[00:01:24] <ellyjones> I think the thing I am currently writing is narrowly edging out my ptrace thread injector for the title of 'worst piece of code I've ever written'
[00:02:58] <kliegs> adlr: with tree just opening now and close to dinner i was planning on holding off submitting that CL until AM
[00:03:09] <ferringb> ellyjones: apache/mod_python hijacker for debugging instances (and doing it cleanly w/out interuptting whatever request that had come in) probably is my worst in terms of hankyness
[00:03:32] <kliegs> adlr: if you want to submit and keep an eye on it that's fine, otherwise I'll do first sign of greenness tomorrow
[00:03:47] <ferringb> useful way to get into a process for doing some heappy inspection though...
[00:04:46] *** Styx has quit IRC
[00:05:04] <ellyjones> I am currently writing something to be preloaded as part of minijail
[00:05:27] <ferringb> mmm. ok, that sounds interesting
[00:05:28] <crosbot> tree became 'Tree is open (thieule: testInterruptedUpdate flaky - chromiumos:18717)'
[00:05:48] <adlr> kliegs: i'll submit
[00:06:57] <adlr> Keybuk: i'm preparing a CL to revert the revert
[00:07:08] *** achuith has quit IRC
[00:10:54] <Keybuk> davidjames: we should make c-b/k-h's version totally more obvious than 0.0.1-r6 like saying 2.6.38 somewhere
[00:11:09] <Keybuk> then packages can do DEPEND=">=chromeos-base/kernel-headers-2.6.38" and such as needed
[00:13:57] *** JasonO has joined #chromium-os
[00:13:57] *** JasonO has joined #chromium-os
[00:16:20] <kliegs> adlr: thanks.  dinner time now
[00:56:19] <crosbot> tree became '?'
[00:56:34] <crosbot> tree became 'Tree is open (thieule: testInterruptedUpdate flaky - chromiumos:18717)'
[01:04:16] *** jrbarnette has joined #chromium-os
[01:04:17] *** ChanServ sets mode: +v jrbarnette
[01:16:51] <davidjames> Keybuk: SGTM
[01:22:44] <Keybuk> davidjames: YGR
[01:22:48] <Keybuk> err, YHR
[01:23:00] *** JasonO has quit IRC
[01:23:55] *** craigdh has joined #chromium-os
[01:23:55] *** ChanServ sets mode: +v craigdh
[01:24:41] *** JasonO has joined #chromium-os
[01:24:41] *** JasonO has joined #chromium-os
[01:25:41] <craigdh> I don't get a commit button on gerrit, do I need some sort of permissions?
[01:27:13] <davidjames> craigdh: Is your CL +2'd and Verified?
[01:27:48] <craigdh> yes
[01:29:00] <craigdh> I also don't have the option to +2 directly, which it seems most do
[01:32:01] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl-binary" from 9b4dd10af7b7ba6c25dfa10e7c2d71385899d41c: thestig at chromium dot org <thestig at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[01:32:47] <crosbot> tree became 'Tree is closed (investigating -> keybuk)'
[01:33:42] <vlaviano> craigdh: I had a similar issue once and it was because of a transition of my chromium account to GA+
[01:33:58] <vlaviano> I had two gerrit identities, once untrusted.  I needed to delete that one to resolve the issue.
[01:34:02] <vlaviano> once->one
[01:34:03] <crosbot> tree became 'Tree is open (tegra2_aebl-binary just failed a repo sync from a couple hours back, issue 18250)'
[01:34:55] <craigdh> ah, this is a @google.com account but I do also have a @chromium.org
[01:35:32] <vlaviano> check out settings->identities in gerrit
[01:35:41] <aaronp_> marcheu: actually, never mind, I found some bugs in my server code that I won't be able to fix before I get back to work on Monday.
[01:35:43] <aaronp_> sorry
[01:36:26] *** vmil86 has quit IRC
[01:36:59] <marcheu> aaronp_: no worries, it's not like client side code is ready and blocked on that
[01:46:05] *** Malmis has quit IRC
[01:46:38] *** Malmis has joined #chromium-os
[01:47:50] *** Greboid has quit IRC
[01:47:50] *** mrseb has quit IRC
[01:47:56] *** marcheu has quit IRC
[01:47:56] *** rjkroege has quit IRC
[01:48:01] *** marcheu has joined #chromium-os
[01:48:20] *** ellyjones has quit IRC
[01:48:42] *** rjkroege has joined #chromium-os
[01:49:10] *** ellyjones has joined #chromium-os
[01:49:10] *** ChanServ sets mode: +v ellyjones
[01:49:55] *** mrseb has joined #chromium-os
[02:07:25] *** JasonO has quit IRC
[02:10:49] *** JasonO has joined #chromium-os
[02:11:35] *** arya_ has joined #chromium-os
[02:12:13] <arya_> Hi, Can anyone tell me where I can aquire intel wireless drivers for chrome os from?
[02:13:20] *** arya_ has quit IRC
[02:15:46] *** lipsinV2 has joined #chromium-os
[02:16:19] *** achuith has joined #chromium-os
[02:16:20] *** ChanServ sets mode: +v achuith
[02:16:42] <aaronp_> marcheu: I lied and managed to get it out for review after all.
[02:16:44] <aaronp_> Have a good weekend!
