[00:08:00] *** perezd has joined #chromium-os
[00:17:15] *** kochi has joined #chromium-os
[00:30:04] *** behdad has quit IRC
[00:33:00] *** sadrul has quit IRC
[00:34:45] *** powderluv_ has joined #chromium-os
[00:34:45] *** ChanServ sets mode: +v powderluv_
[00:37:39] *** powderluv has quit IRC
[00:37:39] *** powderluv_ is now known as powderluv
[00:47:55] *** BladeFreak has joined #chromium-os
[00:48:33] <Fractal> I?m trying to add an input method to Chromium OS, I edited the "chrome/browser/chromeos/input_method/ibus_input_methods.txt" file in the chromium source tree, but it seems I have now to run the script "tools/gen_keyboard_overlay_data/gen_keyboard_overlay_data.py", but I need a @google.com account for that (?)
[00:49:15] <Fractal> Why should I need to work at Google to add an input method?
[00:53:20] <adlr> weird
[00:53:29] <adlr> for some reason, it downloads a spreadsheet
[00:53:33] <adlr> that's private to google.com
[00:53:44] *** superm1 has joined #chromium-os
[00:53:58] <adlr> Fractal: this may be a good one for email, since a lot of the folks that do input stuff are in Japan
[00:53:59] *** mtennant has quit IRC
[01:01:59] <Fractal> adlr: What do you mean? Should I send a mail to someone working on the input stuff?
[01:03:12] <adlr> i'd email chromium-os-dev
[01:04:04] <adlr> or maybe wait a few hours and maybe some Japan folks will come on here
[01:04:17] <Fractal> adlr: Thanks, I will do this
[01:06:42] <Fractal> Hmm it?s 1am here, I?ll be sleeping in a few hours ;-)
[01:19:41] *** Adys has quit IRC
[01:42:56] *** Sarten-X has quit IRC
[01:45:30] *** lipsinV2 has joined #chromium-os
[02:01:56] *** rbyers has joined #chromium-os
[02:06:31] *** Sarten-X has joined #chromium-os
[02:15:14] *** leavittx has quit IRC
[02:16:07] <Fractal> I?m trying to replace the Chrome OS kernel by a Chromium OS kernel (but keeping the Chrome OS root fs), I?ve copied the modules from Chromium OS to the Chrome OS root fs, and everything seems to work, except the touchpad (multi touch does not work)
[02:16:34] <Fractal> I don?t see anything wrong in dmesg or Xorg.0.log, the syntp driver seems to be loaded
[02:16:42] <Fractal> What am I missing?
[02:17:56] *** 18VAAFUBP has joined #chromium-os
[02:17:56] *** ChanServ sets mode: +v 18VAAFUBP
[02:19:25] *** 18VAAFUBP has quit IRC
[02:30:31] <ojn> ... must ... resist ... qemacs vs vim debate. :-)
[02:31:12] <ojn> Fractal, there are no kernel components to the closed trackpad driver. what branch of the rootfs do you have? R13?
[02:31:35] <ojn> Fractal, this is on chromeos hardware, right?
[02:33:33] *** saintlou has quit IRC
[02:33:39] *** saintlou has joined #chromium-os
[02:33:39] *** ChanServ sets mode: +v saintlou
[02:37:01] *** wfrichar has quit IRC
[02:42:01] *** Bigbear1 has joined #chromium-os
[02:45:47] <adlr> Fractal: random guess... insmod serio_raw?
[02:51:42] *** rbyers has quit IRC
[02:52:53] <darinski> MTV sheriffs going off duty... be gentle with the tree.
[02:56:46] * zbehan hides the sledgehammer
[03:04:09] *** davidjames has quit IRC
[03:09:14] <Fractal> adlr: interesting, the serio_raw module only exists for the Chrome OS kernel (2.6.32.26+drm33.12), and for the Chromium OS kernel (2.6.38.3+) there are ct82c710.ko, pcips2.ko and serport.ko in the directory where serio_raw.ko should be
[03:09:39] <adlr> huh, maybe you need to enable it?
[03:10:23] <Fractal> adlr: Yes, perhaps, but it works fine in the full Chromium OS image
[03:11:00] <Fractal> How can I enable it?
[03:11:43] <adlr> not sure offhand, i assume it's a kernel module you can enable via make menuconfig, but i don'tknow how to fit that in w/ our build system
[03:12:34] <Fractal> ojn: yes, it?s on a Samsung Series 5 Chromebook. I don?t know the branch of the rootfs, but it?s the most recent official Google rootfs
[03:13:12] <marcheu> go to src/third_party/kernel/files and run chromeos/scripts/kernelconfig editconfig
[03:13:24] <ojn> Fractal, Ah, switch to dev channel and you should get a newer rootfs that has the modules you need
[03:13:30] <marcheu> that'll give you one menuconfig per config, one of them is chromium os
[03:14:50] <Fractal> ojn: for the modules, I took them from my last Chromium OS build (I synced and built it earlier today)
[03:15:22] <ojn> Oh, serio_raw is built-in on .38, not a module
[03:15:28] *** mtennant has joined #chromium-os
[03:15:28] *** ChanServ sets mode: +v mtennant
[03:16:02] <mtennant> California sheriff signing off for the day
[03:16:19] *** mtennant has quit IRC
[03:16:45] <Fractal> ojn: But I?m running the .38 kernel, so it should work
[03:17:04] <ojn> I would think so yeah
[03:17:04] <Fractal> Is there a way to know if serio_raw is working on my kernel?
