**** BEGIN LOGGING AT Tue Jan 15 02:59:58 2013
Jan 15 06:18:28 <blloyd_> what determines if grub is included in an image or not?
Jan 15 08:20:23 <mckoan> good morning
Jan 15 14:31:01 <Crofton|work> zeddii, if I use liunx-..-custom, what is the best way to create the configuration fragemnts?
Jan 15 14:39:20 <zeddii> Crofton|work. It depends on the configuration baseline. If you are starting from your own configuration and not a one on a linux-yocto-* meta branch, it's a process of factoring out the options that you want to share for boards (i..e the policy) and options you want boards to define (the hardware options). How you come up with the baseline defconfig you start with is up to you (well, everything is, but you know what I mean) and whether or not you trust default
Jan 15 14:39:20 <zeddii> s (i.e. savedefconfig), or want to explicitly control more options.
Jan 15 14:41:52 <Crofton|work> ok
Jan 15 14:41:57 <Crofton|work> that is what I thought :)
Jan 15 14:42:09 <ant_work> zeddii: hi. Is savedefconfig's mini-config supported now?
Jan 15 14:42:21 <zeddii> you could feed it in as a defconfig.
Jan 15 14:42:27 <Crofton|work> obviously, I am trying to figure out how to work with the evil vendor as a starting point
Jan 15 14:42:27 <zeddii> I don't use them myself.
Jan 15 14:42:47 <Crofton|work> and be able to modify as needed for specific use cases
Jan 15 14:42:47 <ant_work> I remember it didn't work not long ago :/
Jan 15 14:43:20 <zeddii> oh really. I'm closing in on some new configuration changes for yocto 1.4, I can throw some extra tests into the mix.
Jan 15 14:43:51 <ant_work> zeddii: byw, I'm testing vanilla 3.8-rc3. Do you have any recipe floating with the same SRCREV you're using?
Jan 15 14:44:35 <zeddii> using for what ? general testing, my configuration change work ? or something else ?
Jan 15 14:44:38 <Crofton|work> I think as a first pass, I will start with the vendor config
Jan 15 14:44:49 <Crofton|work> and add fragments needed to modify it
Jan 15 14:44:50 <ant_work> to test kernel config tools output on my defconfig
Jan 15 14:45:13 <ant_work> and see if there are new mismatches/errors compared to 3.4
Jan 15 14:45:22 <zeddii> Crofton|work, that's a good baseline, since you know it works. I have some scripts that depending on where you go can help chop things down.
Jan 15 14:45:34 <zeddii> ant_work, all my 3.8 work is on linux-yocto-dev at the moment.
Jan 15 14:45:35 <ant_work> (errors judging the fragments are non-harware ;)
Jan 15 14:46:20 <Crofton|work> just getting some changes in via fgragments, and making it easy for people to make kernels work for different use cases will be a start
Jan 15 14:46:28 <Crofton|work> witout dupinig the config
Jan 15 14:46:38 <zeddii> agreed.
Jan 15 14:46:42 <ant_work> Crofton|work: best is to use the output fils you get in /meta to purge the vendor config
Jan 15 14:46:54 <ant_work> i.e. trust kernel-tools :)
Jan 15 14:47:07 <Crofton|work> trust no one :)
Jan 15 14:47:27 <ant_work> that's for the split hw-not hw
Jan 15 14:49:55 <ant_work> Crofton|work: see our example in meta-handheld/tree/recipes-kernel/linux/linux-yocto-3.4. Still WIP , hope to polish it for 3.8
Jan 15 14:50:05 <Crofton|work> ok
Jan 15 14:50:25 <ant_work> there is very little in SRC_URI
Jan 15 14:51:01 <ant_work> with the help of zeddii I could chain-include all the files
Jan 15 14:51:53 * zeddii has some new things for 1.4 that should help as well. and I'm actually getting the cycles to work on them now. Hope that trend holds ;)
Jan 15 14:54:17 <Crofton|work> ant_work, how are you splitting the config intor the fragments?
Jan 15 14:54:54 <gbrennon_> hello everyone
Jan 15 14:54:58 <ant_work> yes, the idea is to share most of them between the machines
Jan 15 14:56:04 <gbrennon_> i'm new to this world of embedded linux. i've installed the poky through yocto project in a imx28 and now i want to install gcc to install openssh. how can i do this?
Jan 15 15:02:44 <ant_work> Crofton|work: the idea is to create specific files for a) machine-hw b) arch-hw c) common features, this last one can be very granular
Jan 15 15:04:00 <Crofton|work> machine is specific device, such as a specific dev board?
