:24:44] *** behdad has joined #chromium-os
[00:25:06] *** bdaugher has joined #chromium-os
[00:27:57] <kliegs> sosa: if you care I found out I'm not using 4.6.0 because binary packages aren't available and setup_board won't update the toolchain unless you pass --nousepkg if binaries aren't available
[00:28:01] *** saintlou has quit IRC
[00:30:38] *** dianders has left #chromium-os
[00:33:44] *** saintlou has joined #chromium-os
[00:33:44] *** ChanServ sets mode: +v saintlou
[00:35:27] *** gabeblack1 has joined #chromium-os
[00:35:27] *** ChanServ sets mode: +v gabeblack1
[00:36:12] *** tbarzic_ has joined #chromium-os
[00:36:29] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from b935976e2ac9e5290daca3e3cd3ec444addb2cb6: Eric Shienbrood <ers at chromium dot org>)'
[00:39:05] <gauravsh> sque: ^^^^ you are it. :)
[00:39:06] <jamescook> [  FAILED  ] PowerSupplyTest.TestDischargingWithHysteresis
[00:41:31] <jamescook> pinged him on IM
[00:42:22] <crosbot> tree became 'Tree is closed (x86 generic PFQ -> PowerSupplyTest, looking for sque )'
[00:42:43] <jamescook> preparing a revert
[00:43:04] <kliegs> i feel like dramatic music should be starting in the background
[00:43:22] *** nick761 has quit IRC
[00:43:25] *** saintlou has quit IRC
[00:43:49] *** nick761 has joined #chromium-os
[00:45:09] <gabeblack1> should the tree be opened while you prepare the revert?
[00:45:46] <jamescook> not yet, I'm just going to revert
[00:46:12] *** McMAGIC--Copy has joined #chromium-os
[00:46:30] <jamescook> reverted
[00:48:20] *** saintlou has joined #chromium-os
[00:48:20] *** ChanServ sets mode: +v saintlou
[00:49:32] <crosbot> tree became 'Tree is open (x86 generic PFQ -> PowerSupplyTest, reverted, cycling green )'
[00:55:33] <jamescook> I killed the build that was going to fail - was that the right thing to do?
[00:57:12] <kliegs> jamescook: its usually depends on how much time it has left
[00:57:26] <jamescook> It doesn't seem to be auto-restarting, it's been 5 min
[00:57:28] <kliegs> most of the time stopping a build is ok, but there are spots where it can get hung (killing during code syncs)
[00:57:30] <jamescook> Do I need to force restart?
[00:57:45] <kliegs> which builder?  it should resume normally if there was a reason to (like to pickup your revert)
[00:57:59] <jamescook> x86 generic PFQ
[00:58:04] <kliegs> I normally avoid stopping bots for <20 minutes.  more than that my judgement
[00:58:46] <kliegs> jamescook: hmm. it does look confused - go ahead and start it manually
[00:59:28] <jamescook> ok, done
[00:59:29] <jamescook> thanks
[00:59:52] *** snickersnack has joined #chromium-os
[01:03:25] *** McMAGIC--Copy has quit IRC
[01:04:24] *** McMAGIC--Copy has joined #chromium-os
[01:05:01] *** unreal has quit IRC
[01:05:47] *** unreal has joined #chromium-os
[01:06:38] <kliegs> jamescook: might be worth filing a bug on it not restarting - from looking at the waterfall your revert should've triggered a build
[01:06:54] <jamescook> ok
[01:06:57] <kliegs> was just under 9 minutes from kill to force kick - pretty sure a build should start in that time
[01:07:34] <kliegs> make sure to mention that the yellow status line for the buildbot was marked as 'building' - it might help with the debugging
[01:08:35] *** snickersnack has quit IRC
[01:11:20] *** snickersnack has joined #chromium-os
[01:11:59] *** gabeblack1 has quit IRC
[01:12:29] *** aaronp has joined #chromium-os
[01:17:23] *** FusionX has quit IRC
[01:18:42] <gauravsh> folks who work on chrome, I am getting an error when I try to cros_workon start chromeos-chrome
[01:18:45] <gauravsh> http://pastebin.ca/2086579
[01:19:13] <gauravsh> it was working fine a few hours ago (I am following the new chrome for chromeos dev workflow using cros_workon_make)
[01:19:21] *** radhermit has joined #chromium-os
[01:19:26] <gauravsh> "Failed command "['git', 'checkout', '09efa79710f20bdc342bd9389b10fb66ba5ffa12']" with extra env None"
[01:19:40] <gauravsh> while pinning project src/breadpad/src
[01:19:42] <gauravsh> ideas?
[01:20:18] <ferringb> might want to verify that rev exists in the repo
[01:20:40] <ferringb> if it doesn't, that raises questions about a locked rev manifest somehow getting used ahead of the local tree (race)
[01:20:45] <ferringb> ...or I'm guessing. ;)
[01:21:19] <kliegs> gauravsh: is it possible you've repo sync'd your overlay directory but not the rest of your checkout?
[01:21:56] <kliegs> actually I think the pinning it does comes from DEPS.git - its possible that's ahead
[01:22:10] <kliegs> i'd try doing a repo sync and seeing if that fixes it
[01:22:31] *** saintlou has quit IRC
[01:22:33] <gauravsh> i think what I did was to just sync chromium source. let me try repo sync-ing everything again
[01:22:36] <kliegs> I also recommend running repo sync twice - I've had it fail on me at points but the error was too far up it'd scrolled off the screen
[01:22:45] <kliegs> ahh - that's probably why
[01:22:50] *** FusionX has joined #chromium-os
[01:22:58] *** frankf_ has joined #chromium-os
[01:22:58] *** ChanServ sets mode: +v frankf_
[01:23:09] <kliegs> the chromium source is actually many repos all tied together. so if you only sync the top level it won't be able to find the dependent repos
[01:23:13] *** lipsinV2 has joined #chromium-os
[01:23:52] <gauravsh> so if I am in the chromiumos/chromium directory. i can't just do a "repo sync ."?
[01:28:59] <gauravsh> kliegs: thanks, doing a repo sync fixed the issue.
[01:30:06] <kliegs> gauravsh: great. and sorry had tabbed out.
[01:30:21] <kliegs> doing repo sync . won't work in the chromium directory - it relies on too many other files
[01:31:11] <kliegs> well - it might work once checked out, I think once you check it out it relies on you manually syncing - but I'm not 100% certain on that
[01:31:26] <kliegs> i don't have time to review the source tonight -c oming up on 8 pm and I need dinner.  rcui might be able to help
[01:31:35] <kliegs> or feel free to ping me tomorrow if you haven't found an answer and still need one
[01:31:41] *** behdad has quit IRC
[01:32:59] <gauravsh> kliegs: that's fine. it definitely doesn't get everything. i will talk to rcui about how dependencies are handled.