[03:18:17] <marcheu> Fractal: I'm wondering though, you built chromiumos right? if so you won't get the syntp driver but instead will get synaptics or ps/2
[03:18:32] <marcheu> can you pastebin your X log?
[03:19:41] <Fractal> marcheu: Yep, I built Chromium OS, then I dd?ed the kernel partition in the Chrome OS KERN_A, copied /lib/modules/* from Chromium OS to Chrome OS, and I?m trying to run the Chrome OS rootfs with the Chromium OS kernel
[03:20:13] <marcheu> pastebin your X log, lets see which driver you're running first
[03:21:47] <Fractal> I?m trying, but I only have the left mouse button ^^
[03:22:54] <marcheu> atl-click
[03:23:54] *** sbyer is now known as sbyer_afk
[03:24:37] <Fractal> Here is it : http://pastebin.archlinux.fr/433235
[03:25:49] <marcheu> yeah synaptics init is failing
[03:26:03] <marcheu> do you have the syntpenh daemon running? it should be a 25mb process, impossible to miss :p
[03:27:05] *** thieule has quit IRC
[03:27:18] <Fractal> No, "ps -e | grep synt" gives no result
[03:28:17] <marcheu> hmm you'll need that, but I'm not really sure how to start it
[03:28:52] <marcheu> failing that, you can replace syntp with synaptics in your xorg.conf, and that will use the open source synaptics driver
[03:33:31] <Fractal> marcheu: thanks, replacing syntp with synaptics works. By the way, why would syntp be better? I only found that drag-and-drop with two fingers does not work with synaptics, but that?s not very convenient anyway
[03:34:04] <marcheu> yeah syntp is a bit different, it has drag & drop but doesn't do coasting
[03:34:09] <marcheu> compared to synaptics
[03:34:51] *** rbyers has joined #chromium-os
[03:34:57] <Fractal> what is coasting?
[03:35:27] *** rbyers has quit IRC
[03:35:45] <marcheu> when you do two finger scroll, it keeps scrolling after your fingers go off the pad
[03:35:52] <marcheu> for a little bit
[03:36:20] <marcheu> also synaptics lets me scroll with a horizontal thumb, which I prefer
[03:48:47] *** sadrul has joined #chromium-os
[03:48:47] *** ChanServ sets mode: +v sadrul
[03:58:03] *** D|sT has quit IRC
[04:06:28] *** jrbarnette has quit IRC
[04:10:59] *** D|sT has joined #chromium-os
[04:13:06] *** sergiu has quit IRC
[04:14:13] *** Bigbear1 has left #chromium-os
[04:20:33] *** saintlou has quit IRC
[04:51:34] *** jochen__ has quit IRC
[04:51:39] *** jochen__ has joined #chromium-os
[04:51:39] *** ChanServ sets mode: +v jochen__
[04:58:49] *** D|sT has quit IRC
[05:05:11] *** JasonO has quit IRC
[05:06:50] *** D|sT has joined #chromium-os
[05:19:13] *** jrbarnette has joined #chromium-os
[05:19:32] *** jrbarnette_ has joined #chromium-os
[05:23:49] *** jrbarnette has quit IRC
[05:23:50] *** jrbarnette_ is now known as jrbarnette
[05:24:50] *** achuith has quit IRC
[05:24:58] *** achuith has joined #chromium-os
[05:24:58] *** ChanServ sets mode: +v achuith
[05:25:15] *** oc80 has quit IRC
[05:26:45] *** oc80z has joined #chromium-os
[05:50:44] *** jrbarnette_ has joined #chromium-os
[05:54:10] *** jrbarnette has quit IRC
[05:55:16] *** jrbarnette_ has quit IRC
[06:25:57] *** Fractal has quit IRC
[06:43:22] *** powderluv has quit IRC
[06:50:45] *** powderluv has joined #chromium-os
[06:50:46] *** ChanServ sets mode: +v powderluv
[06:55:27] *** powderluv has quit IRC
[07:08:48] *** cooled has quit IRC
[07:15:21] *** perezd has quit IRC
[07:27:24] *** achuith has quit IRC
[07:27:32] *** achuith has joined #chromium-os
[07:27:32] *** ChanServ sets mode: +v achuith
[08:19:00] *** sadrul has quit IRC
[08:41:56] *** sadrul has joined #chromium-os
[08:55:44] *** Clark008 has quit IRC
[08:58:53] *** jujugre has joined #chromium-os
[09:23:13] *** yosafbridge has quit IRC
[09:24:57] *** yosafbridge has joined #chromium-os
[09:26:53] *** borios has joined #chromium-os
[09:35:29] *** vmil86 has joined #chromium-os
[09:58:29] *** Styx has joined #chromium-os
[10:06:32] *** jujugre has left #chromium-os
[10:08:03] *** patcito has quit IRC
[10:08:20] *** jujugre has joined #chromium-os
[10:10:20] *** kochi has quit IRC
[10:11:51] *** Fractal has joined #chromium-os
[10:23:16] *** borios_ has joined #chromium-os
[10:23:17] *** borios has quit IRC
[10:28:13] *** kochi has joined #chromium-os
[10:35:47] *** Sergiu has joined #chromium-os
[10:41:42] *** Sergiu has quit IRC
[10:52:47] *** mazda has quit IRC
[10:54:17] *** mazda has joined #chromium-os
[11:13:44] *** leavittx has joined #chromium-os
[11:14:12] *** leavittx has quit IRC
[11:20:18] *** unreal has quit IRC
[11:22:41] *** unreal has joined #chromium-os
[11:54:13] *** borios_ has quit IRC
[12:08:52] *** borios has joined #chromium-os
[12:10:16] *** borios has quit IRC
[12:26:50] *** BladeFreak has quit IRC
[12:30:43] *** BladeFreak has joined #chromium-os
[12:38:33] *** lipsinV2 has quit IRC
[12:46:50] *** lipsinV2 has joined #chromium-os
[12:52:03] *** jujugre has quit IRC
[13:56:48] *** jujugre has joined #chromium-os
[14:17:45] *** pastarmovj has joined #chromium-os
[14:18:54] <pastarmovj> Hi any chromeos sheriffs around? Which bot are we waiting to cycle green now?