Jan 15 15:04:21 <ant_work> our recipe is for 6 different machines/boards
Jan 15 15:04:34 <ant_work> 5 pxa and one sa1100
Jan 15 15:05:28 <Crofton|work> so you have pxa and sa specific stuff
Jan 15 15:05:34 <Crofton|work> and board specific stuff
Jan 15 15:05:38 <Crofton|work> and features stuff
Jan 15 15:05:46 <ant_work> yes,that's the idea
Jan 15 15:06:44 <ant_work> more features can be added including other cfg fragments/features
Jan 15 15:07:43 <ant_work> like i.e. usb
Jan 15 15:09:45 <ant_work> zeddii: ideally we'd like to have the patches in a specific dir under $PV. Iiirc I've tried under /patches but files were not found...
Jan 15 15:17:19 <zeddii> ant_work, do you mean the fragments/scc files ? or actual kernel patches ? Just trying to make sure my memory isn't failing me.
Jan 15 15:18:08 <ant_work> the kernel patches
Jan 15 15:20:39 <zeddii> ok.
Jan 15 15:20:51 * zeddii roots around in his inbox for a refresher
Jan 15 15:21:16 <ant_work> well, see that  http://cgit.openembedded.org/meta-handheld/tree/recipes-kernel/linux/linux-yocto-3.4/spitz/spitz.scc
Jan 15 15:22:13 <ant_work> iirc...sorry caqn't be precise now... I've tried "patch ../" like "include ../" in the .scc
Jan 15 15:23:02 <ant_work> probably I did "patch ../patches/blah"
Jan 15 15:23:51 <zeddii> ah right. I remember now. I have some notes, we'll get it fixed up. if you want, drop me an email with your preferred layout, and I'll look and see about ensureing it works, unless there's some horrible technical reason :) I'm right in that code at the moment, so the timing is good.
Jan 15 15:24:21 <ant_work> sure, thx
Jan 15 15:25:12 <ant_work> it's about common patches, specific patches can/should live in /$machine imho
Jan 15 15:26:01 * zeddii nods
Jan 15 15:29:29 <panda84kde> Hi guys. hi otavio
Jan 15 15:30:33 <panda84kde> suppose I have a Nitrogen6x board, whose configuration is meta-fsl-arm-extra/conf/machine/nitrogen6x.conf
Jan 15 15:31:36 <panda84kde> suppose I have a working build of Yocto for my board
Jan 15 15:34:00 <panda84kde> suppose I successfully built a custom kernel based on Nitroge6x manufacturer git (https://github.com/boundarydevices/linux-imx6), instead of the default shipped by Yocto for Freescale boards
Jan 15 15:35:26 <panda84kde> is it possible to customize kernel receipe in meta-fsl-arm-extra layer only for that board so that kernel source gets downloaded from a different souce?
Jan 15 15:37:41 <panda84kde> I mean: it's clear to me that source uri is defined in meta-fsl-arm/recipes-kernel/linux/linux-imx.inc, but I don't know what's the better way to overwrite that value only for Nitrogen.
Jan 15 15:51:23 <panda84kde> mmm.... I see this in linux-imx_3.0.35.bb:
Jan 15 15:51:25 <panda84kde> > # iMX6Q SabreLITE changes done by Boundary Devices
Jan 15 15:51:26 <panda84kde> > SRC_URI_append_imx6qsabrelite
Jan 15 15:51:54 <panda84kde> probably I should set something like SRC_URI_nitrogen6x. I'll try.
Jan 15 15:52:09 <kergoth> yes, that's how overrides work, see the poky and bitbake manuals for details :)
Jan 15 15:53:46 <panda84kde> thanks kergoth
Jan 15 15:56:53 <darknighte> YPTM: Song_Liu : Sean has joined the call
Jan 15 15:57:31 <fray> YPTM: Mark's on the call
Jan 15 15:57:59 <kergoth> YPTM: CHris Larson on the call
Jan 15 16:00:07 <zenlinux> YPTM: Scott Garman is on the call
Jan 15 16:00:08 <tomz1> YPTM: Tom Z on the call
Jan 15 16:00:09 <Song_Liu> YPTM: welcome to the technical team meeting, everyone. Please let me know who's on the bridge. Thanks!
Jan 15 16:00:19 <strassek> YPTM: Kevin Strasser is dialing in
Jan 15 16:00:47 <scottrif> YPTM:  Scott Rifenbark joined
Jan 15 16:00:51 <chris-kaminski> YPTM: Chris Kaminski from Tobias & Tobias on the bridge
Jan 15 16:00:56 <zeddii> YPTM: Bruce Ashfield on the call.
