**** BEGIN LOGGING AT Mon Apr 30 02:59:58 2012
Apr 30 08:01:03 <ant_work> _tasslehoff_: heh, your msgs are dated 1st Mai
Apr 30 08:05:11 <_tasslehoff_> ant_work: damnit! I really thought I had fixed this now :)
Apr 30 08:05:20 <_tasslehoff_> bloody 12.04
Apr 30 09:05:09 <kenws> good morning
Apr 30 09:06:31 * kenws wonders how the matchbox window manager gets started on core-image-sato...
Apr 30 09:36:13 <JaMa> otavio: ping
Apr 30 11:21:55 <garble> Hey, is it possible, in a recipe, to get the PV of another package and use that in a string? If so how? 
Apr 30 12:17:54 <otavio> JaMa: oi
Apr 30 12:21:41 <erwt> hello
Apr 30 12:21:52 <erwt> any udev experts here ?
Apr 30 12:23:26 <erwt> i want to stop  auto mounting of pendrive during boot up time ,i want to enable this feature after my file sytem is completely up .
Apr 30 12:24:18 <erwt> can some one tell me how do i delay or stop this auto  mounting through udev  
Apr 30 12:28:16 <JaMa> otavio: do you have fix for firefox + gcc-4.7? or are you waiting for someone to send patch for meta-mozilla?
Apr 30 12:28:35 <otavio> JaMa: I am waiting
Apr 30 12:29:06 <otavio> JaMa: I am finishing a product release and I didn't update the oe-core repo here for a while
Apr 30 12:29:15 <JaMa> ok
Apr 30 12:29:30 <otavio> JaMa: if you can ... try to update firefox as well; it would be awesome!
Apr 30 12:30:01 * JaMa also quite busy.. so hopefully someone will be faster
Apr 30 12:30:30 <JaMa> it didn't work with gcc-4.6 and gold anyway...
Apr 30 12:31:37 <otavio> JaMa: really?
Apr 30 12:31:41 <otavio> JaMa: grr
Apr 30 12:32:03 <JaMa> I was told it just segfaults a lot..
Apr 30 12:32:28 <JaMa> but I'm not using it, GNUtoo was testing it
Apr 30 14:59:34 <pb_> morning all
Apr 30 15:00:45 <florian> hey pb_ 
Apr 30 16:08:41 <awsation> Hi, I am trying to bitbake the monit recipe found on git.openembedded.org. It is successfull with EXTRA_OECONF = "--without-ssl"
Apr 30 16:08:51 <awsation> but when I try EXTRA_OECONF = "--with-ssl-lib-dir=${STAGING_LIBDIR} --with-ssl-incl-dir=${STAGING_INCDIR}"
Apr 30 16:09:05 <awsation> it fails with this error:
Apr 30 16:09:40 <awsation> | Couldn't find your SSL library files.
Apr 30 16:09:40 <awsation> | Use --with-ssl-lib-dir option to fix this problem or disable the
Apr 30 16:09:40 <awsation> | SSL support with --without-ssl
Apr 30 16:09:40 <awsation> | 
Apr 30 16:09:40 <awsation> | ERROR: oe_runconf failed
Apr 30 16:09:41 <awsation> NOTE: package monit-4.10.1-r2: task do_configure: Failed
Apr 30 16:09:54 <awsation> any ideas?
Apr 30 17:27:52 <khem> awsation: is openssl staged ?
Apr 30 17:28:01 <khem> awsation: you need to add openssl to DEPENDS 
Apr 30 17:31:59 <awsation> I do have  DEPENDS = "openssl"
Apr 30 17:33:38 <awsation> looking further into the build log, I also see the following: 
Apr 30 17:34:43 <awsation> | checking for _FILE_OFFSET_BITS value needed for large files... 64
Apr 30 17:34:43 <awsation> | checking for SSL support... enabled
Apr 30 17:34:44 <awsation> | checking for SSL include directory... /home/user/build/tmp/qemu86-core-statmp/sysroots/qemux86/usr/include
Apr 30 17:34:44 <awsation> | checking for SSL library directory... Not found
Apr 30 17:34:44 <awsation> | 
Apr 30 17:34:44 <awsation> | Couldn't find your SSL library files.