[14:28:56] *** dubroy has joined #chromium-os
[14:28:56] *** ChanServ sets mode: +v dubroy
[14:29:12] <pastarmovj> .
[14:32:32] *** dubroy_ has joined #chromium-os
[14:33:13] *** dubroy has quit IRC
[14:33:14] *** dubroy_ is now known as dubroy
[14:47:32] *** rharrison_chrome has joined #chromium-os
[14:47:32] *** ChanServ sets mode: +v rharrison_chrome
[14:48:52] *** cwolfe has quit IRC
[14:49:02] *** cwolfe has joined #chromium-os
[14:49:02] *** ChanServ sets mode: +v cwolfe
[14:53:47] *** Fractal has quit IRC
[15:05:53] *** sadrul has quit IRC
[15:09:54] *** quiccker has quit IRC
[15:13:05] *** quiccker has joined #chromium-os
[15:19:33] *** Fractal has joined #chromium-os
[15:35:42] <jglasgow> pastarmovj: Just arrived as sheriff.  Let me see what's up.
[15:38:54] <pastarmovj> great thanks :)
[15:51:33] <pastarmovj> jglasgow: any news on the bots?
[15:53:12] <jglasgow> I don't see much other than Hungte's comment that tegra2_aebl-binary should cycle green.  I haven't yet dug up why he thinks so -- but since he throttled the tree, I am comfortable with you committing, if that is what you are looking to do.
[15:54:52] <jglasgow> It does not look to me like the tegra2_abel-binary builder is actually building.
[15:55:02] <jglasgow> I'm trying to sort it out.
[15:56:51] <pastarmovj> ok i will commit a small change to libchromeos which should be safe
[15:57:18] <pastarmovj> in 10-15minutes though
[15:57:46] <jglasgow> okay.
[15:59:01] *** JasonO has joined #chromium-os
[15:59:01] *** JasonO has joined #chromium-os
[16:00:24] *** Styx has quit IRC
[16:03:00] *** behdad has joined #chromium-os
[16:03:01] *** ChanServ sets mode: +v behdad
[16:05:42] *** Styx has joined #chromium-os
[16:06:58] <jglasgow> It is unclear to me why tegra2_aebl-binary says "building 6 pending", but the waterfall shows nothing since 4:44PDT.  Can anybody explain?
[16:09:09] <cjb> Hi, sorry if I'm doing the wrong thing by asking, but:
[16:09:28] <cjb> I'm trying to run a tegra2 build, and alsa-utils is failing with:
[16:09:29] <cjb> checking for new_panel in -lpanel... no
[16:09:29] <cjb> configure: error: panel library not found
[16:12:16] *** sadrul has joined #chromium-os
[16:12:16] *** ChanServ sets mode: +v sadrul
[16:24:03] <kliegs> cjb: I think I saw a bug like that ~1 week ago where the toolchain wasn't properly setting some cross compiler options
[16:24:16] <kliegs> when did you create your chroot?
[16:24:57] <cjb> about a week ago :)
[16:25:03] <kliegs> cjb: And if you look at the actual configure log in detail from the build does it mention an illegal object format in /lib64/libsomething (I forget which actual lib, but its in /lib64)
[16:25:33] <kliegs> I *think* if you sync and run setup_board it should fix it
[16:25:36] <cjb> sure enough
[16:25:39] <cjb> /lib64/libgcc_s.so.1: file not recognized: File format not recognized
[16:25:46] <kliegs> but I'm not sure as some toolchain changes sometimes need you to rebuild the chroot
[16:26:17] <cjb> thanks!  what command do you mean by "sync"?
[16:26:19] <kliegs> yah - there was a bug then where the cross compiler would accidentally include host library paths so it picked up the wrong library
[16:26:22] <kliegs> 'repo sync'
[16:26:26] <cjb> great
[16:26:27] <kliegs> or 'repo sync -j32'
[16:26:50] <cjb> 'k.  outside the chroot, in ~/chromiumos/, I suppose
[16:27:11] <kliegs> I run it ouside the chroot in .../chromiumos.  But I *think* you can run it inside as well
[16:27:49] <cjb> thanks again.  outside's fine.
[16:28:53] <kliegs> your welcome.  and glad to see you still working on this
[16:29:11] <kliegs> new hardware / architectures can be a bit tough to bring up
[16:29:25] <cjb> good to hear it's not just me :-)
[16:30:03] <kliegs> nope.  always a lot of corner cases.  plus if you're just getting into cros as well the learning curve on that too
[16:33:03] *** borios has joined #chromium-os
[16:41:28] <jglasgow> kliegs: you are a trooper?  Can you help me figure out why tegra2_aebl-binary does not seem to be doing anything?