Jan 15 16:01:01 <Corneliu> YPTM: Cornel joined
Jan 15 16:01:40 <sgw1> YPTM: Saul is on
Jan 15 16:02:03 <nitink> YPTM: nitin is on the bridge
Jan 15 16:02:17 <jmpdelos> YPTM: Jeff Polk is on
Jan 15 16:03:05 <Song_Liu> YPTM: Any opens?
Jan 15 16:05:33 <Song_Liu> Bug 2821 - libowl do_fetch error Bug 3714 - Fetch fail linux-libc-headers-3.1 Bug 3715 - Fetch fail linux-korg-3.1 Bug 3453 - lib32 connman-gnome built for qemux86-64 - ipk Bug 2695 - [HOB]toolchain arch in settings is not saved Bug 3001 – Hob recipe build from terminal fails Bug 3686 - errors when building headless build for eclipse-plugin 1.2.2 Bug 3693 - BUILDDIR error - when selecting kernel version in yocto bsp tool for denzi
Jan 15 16:05:38 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=2821 normal, Medium, 1.4 M1, laurentiu.serban, NEEDINFO , libowl do_fetch error
Jan 15 16:05:39 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3714 normal, Medium, 1.2.2, saul.wold, NEW , X32: hjl-tools missing from github.com for x32 kernel builds
Jan 15 16:05:40 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3715 normal, Undecided, ---, richard.purdie, RESOLVED DUPLICATE, Fetch fail linux-korg-3.1
Jan 15 16:05:41 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3453 normal, Medium, 1.4, radu.moisan, IN PROGRESS DESIGN , Build fails when building multilib with ipk packaging
Jan 15 16:05:42 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=2695 normal, Medium, 1.3 M4, bogdan.a.marinescu, VERIFIED FIXED, [HOB]toolchain arch in settings is not saved
Jan 15 16:05:43 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3001 normal, Undecided, ---, jessica.zhang, RESOLVED NOTABUG, bblayers-hob.conf does not correspond with bblayers.conf
Jan 15 16:05:44 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3686 normal, Medium+, 1.2.2, jessica.zhang, RESOLVED WORKSFORME, errors when building headless build for eclipse-plugin 1.2.2
Jan 15 16:05:45 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3693 normal, Medium, 1.2.2, ioanax.grigoropol, NEW , BUILDDIR error - when selecting kernel version in yocto bsp tool for denzil
Jan 15 16:07:21 <davest> YPTM: I'm online
Jan 15 16:12:26 <Zagor> YPTM: Björn entered the call
Jan 15 16:15:58 <Zagor> my runqemu typo fix should perhaps go into 1.3.1 too
Jan 15 16:16:04 <Zagor> (if it didn't already)
Jan 15 16:16:52 <sgw1> Zagor: please send it a again with a [for Danny] tag so we don't forget this
Jan 15 16:16:57 <Zagor> I have an open: x-tests vs x-ptest (bug #3296)
Jan 15 16:16:57 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3296 enhancement, Medium, 1.4 M1, kai.kang, VERIFIED FIXED, [QA/Test] Add a perl-test recipe to test perl
Jan 15 16:17:01 <Song_Liu> https://wiki.yoctoproject.org/wiki/Yocto_Project_v1.4_Status#Overview
Jan 15 16:17:03 <Zagor> sgw1: ok
Jan 15 16:25:43 <darknighte> YPTM: it's the sgw1 show today!
Jan 15 16:27:34 <sgw1> darknighte: don't mean it to be, just people are awol
Jan 15 16:27:56 <sgw1> not really awol, some are known to be gone!
Jan 15 16:28:58 <fray> ;)
Jan 15 16:29:12 <darknighte> sgw1: I've been gone a while, but I'm still here.  Well, physically.
Jan 15 16:29:53 <sgw1> darknighte: apparently mentally also ;-)
Jan 15 16:30:33 <darknighte> sgw1: what were we talking about again? ;)
Jan 15 16:31:41 <Song_Liu> YPTM: Chris contact (Tobias&Tobias) Chris Kaminski [chris.kaminski@tobias.tv]
Jan 15 16:33:24 <darknighte> YPTM: https://lists.yoctoproject.org/listinfo/webhob
Jan 15 16:33:29 <darknighte> davest: ^^
Jan 15 16:33:54 <sgw1> darknighte: your faster than me!