[01:35:09] *** vandebo has quit IRC
[01:35:44] *** vandebo has joined #chromium-os
[01:35:45] *** ChanServ sets mode: +v vandebo
[01:38:01] *** behdad has joined #chromium-os
[01:38:53] *** FusionX has quit IRC
[01:40:53] *** vandebo has quit IRC
[01:41:02] *** vandebo has joined #chromium-os
[01:41:02] *** ChanServ sets mode: +v vandebo
[01:41:04] <kliegs> gauravsh: chrome is definitely complicated to build - a lot of third party dependencies makes things a bit confusing
[01:41:23] <kliegs> not sure if you've been following the discussion on the new git workflow thread on chromium-dev or not
[01:42:51] *** FusionX has joined #chromium-os
[01:47:40] *** FusionX has quit IRC
[01:48:00] <gauravsh> yes, I am aware. i usually build outside the chroot. so it's using gclient et al. for source updates.
[01:49:00] *** jamescook is now known as jamescookAFK
[01:49:08] <kliegs> i'll be interested to see how the GERRIT_SOURCE works out.  It feels a bit more complicated than LOCAL_SOURCE right now but hopefully will become better
[01:49:13] *** SeligArkin has quit IRC
[01:49:25] *** unreal has quit IRC
[01:49:39] *** unreal has joined #chromium-os
[01:50:59] *** behdad has quit IRC
[01:51:40] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from c82530dc8f63b10f648ae8c5f300ea11fe02b446: Andrew de los Reyes <adlr at chromium dot org>, Brian Harring <ferringb at chromium dot org>)'
[01:53:00] *** jamescookAFK is now known as jamescook
[01:53:02] <jamescook> looking
[01:53:24] *** SeligArkin has joined #chromium-os
[01:53:29] <gabeblack> me too
[01:54:16] <gabeblack> two things look suspicious
[01:54:17] <gabeblack> INFO: RunCommand: ['git', 'branch', '-D', 'temp_auto_checkin_branch'] in /tmp/manifests error: branch 'temp_auto_checkin_branch' not found
[01:54:27] <jamescook> ERROR: No update cache found. Please run cros_generate_update_payloads before running this harness.
[01:54:36] <gabeblack> exactly
[01:54:41] <jamescook> I wonder if this happened because I quit/force restarted this build
[01:54:50] <gabeblack> no idea
[01:55:23] <gabeblack> it sounds plausible though
[01:55:50] <gabeblack> adlr had a few checkins, though, and he tends to work on autoupdate related things
[01:55:58] <gabeblack> I'll go check to see if he knows what the problem is
[01:56:21] *** FusionX has joined #chromium-os
[01:57:49] *** saintlou_ has joined #chromium-os
[01:58:48] <jamescook> gabeblack: I'm tempted to reopen and see what happens
[01:58:49] *** unreal has quit IRC
[02:00:37] * ferringb mutters
[02:00:42] *** nirnimesh has quit IRC
[02:01:23] <crosbot> tree became 'Tree is closed (x86 generic PFQ -> VMTest update failure? jamescook/gabeblack looking)'
[02:02:50] <jamescook> sosa: Does "ERROR: No update cache found. Please run cros_generate_update_payloads before running this harness" sound familiar?
[02:03:08] <ferringb> jamescook: offhand, no version of portage in the tree knows of my change... it's basically dead weight till I90ca55a7 lands
[02:03:16] <ferringb> so I'd be very, very *very* impressed if that did it ;)
[02:03:36] <jamescook> Yup
[02:03:48] * ferringb keeps pointing the finger elsewhere meanwhile ;)
[02:05:14] <gabeblack> we're going to revert that change
[02:05:50] *** nirnimesh has joined #chromium-os
[02:05:51] *** ChanServ sets mode: +v nirnimesh
[02:05:52] <jamescook> OK, looks like VMTest failed again
[02:06:02] <gabeblack> or maybe not?
[02:06:04] <jamescook> which change, again?
[02:06:11] <gabeblack> I think there going to try to fix it
[02:06:19] <jamescook> Who is looking?
[02:07:05] <gabeblack> I've been told sosa checked in a big change but accidentally left part of it out
[02:07:10] <jamescook> aha
[02:07:23] <sosa> yup
[02:07:25] <sosa> working on it
[02:07:33] <jamescook> want us to keep closed?
[02:07:35] <sosa> missed a un-revert
[02:07:36] <sosa> yes please
[02:07:42] <sosa> should take only a cpl minutes
[02:08:02] <crosbot> tree became 'Tree is closed (x86 generic PFQ -> autoupdate, sosa fixing)'
[02:09:23] *** behdad has joined #chromium-os
[02:11:01] <sosa> fix pushed
[02:11:53] <ferringb> sosa: curious, do you have direct push rights for the portage snapshot repo?
[02:12:14] <sosa> yes.
[02:13:00] <jamescook> Shall I reopen?
[02:13:06] <gabeblack> I think so
[02:13:24] <crosbot> tree became 'Tree is open (x86 generic PFQ -> autoupdate, sosa pushed fix)'
[02:13:30] <jamescook> I need to head out soon, carpool
[02:13:42] <gabeblack> ok
[02:13:55] <gabeblack> I'll be sheriffing until 7:00 or 7:30
[02:14:44] <ferringb> sosa: 'k.  rev 8708 (loads rather... slowly, since it's a full cache) I'm looking to get merged after 8009
[02:14:59] <ferringb> sosa: mainly asking since the normal controls seem to be missing (+2 specifically)
[02:16:38] *** wfrichar has quit IRC
[02:16:47] <sosa> you want me to commit 8009 + 8708 right after to be atomic-like?
[02:16:55] <ferringb> nah
[02:17:00] <ferringb> 8708 is pointless w/out 8009
[02:17:13] <ferringb> it's more that 8708 doesn't look like it can be pushed in via normal +2/CR route
[02:17:26] <sosa> ah yes
[02:17:30] <sosa> hum
[02:17:34] <ferringb> so... what's the norm for merging to that repo?  just bribe someone who hsa push rights? ;)
[02:17:37] <ferringb> *has
[02:18:05] <sosa> i actually don't know, did davidjames not know?  msb?