[16:42:21] <jglasgow> Never mind. Seems to be running now.
[16:42:32] <kliegs> jglasgow: glad to help :) and yes I'm a trooper
[16:42:41] <kliegs> but I like it when problems solve themselves
[16:42:55] <jglasgow> We'll I did try restarting it, but maybe I wasn't patient enough.
[16:44:26] <jglasgow> Do you know how long after issuing a command like:
[16:44:26] <jglasgow> ./buildbot_cli.py --prefix-url="i/chromiumos/" -b "tegra2_aebl-binary"  -r "hungte says this should cycle green, but needs to be started"
[16:44:26] <jglasgow> the waterfall will reflect the builder starting?
[16:45:16] <jglasgow> I guess looking at my history and comparing to the waterfall its about 5 minutes.
[16:54:31] *** jujugre has left #chromium-os
[16:55:23] <kliegs> jglasgow: most of the bots reboot the OS after they stop. so that could have slowed down the cycle if you were watching closely
[16:57:36] *** jrbarnette has joined #chromium-os
[17:03:34] *** jrbarnette_ has joined #chromium-os
[17:05:56] *** jrbarnette has quit IRC
[17:05:57] *** jrbarnette_ is now known as jrbarnette
[17:06:26] <kliegs> sosa: ping
[17:06:41] <kliegs> jglasgow: just making sure I didn't misread - things are going as expected?  or do you still need me to look at something?
[17:40:18] *** borios has quit IRC
[17:43:57] *** saintlou has joined #chromium-os
[17:43:57] *** ChanServ sets mode: +v saintlou
[18:00:24] *** wfrichar has joined #chromium-os
[18:00:24] *** ChanServ sets mode: +v wfrichar
[18:04:33] *** BladeFreak has quit IRC
[18:18:19] *** jrbarnette_ has joined #chromium-os
[18:20:40] *** jrbarnette__ has joined #chromium-os
[18:21:43] *** jrbarnette has quit IRC
[18:21:44] *** jrbarnette__ is now known as jrbarnette
[18:24:38] *** jrbarnette_ has quit IRC
[18:28:58] *** BladeFreak has joined #chromium-os
[18:29:50] *** sbyer_afk is now known as sbyer
[18:30:52] *** mtennant has joined #chromium-os
[18:30:52] *** ChanServ sets mode: +v mtennant
[18:31:50] *** McMAGIC--Copy has quit IRC
[18:32:09] *** McMAGIC--Copy has joined #chromium-os
[18:39:37] <mtennant> California build sheriff on duty
[18:55:47] *** perezd has joined #chromium-os
[18:57:29] *** Adys has joined #chromium-os
[19:00:54] <ellyjones> lemma: \lim{time -> \inf}|{ p | package(p) ^ crosworkoned(p) }|/|{ p | package(p)}| -> 1
[19:11:04] *** davidjames has joined #chromium-os
[19:11:04] *** ChanServ sets mode: +v davidjames
[19:12:42] *** magn3ts has quit IRC
[19:14:38] <pastarmovj> I did the commit that breaks the autotests I will revert that commit until I find out why does this happen (6dba0839363a0e5a3adce90766bd3061ad86c257)
[19:15:10] <mtennant> Thanks for the update.
[19:15:11] <Fractal> I?m trying to use an embedded initramfs on my Chromium OS kernel, almost everything works fine, my (busybox) shell script /init is called by the kernel and I can mount the real rootfs, *but* the "read x" command in my /init does not work, it does not block and the variable x stay empty
[19:15:52] <Fractal> And if I add "rd=/bin/sh" to the kernel command line, it does not even boot
[19:16:04] <Fractal> * "rdinit=/bin/sh"
[19:16:08] <mtennant> pastarmovj:  let us know when you've completed the revert
[19:16:39] <cmasone> pastarmovj: the infrequent chrome crash?
[19:16:58] <mtennant> cmasone:  no, it looks different
[19:17:21] *** Styx has quit IRC
[19:17:36] <pastarmovj> yes now it should respect the policy blob instead of the consent file and it seems that it can not find the policy files i copy for the tests
[19:17:58] <cmasone> mtennant: where are the faiulures?
[19:18:24] <cmasone> mtennant: and why is the tree open?
[19:18:26] <pastarmovj> ok i reverted the CL and will re-land it once I get the autotests in better shape
[19:18:52] <cmasone> pastarmovj: which autotests?
[19:19:08] <davidjames> cmasone: http://build.chromium.org/p/chromiumos/builders/x86%20pineview%20binary/builds/5916/steps/Test/logs/stdio
[19:19:25] <cmasone> davidjames: oh the usercrash one?
[19:19:41] <pastarmovj> logging_UserCrash
[19:20:18] <davidjames> pastarmovj: Ah good, so your revert fixes that?
[19:20:25] <pastarmovj> the problem is that my CL makes metrics respect the policy blob so for this  test to pass I copy a manually crafted policy file that contains either metrics on or off
[19:20:47] *** falken has joined #chromium-os
[19:20:47] *** ChanServ sets mode: +v falken
[19:21:01] <davidjames> pastarmovj: Hmm, any reason why preflight queue passed on your CL but PFQ didn't?