Jan 15 16:34:02 * darknighte goes woot woot
Jan 15 16:34:43 <Zagor> Song_Liu: I have an open: x-tests vs x-ptest (re: bug #3296)
Jan 15 16:34:44 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=3296 enhancement, Medium, 1.4 M1, kai.kang, VERIFIED FIXED, [QA/Test] Add a perl-test recipe to test perl
Jan 15 16:35:05 <scottrif> Yocto Project Linux Kernel Development Manual at http://www.yoctoproject.org/docs/1.4/kernel-dev/kernel-dev.html
Jan 15 16:40:13 <Zagor> https://wiki.yoctoproject.org/wiki/Ptest
Jan 15 16:42:01 <Corneliu> Testopia Wiki page: https://wiki.yoctoproject.org/wiki/Testopia
Jan 15 16:42:31 <Corneliu> Testopia HOWTO page(work in progress): https://wiki.yoctoproject.org/wiki/Testopia_HOWTO
Jan 15 16:43:30 <blloyd> gbrennon_: while it is possible to do what you asked, it sounds like a different process may be what you want.  For instance, adding IMAGE_FEATURES_append := " ssh-server-openssh" and building a new image and putting it on the device would give you openssh.
Jan 15 16:44:31 <Zagor> Corneliu: can Testopia handle automated tests, such as nightly tests ran by autobuilder?
Jan 15 16:44:39 <blloyd> ^or add EXTRA_IMAGE_FEATURES := "ssh-server-openssh" to your local.conf to put in all.
Jan 15 16:45:32 <Corneliu> Zagor: I belive it can, it has a framework for that
Jan 15 16:46:08 <Corneliu> Zagor: It says it can in the testopia Manual
Jan 15 16:46:40 <Zagor> Corneliu: link?
Jan 15 16:46:41 <Corneliu> Zagor: we plan to use that in order to fully automate testing from test running to test reporting and bug reporting
Jan 15 16:47:04 <Corneliu> http://landfill.mozilla.org/testopia2/testopia/doc/Manual.pdf
Jan 15 16:48:16 <scottrif> halstead: Doc page needs some work also
Jan 15 16:49:06 <Corneliu> Zagor: check the "Using Testopia With Automated Test Scripts" section
Jan 15 16:49:26 <fray> YPTM: I need to leave the call early..
Jan 15 16:52:11 <Zagor> Corneliu: hmm, can TestRuns only store "Targetted Pass percentage" and not result for each test case?
Jan 15 16:52:12 <blloyd> I am having trouble where my ext3 image files do not have grub installed.  It is in some and not others. Where should I be looking to find why this is/isn't installed?  I'm sure it's something stupid I'm forgetting.
Jan 15 16:52:26 <Song_Liu> YPTM: thank you all for joining the meeting. Have a nice day or evening!
Jan 15 16:52:27 <Zagor> Corneliu: (http://landfill.bugzilla.org/testopia2/docs/html/api/Bugzilla/WebService/Testopia/TestRun.html)
Jan 15 16:54:00 <sgw1> Song_Liu: thank you!
Jan 15 17:01:19 <blloyd> Song_Liu: there used to be meeting invites sent via Yocto mailing list.  I can't find any for this year.
Jan 15 17:02:54 <blloyd> nm, found via keyword agenda.
Jan 15 17:03:28 <blloyd> Evolution must have trashed it's indexes again.
Jan 15 17:31:32 <blloyd> anyways, anyone have any suggestions why a IMAGE.ext3 file would not have grub installed into it?  I have a custom image based on core-image-sato which has it, but core-image-sato-dev doesn't have it.
Jan 15 17:32:13 <kergoth> check IMAGE_INSTALL, as usual with any image, i'd say..
Jan 15 17:40:58 <ftonello> guys, im having a warning from Qt about iconv_open
Jan 15 17:41:12 <ftonello> shouldnt eglibc provides iconv support?!
Jan 15 17:47:41 <blloyd> kergoth: neither one refers to it, but one has it and the other doesn't.  The bootloader seems to be a special case and I haven't figured it out yet.
Jan 15 17:49:20 <kergoth> ia32-base pulls it in via 12:MACHINE_ESSENTIAL_EXTRA_RDEPENDS += "grub"
Jan 15 17:49:23 <kergoth> from a quick grep
Jan 15 17:50:20 <blloyd> actually, both reference it from MACHINE_ESSENTIAL_EXTRA_RDEPENDS="grub".
Jan 15 17:50:46 <blloyd> just found where, but it's still not installed every time, which is what I'm trying to figure out.