[02:18:16] <ferringb> afaik anush has push rights
[02:18:31] <ferringb> but I presume that repo was turned down on rights to keep people from modifying it (which is good, exempting my patch)
[02:21:12] *** BladeFreak has quit IRC
[02:22:11] *** unreal has joined #chromium-os
[02:23:21] <jamescook> see y'all tomorrow, sheriff #1 signing off
[02:23:26] <sosa> temporarily given admin's the ability to +2 that repo
[02:23:39] <sosa> ^^^@ferringb
[02:23:59] <sosa> brb --rebotting
[02:24:03] <sosa> rebooting*
[02:24:08] *** sosa has quit IRC
[02:24:12] *** SeligArkin has quit IRC
[02:24:14] *** jamescook has quit IRC
[02:27:51] *** SeligArkin has joined #chromium-os
[02:31:56] *** sosa has joined #chromium-os
[02:31:56] *** ChanServ sets mode: +v sosa
[02:33:53] *** sadrul has joined #chromium-os
[02:35:38] *** aaronp has quit IRC
[02:36:16] *** sosa has quit IRC
[02:36:47] *** sosa has joined #chromium-os
[02:36:47] *** ChanServ sets mode: +v sosa
[02:37:19] *** BladeFreak has joined #chromium-os
[02:37:23] *** sosa has quit IRC
[02:37:29] *** BladeFreak has quit IRC
[02:37:29] *** BladeFreak has joined #chromium-os
[02:40:02] *** sosa has joined #chromium-os
[02:40:02] *** ChanServ sets mode: +v sosa
[02:44:45] *** aaronp has joined #chromium-os
[02:49:24] *** dhendrix has joined #chromium-os
[02:49:25] *** ChanServ sets mode: +v dhendrix
[02:51:14] *** FusionX has quit IRC
[02:55:28] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_kaen canary" from None: )'
[02:55:52] *** FusionX has joined #chromium-os
[02:57:34] <gabeblack> There seem to be two build failures here
[02:57:51] <gabeblack> glog-0.3.1-r1: src/mock-log.h:43:25: fatal error: gmock/gmock.h: No such file or directory
[02:58:16] <gabeblack> or more completely
[02:58:17] <gabeblack> glog-0.3.1-r1: In file included from src/mock-log_test.cc:34:0: glog-0.3.1-r1: src/mock-log.h:43:25: fatal error: gmock/gmock.h: No such file or directory
[02:58:34] <gabeblack> And then some failures in u-boot
[02:59:41] <gabeblack> I'll start by looking at the u-boot error since I'm familiar with that source
[02:59:52] <gabeblack> chromeos-u-boot-0.0.1-r240: tegra2_mmc.c: In function 'mmc_send_cmd':
[02:59:52] <gabeblack> chromeos-u-boot-0.0.1-r240: tegra2_mmc.c:350:3: error: 'mask' may be used uninitialized in this function [-Werror=maybe-uninitialized]
[02:59:52] <gabeblack> chromeos-u-boot-0.0.1-r240: cc1: all warnings being treated as errors
[03:00:22] <gabeblack> according to the waterfall there don't seem to be any u-boot related checkins in the recent past
[03:01:29] <gabeblack> there were two changes from wad having to do with tests, but at first glance they don't seem to be related
[03:03:09] <marcheu> http://git.chromium.org/gitweb/?p=chromiumos/third_party/u-boot.git;a=commitdiff;h=5cac8a8221c9470c75f81399e3a92eeda28a9851
[03:05:37] *** aaronp has quit IRC
[03:06:42] *** aaronp has joined #chromium-os
[03:07:46] *** chocobo__ has quit IRC
[03:11:09] *** JamesBlonde has joined #chromium-os
[03:12:59] *** vandebo has quit IRC
[03:12:59] *** ScriptRipper has quit IRC
[03:12:59] *** zmedico has quit IRC
[03:13:00] *** deshantm has quit IRC
[03:13:00] *** zork_ has quit IRC
[03:13:00] *** Jahei has quit IRC
[03:13:00] *** Stepan has quit IRC
[03:13:01] *** aaronp has quit IRC
[03:13:01] *** rsleevi has quit IRC
[03:13:01] *** McMAGIC--Copy has quit IRC
[03:13:01] *** nick761 has quit IRC
[03:13:02] *** cooled has quit IRC
[03:13:02] *** patcito has quit IRC
[03:13:02] *** TW1920 has quit IRC
[03:13:02] *** gauravsh has quit IRC
[03:13:02] *** Inumedia has quit IRC
[03:13:02] *** Adys has quit IRC
[03:13:02] *** superm1 has quit IRC
[03:13:02] *** sakoman has quit IRC
[03:13:03] *** falken has quit IRC
[03:13:03] *** lool has quit IRC
[03:13:03] *** ngharo has quit IRC
[03:13:03] *** rjkroege has quit IRC
[03:13:04] *** BladeFreak has quit IRC
[03:13:04] *** katier has quit IRC
[03:13:04] *** flackr has quit IRC
[03:13:04] *** zbehan has quit IRC
[03:13:04] *** BThompson has quit IRC
[03:13:04] *** xiyuan has quit IRC
[03:13:05] *** unreal has quit IRC
[03:13:05] *** behdad has quit IRC
[03:13:05] *** rginda_home has quit IRC
[03:13:05] *** Malmis has quit IRC
[03:13:06] *** robarnold has quit IRC
[03:13:06] *** Ruetobas has quit IRC
[03:13:06] *** yoshiki__ has quit IRC
[03:13:06] *** Sarten-X has quit IRC
[03:13:06] *** seanpaul has quit IRC
[03:13:07] *** shankar has quit IRC
[03:13:07] *** adlr has quit IRC
[03:13:07] *** ukai has quit IRC
[03:13:07] *** ellyjones has quit IRC
[03:13:07] *** sadrul has quit IRC
[03:13:08] *** radhermit has quit IRC
[03:13:08] *** pstew has quit IRC
[03:13:08] *** jaharkes has quit IRC
[03:13:08] *** JamesBlonde has quit IRC
[03:13:09] *** tbarzic_ has quit IRC
[03:13:09] *** lipsinV2 has quit IRC
[03:13:10] *** yosafbridge has quit IRC
[03:13:10] *** Calchan has quit IRC
[03:13:10] *** snickersnack has quit IRC
[03:13:10] *** bdaugher has quit IRC
[03:13:10] *** marcheu has quit IRC
[03:13:10] *** nexusz99 has quit IRC
[03:13:10] *** stalled has quit IRC
[03:13:11] *** bryeung has quit IRC
[03:13:11] *** ttuttle|work has quit IRC
[03:13:11] *** kees has quit IRC
[03:13:11] *** solar has quit IRC
[03:13:11] *** steev has quit IRC
[03:13:16] <gabeblack> did we change compilers recently?
[03:13:30] <ferringb> around 1pm aparetly
[03:13:31] <ferringb> 4.6
[03:13:39] <gabeblack> ah, ok
[03:13:47] <gabeblack> do you know how I can update my chroot?