[19:21:03] <pastarmovj> my revert makes metrics work with the consent file and i will look into fixing the autotests and once I know they work i will commit fix for them and reland this CL
[19:21:04] <mtennant> cmasone:  the bot hasn't actually finished, yet.  but I'll close the tree myself
[19:21:09] <davidjames> pastarmovj: er, but pineview didn't?
[19:21:26] <mtennant> pastarmovj: is there a bug for this?
[19:21:30] <pastarmovj> do you think this could have been flaky?
[19:21:38] <jrbarnette> fractal: for the "read x", what is stdin?
[19:21:49] <cmasone> pastarmovj: no, I think that the bug is real
[19:22:38] <cmasone> pastarmovj: there's other logic that fakes out ownership by dropping policy blobs in the right place in the tests...make sure you're not running afoul of that
[19:23:14] <pastarmovj> are these tests run in parallel or sequentially
[19:23:40] <pastarmovj> i thought they run sequentially so i have more or less control over the system until the test is running
[19:24:04] <davidjames> pastarmovj: Tests are run sequentially right now I believe, but not good to depend on that as we're hoping to run in parallel :)
[19:24:19] <Fractal> jrbarnette: hm, I don?t know, I just wrote "read x" in the script, and I mounted a devtmpfs at /dev
[19:24:24] <cmasone> davidjames: well, we're not going to run tests in parallel on the same machine at the same time
[19:24:56] *** lipsinV2 has quit IRC
[19:24:57] <jrbarnette> fractal:  there's a good chance your default stdin is /dev/null
[19:25:02] <davidjames> cmasone: Right, separate KVMs
[19:25:23] <cmasone> pastarmovj: the tests are run sequentially, yes.  But depending on what class your test inherits from, that test might have setup logic that you're not aware of.  Did you run this test on your own?
[19:25:55] <pastarmovj> yes i did i acutally commited a test to autotests yesterday that should have prepared them for this commit
[19:26:03] <Fractal> jrbarnette: ok, so I should write something like "read x < /dev/console", and "echo y > /dev/console" ?
[19:26:25] <pastarmovj> sorry s/committed a test/committed a change/
[19:26:28] <Fractal> jrbarnette: Or perhaps /dev/tty1?
[19:26:33] *** saintlou has quit IRC
[19:26:53] <jrbarnette> fractal: i recommend /dev/tty1
[19:27:11] <jrbarnette> fractal:  there's some sample initramfs code similar to what you're talking about
[19:27:22] <jrbarnette> fractal: see src/platform/initramfs
[19:28:13] <darinski> mtennant: we should probably stop/restart builders that are destined to fail?
[19:28:13] <jrbarnette> fractal: in particular, the factory code starts w/ initramfs, then mounts root, etc., and does switch_root
[19:29:04] <pastarmovj> 19:33:09 INFO |                 GOOD    logging_UserCrash       logging_UserCrash       timestamp=1311355989    localtime=Jul 22 19:33:09       completed successfully
[19:29:10] <Fractal> jrbarnette: Thanks, I?ll take a look :-)
[19:29:30] <cmasone> pastarmovj: run the whole smoke suite?  There shouldn't be any bad interactions, looking at the code, but...
[19:29:33] *** saintlou has joined #chromium-os
[19:29:33] *** ChanServ sets mode: +v saintlou
[19:29:45] <cmasone> pastarmovj: you can just run_remote_tests.sh .... suite_Smoke
[19:29:54] *** jrbarnette_ has joined #chromium-os
[19:29:54] *** jrbarnette has quit IRC
[19:29:55] *** jrbarnette_ is now known as jrbarnette
[19:29:56] <pastarmovj> ok will try that
[19:30:00] <mtennant> darinksi:  I think we're ok.  the non-pfq bots never got that change
[19:30:10] *** jrbarnette_ has joined #chromium-os
[19:30:58] *** behdad has quit IRC
[19:31:54] *** davidjames has quit IRC
[19:34:24] <pastarmovj> Total PASS: 17/17 (100%)
[19:34:30] *** jrbarnette_ has quit IRC
[19:34:30] *** jrbarnette has quit IRC
[19:38:55] <pastarmovj> cmasone: how can I see the exact configuration a buildbot is running so that i can reproduce it locally?
[19:39:46] <darinski> pastarmovj: maybe you need to run it multiple times
[19:40:13] <darinski> it seems that some bots failed, some didn't -- so it might be intermittent based on context
[19:40:38] <darinski> i think this test is dependent on the contents of /var/log/messages, etc... and has similar issues before
[19:41:21] <pastarmovj> darinski:yes but this is also why i want to be as close to the bot configuration as possible i have ran this test around 4 or 5 times now and I don't see it failing yet
[19:41:55] <darinski> you might want to get the logs from the failure on the bots and take a look
[19:43:19] <darinski> otherwise, this is what the bot ran: INFO: PROGRAM(cbuildbot) -> RunCommand: ['bin/cros_au_test_harness', '--no_graphics', '--no_delta', '--board=x86-pineview', '--test_prefix=SimpleTest', '--verbose', '--base_image=/b/cbuild/src/build/images/x86-pineview/latest-cbuildbot/chromiumos_test_image.bin', '--target_image=/b/cbuild/src/build/images/x86-pineview/latest-cbuildbot/chromiumos_test_image.bin', '--test_results_root=/b/cbuild/chr
[19:43:19] <darinski> oot/tmp/cbuildbothu73oz/test_harness'] in dir /b/cbuild/src/scripts[0m
[19:45:18] <pastarmovj> yep thanks for the hints i have the suspicion that emerging the autotests instead of running them from the host might make a difference
[19:50:58] *** jrbarnette has joined #chromium-os
[19:50:58] *** ChanServ sets mode: +v jrbarnette
[19:59:58] <pastarmovj> I think you can open the tree now
[20:00:08] <pastarmovj> the x86-pine bot has cycled green
[20:01:02] <mtennant> pastarmovj:  thanks, I already did.  but it got auto-closed again
[20:01:11] <mtennant> chrome seg fault is back.  will add to bug
[20:01:21] <darinski> yes, please do
[20:01:24] <darinski> i reopened the tree
[20:01:52] <mtennant> ok
[20:12:26] <mtennant> looks like it closed again.  stand by
[20:18:45] <darinski> seems like an intermittent network issue, tree opened, will watch carefully...
