**** BEGIN LOGGING AT Wed Feb 01 02:59:57 2012 Feb 01 09:59:00 morning all Feb 01 09:59:12 mornin Feb 01 10:00:41 good morning Feb 01 10:13:46 hi all Feb 01 10:14:10 hi pb_ Feb 01 10:36:59 good morning Feb 01 10:49:30 morning Feb 01 11:00:27 Hello. Could someone possibly explain where a basic 'inherit scons' recipe looks for its SConstruct file? Feb 01 14:46:19 http://www.linuxfoundation.org/news-media/blogs/browse/2012/01/free-embedded-linux-training-yocto-developer-day-february-14th Feb 01 14:46:41 hmm, I think this describes our relationship with Yocto fairly well: Feb 01 14:46:57 The project is supported by major embedded hardware vendors, embedded Linux operating system vendors, the OpenEmbedded Project, and many other organizations, with a governance structure based on the open source tenets of transparency and meritocracy. Feb 01 15:03:28 03Paul Eggleton  07master * r43b473275d 10bitbake.git/ (bin/bitbake-layers lib/bb/cooker.py lib/bb/providers.py): (log message trimmed) Feb 01 15:03:28 bitbake-layers: improve show-overlayed output Feb 01 15:03:28 Make the following improvements to the show-overlayed subcommand: Feb 01 15:03:28 * Show recipes that are overlayed when the version is higher or lower, Feb 01 15:03:28 not just when it is the same. This gives a much better picture of the Feb 01 15:03:29 influence each layer is having over the metadata used for building. Feb 01 15:03:30 This can be disabled with the -s option if you just want to see Feb 01 15:03:30 03Paul Eggleton  07master * rf6493e4bad 10bitbake.git/bin/bitbake-layers: (log message trimmed) Feb 01 15:03:31 bitbake-layers: list overlayed classes in show-overlayed Feb 01 15:03:31 Classes (.bbclass files) can be overlayed in a layer although they are Feb 01 15:03:32 currently located by BitBake in a different way (via BBPATH instead of Feb 01 15:03:32 using layer priority) and thus it is useful to be able to see when this Feb 01 15:03:33 is in effect and which layer's class is actually being used. Feb 01 15:03:33 Signed-off-by: Paul Eggleton Feb 01 15:03:34 03Paul Eggleton  07master * r05e86ba966 10bitbake.git/bin/bitbake-layers: (log message trimmed) Feb 01 15:03:34 bitbake-layers: add show-recipes subcommand Feb 01 15:03:46 bitbake-layers: add copyright notice Feb 01 15:03:46 This ought to have been added earlier. (I consulted with Chris Larson on Feb 01 15:03:47 (17 lines omitted) Feb 01 18:11:05 re Feb 01 18:37:31 has anyone noticed any problems with busybox udhcpc? on my system its sometimes started several times in parallel Feb 01 18:37:44 I am debugging the scripts, but not yet sure where it is coming from Feb 01 18:37:59 build is OE Core / Angstrom based Feb 01 18:38:06 multiple interfaces set to auto? :) Feb 01 18:38:26 kergoth: nope Feb 01 18:38:31 that's the confusing part :) Feb 01 18:39:23 both are started to for eth0 Feb 01 18:39:28 damn, was hoping it was an easy one :) Feb 01 18:39:30 dunno Feb 01 18:40:09 I could not yet figure out who exactly is supposed to be starting the dhcp client Feb 01 18:40:18 ifup? Feb 01 18:40:41 most likely, yeah Feb 01 18:45:41 is Alex Lennon here? Feb 01 19:12:48 anyone know how to get distutils to pass through some LDFLAGS? Feb 01 19:12:56 i've never used distutils at all Feb 01 19:13:23 ldflags= ? Feb 01 19:54:10 Hi guys, I'm ready to start hacking on some of the kernel source files. What is the preferred way to do this? Figure out what revision oe is pulling for my image, clone repo, checkout that revision that and start editing, then change the src uri from the web to my working copy? Feb 01 19:54:47 Or perhaps just hack on it in the work directory until I've got something that works Feb 01 19:55:34 spacecolonyone: if you hack in the workdir or devshell, be carefull not to -c clean your changes Feb 01 19:56:40 right. eventually I'll still need to clone and c/o to diff against, generate patch and add that to my custom recipe, no? Feb 01 19:58:09 spacecolonyone: build a meta-toolchain for concerned arch install it on your machine and use this SDK to do your work is another option Feb 01 19:58:41 and then when you commit those changes to your kernel git tree you can keep updating the SRCREV in the recipe Feb 01 19:58:49 well it depends on how you want to approach it I guess, not seldom if I have to fix something in an existing recipe I hack in -c devshell (it's convenient because you have all the scripts to compile and install the package) and when I feel that it works well enough I diff my changes to the original package, then copy the diff to the recipe and make sure it gets applied, and rebuild the whole thing then Feb 01 19:58:50 and keep rebuilding rfs+kernel Feb 01 20:01:11 khem I'm a bit concerned that that route means I'll be learning two tool chains, also since I'm building this for an astronomical instrument that someone else may oneday support I'd like to keep the number of steps to a minimum. E.g. I want to keep it so I or anyone in the future can simply do bitbake my-image Feb 01 20:02:19 koen: ping Feb 01 20:02:30 Jin^eLD: that should promising, when you say hack in -c devshell you are referring to