Jan 15 17:52:03 <kergoth> further grepping should tell you that that variable is used in packagegroup-core-boot
Jan 15 17:52:13 <kergoth> so check the contents of that ipk, and make sure its pulled into both images
Jan 15 17:52:44 <kergoth> opkgfiles () { ar p $1 data.tar.gz | tar -tvz }
Jan 15 17:52:53 <kergoth> opkgfilse path/to/ipk
Jan 15 17:52:59 <kergoth> or just inspect buildhistory
Jan 15 17:53:10 <kergoth> erm, wrong function
Jan 15 17:53:24 <kergoth> opkginfo () { ar p $1 control.tar.gz | tar -zxO ./control }
Jan 15 17:53:30 <kergoth> that'll dump the metadata from an ipk
Jan 15 18:18:30 <blloyd> thank you kergoth.
Jan 15 18:18:59 <kergoth> np
Jan 15 18:19:27 <blloyd> I think I have enough pointers to finish untangling it.
Jan 15 18:19:35 <kergoth> cool, good luck :)
Jan 15 18:19:57 <kergoth> there's also bitbake -g of course, but that doesn't tell you *why* something got pulled in,v ia what mechanism
Jan 15 18:21:06 <kergoth> actually, you might be able to use the new WIP dependinfo subcommand of my bb tool too - https://github.com/kergoth/bb#readme
Jan 15 18:21:09 <kergoth> https://gist.github.com/4455568 shows its output
Jan 15 18:21:16 <blloyd> nope, it doesn't.
Jan 15 18:21:30 <blloyd> might look at that.
Jan 15 18:23:17 <blloyd> ok, I'm confused.  task-core-boot has RDEPENDS_task-core-boot = " ... ${MACHINE_ESSENTIAL_EXTRA_RDEPENDS}" and both images have MACHINE_ESSENTIAL_EXTRA_RDEPENDS="grub".  But an rpm query shows one with and one without grub.
Jan 15 18:27:03 <kergoth> wtf, i jsut set incopmatible license and the build blew up with a stacktrace complmaining about lists not being hashable :(
Jan 15 18:32:04 <fatal^> trying yocto for the first time.... dbus-glib do_install has been stuck with gtkdoc-rebase eating all cpu power for an hour or so now.... I know gtkdoc is usually slow, but this is kind of insane right?
Jan 15 18:34:50 <blloyd> pulled bb to try it.  ImportError: No module named tinfoil
Jan 15 18:36:05 <kergoth> either old bitbake, or you didn't run it after sourcing your setup scripts
Jan 15 18:38:13 <blloyd> old bitbake.  denzil
Jan 15 18:38:35 <blloyd> how current of a bitbake is required?
Jan 15 18:41:14 <blloyd> no tinfoil in denzil so newer than that.  :(
Jan 15 18:42:07 <kergoth> ah, denzil, that's pretty old now. i'll have to look into porting bb to the older bitbake
Jan 15 18:42:34 <kergoth> ah well, guess that's not an option for this, but i'm sure you'll geti t figured out
Jan 15 18:49:34 <fatal^> strace to the rescue... gtkdoc-rebase was stuck in and endless loop because of a weird symlink...
Jan 15 18:50:37 <blloyd> Since denzil is no longer main, I'll be looking at forward porting my stuff to the current baseline.
Jan 15 18:54:20 <Crofton|work> zeddii, would it be reasonable to focus on one kernel config and doing board customization via device tree bits?
Jan 15 18:57:26 <blloyd> so is danny current stable?
Jan 15 19:00:53 * kergoth sighs
Jan 15 19:01:01 <kergoth> blloyd: yes, that's the most recent release
Jan 15 19:01:13 <kergoth> based on the most recent release, i shoul say
Jan 15 19:01:45 <kergoth> so, as far as i can tell, there's no gplv2 binutils in oe-core anymore..
Jan 15 19:07:56 <blloyd> how come there isn't a danny branch in poky-extras?
Jan 15 19:09:54 <zeddii> Crofton|work, I'd call that reasonable.
Jan 15 19:19:55 <cristianiorga> walters: hello
Jan 15 19:24:23 <adalton> Hi all.  I'm interested in using yocto to build a root filesystem for an lxc vm.  I noticed that there's a meta-virtualization layer with an lxc recipe.  Is that for creating lxc VMs or creating root filesystems that include lxc tools? :-)
Jan 15 19:25:16 <walters> cristianiorga, hello
Jan 15 19:25:48 <cristianiorga> walters, I try to do an upgrade to polkit 0.110
Jan 15 19:26:03 <adalton> And, if it's the latter, any pointers to creating a rootfs suitable for lxc?
Jan 15 19:26:14 <cristianiorga> walters, but it seems to have a hard dependency on mozjs185
Jan 15 19:26:16 <walters> cristianiorga, and you ran into the spidermonkey dep, i'm guessing?