Apr 30 17:34:46 <awsation> | Use --with-ssl-lib-dir option to fix this problem or disable the
Apr 30 17:34:52 <awsation> | SSL support with --without-ssl
Apr 30 17:34:54 <awsation> | 
Apr 30 17:34:56 <awsation> | ERROR: oe_runconf failed
Apr 30 17:37:11 <awsation> the ssl lib cannot be found in ${STAGING_LIBDIR}
Apr 30 20:49:17 <GNUtoo-desktop> hi khem 
Apr 30 20:50:07 <GNUtoo-desktop> is that normal for armv4: http://paste.debian.net/166740/
Apr 30 20:53:54 <fray> has anyone tried to use boost (or anything that uses boost) with master of oe-core?
Apr 30 20:53:59 <fray> I'm getting failures in the STL
Apr 30 21:09:40 <khem> GNUtoo-desktop: whats up
Apr 30 21:10:55 <pb_> re
Apr 30 21:11:24 <pb_> Tartarus: it is a shame you are not standing for re-election to the tsc.
Apr 30 21:11:38 <GNUtoo-desktop> khem, hi I wondered if the enabled on neon was a problem
Apr 30 21:11:50 <Tartarus> Yeah, but just don't have the time to give OE stuff the attention it deserves from someone on the TSC
Apr 30 21:12:01 <khem> GNUtoo-desktop: if you pass correct mtune march it should take care of that
Apr 30 21:12:06 <GNUtoo-desktop> ok
Apr 30 21:12:08 <GNUtoo-desktop> thanks a lot
Apr 30 21:12:15 <Tartarus> it's hard work being a trained monkey you know :)
Apr 30 21:12:17 <khem> GNUtoo-desktop: we dont configure for a given arch per say
Apr 30 21:12:26 <pb_> Tartarus: yeah, that is too bad.
Apr 30 21:12:47 <GNUtoo-desktop> ah ok, I tought so
Apr 30 21:12:47 <khem> if we used -with-tune --with-arch then it will havce different defaults
Apr 30 21:13:07 <GNUtoo-desktop> ahhh that explain why the compiler can output for armv7
Apr 30 21:14:21 <Jay7> GNUtoo-desktop: ping
Apr 30 21:14:30 <GNUtoo-desktop> Jay7, pong
Apr 30 21:14:40 <Jay7> GNUtoo-desktop: is nokia n900 bundled with cable suitable to reflash?
Apr 30 21:14:48 <khem> pb_: once you mentioned you were using mips32 soc
Apr 30 21:14:53 <Jay7> i.e. no special system cable is needed?
Apr 30 21:15:01 <GNUtoo-desktop> Jay7, usually yes but note that the first bootloader(x-loader) is signed
Apr 30 21:15:03 <khem> pb_: is it something of the shelf ?
Apr 30 21:15:04 <GNUtoo-desktop> then comes nolo
Apr 30 21:15:14 <GNUtoo-desktop> which isn't signed and then comes the kernel
Apr 30 21:15:44 <GNUtoo-desktop> so what you flash is the kernel or whatever that takes its place(like a chainloaded uboot)
Apr 30 21:15:50 * Jay7 is thinking about getting n900 into collection :)
Apr 30 21:16:08 <pb_> khem: sort of.  it's a broadcom SoC.  the particular IC we are using is not an off-the-shelf part but I think the same core is in quite a lot of their other chips.
Apr 30 21:16:09 <GNUtoo-desktop> Jay7, do you know alsa?
Apr 30 21:16:21 <Jay7> GNUtoo-desktop: easy answer is no :)
Apr 30 21:16:44 <GNUtoo-desktop> ouch because in SHR we have sound quality issues during calls + the port is not finished
Apr 30 21:16:50 <khem> pb_: ok.