[03:13:53] <gabeblack> I think the update broke u-boot
[03:14:01] <gabeblack> and this glog thing, apparently
[03:14:10] <gabeblack> there are compiler warnings which are being promoted to errors
[03:14:11] <ferringb> should be able to just pop out of the chroot, and back in, and have it do necessary upgrades (if not, it should)
[03:14:14] <gabeblack> but I'm not getting them
[03:14:17] <ferringb> yeah, new gcc's do that ;)
[03:14:24] <gabeblack> ok, I'll give that a shot
[03:14:31] <gabeblack> well
[03:14:44] *** Greboid has quit IRC
[03:14:46] <gabeblack> actually, the glog thing may be genuinely broken since it couldn't find a header file
[03:15:00] *** Greboid has joined #chromium-os
[03:15:04] <gabeblack> but I'll see what I can do with u-boot
[03:15:11] <gabeblack> if anyone else would like to look at glog in the meantime, that would be helpful
[03:15:57] *** JamesBlonde has joined #chromium-os
[03:15:57] *** aaronp has joined #chromium-os
[03:15:57] *** BladeFreak has joined #chromium-os
[03:15:57] *** sadrul has joined #chromium-os
[03:15:57] *** unreal has joined #chromium-os
[03:15:57] *** behdad has joined #chromium-os
[03:15:57] *** vandebo has joined #chromium-os
[03:15:57] *** lipsinV2 has joined #chromium-os
[03:15:57] *** radhermit has joined #chromium-os
[03:15:57] *** snickersnack has joined #chromium-os
[03:15:57] *** McMAGIC--Copy has joined #chromium-os
[03:15:57] *** nick761 has joined #chromium-os
[03:15:57] *** tbarzic_ has joined #chromium-os
[03:15:57] *** bdaugher has joined #chromium-os
[03:15:57] *** marcheu has joined #chromium-os
[03:15:57] *** rsleevi has joined #chromium-os
[03:15:57] *** rginda_home has joined #chromium-os
[03:15:57] *** Ruetobas has joined #chromium-os
[03:15:57] *** Malmis has joined #chromium-os
[03:15:57] *** pstew has joined #chromium-os
[03:15:57] *** yoshiki__ has joined #chromium-os
[03:15:57] *** ScriptRipper has joined #chromium-os
[03:15:57] *** cooled has joined #chromium-os
[03:15:57] *** patcito has joined #chromium-os
[03:15:57] *** nexusz99 has joined #chromium-os
[03:15:57] *** TW1920 has joined #chromium-os
[03:15:57] *** ukai has joined #chromium-os
[03:15:57] *** ellyjones has joined #chromium-os
[03:15:57] *** gibson.freenode.net sets mode: +vvvv vandebo pstew ukai ellyjones
[03:15:57] *** adlr has joined #chromium-os
[03:15:57] *** shankar has joined #chromium-os
[03:15:57] *** seanpaul has joined #chromium-os
[03:15:57] *** robarnold has joined #chromium-os
[03:15:57] *** Sarten-X has joined #chromium-os
[03:15:57] *** katier has joined #chromium-os
[03:15:57] *** flackr has joined #chromium-os
[03:15:57] *** zbehan has joined #chromium-os
[03:15:57] *** BThompson has joined #chromium-os
[03:15:57] *** gibson.freenode.net sets mode: +vvvv adlr seanpaul katier BThompson
[03:15:57] *** xiyuan has joined #chromium-os
[03:15:57] *** gauravsh has joined #chromium-os
[03:15:57] *** Inumedia has joined #chromium-os
[03:15:57] *** zmedico has joined #chromium-os
[03:15:57] *** stalled has joined #chromium-os
[03:15:57] *** Adys has joined #chromium-os
[03:15:57] *** superm1 has joined #chromium-os
[03:15:57] *** bryeung has joined #chromium-os
[03:15:57] *** gibson.freenode.net sets mode: +vv xiyuan gauravsh
[03:15:57] *** sakoman has joined #chromium-os
[03:15:57] *** falken has joined #chromium-os
[03:15:57] *** rjkroege has joined #chromium-os
[03:15:57] *** ngharo has joined #chromium-os
[03:15:57] *** lool has joined #chromium-os
[03:15:57] *** ttuttle|work has joined #chromium-os
[03:15:57] *** jaharkes has joined #chromium-os
[03:15:57] *** deshantm has joined #chromium-os
[03:15:57] *** kees has joined #chromium-os
[03:15:57] *** zork_ has joined #chromium-os
[03:15:57] *** Jahei has joined #chromium-os
[03:15:57] *** solar has joined #chromium-os
[03:15:57] *** Stepan has joined #chromium-os
[03:15:57] *** yosafbridge has joined #chromium-os
[03:15:57] *** Calchan has joined #chromium-os
[03:15:57] *** steev has joined #chromium-os
[03:16:14] *** benchan has quit IRC
[03:17:20] <gabeblack> my chroot didn't unmount cleanly
[03:17:22] <gauravsh> aebl and kaen have gone red too with what looks like the same failure. revert diander's change?
[03:17:26] <gauravsh> http://gerrit.chromium.org/gerrit/#change,8469
[03:17:26] <gabeblack> and didn't update
[03:17:47] <gabeblack> His change is involved, but I'm not sure that reverting it will help
[03:17:52] <gabeblack> if the compiler change is the real problem
[03:18:01] <marcheu> yeah I was wondering about that
[03:18:13] <gabeblack> if I can reproduce it, it should hopefully be easy to fix
[03:18:32] <gabeblack> but to do that, I need to figure out how to update my chroot
[03:18:47] <marcheu> what's up with it?
[03:19:05] <gabeblack> I'm going to do a build_packages since that says something about updating the chroot
[03:19:27] <gabeblack> unless someone has a better idea?
[03:19:42] <gabeblack> Apparently I don't have the new 4.6 compiler we moved to
[03:20:11] <gabeblack> ok, it looks like it's updating my toolchain
[03:24:10] *** JamesBlonde has quit IRC
[03:25:41] <gabeblack> At least one of these is a false positive
[03:25:50] <gabeblack> I think 4.6.1 cleans up some of 4.6.0's false positives
[03:27:19] <gabeblack> oh, apparently there was only one instance
[03:28:48] <gauravsh> yeah, it is a false positive, but mask should still be initialized to a default.
[03:30:15] *** darinski has joined #chromium-os
[03:30:42] *** ChanServ sets mode: +v darinski
[03:30:42] *** Keybuk has quit IRC
[03:31:01] <crosbot> tree became 'Tree is closed (uboot ARM compilation -> gabeblack investigating)'
[03:31:23] <darinski> anybody reverting http://gerrit.chromium.org/gerrit/#change,8469 ?
[03:31:45] <gabeblack> ok, so u-boot is fixed now
[03:32:12] <gabeblack> next is the glog failure
[03:33:48] <gauravsh> i didn't see that failure on the other red arm builders
[03:34:16] <gabeblack> The commit message of this change makes me suspect it's involved, although it was committed a while ago
[03:34:23] <gabeblack> http://gerrit.chromium.org/gerrit/#change,7035
[03:35:49] <gauravsh> yeah, that's the last change but this was more than a month ago, so very unlikely. Unless it's again the new toolchain.
[03:35:59] <gabeblack> yep
[03:36:10] <gabeblack> it actually just built for me, which is strange
[03:36:21] <gabeblack> maybe some sort of filesystem/syncing problem?
[03:36:26] <gauravsh> it's not happening on all the ARM builders.
[03:36:35] <gabeblack> I don't totally believe that, but I'm not sure what else it might be
[03:36:42] <gabeblack> I'll reopen the tree
[03:37:08] <gauravsh> i suggest forcing a build on the builder which had this error, and waiting to see if it goes green. it is likely that it was a side-effect of the uboot compile failure.
[03:37:41] <crosbot> tree became 'Tree is open (uboot compilation fixed, glog failure consider flaky)'
[03:37:46] *** Sarten-X has quit IRC
[03:37:57] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2 seaboard full" from b48469c464da5749c920350e232b205c3ee974a2: Stefan Reinauer <reinauer at chromium dot org>, tbarzic <tbarzic at google dot com>)'
[03:38:27] <gabeblack> I don't know how to do that, but if you do go ahead
[03:39:07] <gauravsh> i don't know which builder it failed. the sheriff faq has the necessary pointers.