[02:16:50] *** achuith has quit IRC
[02:16:51] *** aaronp_ has quit IRC
[02:16:58] *** achuith has joined #chromium-os
[02:16:58] *** ChanServ sets mode: +v achuith
[02:35:15] *** stalled has quit IRC
[02:39:43] *** wfrichar has joined #chromium-os
[02:39:43] *** ChanServ sets mode: +v wfrichar
[02:40:09] <Hexxeh> Is there a working on-screen keyboard for CrOS yet? Part of USE=touchui maybe?
[02:42:51] <Keybuk> Completed x11-drivers/xf86-input-cmt-0.0.1-r20 (in 0m3.9s)
[02:42:56] <Keybuk> adlr: ^ completed on the dodgy buildbot
[02:46:03] *** stevenjb has quit IRC
[02:48:11] *** D|sT has quit IRC
[02:48:25] *** stalled has joined #chromium-os
[02:49:34] *** wfrichar has quit IRC
[02:49:52] *** stevenjb has joined #chromium-os
[02:50:06] <Keybuk> kinaba: ping
[02:50:06] *** stevenjb has quit IRC
[03:02:54] *** JasonO has quit IRC
[03:04:28] <Keybuk> sheriff out
[03:04:29] *** Keybuk has quit IRC
[03:07:01] *** JasonO has joined #chromium-os
[03:20:25] *** saintlou has quit IRC
[03:34:24] *** jrbarnette has quit IRC
[03:48:56] *** xiyuan has joined #chromium-os
[03:48:56] *** ChanServ sets mode: +v xiyuan
[03:54:57] *** gfrog has joined #chromium-os
[03:59:11] *** behdad has quit IRC
[04:10:09] *** JasonO has quit IRC
[04:17:02] *** grundler has quit IRC
[04:36:36] *** behdad has joined #chromium-os
[04:47:26] *** kerz has quit IRC
[04:51:08] *** BThompsonGR has joined #chromium-os
[05:00:34] *** behdad has quit IRC
[05:08:56] *** deshantm has quit IRC
[05:09:22] *** jochen__ has quit IRC
[05:09:28] *** jochen__ has joined #chromium-os
[05:09:28] *** ChanServ sets mode: +v jochen__
[05:11:42] *** deshantm has joined #chromium-os
[05:44:56] *** BladeFreak has joined #chromium-os
[06:06:47] *** petermayo has quit IRC
[06:12:01] *** BThompsonGR has quit IRC
[06:20:56] *** vmil86 has joined #chromium-os
[06:24:21] *** McMAGIC--Copy has quit IRC
[06:25:43] *** McMAGIC--Copy has joined #chromium-os
[06:41:11] *** D|sT has joined #chromium-os
[08:19:55] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_kaen-binary" from d2c695ea51a9a0aef3c29788147908f88339adc5: hbono at chromium dot org <hbono at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[08:28:50] <crosbot> tree became 'Tree is open (kinaba: repo sync failure of issue 18250. the bot turned green now)'
[08:45:55] *** Styx has joined #chromium-os
[09:01:06] *** vapier has joined #chromium-os
[09:03:12] *** SoulRaven has joined #chromium-os
[10:11:33] *** patcito has quit IRC
[10:12:45] *** Sergiu_ has joined #chromium-os
[10:13:48] *** Sergiu_ has quit IRC
[10:14:07] *** Sergiu_ has joined #chromium-os
[10:33:05] *** Sergiu_ has quit IRC
[11:02:17] <crosbot> tree became '?'
[11:02:32] <crosbot> tree became 'Tree is open (kinaba: repo sync failure of issue 18250. the bot turned green now)'
[11:11:44] *** BladeFreak has quit IRC
[11:39:52] *** kinaba has left #chromium-os
[12:12:18] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-alex-binary" from e5945a0305e0fb7168197132806e575e9e49ba7f: _third_party_ at chromium dot org, bashi at chromium dot org <bashi at chromium dot org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>)'
[14:52:01] *** rbyers has joined #chromium-os
[14:53:49] *** rbyers has left #chromium-os
[14:55:04] *** rbyers_ has quit IRC
[14:55:18] *** rbyers_ has joined #chromium-os
[14:55:18] *** ChanServ sets mode: +v rbyers_
[15:02:47] *** rbyers_ is now known as rbyers
[15:17:45] *** rharrison_chrome has joined #chromium-os
[15:17:46] *** ChanServ sets mode: +v rharrison_chrome
[15:17:57] <rharrison_chrome> East Coast Sheriff checking in
[15:18:47] <crosbot> tree became 'Tree is closed (investigating -> rharrison)'
[15:19:08] *** kemono has joined #chromium-os
[15:24:23] <crosbot> tree became 'Tree is open (looks like flake in chromeos_login, will continue to investigate -> rharrison)'
[15:36:54] <crosbot> tree became 'Tree is open (chromeos_login failure -> crosbug:18757)'
[15:54:00] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb-binary" from dcc8cdf9f2e9e6cfa090cb0eb168c45992ca74db: hamaji at chromium dot org <hamaji at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, timurrrr at chromium dot org <timurrrr at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>)'
[15:57:25] *** jochen___ has joined #chromium-os
[15:57:25] *** ChanServ sets mode: +v jochen___
[15:58:52] <crosbot> tree became 'Tree is closed (investigating -> rharrison)'
[15:59:16] *** jochen__ has quit IRC
[15:59:16] *** jochen___ is now known as jochen__
[16:02:41] <rharrison_chrome> looks like the issue is in the testing framework, I don't think the blamelist CLs are responsible
[16:04:13] <rharrison_chrome> is anyone on the channel familiar with ./cros_run_vm_test?