Apr 30 21:17:00 <khem> pb_: are you using oe on it ?
Apr 30 21:17:06 <GNUtoo-desktop> but JaMa uses it not as a phone and he's happy with it
Apr 30 21:17:18 <khem> pb_: I was wondering what would be a good reference board for mips 
Apr 30 21:17:30 <Jay7> GNUtoo-desktop: you may propose it as CEWG project
Apr 30 21:17:30 <khem> do u have any suggestions
Apr 30 21:17:43 <Jay7> (former CELF)
Apr 30 21:18:36 <GNUtoo-desktop> Jay7, ah ok
Apr 30 21:19:00 <pb_> khem: yeah, we are using oe on it and it seems to be working fairly well.
Apr 30 21:19:07 <Jay7> this year proposal deadline is over but who knows.. :)
Apr 30 21:19:26 <khem> pb_: ok.
Apr 30 21:27:44 <kenws> Hi there, I just noticed that GCC 4.7 and Qt 4.8 have landed in master. Is the tree unfrozen again and if so, is there any release tag or something?
Apr 30 21:27:55 <fray> yes and yes
Apr 30 21:28:10 <fray> "denzil" is the release branch
Apr 30 21:28:19 <kenws> ah
Apr 30 21:28:32 <fray> I'm not sure if there is a release tag yet.. but there should be any day..
Apr 30 21:28:37 <fray> I would expect it to be 2012-1
Apr 30 21:29:06 <khem> kenws: now you can merge all my work into meta-linaro
Apr 30 21:29:08 <kenws> I just issued a git fetch and checked http://cgit.openembedded.org/openembedded-core/refs/ - nothing there yet
Apr 30 21:29:10 <khem> from my layer
Apr 30 21:29:23 <khem> kenws: its a different branch
Apr 30 21:30:06 <kenws> khem: yep, I'll do so. It may take a little longer because I want to have a meta-linaro support the release
Apr 30 21:30:25 <khem> kenws: sure 
Apr 30 21:30:30 <khem> tag it denzil
Apr 30 21:30:43 <kenws> I've just started to build the sato and qt4e images targeting mips and ppc using the linaro 4.6 toolchain
Apr 30 21:31:00 <kenws> khem: the tag or the branch?
Apr 30 21:31:30 <khem> kenws: I was suggested a branch name for you 
Apr 30 21:31:40 <khem> which will then work with denzil release of OE-Core
Apr 30 21:32:02 <fray> in the oe-core repository denzil is a branch..  I don't yet see a tag
Apr 30 21:32:03 <kenws> aah, got it - thx : )
Apr 30 21:32:23 <kenws> if there any documentation about the oe-core release methodology?
Apr 30 21:32:29 <kenws> s/if/is/
Apr 30 21:32:57 <fray> there is some.. look for a post from Song Liu on the oe-core list..
Apr 30 21:33:06 <fray> (it's talking Yocto Project, but it's basically the same thing)
Apr 30 21:33:34 <fray> over the next couple of weeks, I have the action (as part of the OE TSC) to better document the release process and have an ongoing status indicator of what phase we're in, and when the next phase is likely to begin
Apr 30 21:33:37 <khem> kenws: releases are 6 months apart April and October roughly
Apr 30 21:33:55 <fray> yup.. thats the core framework to work under..
Apr 30 21:34:07 <khem> which also coincide with ubuntu :) hopefully makes your life easier
Apr 30 21:34:08 <fray> since we just released.. May though August is likely to be all development..
Apr 30 21:34:08 <kenws> I've read that oe-core is following the yocto 6mnth release cycle but haven't noticed that the denzil branch is going to be the release (until now)
Apr 30 21:34:17 <fray> Sept - Oct progressive freeze to release
Apr 30 21:34:28 <kenws> I see : )
Apr 30 21:34:34 <khem> kenws: you should pay attention to ml :)
Apr 30 21:34:45 <fray> kenws, I don't think an announcement has gone out yet that it's been released.. but it's -very- close..