[03:39:22] <gabeblack> seaboard is the same u-boot failure
[03:40:00] <crosbot> tree became 'Tree is open (seaboard failure is fixed u-boot problem)'
[03:42:14] <gabeblack> I'm going to get some dinner
[03:42:16] <gabeblack> bbiab
[03:43:19] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy canary" from None: )'
[03:44:49] <gabeblack> ok, maybe I'll wait a few minutes for dinner
[03:45:06] *** Sarten-X has joined #chromium-os
[03:47:04] <gabeblack> This is a coreboot build breakage, apparently linked to gcc 4.6
[03:47:15] *** sbyer is now known as sbyer_gone_home
[03:47:40] <gabeblack> I just talked to reinauer, our coreboot guru, and he already has a fix ready
[03:47:48] <gabeblack> we'll be getting it into the tree momentarily
[03:48:26] <crosbot> tree became 'Tree is closed (yjlou is looking. "lumpy canary": coreboot build failed)'
[03:49:21] *** saintlou has joined #chromium-os
[03:50:36] *** srao has joined #chromium-os
[03:50:36] *** ChanServ sets mode: +v srao
[03:51:01] *** saintlou_ has quit IRC
[03:51:20] *** darinski has quit IRC
[03:54:50] *** stevenjb has quit IRC
[03:55:07] <gabeblack> yjlou: We have a fix on the way
[03:55:13] <gabeblack> it's another compiler problem
[03:56:30] *** dennisjeffrey has quit IRC
[03:57:19] <gabeblack> ok, the tree should be fixed
[03:57:59] *** petermayo has quit IRC
[03:58:10] <crosbot> tree became 'Tree is open (coreboot build failure under gcc 4.6 has been fixed)'
[03:58:38] <gabeblack> ok, now I'm really going to dinner :-)
[03:58:41] *** petermayo has joined #chromium-os
[03:58:41] *** ChanServ sets mode: +v petermayo
[03:58:44] <gabeblack> good luck, and I'll bbl
[03:59:11] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 pineview full" from fa80156788a386afe207a1286b6390ccb78a8ee2: Matt Tennant <mtennant at chromium dot org>, Stefan Reinauer <reinauer at chromium dot org>, tbarzic <tbarzic at google dot com>)'
[04:02:25] *** saintlou has quit IRC
[04:04:00] *** sadrul has quit IRC
[04:06:51] <crosbot> tree became 'Tree is throttled (Ahmad changed the gcc to 4.6.0. core-boot is just fixed. u-boot seems still failed)'
[04:14:04] *** sadrul has joined #chromium-os
[04:14:04] *** ChanServ sets mode: +v sadrul
[04:22:50] *** grundler has quit IRC
[04:29:22] *** magn3ts has joined #chromium-os
[04:52:31] <gabeblack> I'm back
[04:53:18] <gabeblack> these other u-boot failures look like the same thing
[04:56:15] <crosbot> tree became 'Tree is open (the u-boot build failure for the ARM builds has been fixed already)'
[04:57:16] *** gabeblack has left #chromium-os
[04:57:30] *** gabeblack has joined #chromium-os
[04:57:31] *** ChanServ sets mode: +v gabeblack
[05:08:13] *** petermayo has left #chromium-os
[05:09:49] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex canary" from None: )'
[05:13:08] <crosbot> tree became 'Tree is closed (yjlou is on)'
[05:16:55] *** behdad has quit IRC
[05:17:34] *** behdad has joined #chromium-os
[05:17:34] *** ChanServ sets mode: +v behdad
[05:23:40] *** magn3ts has quit IRC
[05:38:47] *** magn3ts has joined #chromium-os
[05:49:38] *** Kyngdom has joined #chromium-os
[05:55:19] *** BladeFreak has quit IRC
[05:56:37] *** yoshiki__ has quit IRC
[06:12:17] *** behdad has quit IRC
[06:14:34] <crosbot> tree became 'Tree is open (seems VMTest is failed, not related to anyone's CL. Mailed to troopers@ to investigate)'
[06:29:55] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from 036ebbe766db94764b47577af8fb7d40eb5688f1: Stefan Reinauer <reinauer at chromium dot org>)'
[06:31:58] <crosbot> tree became 'Tree is open (crosbug.com/20323)'
[06:59:00] *** abhi_ has joined #chromium-os
[07:00:47] *** abhi_ has left #chromium-os
[07:01:58] *** unreal has quit IRC
[07:10:03] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb_he canary" from None: )'
[07:58:03] *** silverroots has joined #chromium-os
[07:58:59] <silverroots> how to enable chrome browser logging on chrome OS? i can see UI logs, window manager logs...but chrome folder is empty in /var/log/
[08:00:45] *** radhermit has quit IRC
[08:01:54] *** radhermit has joined #chromium-os
[08:02:18] *** silverroots1 has joined #chromium-os
[08:03:36] *** aaronp has quit IRC
[08:04:14] *** silverroots has quit IRC
[08:08:55] *** okamikin has joined #chromium-os
[08:09:53] <okamikin> can I request compatibility for a laptops built in wifi card here?
[08:11:22] *** silverroots1 has quit IRC
[08:16:46] *** okamikin has quit IRC
[08:19:12] *** radhermit has quit IRC
[08:26:29] *** radhermit has joined #chromium-os
[08:41:16] <crosbot> tree became 'Tree is open (VM is flaky)'
[09:07:16] *** Mycro has joined #chromium-os
[09:10:55] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl canary" from None: )'
[09:14:52] *** Mycro has quit IRC
[09:18:57] *** patcito has quit IRC
[09:19:33] *** unreal has joined #chromium-os
[09:19:37] <crosbot> tree became 'Tree is open (tegra2_aebl canary is using old r241, which bug should be fixed in new version)'
[09:28:20] *** snickersnack has quit IRC
[09:29:41] *** wbednarski has joined #chromium-os
[10:21:38] *** darinski has joined #chromium-os
[10:21:38] *** ChanServ sets mode: +v darinski
[10:37:00] *** darinski has quit IRC
[10:37:13] *** darinski has joined #chromium-os
[10:37:14] *** ChanServ sets mode: +v darinski
[10:55:03] *** Adys has quit IRC
[11:03:49] *** radhermit has quit IRC
[11:06:34] *** Adys has joined #chromium-os
[11:11:29] *** radhermit has joined #chromium-os
[11:16:41] *** chocobo__ has joined #chromium-os
[11:16:41] *** ChanServ sets mode: +v chocobo__
[11:16:47] *** Adys has quit IRC
[11:17:00] *** chocobo___ has joined #chromium-os
[11:17:00] *** ChanServ sets mode: +v chocobo___
[11:19:40] *** Adys has joined #chromium-os
[11:21:25] *** chocobo__ has quit IRC
[11:21:25] *** chocobo___ is now known as chocobo__
[11:32:18] *** mmocny has quit IRC
[11:32:19] *** rosso_ has joined #chromium-os
[11:33:16] *** mmocny has joined #chromium-os
[11:33:16] *** ChanServ sets mode: +v mmocny
[12:01:08] *** zirui has joined #chromium-os
[12:02:13] <zirui> Hi, I encountered the server-closed exception when installing depot_tools. Is there a way around this?
[12:04:00] <zirui> HELLO! This is the first time I come here. Is there anyone? Could you respond?