[16:17:30] <crosbot> tree became 'Tree is closed (pinged troopers about issue -> rharrison)'
[16:20:09] <dianders> rharrison_chrome: sig11 in Chrome in suite_Smoke/login_CryptohomeMounted?
[16:22:29] <rharrison_chrome> dianders: That looks like it is being ignored, that test is passing. The issue that caused redness looks like it is the updater test.
[16:25:34] <dianders> rharrison_chrome: Sigh...  ;)
[16:28:30] *** JasonO has joined #chromium-os
[16:30:39] *** rbyers_ has joined #chromium-os
[16:31:30] <rharrison_chrome> dianders: Just talked with kliegs, it is the sig 11 that is causing the redness, not something else. This a known transient issue
[16:31:45] *** rbyers_ has quit IRC
[16:31:50] <rharrison_chrome> I will reopen the tree momentarily
[16:32:12] <dianders> rharrison_chrome: OK, good to know.  ...was wondering that, but always hard to figure out for sure.
[16:33:54] *** rbyers_ has joined #chromium-os
[16:33:55] *** ChanServ sets mode: +v rbyers_
[16:35:08] *** rbyers_ has quit IRC
[16:35:37] <crosbot> tree became 'Tree is open (redness due to flake -> rharrison)'
[16:37:50] *** rbyers_ has joined #chromium-os
[16:51:05] *** sadrul has quit IRC
[16:57:41] <crosbot> tree became 'Tree is open'
[17:03:42] <ellyjones> morning all
[17:07:01] <benchan> [5~[5~/c
[17:07:01] <benchan> //c//
[17:07:17] *** benchan has quit IRC
[17:07:41] *** benchan has joined #chromium-os
[17:07:42] *** ChanServ sets mode: +v benchan
[17:07:46] <ellyjones> benchan: I believe that a) you are technically correct about sysconf(3) and b) anyone whose sysconf() returns a negative _SC_GETPW_R_SIZE_MAX deserves what they get
[17:08:08] *** benchan is now known as Guest15930
[17:08:19] *** benchan_ has joined #chromium-os
[17:08:20] *** ChanServ sets mode: +v benchan_
[17:08:51] <ellyjones> it is analogous to having INT_MAX defined to a negative value or whatever
[17:10:10] <benchan_> ellyjones: I looked at the glibc source code, but haven't found where it may return a negative value for _SC_GETGR_R_SIZE_MAX, at least on Linux
[17:10:38] <ellyjones> as far as I'm concerned those are constants provided by the system and if they're wrong the system deserves it
[17:11:48] <benchan_> ellyjones: that's true too. alternatively, we can just resize the buffer when errno == ERANGE
[17:12:35] <ellyjones> if errno == ERANGE we should just abort
[17:13:00] <ellyjones> "The maximum needed size for buf can be found using sysconf(3) with the argument _SC_GET{PW,GR}_R_SIZE_MAX."
[17:13:05] <ellyjones> so increasing it cannot possibly help
[17:14:11] <benchan_> ellyjones: I did remember I saw a discussion before that sysconf returns a smaller size on amd64, not sure if that's fixed or not
[17:15:23] <ellyjones> I will believe this is a problem if you can find a system where sysconf(3) returns a negative size :P otherwise I think this line of reasoning is just too paranoid
[17:16:02] *** varunjain has quit IRC
[17:16:11] *** varunjain has joined #chromium-os
[17:16:12] *** ChanServ sets mode: +v varunjain
[17:16:35] <benchan_> ellyjones: I think this is what I saw last time: http://lists.debian.org/debian-glibc/2009/04/msg00106.html
[17:17:43] <benchan_> ellyjones: I didn't dig into that, maybe that's fixed or a non-issue
[17:17:51] *** IamTrying has joined #chromium-os
[17:18:00] <ellyjones> argh, it's a _manpage_ bug?
[17:18:03] <IamTrying> Chromium os is hacked? e.g: http://venturebeat.com/2011/08/03/security-researchers-hack-googles-chrome-os/
[17:18:10] <ellyjones> (quoth http://lists.debian.org/debian-glibc/2009/04/msg00110.html)
[17:18:34] <ellyjones> IamTrying: "They found a flaw in ScratchPad, a preinstalled extension to the Chrome OS that lets people take notes and save them to cloud-based Google Docs."
[17:18:36] *** sadrul has joined #chromium-os
[17:19:05] <IamTrying> ellyjones, thats a shame hack no?
[17:19:16] <ellyjones> I can't understand your sentence
[17:19:18] *** petermayo has joined #chromium-os
[17:19:18] *** ChanServ sets mode: +v petermayo
[17:19:26] <benchan_> ellyjones: not sure, the example in getpwnam_r manpage only shows the negative case, not the errno case
[17:19:55] <ellyjones> benchan_: you are right, but I wish you weren't right :P
[17:20:06] <IamTrying> ellyjones, how can Google OS be hacked? All our accounts will be in problem. Which is shame i am not safe if i use Google OS?
[17:20:08] <ellyjones> every time I deal with getpwnam_r and friends I wish I was using getpwnam
[17:20:34] <ellyjones> IamTrying: what? someone found a security flaw in an extension - it's like finding a security flaw in a program that runs on windows or whatever
[17:20:48] <ellyjones> IamTrying: except that security flaws in extensions can't persist after you uninstall the extension
[17:22:18] <IamTrying> ellyjones, Does that mean Android OS (division of Google OS) has this same hacked functionality?