Apr 30 21:35:00 <khem> its being released today I think
Apr 30 21:35:03 <fray> RP sent out an email about the Denzil branch though, and opening of master..  (at least i thought he did)
Apr 30 21:35:09 <fray> khem, yup AFAIK today or tomorow
Apr 30 21:35:15 <khem> ok
Apr 30 21:36:07 <kenws> hm, if I would have known this earlier (about the denzil branch) I could have tested it earlier. But you are right, I'm not following the ml closely - I'm usually just grepping through the archive...
Apr 30 21:36:37 <khem> kenws: its ok 
Apr 30 21:36:54 <khem> you can still do it. layers release doesnt have to be same day
Apr 30 21:37:04 <fray> kenws, that is why we (the TSC) have decided we need a status page explaining where we are..
Apr 30 21:37:16 <khem> as long as you test and tag the layer with corresponding OE-Core release you should be good to go
Apr 30 21:37:25 <fray> there were a lot of complaints about the freeze to release part of the process... people were unprepared for a slow down in the acceptance of changes
Apr 30 21:37:56 <khem> I always have proposed to slow down and release 
Apr 30 21:38:09 <khem> it just works for better quality
Apr 30 21:39:10 <kenws> khem: yeah sure, but meta-linaro relies on functionality provided by oe core and some of it like the binary toolchain stuff is rarely used/tested
Apr 30 21:40:22 <fray> kenws, be aware if you are working on binary toolchains that folks from Mentor and Wind River are doing similar work in preparing, distributing and verifying configuration of binary toolchains...
Apr 30 21:40:26 <kenws> I guess it's not that far from what I'm using right now - so it should be all good : )
Apr 30 21:40:38 <fray> (initially with the existing csl_toolchain class) but many of the concepts are expected to become more universal
Apr 30 21:40:39 <kenws> yep, kergoth is doing some great work there
Apr 30 21:41:15 <fray> as WR we're focusing on validation and tuning control.. we want to be sure that our customers (when using the binary toolchains) don't stray into unsupported tunes.. since they are likely to get errors or incorrect compilations..
Apr 30 21:41:26 <kenws> I wish I could build the libc instead of copying the one provided by the binary toolchain. but I haven't found the time to dig into it
Apr 30 21:41:47 <fray> that is another thing we will be working on.  I thought there was already a switch, but maybe we were talking about adding one?
Apr 30 21:42:15 <fray> but the default is use the binary, switched it will enable build using the regular recipe sources [or a replacement alternative]
Apr 30 21:42:23 <kenws> fray: afair kergoth wanted the same a few weeks ago and we both thought it wasn't there yet
Apr 30 21:43:07 <kenws> yes, having a switch would be great
Apr 30 21:43:11 <fray> I've been only tracking the work on the side.. "seebs" is the one working on that for WR
Apr 30 21:43:33 <kenws> ok, good to know there are others out there : )
Apr 30 21:44:30 <khem> kenws: actually any components for that matter should be pluggable
Apr 30 21:44:48 <khem> in theory if it works in practice or not is another thing
Apr 30 21:45:01 <fray> khem, ya, I agree.. the binary toolchains are a fairly special case though -- since they are critical for the system as a whole to work
Apr 30 21:45:11 <khem> but if someone said use binutils from internal toolchain and only gcc from external toolchain then that should be possible
Apr 30 21:45:50 <kenws> khem: using the "provides" mechanism?
Apr 30 21:45:52 <khem> so infact replacing target bits. e.g. libc should be a bit simpler
Apr 30 21:46:07 <fray> khem ya, it should work.. but thats one thing I don't think will right now  ;)
Apr 30 21:46:17 <khem> true 
Apr 30 21:46:37 <khem> kergoth's work and seebs work will pave path for it though
Apr 30 23:02:16 <JeremyClarkson> Hey all, newbie question here.  I'm trying to add a custom recipe to a custom console image build target, which builds a kernel module.  I'm wondering how I can set up the do_install task to add the built module (.ko file) to the sys/modules directory. of the rootfs.  I've done a lot of reading, but I haven't found the answer to my question.  I'd appreciate any pointers/help.