[12:19:49] *** unreal has quit IRC
[12:20:38] *** unreal has joined #chromium-os
[12:32:44] *** darinski has quit IRC
[13:16:22] *** ukai has quit IRC
[13:19:25] *** zork_ has quit IRC
[13:19:34] *** falken has quit IRC
[13:20:58] *** zork_ has joined #chromium-os
[13:21:21] *** falken has joined #chromium-os
[13:27:07] *** kochi has quit IRC
[13:29:57] *** darinski has joined #chromium-os
[13:30:10] *** ChanServ sets mode: +v darinski
[13:30:25] *** kochi has joined #chromium-os
[13:42:50] *** ukai has joined #chromium-os
[13:42:50] *** ChanServ sets mode: +v ukai
[13:46:12] <crosbot> tree became '?'
[13:46:27] <crosbot> tree became 'Tree is open (tegra2_aebl canary is using old r241, which bug should be fixed in new version)'
[14:07:33] *** falken has quit IRC
[14:07:35] *** zork_ has quit IRC
[14:07:40] *** falken has joined #chromium-os
[14:07:47] *** zork_ has joined #chromium-os
[14:32:05] *** unreal has quit IRC
[14:36:37] *** unreal has joined #chromium-os
[14:46:39] *** unreal has quit IRC
[14:47:38] *** unreal has joined #chromium-os
[14:47:39] *** unreal has joined #chromium-os
[14:51:49] *** McMAGIC--Copy has quit IRC
[14:52:22] *** darinski has quit IRC
[14:52:54] *** McMAGIC--Copy has joined #chromium-os
[15:03:08] *** mmocny has quit IRC
[15:03:30] *** mmocny has joined #chromium-os
[15:03:30] *** ChanServ sets mode: +v mmocny
[15:04:01] *** darinski has joined #chromium-os
[15:15:32] *** SeligArkin has quit IRC
[15:16:12] *** rharrison_chrome has joined #chromium-os
[15:16:13] *** ChanServ sets mode: +v rharrison_chrome
[15:18:35] *** SeligArkin has joined #chromium-os
[15:22:41] *** behdad has joined #chromium-os
[15:37:50] *** darinski has quit IRC
[15:37:52] *** behdad has quit IRC
[15:38:57] *** behdad has joined #chromium-os
[15:38:57] *** ChanServ sets mode: +v behdad
[15:40:16] *** unreal has quit IRC
[15:40:43] *** unreal has joined #chromium-os
[15:58:29] *** Ruetobas has quit IRC
[15:58:45] *** Ruetobas has joined #chromium-os
[16:05:07] *** darinski has joined #chromium-os
[16:07:23] *** darinski_ has joined #chromium-os
[16:11:20] <kliegs> zirui: I'm not familiar with that exception. are you still seeing it?
[16:11:21] *** darinski has quit IRC
[16:23:41] *** aaronp has joined #chromium-os
[16:24:32] *** jennb has quit IRC
[16:26:50] *** wfrichar has joined #chromium-os
[16:26:50] *** ChanServ sets mode: +v wfrichar
[16:42:27] *** behdad1 has joined #chromium-os
[16:43:52] *** behdad has quit IRC
[17:03:56] *** rharrison_chrome has quit IRC
[17:05:24] *** wbednars_ has joined #chromium-os
[17:08:56] *** wbednarski has quit IRC
[17:09:01] *** rharrison_chrome has joined #chromium-os
[17:09:01] *** ChanServ sets mode: +v rharrison_chrome
[17:12:56] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "stumpy canary" from None: )'
[17:14:50] *** bshe has quit IRC
[17:19:30] *** darinski has joined #chromium-os
[17:20:13] *** darinski_ has quit IRC
[17:23:29] *** wbednarski has joined #chromium-os
[17:25:23] *** grundler has joined #chromium-os
[17:25:23] *** ChanServ sets mode: +v grundler
[17:50:41] *** darinski has quit IRC
[18:08:17] <crosbot> tree became 'Tree is open (VMTest looks flaky)'
[18:09:43] <cmasone> looks flaky isn't really actionable :-)  What issue is it a reproduction of?
[18:12:41] *** lipsinV2 has quit IRC
[18:15:10] *** rosso_ has quit IRC
[18:32:32] *** satorux_ has quit IRC
[18:43:08] *** sbyer_gone_home is now known as sbyer
[18:44:06] *** Keybuk has joined #chromium-os
[18:44:06] *** ChanServ sets mode: +v Keybuk
[18:45:34] *** mmocny has quit IRC
[18:46:40] *** mmocny has joined #chromium-os
[18:46:40] *** ChanServ sets mode: +v mmocny
[18:47:21] *** dennisjeffrey has joined #chromium-os
[18:47:22] *** ChanServ sets mode: +v dennisjeffrey
[18:59:01] *** davidjames has joined #chromium-os
[18:59:02] *** ChanServ sets mode: +v davidjames
[19:00:22] *** behdad1 has quit IRC
[19:05:13] *** patcito has joined #chromium-os
[19:05:36] *** tbarzic_ has quit IRC
[19:09:54] <kees> morning!
[19:11:49] <ellyjones> afternoon, heathen :)
[19:12:13] *** nexusz99 has quit IRC
[19:23:04] <kees> ellyjones: s/heathen/left-coaster/ :)
[19:23:35] <kees> I should just attempt "start-of-my-work-day-greetings!" :)
[19:23:36] *** ChanServ sets mode: +v ojn
[19:25:10] *** wbednarski has quit IRC
[19:25:35] *** jamescook has joined #chromium-os
[19:25:35] *** ChanServ sets mode: +v jamescook
[19:25:43] *** wbednarski has joined #chromium-os
[19:26:23] <jamescook> Build sheriff reporting for duty
[19:30:22] *** wbednarski has quit IRC
[19:42:48] <grundler> kees: or teach crosbot to answer you back nicely faster than elly can :)
[19:44:12] <kees> grundler: hehe
[19:45:01] *** jennyz has joined #chromium-os
[19:45:02] *** ChanServ sets mode: +v jennyz
[19:45:06] <kees> grundler: I had cmasone already complaining about the "good" part of "good morning". :)  "<TZ-appropriate-greeting>!"
[19:53:11] <jamescook> if user.find("masone") != -1: greeting += ", chump"  :-)
[19:54:18] <kees> hehe
[19:55:16] *** sbyer is now known as sbyer_afk
[19:56:31] *** frankf_ has quit IRC
[20:00:41] *** wbednarski has joined #chromium-os
[20:10:10] *** sbyer_afk is now known as sbyer
[20:11:11] *** jamescook is now known as jamescookAFK
[20:13:26] <kees> okay... I have a portage question. redpig has suggested perhaps sosa might know :)
[20:13:46] <kees> I noticed that autotest-tests's platform_ToolchainOptions installs its own binutils.
[20:13:55] <kees> I wanted to fix that to use sys-devel/binutils instead
[20:14:09] <kees> but I hit: http://pastebin.ubuntu.com/702366/
[20:14:36] <kees> and digging in showed me: http://pastebin.ubuntu.com/702365/
[20:14:43] <kees> and now I'm stuck :P
[20:19:47] *** jamescookAFK is now known as jamescook
[20:20:25] <redpig> the silence is palpable :p
[20:20:40] * ttuttle|work crickets.
[20:21:14] *** unreal has quit IRC
[20:21:28] *** unreal has joined #chromium-os
[20:21:41] * kees stares at Keybuk
[20:23:32] <Keybuk> hmm?