[17:22:24] <ellyjones> uh what
[17:22:32] <ellyjones> android and chromium OS are completely separate codebases
[17:22:39] <ellyjones> designed and written by separate groups of people
[17:23:02] <benchan_> ellyjones: I also wish I didn't see that example on the manpage... I'm getting the feeling that source code becomes more trustworthy than manpage/documentation
[17:23:28] <ellyjones> if (bufsize == -1) bufsize = 16384;
[17:23:31] <ellyjones> thanks, manpage!
[17:23:46] <ellyjones> "/* Should be more than enough */"
[17:24:27] <ellyjones> so we need to loop on getpwnam_r, testing errno for ERANGE (and ENOMEM?)
[17:24:29] <IamTrying> ellyjones, what is the core of Google OS? Linux kernel or its there modified linux kernel? If that is a hack does it mean all Linux has that hole?
[17:24:46] <ellyjones> IamTrying: what
[17:25:01] <ellyjones> IamTrying: that article is talking about a security flaw in a chrome extension that happens to come preinstalled on Chrome OS
[17:25:17] <ellyjones> IamTrying: it has nothing at all to do with Linux; the same flaw would affect that extension installed on Chrome running on Windows
[17:26:22] <IamTrying> ellyjones, ok ic, but to me its like a shame using Google OS, which is hacked (i first did not look into that deep, my impression was its hacked just like Windows)
[17:27:13] <benchan_> ellyjones: I guess ENOMEM implies abort()?
[17:29:10] <ellyjones> benchan_: ERANGE should imply abort()! :P
[17:29:39] <ellyjones> IamTrying: sorry, I can't explain this in sufficient detail, but suffice it to say that that article makes a big deal of a small thing
[17:30:22] <IamTrying> ellyjones, true but, i get always confused, Why google needed google OS? Cant they simply use Linux/Debian ?
[17:30:23] <benchan_> ellyjones: yes, isn't getpwnam sufficient in this case? does it really need to be reentrant?
[17:31:14] <ellyjones> IamTrying: perhaps you should watch the explanatory videos on http://www.chromium.org/chromium-os
[17:31:30] <ellyjones> benchan_: you sound like me three days ago :P redpig wants me to use getpwnam_r()
[17:32:51] <benchan_> ellyjones: I guess it feels more comfortable to make things under our control :)
[17:33:17] <ellyjones> I actually feel much less comfortable about the ~25 lines of extra code involved in using getpwnam_r here
[17:33:30] <ellyjones> and it doesn't _actually_ make it reentrant since they use a static file descriptor under the hood
[17:35:11] <benchan_> ellyjones: oh, I guess for consistency, _r is probably still preferred
[17:36:05] <IamTrying> What is the multimedia framework they use in Google OS?
[17:36:15] <IamTrying> Do we have Java in Google OS?
[17:36:32] *** behdad has joined #chromium-os
[17:36:58] <ellyjones> there is no such thing as "google os"
[17:37:23] <IamTrying> sorry Chromium OS == Google OS i mean
[17:37:28] * benchan_ heading to breakfast
[17:39:59] <IamTrying> If i want to write MyownOS, from Chromium OS. Can i do that? And customize with my own UI etc?
[17:40:31] <ellyjones> chromium OS is entirely open-source
[17:40:44] <ellyjones> as long as you abide by the terms of the BSD-style license we ship it under, you may do as you please with it
[17:43:40] <IamTrying> ellyjones, thats cool.  If i accept the license of BSD. Can i do what ever i want, specially like i change the whole "Chromium OS" and rename it to "MyOS"
[17:44:18] <ellyjones> the license lays out exactly what you may and may not do
[17:44:34] <ellyjones> bear in mind that some parts of the OS are third-party projects under different licenses which you may also need to comply with
[17:46:19] <IamTrying> Thats fine, as long as i have a "Boot > Login > Desktop (no software nothing)" All will be my web softwares
[17:46:46] <IamTrying> Boot > Login > Desktop (all myOS branding) if i can then its excellent OS.
[17:47:23] <ellyjones> you should look at the source tree, then - see http://bit.ly/chromeos-faq
[17:48:58] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot_master" on "x86 generic PFQ" from 25a908e4cd402e5535823f54bd8af76e1ebf4b2d: agl at chromium dot org <agl at chromium dot org@a4e77c2c-9104-11de-800e-5b313e0d2bf3>)'
[17:49:10] <ellyjones> hrm
[17:49:42] <IamTrying> ellyjones, is it then nice idea to simply use Debian linux? To make own branded OS? (to avoid those license issues) Chromium OS would may have!
[17:49:59] <ellyjones> we're building _llvm_??
[17:50:11] <ellyjones> IamTrying: sorry, I can't really answer your questions, I have work to do ^^;
[17:50:49] <ellyjones> ok, x86-generic-pfq can't fetch the llvm tarball or something
[17:51:35] <ellyjones> I'm going to reopen the tree I think
[17:52:21] <rharrison_chrome> ok
[17:53:00] <ellyjones> yeah, I can fetch the tarball
[17:53:04] <ellyjones> I assume we just sucked temporarily
[17:53:05] *** jrbarnette has joined #chromium-os
[17:53:34] <crosbot> tree became 'Tree is open (ellyjones: x86-generic-pfq failed to fetch from llvm.org; ineffable mysteries of the universe blamed)'
[17:53:34] <ellyjones> done
[17:53:52] <rharrison_chrome> Do you know if there is an existing bug for issues with fetching tarballs?
[17:54:23] <ellyjones> like 'internet sometimes sucks'?