Jan 15 19:26:28 <cristianiorga> walters, riiight
Jan 15 19:26:31 <cristianiorga> walters, :-)
Jan 15 19:27:10 <cristianiorga> walters, now... there are some patches for mozjs185, some experimental stuff
Jan 15 19:27:28 <cristianiorga> walters, but to me is seems weird that polkit needs spidermonkey
Jan 15 19:27:35 <walters> cristianiorga, so there is http://cgit.freedesktop.org/polkit/commit/?id=34cb6353b9affd6c04ea480df5fc39d0ca72319d
Jan 15 19:27:44 <cristianiorga> walters, cannot this be avoided?
Jan 15 19:28:16 <walters> cristianiorga, as far as the background on that...the old .pkla format kept getting pressure to become a programming language
Jan 15 19:28:34 <walters> which would turn into a monstrosity like udev rules
Jan 15 19:28:44 <cristianiorga> walters, so they chose JS?
Jan 15 19:28:48 <walters> right
Jan 15 19:29:38 <walters> i know of at least one large corporation whose admins have written custom JS rules already for their linux desktop deployments
Jan 15 19:29:55 <cristianiorga> walters, i did the upgrade to 0.110, which should include that patch
Jan 15 19:30:25 <cristianiorga> walters, still does not work
Jan 15 19:31:58 <walters> what are the symptoms of the failure?
Jan 15 19:33:21 <cristianiorga> pheew
Jan 15 19:33:26 <cristianiorga> 1 min, please
Jan 15 19:38:41 <cristianiorga> http://pastebin.com/erpuR33f
Jan 15 19:40:36 <cristianiorga> walters, that pastebin was for you
Jan 15 19:42:23 <walters> cristianiorga, ah well, it's still a build dependency at the moment.  I'm not actually sure if anyone has submitted a patch for conditional compilation or not
Jan 15 19:42:42 <cristianiorga> it didn't
Jan 15 19:42:53 <cristianiorga> well, at least it is not submitted
Jan 15 19:43:16 <cristianiorga> haven't really checked the mailing lists
Jan 15 19:43:38 <cristianiorga> anyhow, from your insights, mozjs185 dep is here to stay
Jan 15 19:44:02 <cristianiorga> so, what better time to update spidermonkey than this one??
Jan 15 19:44:03 <cristianiorga> :-))
Jan 15 19:44:51 <cristianiorga> I reckon that people will request this feature at some point in YP future, so this needs to be done
Jan 15 19:45:13 <Crofton|work> has anyone watched this?
Jan 15 19:45:14 <Crofton|work> https://training.linuxfoundation.org/free-linux-training/linux-tutorials-and-videos/a-managers-guide-to-building-a-product-with-embedded-linux
Jan 15 19:45:22 <cristianiorga> (wrinting policy scripts in JS, that is)
Jan 15 19:57:48 <adalton> Anyone have any suggestions for building an lxc root filesystem?
Jan 15 20:08:14 <ftonello> kergoth: I think I found out about the .so link name problem.. this library doesn't have a soname ELF tag..
Jan 15 20:08:24 <blloyd> is there a document covering migrating from denzil to danny build systems?
Jan 15 20:09:04 <kergoth> ftonello: that would be a problem, yes
Jan 15 20:40:32 <blloyd> guess that's not too bad.  Only took 2 hours to move from denzil to danny.  Now if only it is a valid move....
Jan 15 21:04:49 <adalton> Question about the openembedded-core repo.  I notice is has much of what comes with yocto itself.  Are these files intended to replace the ones that came with yocto, or is this some sort of layer?
Jan 15 21:04:56 <adalton> (sorry -- I'm very new to yocto)
Jan 15 21:07:33 <zenlinux_> adalton, yep, oe-core is a layer - the fundamental layer that yocto ships
Jan 15 21:08:10 <zenlinux_> When a Yocto release is made, we take a snapshot of oe-core, and that is included in the metadata you get with a yocto release tarball
Jan 15 21:08:28 <adalton> ah, ok -- that make sense
Jan 15 21:09:43 <adalton> let me ask you this: does it make sense to use yocto to build a rootfs for a lxc vm?
Jan 15 21:10:15 <adalton> I found a meta-virtualization layer that includes an lxc recipe, but I'm not sure if that's to build a lxc vm or to build a distro that supports running lxc vms
Jan 15 21:10:35 <walters> adalton, probably, yes; it's quite a flexible system
Jan 15 21:11:26 <adalton> that depended on bridge-utils, so I installed the meta-openembedded/meta-oe layer (that provided bridge-utils), but then that needed gtk-immodules-cache.bbclass
Jan 15 21:11:40 <adalton> which lead me to the openembedded-core question
Jan 15 21:12:06 <adalton> seems that class is in the repo, but wasn't in the last release
Jan 15 21:12:53 <walters> adalton, try removing x11 from DISTRO_FEATURES if it's there now
Jan 15 22:11:47 <adalton> walters: thanks for the suggestion; how do I remove x11 from DISTRO_FEATURES?