[20:22:13] *** sbyer is now known as sbyer_afk
[20:24:39] *** falken has quit IRC
[20:27:37] *** JasonO has quit IRC
[20:28:16] <ellyjones> wfrichar: ping
[20:37:55] <darinski> tree is open, intermittent network failures bug filed -- crbug.com/90223
[20:38:00] *** Malmis_ has joined #chromium-os
[20:38:10] *** falken has joined #chromium-os
[20:38:10] *** ChanServ sets mode: +v falken
[20:38:53] *** mtennant has quit IRC
[20:39:40] *** petermayo has joined #chromium-os
[20:39:40] *** ChanServ sets mode: +v petermayo
[20:40:27] *** Malmis has quit IRC
[20:45:37] *** Inumedia_ has joined #chromium-os
[20:47:58] *** Inumedia has quit IRC
[20:49:25] *** Malmis has joined #chromium-os
[20:49:46] *** Malmis_ has quit IRC
[20:54:03] *** tbarzic__ has joined #chromium-os
[20:58:55] *** sbyer_afk is now known as sbyer
[20:59:54] *** saintlou has quit IRC
[21:02:03] *** saintlou has joined #chromium-os
[21:02:03] *** ChanServ sets mode: +v saintlou
[21:03:36] *** saintlou has quit IRC
[21:09:55] *** 18VAAFZXD has joined #chromium-os
[21:09:56] *** ChanServ sets mode: +v 18VAAFZXD
[21:13:12] *** 18VAAFZXD has quit IRC
[21:13:23] *** achuith has quit IRC
[21:13:26] *** mtennant has joined #chromium-os
[21:13:27] *** ChanServ sets mode: +v mtennant
[21:13:36] *** achuith has joined #chromium-os
[21:13:36] *** ChanServ sets mode: +v achuith
[21:18:25] *** achuith has quit IRC
[21:18:29] <wfrichar> ellyjones: what's up?
[21:18:34] *** achuith has joined #chromium-os
[21:18:34] *** ChanServ sets mode: +v achuith
[21:19:01] <ellyjones> wfrichar: "Wrong Image Format for C Command"?
[21:19:12] <wfrichar> ??
[21:19:27] <ellyjones> I just tried booting a new image on an arm platform
[21:19:38] <ellyjones> received an error I have never seen before
[21:19:41] <ellyjones> namely that :)
[21:19:47] <wfrichar> Not mine. I've never booted ARM.
[21:20:01] <ellyjones> hm, okay
[21:20:08] *** patcito has joined #chromium-os
[21:20:10] <cjb> ellyjones: oh, that's a uboot error
[21:20:11] *** Malmis_ has joined #chromium-os
[21:20:28] <ellyjones> what does it mean?
[21:20:37] <wfrichar> Beats me.
[21:20:44] <wfrichar> 'git blame' is your friend.
[21:20:54] <cjb> it might be looking for a kernel image at the time
[21:21:17] <cjb> which means that it tried to find the contents of a .uImage file at a specific spot in RAM but didn't find it there
[21:21:21] <cjb> but it might be something else, too
[21:21:47] <cjb> anyway, I'd expect this is an extremely-early-boot kind of thing, something basic wrong with the boot setup
[21:21:56] <ellyjones> indeed, I would expect that too
[21:22:28] *** Malmis has quit IRC
[21:28:38] *** dtu has quit IRC
[21:32:19] *** petermayo has quit IRC
[21:33:28] *** Malmis_ has quit IRC
[21:33:47] *** espeed has joined #chromium-os
[21:34:57] *** saintlou has joined #chromium-os
[21:34:57] *** ChanServ sets mode: +v saintlou
[21:35:41] *** Malmis has joined #chromium-os
[21:38:58] *** cjb has quit IRC
[21:42:23] *** Malmis has quit IRC
[21:45:18] *** achuith has quit IRC
[21:45:25] *** achuith has joined #chromium-os
[21:45:25] *** ChanServ sets mode: +v achuith
[21:45:40] <gauravsh> ttuttle: haha, you beat me to it.
[21:46:08] *** Malmis has joined #chromium-os
[21:47:48] *** achuith has quit IRC
[21:47:53] *** cjb has joined #chromium-os
[21:48:00] *** behdad has joined #chromium-os
[21:48:00] *** ChanServ sets mode: +v behdad
[21:48:22] *** achuith has joined #chromium-os
[21:48:22] *** ChanServ sets mode: +v achuith
[21:48:58] *** mtennant has quit IRC
[21:55:26] *** arun_ has joined #chromium-os
[21:55:26] *** arun_ has joined #chromium-os
[21:57:23] *** rbyers has joined #chromium-os
[22:01:43] *** Adys has quit IRC
[22:05:22] *** Hamlin_ has quit IRC
[22:14:16] *** Hamlin has joined #chromium-os
[22:15:26] <cjb> neat.  I have a build that I can chroot into and run binaries from on my device.