[17:54:39] <ellyjones> http://code.google.com/p/chromium-os/issues/detail?id=18415 seems broadly similar
[17:54:50] <ellyjones> actually, the fact that it's trying to fetch from llvm.org instead of localmirror is a bug
[17:55:00] <ellyjones> I'll file that
[17:55:25] <rharrison_chrome> That is what I was thinking, the builders should be using a local copy incase the internet does suck :-) Thanks
[17:56:22] <crosbot> tree became 'Tree is open (ellyjones: x86-generic-pfq failed to fetch from llvm.org; ineffable mysteries of the universe blamed; crosbug.com/18758)'
[18:01:23] *** oc80z has quit IRC
[18:01:23] *** oc80z has joined #chromium-os
[18:02:28] *** jrbarnette_ has joined #chromium-os
[18:05:30] *** jrbarnette has quit IRC
[18:05:31] *** jrbarnette_ is now known as jrbarnette
[18:08:41] *** benchan_ has quit IRC
[18:08:48] *** benchan has joined #chromium-os
[18:08:48] *** ChanServ sets mode: +v benchan
[18:08:50] *** benchan has quit IRC
[18:17:42] *** saintlou has joined #chromium-os
[18:17:43] *** ChanServ sets mode: +v saintlou
[18:18:48] <ellyjones> who's a good person for arm video issues?
[18:19:50] <cjb> ellyjones: perhaps marcheu?
[18:19:57] <ellyjones> marcheu: you around? :)
[18:20:16] <ellyjones> my kaen shows snow on the LCD as soon as the kernel boots, even though userspace seems healthy
[18:23:07] <vpalatin> ellyjones: before or after the splashscreen  displayed on the framebuffer by ply-image ?
[18:24:26] <ellyjones> after the splashscreen
[18:24:29] <ellyjones> the flow happening is this:
[18:24:41] <ellyjones> 1) boot machine with no stick; get the regular dev-mode screen
[18:24:51] <ellyjones> 2) hit space, insert recovery stick
[18:25:05] <ellyjones> 3) kernel boots most of the way with debug output, then the entire screen gets filled with snow
[18:25:14] <ellyjones> after 3, the kernel and userspace are both still alive and well, as far as I can tell
[18:26:34] <ellyjones> (in fact, I can tell the recovery is running, since I occasionally see fs mount messages)
[18:26:50] <ellyjones> the snow is stable (i.e., doesn't change, and seems to be the same across reboots)
[18:26:59] <vpalatin> ah, it's on the recovery path
[18:27:32] <vpalatin> I don't even exactly know what we are supposed to display there : text from fb console ?
[18:28:31] <ellyjones> yeah, then some UI like "hey, we're about to wipe your machine"
[18:31:06] *** IamTrying has left #chromium-os
[18:31:19] <ellyjones> (also does the tpm on kaen always require a hard reset after recovery? just a soft reset seems to leave the machine unbootable)
[18:32:32] *** wfrichar has joined #chromium-os
[18:32:33] *** ChanServ sets mode: +v wfrichar
[18:36:24] *** benchan_ has joined #chromium-os
[18:36:25] *** ChanServ sets mode: +v benchan_
[18:40:02] *** benchan_ has left #chromium-os
[18:40:17] *** kerz has joined #chromium-os
[18:40:18] *** ChanServ sets mode: +v kerz
[18:40:43] *** benchan_ has joined #chromium-os
[18:40:43] *** ChanServ sets mode: +v benchan_
[18:44:24] *** benchan_ has quit IRC
[18:44:35] *** benchan_ has joined #chromium-os
[18:44:36] *** ChanServ sets mode: +v benchan_
[18:45:50] *** benchan_ has quit IRC
[18:46:06] *** benchan has joined #chromium-os
[18:46:06] *** ChanServ sets mode: +v benchan
[18:55:24] <crosbot> tree became 'Tree is open (ellyjones: x86-generic-pfq failed to fetch from Google Storage; ineffable mysteries of the universe blamed; crosbug.com/18758)'
[18:56:24] <ellyjones> crosbot: botsnack
[18:56:24] <crosbot> ellyjones: :)
[18:57:05] <adlr> what's botsnack?
[18:57:18] <davidjames> ellyjones: Took a look at the PFQ and it was actually downloading a prebuilt for llvm there
[18:57:39] <ellyjones> adlr: it's a snack for a bot :P
[18:57:39] <davidjames> ellyjones: So that was definitely coming from Google Storage. Note as well that the llvm tarball also comes from Google Storage
[18:57:48] <ellyjones> davidjames: I dunno why it failed to download, then
[18:58:38] <davidjames> ellyjones: Yes, especially given how many times it retries, and also given that we downloaded a ton of other packages from the same host in the same build
[18:58:44] <ellyjones> well
[18:58:46] *** craigdh has quit IRC
[18:58:54] <adlr> crosbot: botsnack
[18:58:54] <crosbot> adlr: :)
[18:58:56] <ellyjones> you can retitle the bug and reassign it :P but it should stay open until we figure out wtf
[18:59:07] <adlr> hahahahahaha
[18:59:13] <ellyjones> adlr: it's like a scooby snack
[19:00:06] <davidjames> ellyjones: I believe that different files in Google Storage redirect to different backends, so I guess the specific backend that was hosting llvm was down at the time (and was persistently down, given that we retried after a while)
[19:00:35] <ellyjones> :(
[19:00:54] <ellyjones> I just want to download nine megabytes, &c
[19:03:11] *** ScriptRipper has quit IRC
[19:09:34] *** lipsinV2 has quit IRC
[19:11:27] *** behdad has quit IRC
[19:11:59] *** behdad has joined #chromium-os
[19:11:59] *** ChanServ sets mode: +v behdad
[19:25:04] *** jrbarnette_ has joined #chromium-os
[19:26:56] *** jrbarnette has quit IRC
[19:26:56] *** jrbarnette_ is now known as jrbarnette
[19:27:03] *** sadrul1 has joined #chromium-os
[19:27:03] *** ChanServ sets mode: +v sadrul1
[19:29:42] <marcheu> ellyjones: hmm, I'm not sure how that'd happen. does running ply-image again by hand later works?