Apr 30 23:04:06 <khem> JeremyClarkson: are you usinf OE-Core based setup
Apr 30 23:04:45 <khem> you can add inherit module
Apr 30 23:04:46 <JeremyClarkson> I'm using a gumstix board, and this (http://gumstix.org/software-development/open-embedded/61-using-the-open-embedded-build-system.html) guide.  (I'm not sure)
Apr 30 23:04:50 <khem> to your recipe
Apr 30 23:05:02 <khem> export KERNEL_SRC="${STAGING_KERNEL_DIR}"
Apr 30 23:05:09 <JeremyClarkson> ok, I currently inherit module
Apr 30 23:05:30 <khem> so now you want to add the module to your image right ?
Apr 30 23:05:35 <JeremyClarkson> yes
Apr 30 23:05:40 <khem> its building fine I suppose
Apr 30 23:05:42 <JeremyClarkson> looks like I need to do the export
Apr 30 23:05:47 <JeremyClarkson> yep, building fine
Apr 30 23:05:54 <JeremyClarkson> what does that export line do?
Apr 30 23:06:18 <khem> oh it finds the kernel sources 
Apr 30 23:06:25 <khem> but I I guess you have it right 
Apr 30 23:06:36 <JeremyClarkson> yeah
Apr 30 23:07:09 <JeremyClarkson> if it helps, this is my current install task:
Apr 30 23:07:25 <JeremyClarkson> do_install () {
Apr 30 23:07:25 <JeremyClarkson> 	install -d ${D}${base_libdir}/modules/${KERNEL_VERSION}/kernel/drivers/test
Apr 30 23:07:26 <JeremyClarkson> 	install -m 0644 ${S}/fsp-accessory*${KERNEL_OBJECT_SUFFIX} ${D}${base_libdir}/modules/$  {KERNEL_VERSION}/kernel/drivers/test
Apr 30 23:07:27 <JeremyClarkson> }
Apr 30 23:07:56 <JeremyClarkson> does that seem like it should work?
Apr 30 23:08:15 <JeremyClarkson> (and again, I really appreciate the help)
Apr 30 23:08:21 <khem> mostly you should not need do_install
Apr 30 23:08:34 <khem> module bbclass should take care of it
Apr 30 23:09:03 <khem> RDEPENDS_pn-<your-image-name>_append = " kernel-module-<yourmodule>"
Apr 30 23:09:03 <JeremyClarkson> what does module bbclass do/mean
Apr 30 23:09:22 <khem> if you add that above to your local.conf file 
Apr 30 23:09:31 <khem> it should then get into image dependency
Apr 30 23:09:39 <khem> and get pulled in when building the image
Apr 30 23:10:18 <khem> first you build it and second part is adding it to image 
Apr 30 23:10:48 <JeremyClarkson> awesome.  So by adding the dependency statement to my local.conf, I can remove the do_install task?
Apr 30 23:11:06 <khem> well, I dont know to be sure 
Apr 30 23:11:17 <khem> but module class will have a install task too 
Apr 30 23:11:22 <khem> which generally is sufficient 
Apr 30 23:11:33 <khem> unless your module build system needs some special things
Apr 30 23:11:46 <khem> then you can either override the install task like you were doing
Apr 30 23:12:07 <khem> or you can also tweak it by appending and prepending to the original install task
Apr 30 23:12:43 <JeremyClarkson> I don't believe it does.  I'll comment out the install task and test it.  Also, what targets do I need to clean before rebuilding for a kernel module?  (I'm a bit confused with how bitbake handles rebuilding)
Apr 30 23:13:02 <khem> just clean your module
Apr 30 23:13:50 <JeremyClarkson> excellent.  I was doing a linux-omap3 clean to be safe (which takes a wuile to rebuild)
**** ENDING LOGGING AT Tue May 01 03:00:01 2012