[22:15:52] <cjb> but trying to boot it from my own non-cros kernel gets me a hang.  I wonder what kernel dependencies there are.
[22:16:16] <cjb> I know it gets as far as mounting C-ROOT/ and running sbin/init, since there'd be a panic otherwise, but it looks like nothing makes it to /dev/fb0 after that
[22:17:26] <kliegs> cjb: might be some of the verity stuff
[22:18:01] <kliegs> ellyjones: have you updated firmware recently? the kernel image format changed in the last week so if your firmware isn't current it won't find it
[22:18:18] <ellyjones> kliegs: that sounds legit
[22:18:32] <kliegs> see email from Olof last Friday I believe - I can find subject/details if you need
[22:18:56] <ellyjones> found it
[22:21:13] <cjb> kliegs: ah, I don't know what that is.  Can I get the initscripts to give me early boot debugging over serial?
[22:21:44] <ellyjones> they always do if you have serial, I think
[22:21:47] *** D|sT has quit IRC
[22:22:11] <cjb> cool.  does it inherit the kernel console=, or is there somewhere else I need to tell it to use a given port?  (ttyS2 in this case)
[22:24:00] * ellyjones does not know
[22:24:28] <cjb> ok, no problem :)
[22:30:01] *** Adys has joined #chromium-os
[22:37:03] <cjb> heh, oh dear:
[22:37:05] <cjb> [  662.681113] Freeing init memory: 116K
[22:37:05] <cjb> [  665.092564] Unhandled prefetch abort: debug event (0x002) at 0x40496a64
[22:37:05] <cjb> [  665.334343] Unhandled prefetch abort: debug event (0x002) at 0x404f7a64
[22:37:08] <cjb> [  665.558636] Unhandled prefetch abort: debug event (0x002) at 0x4047ca64
[22:38:39] <cjb> and then just:
[22:38:40] <cjb> [  703.764336] init: tty2 main process (739) terminated with status 1
[22:38:40] <cjb> [  703.770958] init: tty2 main process ended, respawning
[22:38:40] <cjb> [  703.791062] init: ttyS0 main process (741) terminated with status 1
[22:39:26] <cjb> oh, cool!  I have developer console.
[22:39:50] <cjb> Although "chronos" asks for a password for some reason.
[22:41:37] <redpig> cjb: you may need to flip the dev switch if your device has one or add "cros_debug" to your kernel commandline
[22:41:49] <cjb> redpig: heh, it does not -- it's an OLPC laptop
[22:41:50] <redpig> I haven't been following your discussions closely enouhg to be able ot help better :p
[22:42:10] <redpig> cjb: cool - yeah add " cros_debug" to your kernel commandline or bake a password into your build
[22:42:16] <redpig> otherwise chronos/root will be blocked
[22:42:48] <cjb> great, doing so now
[22:44:49] <cjb> hmph, booted with cros_debug and it still wants a password
[22:45:06] <cjb> [    0.000000] Kernel command line: root=/dev/sda3 rootwait console=ttyS2,115200 cros_debug
[22:45:52] <cjb> redpig: is that kernel arg interpreted by the initscripts, or the kernel?
[22:45:59] <cjb> ('cause I'm not running a cros kernel.)
[22:46:24] <redpig> cjb: by the initscripts and pam.d stuff
[22:46:32] <cjb> 'k.  So it should be working but isn't.
[22:46:34] <redpig> hrm try it like this
[22:46:40] <redpig> "cros_debug somethingmeaningless"
[22:46:59] <redpig> there was a bug where cros_debug couldn't be the last entry, and maybe it isn't fixed yet :)
[22:47:17] <redpig> (or just move it before console= or whatever)
[22:47:38] <cjb> cool, trying
[22:48:02] <wfrichar> cjb: you should be able to use src/platform/vboot_reference/scripts/image_signing/set_chronos_password.sh to modify your image, and src/platform/vboot_reference/scripts/image_signing/sign_official_build.sh to re-sign it, so you don't have to rebuild.
[22:48:12] <ojn> or do a build_image_for_test before you write it out, then password for chronos and root will be test0000
[22:48:42] <wfrichar> If you want to run build_image again, you can just use src/scripts/set_shared_user_password.sh to set the default chronos password for all builds.
[22:49:33] <ellyjones> sheesh... outside temp in CAM is 102 right now
[22:49:46] <ojn> ellyjones, still a nice 75 here
[22:49:53] <ellyjones> you and your stable climate
[22:50:03] <redpig> ellyjones: _ouch_ only 90f here
[22:50:06] <wfrichar> Well, it's a little warmer *outside*. It's 75 in my office.
[22:50:15] <ellyjones> it's probably up near 85 or so where I sit
[22:50:33] <wfrichar> ugh.
[22:50:40] <ellyjones> (I,I http://www.thefuckingweather.com/?zipcode=02139)
[22:52:02] <redpig> nice
[22:53:39] * cjb also hiding from the heat in Cambridge MA.