[20:23:52] <kees> Keybuk: any ideas on the portage mystery above?
[20:23:58] <Keybuk> no, but let me know when you find out :)
[20:24:02] <kees> haha
[20:25:24] <redpig> the package.provided stuff seemed messy last I looked because if it was allowed it could cause weirdness during crossdev builds
[20:25:30] <redpig> but I may have just not understood it well enough
[20:27:46] *** sbyer has quit IRC
[20:33:51] *** sbyer has joined #chromium-os
[20:33:51] *** ChanServ sets mode: +v sbyer
[20:36:13] <marcheu> kees: binutils isn't packaged, it's installed by setup_board when it setups the toolchain
[20:36:20] *** jochen__ has quit IRC
[20:36:25] *** jochen__ has joined #chromium-os
[20:36:25] *** ChanServ sets mode: +v jochen__
[20:42:35] <kees> marcheu: what would it take to get it properly installed on the device?
[20:42:58] *** maruel has joined #chromium-os
[20:42:58] *** ChanServ sets mode: +v maruel
[20:43:11] <maruel> expect some turbulence on the waterfall for the current cycle
[20:43:17] <maruel> (my fault)
[20:44:42] <jamescook> marcheu: I'm sheriff today, anything I need to do?
[20:45:35] <jamescook> er, maruel: I'm sheriff today, anything I need to do?
[20:45:39] <jamescook> damn autocomplete
[20:46:26] <marcheu> kees: look at what setup_board does; now with that said binutils shouldn't install on the device right?
[20:46:38] <marcheu> kees: or does autotest need binutils for some reason
[20:49:46] <gabeblack> I'm a sherrif and I'm here for duty
[20:50:07] <jamescook> Good morning
[20:50:14] <gabeblack> :-)
[20:50:36] <srao> marcheu: kees: you may already know this, but I think the issue with installing binutils is that gentoo refuses to install it while the package is listed in /etc/portage/profile/package.provided (in the chroot)
[20:50:36] <srao> you can remove it from there
[20:51:20] <marcheu> but I think it's in package.provided because setup_board installs it
[20:51:26] <marcheu> as part of the toolchain
[20:51:37] <kees> marcheu: right, I'm making a change to platform_ToolchainOptions and it needs objdump
[20:51:41] <marcheu> my point is you probably don't want to touch that :p
[20:52:13] <kees> marcheu: hrm, but I don't care about the chroot -- I just want it on the device
[20:52:14] <marcheu> i.e. either update the version in setup_board or don't touch it
[20:53:19] <marcheu> really I'd read setup_board thoroughly first to understand the subtleties
[20:53:48] <srao> well I think the test has to "side-load" the package because it wants it on the target, right?
[20:54:01] <marcheu> right
[20:54:06] <kees> sideload?
[20:54:18] <srao> well by sideload I mean bypass the packaging system
[20:54:28] <kees> oh, well, right now that's what it does.
[20:54:35] <srao> yup
[20:54:49] <kees> I'm trying to understand why this is needed. can't it just cross-compile binutils and toss it on the device?
[20:55:13] <marcheu> well that's what it does right
[20:55:46] <kees> er, right now the test itself does the build
[20:56:02] <kees> that seems rather insane given that we have portage for doing such things :)
[20:56:36] <maruel> jamescook: nothing special
[20:56:48] <jamescook> ok
[20:56:49] <marcheu> kees: but if you produce an ARM binutils how do you run it in the chroot now :)
[20:57:19] <kees> marcheu: right, that's what I mean. isn't this what the cross compiler logic is for?
[20:57:34] <kees> marcheu: I don't want to run it on the chroot. this is for a test on the device.
[20:58:07] <marcheu> yeah but then you'd override the cross compiling binutils with a native binutils
[20:58:49] <kees> I must clearly be misunderstanding something.
[20:59:10] <srao> I'm not sure if the reason we put the toolchain into package.provides is because we use a cross compiler, but i suspect it's related
[20:59:10] <kees> there must be a way to build packages for the chroot and for the device separately. otherwise libc could never be built
[21:01:01] <srao> well like marcheu said, setup_board kind of "provides" libc to the gentoo system so that might be the thing to look at
[21:01:19] <marcheu> srao: I think that's precisely the reason it's in there, so tell emerge-target not to build a native one
[21:03:56] *** echelog-1 has joined #chromium-os
[21:04:29] <marcheu> kees: so you basically tell portage "look, I have that package, don't try to create a native version that won't work on my chroot"
[21:04:31] <srao> marcheu: well you can force it to build libc if you pass --nousepkg to setup_board though
[21:04:42] <srao> so it must have some logic to handle that case
[21:04:46] <marcheu> but libc is very different
[21:04:55] <marcheu> libc doesn't cross compile
[21:04:58] <marcheu> gcc and binutils do
[21:05:34] <srao> so, you're saying if i pass --nousepkg to "./setup_board --board=arm-generic" it would fail ?
[21:05:49] <marcheu> I think so right?
[21:05:55] <srao> i guess I should try it
[21:06:01] <srao> that seems broken :-)
[21:06:18] <marcheu> well, cross compiling will never be 100% like native, I mean...
[21:06:37] <srao> yeah, but *someone* has to  build the arm glibc and we're not using arm native builders
[21:06:54] <srao> so it has to be cross compiling it somehow on an x86 builder
[21:07:17] <marcheu> don't we get a pre-made package for that?
[21:07:25] <marcheu> again look at setup_board it does all this
[21:07:28] <srao> sure, but even if that's the case, my point stands
[21:09:12] <srao> in any case, I'm trying it out now :-)
[21:09:25] <kees> I remain confused. either it can build things into two places (chroot and device), or it can't. if it can't, then why isn't glibc destroyed when cross-compiling for ARM? if it can, why can't I build binutils for the device and not wipe out the chroot's binutils?
[21:10:20] <marcheu> kees: because it's in package.provided
[21:11:01] <kees> heh. okay, I am full circle. I will go read more gentoo docs and setup_board
[21:11:20] <marcheu> having to hack on setup_board helped me understand a lot of things
[21:11:30] <marcheu> not that I understand the whole script now :p
[21:13:29] *** gabeblack1 has joined #chromium-os
[21:13:29] *** ChanServ sets mode: +v gabeblack1
[21:37:47] *** gabeblack1 has quit IRC
[21:40:46] <srao> fyi, I think "./setup_board --board=arm-generic --nousepkg --force" worked for me
[21:40:59] <marcheu> what did it do?
[21:41:19] <srao> it built binutils, then gcc, then glibc
[21:41:30] <marcheu> did it make a cross compiling toolchain?
[21:42:03] <marcheu> cause I don't think it can do anything else
[21:42:12] <srao> yeah I think it did but double checking
[21:43:36] <srao> strange, I just re-ran it and it didn't rebuild anything this time
[21:43:54] <marcheu> did you make the switch to 4.6.0 last time?