Jan 15 22:12:22 <adalton> I mean, what the most appropriate way to do it?
Jan 15 22:13:59 <msm> what's the way to turn on a kernel feature if a package requires it?
Jan 15 22:18:22 <adalton> or more generally, if anyone has any pointers for using yocto to build a lxc rootfs, I'd love to hear them.
Jan 15 22:25:31 <ant_home> zeddii: does linux-yocto-dev contain latest kernel cache? (still fetching...)
Jan 15 22:58:45 <kergoth> adalton: to remove something with distro features, you can make your own distro, or override the definition of it with an appropriate override, or you could use http://git.yoctoproject.org/cgit/cgit.cgi/meta-mentor/tree/classes/user_features.bbclass which makes it a bit easier to alter it from local.conf
Jan 15 22:58:50 <kergoth> s/with/from/
Jan 15 23:03:47 <adalton> kergoth: thanks for the pointer
Jan 15 23:20:22 <ftonello> who firsts launch rc 0?
Jan 15 23:21:00 <ftonello> or better, does someone knows which runlevel is what?
Jan 15 23:26:51 <kergoth> as on all linux systems, /etc/inittab controls the behavior of sysvinit, and spawns all the rc scripts
Jan 15 23:27:01 <kergoth> all non-systemd linux systems, that is
Jan 15 23:29:25 <ftonello> kergoth: thanks
Jan 15 23:30:07 <ftonello> kergoth: there is a line: Runlevels 2-5 are multi-user.
Jan 15 23:30:29 <ftonello> i only see 5 been fired.. what about the rest?
Jan 15 23:33:37 <kergoth> i don't have time to teach a linux 101 course here
Jan 15 23:34:57 <JaMa|Away> :)
Jan 15 23:36:31 <JaMa|Away> kergoth: non-systemd, non-upstart, non-openrc systems :)
Jan 15 23:38:09 <kergoth> heh, true
Jan 15 23:40:08 <khem> what no systemd
Jan 15 23:40:25 <khem> that system must be baptised
Jan 15 23:40:39 <ant_home> khem: JaMa|Away
Jan 15 23:40:46 <JaMa|Away> ant_home
Jan 15 23:41:13 <ant_home> hi, any reason for the new patches in patchwork to be archived?
Jan 15 23:41:28 <khem> for future referals ?
Jan 15 23:42:03 <ant_home> see, I'm used to archive and set them as applied * afterwards *
Jan 15 23:42:14 <JaMa|Away> ant_home: I mark them as archived when I move them to right bundle
Jan 15 23:42:20 <JaMa|Away> ant_home: as discussed on ML
Jan 15 23:42:31 <ant_home> ok, I've missed that bit
Jan 15 23:42:44 <JaMa|Away> it's hidden in some thread, mmt
Jan 15 23:42:45 <khem> ant_home: new ones should not be archived
Jan 15 23:43:13 <JaMa|Away> "[oe] OE-Classic patches in patchwork" thread
Jan 15 23:43:22 <khem> most of times git push hook will do the archiving and stuff automatically when they get pushed
Jan 15 23:43:32 <khem> but it misses sometimes
Jan 15 23:43:46 <ant_home> well, it's just the default filters won't let them show immediately
Jan 15 23:46:35 <ant_home> JaMa|Away: btw, even though I run a local PRServer I see ..2+gitAUTOINC+06...
Jan 15 23:48:58 <ant_home> khem: fyi there must have been some quirk in the toolchain the days I was playing with v4bx for binaries built with klcc. Now all is normal w/out touching the flags in the class
Jan 15 23:49:09 <JaMa|Away> ant_home: in cooker log or .ipk filename?
Jan 15 23:49:11 <ant_home> normal = no bx in the binaries
Jan 15 23:49:35 <JaMa|Away> khem: I mark them as Archived when moving them to right bundle
Jan 15 23:50:29 <JaMa|Away> khem: IIRC when some patch is sent to ML twice, git hook usually marks first one find as Accepted (even when it was Superseeded)
Jan 15 23:50:50 <JaMa|Away> khem: can we change git hook to search only for patches with state "New"?
Jan 15 23:51:36 <khem> JaMa|Away: git hook does md5
Jan 15 23:51:39 <khem> of content
Jan 15 23:52:03 <khem> ant_home: ??