[19:30:07] <ellyjones> hm
[19:30:15] <ellyjones> I built a new test image and booted it, and the UI looks fine now
[19:30:16] <ellyjones> weird
[19:30:25] *** sadrul has quit IRC
[19:32:01] *** mrseb has quit IRC
[19:43:33] *** jrbarnette has quit IRC
[19:47:20] <superm1> ellyjones, on the regular test image, did you notice that the "chromium" splash screen was missing too (and replaced by a little bit of snow)?
[19:50:04] <ellyjones> the signin screen appears entirely normal
[19:50:22] <superm1> yeah i mean after the kernel is loaded before you hit the signin screen
[19:50:47] <superm1> along the top of the screen
[19:51:42] <ellyjones> I wasn't watching :( I'll check soon, I am developing something else
[19:58:05] *** chocobo__ has joined #chromium-os
[19:58:05] *** ChanServ sets mode: +v chocobo__
[19:58:12] *** patcito has joined #chromium-os
[19:59:29] *** mrseb has joined #chromium-os
[20:00:31] *** jrbarnette has joined #chromium-os
[20:00:32] *** ChanServ sets mode: +v jrbarnette
[20:01:46] *** Ender has joined #chromium-os
[20:02:28] *** JasonO has quit IRC
[20:07:28] <ellyjones> printf("j %p caps %016lx uid %d %08x %d %d\n", j, j->caps, j->uid, j->uid, j->uid, j->uid);
[20:07:36] <ellyjones> j 0xc8218 caps 0000000000000000 uid 0 000003e8 1000 1000
[20:07:39] <ellyjones> Please discuss.
[20:09:00] *** Keybuk has joined #chromium-os
[20:09:01] *** ChanServ sets mode: +v Keybuk
[20:10:55] *** Ender is now known as JasonO
[20:11:00] <marcheu> ellyjones: j->caps isn't a long?
[20:11:05] *** JasonO has quit IRC
[20:11:05] *** JasonO has joined #chromium-os
[20:11:58] <ellyjones> it's uint64_t
[20:12:10] <marcheu> and your platform is 32 bit?
[20:12:15] <ellyjones> armv7
[20:12:27] <marcheu> so you're trying to stuff a 64 bit value into a 32 bit long?
[20:12:54] <ellyjones> I can try using %llx instead
[20:13:03] <marcheu> yes you can
[20:13:28] <ellyjones> let's see if that helps
[20:14:07] <ellyjones> hrm
[20:14:20] <ellyjones> ok, so I need to make extra sure my printf format specs line up
[20:17:11] <ellyjones> marcheu: good call :)
[20:17:28] <ellyjones> that + the requirement to align 64-bit values to even registers was causing printf to print a random value basically
[20:20:15] <benchan> ellyjones: you may also try "PRIx64" in inttypes.h if it's uint64_t
[20:26:28] * benchan wondering if any IRC client can multiplex between UI and console...
[20:26:38] *** JakeSays has quit IRC
[20:26:43] <ellyjones> benchan: what do you mean?
[20:28:53] <benchan> ellyjones: basically, I want a client that provides a UI frontend when I'm on X and a console frontend when I'm on a screen/tmux session
[20:29:18] <ellyjones> irssi can somewhat do that
[20:29:32] <ellyjones> irssi has a mode called 'irssi-proxy' where it proxies for other IRC clients and keeps your session alive
[20:31:45] <benchan> ellyjones: Does that mean I can connect XChat to an irssi-proxied session?
[20:32:47] <ellyjones> I believe so
[20:33:17] <benchan> ellyjones:  nice, let me try it.  thanks!
[20:33:41] <ellyjones> caveat lector: I have never tried this
[20:35:21] <benchan> get a feeling that it should work...
[20:36:44] *** benchan has quit IRC
[20:37:17] *** JakeSays has joined #chromium-os
[20:41:43] *** benchan has joined #chromium-os
[20:41:43] *** ChanServ sets mode: +v benchan
[20:42:49] *** benchan_ has joined #chromium-os
[20:42:49] *** ChanServ sets mode: +v benchan_
[20:42:55] *** benchan_ has quit IRC
[20:46:28] *** redpig has quit IRC
[20:46:48] *** redpig has joined #chromium-os
[20:46:48] *** ChanServ sets mode: +v redpig
[20:47:28] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_arthur canary" from None: )'
[20:49:15] <crosbot> tree became 'Tree is closed (thieule: Investigating)'
[20:50:38] *** benchan has quit IRC
[20:54:36] <crosbot> tree became 'Tree is open (thieule: Network error - Bad Gateway - when accessing Chrome SVN)'
[20:55:33] *** jennb has joined #chromium-os
[20:55:33] *** ChanServ sets mode: +v jennb
[21:00:14] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86 generic full" from 77104ce9f23c88e4f7cbe9226083929d02db1834: _third_party_ at chromium dot org, bsalomon at google dot com <bsalomon at google dot com@2bbb7eff-a529-9590-31e7-b0007b416f81>, reed at google dot com <reed at google dot com@2bbb7eff-a529-9590-31e7-b0007b416f81>, robertm at google dot com <robertm at google dot com@fcba33aa-ac0c-11dd-b9e7-8d5594d729c2>)'
[21:01:16] <crosbot> tree became 'Tree is closed (thieule: Investigating)'
[21:01:52] *** Hexxeh has quit IRC
[21:02:20] *** Hexxeh has joined #chromium-os
[21:02:26] <rharrison_chrome> thieule: It is 502 Bad Gateway again
[21:02:48] <crosbot> tree became 'Tree is open (thieule: Network error - Bad Gateway - when accessing Chrome SVN)'
[21:02:50] <thieule> rharrison: thanks, reopening
[21:04:02] <rharrison_chrome> thieule: Is this a repeat of the previous closure, ie. both bots got hit by the same outage or is this a new outage?