[22:55:03] <ellyjones> one of my friends in NYC asserts that there it is "102, feels like 110" outside
[22:56:35] <redpig> it was like that here yesterday
[22:56:36] *** rbyers has quit IRC
[22:56:38] <redpig> heat index past 110
[22:56:43] <ellyjones> do not want
[22:56:43] <redpig> thought I was back in texas
[22:56:47] * ellyjones has to walk home :P
[22:57:00] <ellyjones> I guess I'll just stay at work for an extra couple of hours so it cools down
[22:57:00] <redpig> hire a camel
[22:57:12] <ellyjones> haha
[22:57:28] <ellyjones> Liz has our dog, who is a Samoyed (a sled dog) and is like x.x whenever we go outside
[22:58:10] <ojn> ouch, yeah no fun for a cold weather dog in that heat.
[22:58:28] <ellyjones> yes :\
[22:58:29] <ellyjones> oh well
[23:02:04] <cjb> redpig: thanks, that was it -- non-final cros_debug gets me a login
[23:03:52] <arkaitzr> is there a bug for the non-final cros_debug issue?
[23:04:18] <arkaitzr> a bug filed with the details
[23:04:57] <redpig> cjb: excellent!
[23:05:33] <redpig> arkaitzr: not sure - the reason I hadn't before is because crossystem hadn't exhibited it (it was a bug in my old cros_boot_mode code that I was aware of but we killed that)
[23:07:01] <arkaitzr> I get some errors in my Mario for the crossystem output
[23:07:27] <arkaitzr> cros_debug prints correctly
[23:08:40] *** powderluv has joined #chromium-os
[23:08:40] *** ChanServ sets mode: +v powderluv
[23:09:51] <redpig> arkaitzr: hrm what types of errors? I know mario isn't fully crossystem friendly
[23:10:25] <arkaitzr> dev_boot_usb, fmap_base, kern_nv, etc
[23:10:43] <arkaitzr> redpig: they all have (error) instead of a value
[23:11:01] <redpig> arkaitzr: try running it as root once and see what it does?
[23:11:16] <redpig> I think _some_ of those values are post-mario only
[23:11:47] <redpig> but randall would know more.  (I know crossystem digs around in /sys and caches what it finds which si why I asked about running it with privs once)
[23:11:59] <arkaitzr> with root is the same behavior
[23:12:15] <redpig> is the main_fw field populated?
[23:12:38] <arkaitzr> mainfw_type = developer
[23:12:49] <redpig> cool so it lookss like mario compat mode is working
[23:13:12] <redpig> (mario doesn't have any other types of rw firmware but crossystem fakes developer so we could make the difference work well for mario and alex and zgb)
[23:13:23] <arkaitzr> OK
[23:13:33] <wfrichar> arkaitzr: the rest of those fields aren't applicable to the Cr-48. (error) is normal.
[23:14:02] <wfrichar> They apply to BIOS features that were added to Alex and ZGB.
[23:14:43] <arkaitzr> OK so that's unrelated to cros_debug not being processed correctly
[23:15:07] *** mtennant has joined #chromium-os
[23:15:07] <wfrichar> Should be.
[23:15:16] *** ChanServ sets mode: +v mtennant
[23:15:50] *** monteslu has joined #chromium-os
[23:16:56] *** darinski has quit IRC
[23:17:07] <monteslu> Is webGL hardware accelerated on the Samsung Chromebook?
[23:18:48] <monteslu> I don't know if Mesa DRI for intel 3150 GPU is hardware accelerated or not and if chromeOS is even using it if it is
[23:20:09] <Hexxeh> monteslu: check chrome://gpu-internals
[23:20:52] *** behdad has quit IRC
[23:21:42] <monteslu> Hexxeh, thanks!
[23:21:49] <ellyjones> for a good time, forget about kernel's specialness and 'git rebase cros/master'
[23:21:59] <monteslu> of course my chromebook is sitting on my desk at home right now and not here at work :)
[23:22:11] <cjb> cool, getting farther:  "Check failed: egl_display_ != EGL_NO_DISPLAY"
[23:22:21] * cjb will switch from opengles to opengl/mesa
[23:23:18] <redpig> monteslu: "WebGL: Hardware accelerated
[23:23:27] <redpig> is what my samsung chromebook tells me from that page
[23:23:31] <monteslu> yikes
[23:23:38] <monteslu> then it need some serious work
[23:23:49] <monteslu> way slow for hardware
[23:25:02] <Hexxeh> I laugh, my fully specced out iMac 27" gets lower framerates on WebGL tests than my Macbook Air base model on the same tests running ChromeOS :V
[23:25:25] <cjb> monteslu: yeah, I made the mistake of trying some WebGL on it :)
[23:25:29] <cjb> <1fps
[23:25:46] <monteslu> my 3 year old nvidia gpu gets 10 ten times the frame wrate on chrome+fedora
[23:26:36] <monteslu> hah, its actually 6 years old!
[23:26:43] <monteslu> nvidia integrated 8200
[23:27:55] <monteslu> there must be driver issues, the intel 3150 isnt that bad
[23:28:09] <Hexxeh> In my experience it's depended far more on the driver than the underlying hardware
[23:28:19] <monteslu> agreed
[23:28:30] <Hexxeh> nVidia 320M getting twice the framerates of my Radeon HD 5750
[23:30:50] <monteslu> the 2d canvas animation stuff I do on the chromebook seems great though  :)
[23:30:57] *** darinski has joined #chromium-os
[23:30:57] *** ChanServ sets mode: +v darinski





top