[21:43:57] <srao> it definitely built stuff the first time... most had scrolled off the screen
[21:44:03] <srao> yeah It did that
[21:44:20] <marcheu> well then that's why
[21:44:35] <srao> yeah i think you're correct
[21:44:43] <srao> it was building the 4.6 toolchain
[21:48:14] <srao> so i'm not sure where it gets the arm libc from
[21:48:25] <srao> someone bootstraps it for you and you pull down the binaries?
[21:48:54] <marcheu> yes
[21:49:03] <marcheu> at least that's my limited understanding
[21:49:17] <marcheu> IIRC setup_board gets the tarballs, unpacks
[21:49:30] <marcheu> you will want to do a setup_board from scratch and look at /build/x86-generic
[21:49:31] <srao> yeah usually, but I thought --nousepkg overrides that
[21:49:42] <marcheu> AFAIK nousepkg doesn't work with setup_board
[21:49:46] <srao> overrides the pre-built binary blobs
[21:49:56] <srao> i've used it before on x86 targets
[21:50:05] <marcheu> was that before the cros_sdk switch though
[21:50:07] <srao> when I was trying out gcc 4.6 and newer libc a few months ago
[21:50:08] <srao> yes it was
[21:50:30] <marcheu> I think it was different. where is davidjames anyway :)
[21:59:51] *** Hexxeh_ is now known as Hexxeh
[22:01:12] <davidjames> marcheu: --nousepkg works fine with setup_board, but it picks latest version of toolchain, maybe a version that has not been tested with sdk builder yet
[22:01:17] <crosbot> tree became 'Tree is closed (Automatic: "update" on "x86-mario canary")'
[22:01:45] <davidjames> marcheu: setup_board --usepkg picks latest version marked as stable by sdk builder. Announcement yesterday about gcc 4.6.0 switch was premature because it was sent out before the builder had marked it as stable
[22:03:01] <jamescook> gabeblack: you around?
[22:03:07] <jamescook> looking
[22:04:00] <marcheu> davidjames: so, the reason autotest uses a hand-rolled binutils is because we can't build our own, as it would overwrite the {cross compiled binutils, added to package.provided} binutils right?
[22:04:27] <jamescook> davidjames: are you doing something with x86-mario canary?
[22:04:32] <davidjames> marcheu: I don't know why autotest uses a hand-rolled binutils
[22:04:57] <marcheu> well they want to put a native version of objdump on the device
[22:05:15] <davidjames> jamescook: Looks like djmm hit force build, ask him
[22:05:29] <jamescook> sorry, wrong nick
[22:05:31] <marcheu> so my thinking is that it's not possible to do that with the current setup_board scheme because we can only make cross toolchains
[22:07:09] <crosbot> tree became 'Tree is open (x86-mario canary recovering from dead machine)'
[22:08:48] <jamescook> gabeblack: I'm going to step out for lunch, back soon
[22:08:52] *** jamescook is now known as jamescookAFK
[22:09:23] <davidjames> marcheu: Installing a native objdump using emerge sounds a bit tricky, vapier might have ideas on how to do it
[22:09:51] *** sbyer is now known as sbyer_at_lunch
[22:09:53] <marcheu> davidjames: right, you'd need to remove it from package.provided, but we can't do that because it's brought by setup_board
[22:10:02] <vapier> there's no real way to install just objdump without getting all of binutils
[22:10:57] <davidjames> vapier: marcheu would be happy to have all of binutils, but would want it only on test images, and would also want version built for target
[22:11:16] <marcheu> well it's not for me, it's for kees, but same problem :)
[22:11:39] <vapier> ok, that's a problem we can solve ... splitting up binutils is dubious
[22:12:18] *** saintlou has joined #chromium-os
[22:12:18] *** ChanServ sets mode: +v saintlou
[22:15:37] *** unreal has quit IRC
[22:16:22] <vapier> davidjames: here's a trick that you might think ugly ...
[22:16:57] <vapier> we specify sys-devel/binutils-0 in package.provided and >=sys-devel/binutils-1 in chromeos-test
[22:17:26] <vapier> binutils won't normally get installed as we've got a stub for it, but when people really want it, it'll get pulled in
[22:18:57] <davidjames> vapier: Sneaky sneaky :) Is there any non-hacky way to do it?
[22:19:21] <davidjames> gabeblack: zgb_he canary has been red for long time with consistent update failures
[22:19:52] <davidjames> gabeblack: They seem to be reproducible on every build, and were happening before the gcc upgrade so are not related to that
[22:20:22] <vapier> davidjames: it depends on why we have binutils in package.provided in the first place
[22:23:17] <davidjames> vapier: Well it's to indicate we're using a cross toolchain and don't need the target toolchain just to build stuff
[22:23:50] <davidjames> vapier: Target toolchain should only be installed if it's a runtime dep, but not if it's a build dep as build deps really should be HDEPEND
[22:24:20] <vapier> why would it be installed at all
[22:24:24] <vapier> it should be part of the system target
[22:24:37] <vapier> if you don't want it in the cros system target, then we should update the profiles to delete it
[22:31:48] <vapier> i do see a few packages that have a depend on binutils, but i'm pretty sure they're broken and/or should be changed to a blocker
[22:32:28] <vapier> in fact, x-modular.eclass is makin' me angry
[22:32:32] <vapier> i'll bug the x11 guys
[22:56:33] *** jamescookAFK is now known as jamescook
[23:03:03] *** jamescook is now known as jamescookAFK
[23:05:49] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "lumpy canary" from None: )'
[23:07:52] *** sbyer_at_lunch is now known as sbyer
[23:08:53] *** rharrison_chrome has quit IRC
[23:08:55] <gabeblack> The lumpy canary failure looks like that same vmtest failure again
[23:09:55] <crosbot> tree became 'Tree is open (flake: crosbug.com/20323 or crosbug.com/20887)'
[23:18:13] *** mdempsky has joined #chromium-os
[23:18:13] *** ChanServ sets mode: +v mdempsky
[23:40:06] <mdempsky> Did the amd64 toolchain change recently?  Yesterday I was able to build most packages for amd64-generic, but now I keep getting "cannot find -lgcc_s" errors.
[23:42:41] <srao> mdempsky: yes it did
[23:42:58] <srao> mdempsky: there's a known problem, not sure if a bug has been opened yet
[23:43:14] <mdempsky> srao: ah
[23:44:05] <srao> mdempsky: you can hack around it by doing "cd /usr/lib/gcc/x86_64-cros-linux-gnu/4.6.0/gcc/x86_64-cros-linux-gnu/  && mv lib64/libgcc_s.so* 4.6.x-google/ -i && rmdir lib64" (as root)
[23:44:21] <srao> there's some other stuff which seems to be broken though on amd64-generic
[23:44:33] <srao> vapier is investigating
[23:45:37] <mdempsky> Yeah, I was working on fixing some other build failures, then suddenly everything started failing on me. :)
[23:45:51] *** jamescookAFK is now known as jamescook
[23:46:30] <mdempsky> srao: Thanks, that workaround seems to let me build stuff again.
[23:46:45] <srao> thanks routed to vapier :-)
[23:54:04] *** echelog-1 has joined #chromium-os
[23:57:50] <srao> ok I filed crosbug.com/21260 for the 64bit gcc-4.6 issue
[23:58:29] <mdempsky> cool
[23:59:20] *** gabeblack has quit IRC





top