[21:04:11] <Keybuk> my guess is that it's the same outage
[21:04:21] <Keybuk> and it just took longer for this builder to fail
[21:05:49] <rharrison_chrome> That is what I was thinking, though I am a bit concerned that this maybe a symptom of a systemic issue on the backend
[21:08:23] <Keybuk> it could also just mean they restarted the svn server, and it hit every build using svn at that time
[21:09:19] <rharrison_chrome> Would that be a question for the Chrome Infrastructure Team or is that beyond Chrome?
[21:10:55] <ellyjones> I'm going to teach crosbot to automagically recognize common kinds of tree closures :)
[21:11:35] <thieule> chatting with chrome trooper about their status now
[21:11:53] <rharrison_chrome> thieule: thanks for following up on this
[21:24:06] *** ScriptRipper has joined #chromium-os
[21:24:26] *** ScriptRipper has quit IRC
[21:24:26] *** ScriptRipper has joined #chromium-os
[21:24:30] <crosbot> tree became 'Tree is open (thieule: Network error - Bad Gateway - when accessing Chrome SVN, crosbug.com/18776)'
[21:28:05] <superm1> ellyjones, this look like about what you're seeing?: http://crosbug.com/p/5329
[21:28:33] <kliegs> ellyjones: can you teach crosbot to automatically patch bad commits too?
[21:29:18] <ellyjones> kliegs: :)
[21:29:43] <ellyjones> superm1: yes, exactly - I have a screenshot in ~ellyjones/www/oops.jpg
[21:30:17] *** powderluv has joined #chromium-os
[21:30:18] *** ChanServ sets mode: +v powderluv
[21:30:26] <superm1> well at least for now you can VT flip for a workaround it seems
[21:31:41] <Keybuk> srao: setting console= on x86 you can still open /dev/console
[21:33:04] <srao> srao: interesting...
[21:33:06] <srao> err
[21:33:12] <srao> Keybuk: interesting ... :-)
[21:33:22] <Keybuk> nah, not really, that's what I'd expect
[21:33:28] <Keybuk> /dev/console is a virtual device, after all, not a real console
[21:33:31] <srao> well interesting that it doesn't work on arm
[21:33:34] <Keybuk> it's the ARM behaviour I find kooky ;-)
[21:33:37] <srao> yeah
[21:34:00] <srao> I'll root around in the kernel
[21:35:32] *** SoulRaven has quit IRC
[21:43:54] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "tegra2_aebl-binary" from 487c3a152dec6cce48909eb0bb01c22e75bfa40e: agl at chromium dot org <agl at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>, agl at chromium dot org <agl at chromium dot org@a4e77c2c-9104-11de-800e-5b313e0d2bf3>, chrome-admin at google dot com <chrome-admin at google dot com@0039d316-1c4b-4281-b951-d872f2087c98>)'
[21:45:00] <SeligArkin> What commands should I run / copy-paste to try and report a memory leak I've been experiencing for awhile now on my chromebook -- it appears to be a very gradual one as it only happens after a few days of the chromebook being used, but gradually the number of "He's dead, jim!"s rises as time on continues
[21:45:11] <crosbot> tree became 'Tree is closed (investigating -> keybuk)'
[21:46:42] <crosbot> tree became 'Tree is open (repo sync hung without output)'
[21:47:13] *** saintlou has quit IRC
[21:52:56] <srao> any idea why I can do "cros_workon --board=(some x86 board) start perf" and it works, but "cros_workon --board=(some arm board) start perf" fails ?
[21:54:01] <srao> oh... the perf ebuild is missing arm in KEYWORDS... grrr
[22:10:51] <Keybuk> hmm, at one point only perf-next worked on arm
[22:11:06] <Keybuk> it may be that when we updated to 2.6.38 nobody fixed the keywords of perf itself
[22:15:38] <srao> yeah perf-next is fine
[22:15:57] <srao> i filed a bug and uploaded a CL... nice to have a simple bug once in a while :-)
[22:16:16] *** Kernel-Panic has joined #chromium-os
[22:16:24] *** BThompsonGR has joined #chromium-os
[22:17:24] <srao> SeligArkin: what release are you on ?
[22:21:18] *** Keybuk has quit IRC
[22:21:23] *** Keybuk has joined #chromium-os
[22:21:23] *** ChanServ sets mode: +v Keybuk
[22:24:16] *** vmil86 has quit IRC
[22:26:59] *** Kernel-Panic has quit IRC
[22:37:46] <crosbot> tree became 'Tree is closed (Automatic: "cbuildbot" on "x86-zgb canary" from None: )'
[22:39:09] <ellyjones> hrm... to push a risky change on friday afternoon or not...