Jan 15 23:52:13 <khem> so the patch you sent is not needed anymore
Jan 15 23:53:05 <JaMa|Away> khem: and when the content is the same? can it mark only "new"->"accepted"
Jan 15 23:53:12 <ant_home> khem: only the one in the ML
Jan 15 23:53:45 <khem> JaMa|Away: yes its not tracking the chain
Jan 15 23:53:52 <khem> may be we can enhance it
Jan 15 23:54:17 <ant_home> JaMa|Away: I'm patching the kernel, I see that in workdir
Jan 15 23:54:23 <ant_home> let it package
Jan 15 23:54:40 <ant_home> khem: the problem was, the -klibc binaries were not really rebuilt :/
Jan 15 23:54:46 <khem> oh
Jan 15 23:54:52 <khem> sstate ?
Jan 15 23:55:01 <JaMa|Away> khem: that would be great, because a lot of time I manually mark something as "superseeded" when I see same patch sent to ML again
Jan 15 23:55:10 <ant_home> I think we have some issue with git, PR and sstate :/
Jan 15 23:55:30 <khem> how to set the PR server
Jan 15 23:55:34 <JaMa|Away> khem: and then when I'm checking why last version wasn't automatically marked as accepted I find that old patch is no longer superseeded but accepted
Jan 15 23:55:37 <JaMa|Away> ant_home: that's correct
Jan 15 23:55:44 <JaMa|Away> khem: https://wiki.yoctoproject.org/wiki/PR_Service
Jan 15 23:55:55 <khem> JaMa|Away: yes there is no chaining
Jan 15 23:56:11 <khem> JaMa|Away: each patch is treated separately
Jan 15 23:56:29 <khem> there are some enhancements in pw itself to detect duplicate patches
Jan 15 23:56:34 <khem> but they all are heuristics
Jan 15 23:57:11 <JaMa|Away> khem: why not just add filter "State==New" where it's filtering by content md5?
Jan 15 23:57:35 <JaMa|Away> that would be enough for correctly marked patches
Jan 15 23:57:53 <JaMa|Away> ant_home: see https://bugzilla.yoctoproject.org/show_bug.cgi?id=3071
Jan 15 23:57:53 <yocti> Bug 3071: enhancement, Medium, 1.4 M2, constantinx.musca, REOPENED , sstate-cache cannot be shared between builders with different cache/bb_persist_data.sqlite3
Jan 15 23:59:18 <khem> JaMa|Away: the md5 is over patch content not pw metadaya
Jan 16 00:06:30 <JaMa|Away> ant_home: and my reply on http://patchwork.openembedded.org/patch/40403/
Jan 16 00:09:00 <ant_home> at least I have only AUTOINC for those using SRCPV like linux-yocto
Jan 16 00:09:22 <ant_home> constantly
Jan 16 00:13:03 <JaMa|Away> ant_home: what about .ipk filenames?
Jan 16 00:13:20 <JaMa|Away> ant_home: or Version in .ipk metadata (PKGV)
Jan 16 00:13:20 <ant_home> still patching, almost done...
Jan 16 00:35:24 <ant_home> JaMa|Away: correct, the ipk has +git0+ instead of +gitAUTOINC+
Jan 16 00:39:40 <ant_home> gn
Jan 16 01:16:14 <ftonello> kergoth: its not about teaching linux 101.. run levels are not necessary standard..
Jan 16 02:06:43 <kergoth> ftonello: sysvinit has been standard in linux systems since at least before 1994. it's also standard in other non-linux unixes which are based on SystemV (as opposed to BSD). only recently have distro started moving to alternatives like upstart and systemd
Jan 16 02:07:18 <kergoth> to clarify, the runlevels and inittab and all are sysv, sysvinit itself is a specific project
Jan 16 02:29:04 <ftonello> kergoth: http://en.wikipedia.org/wiki/Runlevel it appears that oe follows debian standard
Jan 16 02:34:20 <ftonello> kergoth: the weired thing is that when i update to poky 8, my scripts are not been called by rc
Jan 16 02:34:53 <ftonello> it looks like runlevel 5 is not been fired.. but i know it is because xserver-nodm is there
Jan 16 02:55:56 <EddyLaiTW> hi, I have download and build the yocto 1.3, config the kernel to use "Intel IGB network driver", then boot the image from the USB
Jan 16 02:56:21 <EddyLaiTW> found the IGB driver inside is v3.2.10
Jan 16 02:57:24 <EddyLaiTW> how can I patch it to v4.2 from intel's (sourceforge) website
**** ENDING LOGGING AT Wed Jan 16 02:59:59 2013