[22:39:34] <crosbot> tree became 'Tree is closed (investigating -> keybuk)'
[22:42:51] *** rbyers_ has quit IRC
[22:43:31] <gauravsh> ellyjones: for a moment, i thought this closer was you again. cryptohome test failure...
[22:43:46] <ellyjones> wasn't me :P
[22:43:58] <gauravsh> i know, different test
[22:44:00] <ellyjones> I never break the tree
[22:44:37] <gauravsh> for reasonable small values of 'never' :P
[22:45:25] <crosbot> tree became 'Tree is open (flakey test in cryptohome, opened 18788)'
[22:45:26] <ellyjones> I never broke the tree today!
[22:49:18] <kliegs> ellyjones: that's cause you slept in
[22:49:23] <kliegs> if you'd worked a full day ...
[22:49:36] <ellyjones> nah, I was busy getting fingerprinted by the government this morning :)
[22:49:42] <ellyjones> otherwise I would be breaking the tree as we speak
[22:56:33] <SeligArkin> srao: dev channel -- 14 I can copy the more exact numbers if nessary
[22:59:16] <kliegs> yikes. they fingerprint tree-breakers now?  NSA is really locking down
[23:00:12] *** Greboid has joined #chromium-os
[23:03:07] *** benchan has joined #chromium-os
[23:03:12] *** ChanServ sets mode: +v benchan
[23:04:09] *** Adys has quit IRC
[23:06:28] *** Adys has joined #chromium-os
[23:11:30] *** powderluv has quit IRC
[23:11:33] <ellyjones> !!! All ebuilds that could satisfy "sys-fs/fuse[-kernel_linux]" for /home/ellyjones/trunk/src/build/images/tegra2_kaen/0.15.856.2011_08_05_1657-a1/rootfs/usr/local/ have been masked.
[23:11:37] <ellyjones> this is new.
[23:14:01] *** annekenyon has joined #chromium-os
[23:15:56] <ellyjones> annekenyon: hiya
[23:16:07] *** rharrison_chrome has quit IRC
[23:18:28] *** annekenyon has quit IRC
[23:21:10] <srao> SeligArkin: interesting, I actually have to run but if you could post your info on the list that would be good
[23:22:14] <SeligArkin> srao: what list? The dev list? And part of the reason I popped in here was to find out what information would be most helpful. :P
[23:24:18] *** annekenyon_ has joined #chromium-os
[23:24:45] <srao> SeligArkin: chromiumos-dev is the proper list, what is most helpful is probably things like /proc/meminfo and about:memory
[23:25:09] <SeligArkin> can you access /proc/meminfo from a non-dev-switched device?
[23:25:24] <srao> I think you can from chrome://system
[23:25:40] *** Styx has quit IRC
[23:28:14] <annekenyon_> Hi, I am the following problem when I run sudo emerge: Error: The above package list contains packages which cannot be installed at the same time on the same system.
[23:28:37] <annekenyon_> and it says ""net-misc/dhcpcd" is blocking chromeos-base/hard-host-depends-0.0.1-r92"
[23:29:50] <ellyjones> hrm
[23:29:53] <ellyjones> that is weird
[23:30:33] <ellyjones> I don't see any open bugs for it
[23:30:43] <ellyjones> that is after a clean checkout, then setup_board, then build_packages?
[23:31:40] <annekenyon_> I think so -- the second two definitely, but I did the checkout several days ago, so I can't recall if things went perfectly.
[23:32:17] <ellyjones> try syncing, then :) I have a different failure with ToT
[23:32:46] <ellyjones> chromeos-u-boot-9999: make: *** wait: No child processes.  Stop.
[23:32:46] *** behdad has quit IRC
[23:32:49] <ellyjones> _that's_ new.
[23:33:47] <annekenyon_> ToT?
[23:34:01] <srao> there was a change recently which required me to do a emerge-tegra2_kaen --unmerge chromeos-u-boot
[23:36:01] <ellyjones> annekenyon_: tip of tree
[23:36:19] <ellyjones> x86-video-fbdev failed for me just now...
[23:36:25] <ellyjones> I wonder if this has to do with toolchain work?
[23:37:03] <marcheu> video-fbdev failing how?
[23:37:35] <ellyjones> in fact, _many_ packages are failing to build
[23:38:28] <ellyjones> modemmanager failed with missing asm/sigcontext.h and linux/limits.h via glib, dbus-c++ failed with missing EDEADLK caused by missing linux/errno.h
[23:38:32] <ellyjones> I am missing headers for some reason
[23:38:45] * ellyjones reruns setup_board
[23:41:17] <Keybuk> I changed kernel-headers the other day
[23:41:41] <Keybuk> but I have no idea why that would result in them being missing, as the change was to actually depend on them in places
[23:41:54] <Keybuk> ellyjones: what does emerge-$BOARD -pqv chromeos-base/kernel-headers say?
[23:42:24] <ellyjones> I'll tell you after my current build_packages run ends
[23:52:10] <annekenyon_> the sync worked, but I still have the same error.
[23:53:28] <ellyjones> about dhcpcd? you may need to setup_board againn
[23:55:40] <annekenyon_> ellyjones: did it, still have the same error.
[23:57:19] * ellyjones baffled
[23:57:26] <ellyjones> the MTV people are at a picnic, I think...
[23:57:39] <redpig> omnomnom
[23:58:46] <ellyjones> chromeos-assets-split-0.1.0-r65 collided with chromeos-assets-0.0.1-r77
[23:58:47] <ellyjones> argh
[23:59:02] <ellyjones> I bet there was a PSA I did not read





top