arguments to what? bitbake? I just tried that and it doesnt seem to work Feb 01 20:02:49 spacecolonyone: if you want to try my approach I'd do the following Feb 01 20:03:06 s /should /sounds Feb 01 20:03:12 khem: Are you the one working in kmod for oe-core? Feb 01 20:03:29 make sure you remove rm_work from your conf, just to make sure it does not wipe your workdir too soon Feb 01 20:03:46 then clean your recipe, i.e. bitbake -c cleanall yourkernel Feb 01 20:03:54 theb bitbake -c install yourkernel Feb 01 20:03:59 then bitbake -c devshell yourkernel Feb 01 20:04:13 from this point on you can start hacking; to compile you simply run ../temp/run.do_compile.xxx Feb 01 20:04:17 otavio: yes I am Feb 01 20:04:33 otavio: I am seeing a segfault when generating rootfs Feb 01 20:04:45 if you need to generate the uboot image, etc, just see what other run scripts you have in ../temp - they will do the work for you Feb 01 20:04:57 its kind of strange that it does not dump core so I am not sure if its kmod binary or something else Feb 01 20:05:02 but I know its due to kmod Feb 01 20:05:04 btw when I am speaking of "temp" in this sense I mean workdir/yourarch/yourkernelrecipe/ Feb 01 20:05:08 khem: right Feb 01 20:05:22 otavio: so I am still working on isolating the problem Feb 01 20:05:33 that directory isn't under git control though, right? Feb 01 20:05:43 khem: I am asking because I am working at udev and for 180 we need it. Feb 01 20:06:05 otavio: ok. I think once I get it working we can have it on oe-core Feb 01 20:06:24 but so far it does have some kinks to iron our Feb 01 20:06:27 spacecolonyone: afaik it's not, I remember it used to be in OE classic couple of years ago (at least it did work for svn back then, not sure about git) Feb 01 20:06:34 khem: I'll keep working using 175 as base but I'd like to propose 180 directly as it is going to drop a lot of diverging code between oe-core and meta-oe Feb 01 20:07:08 so then to diff and patch I still need to go manually clone the kernel elsewhere right? Feb 01 20:07:08 khem: once you have something usable, please ping me so I can test the patches and get it on a local branch to play with it Feb 01 20:07:26 otavio: I already have something Feb 01 20:07:43 spacecolonyone: yes, or just copy the sources in workdir to some .orig directory so you have the original for later diff'ing Feb 01 20:07:44 denix: ping Feb 01 20:07:53 otavio: http://git.openembedded.org/openembedded-core-contrib/log/?h=kraj/misc-fixes Feb 01 20:07:58 see the top 3 patches Feb 01 20:08:15 darknighte: why that extra 'e' at the end :) Feb 01 20:08:24 darknighte: does that mean 'embedded' Feb 01 20:08:58 khem: it's always been there, but you can think of it as embedded, if you please. Feb 01 20:09:12 khem: you going to ELC in 2 weeks? Feb 01 20:09:17 yes I do Feb 01 20:09:33 hmm I always thought it was without e Feb 01 20:09:42 ok. so I actually had rm_work in my local.conf, but I was still able to find the file I need to screw with (board-omap3beagle.c) hanging around in a subdir or /work/ Feb 01 20:10:28 khem: but if I can't build a rootfs it doesn't work for me to test Feb 01 20:10:38 khem: so it is a no-go for me at this moment Feb 01 20:10:52 otavio: at least try may be its something crazy in my env Feb 01 20:11:28 can anyone confirm that the meta-ti layer supports both the original beagleboard and the beagleboard xM? Feb 01 20:11:37 khem: will do; but before I need to finish 175 as it is quite close of 180 regarding as the changes needed in oe to handle it Feb 01 20:11:48 khem: mostly because of backward compatibility issues Feb 01 20:11:52 nod Feb 01 20:15:20 darknighte: pong Feb 01 20:15:26 darknighte: I can confirm Feb 01 20:15:43 denix: thought as much, but thought I'd check. Feb 01 21:11:36 where are the OE guys staying for FOSDEM this year? Feb 01 21:12:53 khem: ooh, kmod instead of module-init-tools for images? nifty Feb 01 21:13:09 kergoth: yes :) Feb 01 21:13:25 kergoth: I dont see any reason to have module-init-tools-cross do you ? Feb 01 21:13:43 wouldn't think so, not anymore anyway Feb 01 21:13:51 yeah Feb 01 21:14:14 I have this crazy segfault that I am not sure is in kmod or caused in some other tool by kmod Feb 01 21:14:22 starce is not dumping anything Feb 01 21:14:38 and setting ulimit -c umlimited is not saving core Feb 01 21:14:55 but it says it recieved a segfault Feb 01 21:15:01 this do_rootfs task Feb 01 21:15:29 strange Feb 01 21:15:45 it is indeed Feb 01 21:17:02 but if I use depmod from module-init-tools all works Feb 01 21:17:16 so its definitely something kmodish Feb 01 21:18:05 SIGFOOD Feb 01 22:36:30 evening all Feb 01 22:37:21 hi pb_ Feb 01 22:38:20 Hi pb_ you back home or still in states Feb 01 22:41:13 hi Feb 01 22:41:24 tzdata 2001g is in meta-oe Feb 01 22:41:36 tzdata 2011n is in openembedded-core Feb 01 22:42:14 GNUtoo: there was a patch to consolidate the differences into oe-core Feb 01 22:42:23 after that one if meta-oe should go away Feb 01 22:42:33 ok Feb 01 22:42:37 because we have that bug: Feb 01 22:42:50 http://www.shr-project.org/trac/ticket/1549 Feb 01 22:42:51 cheers guys, I'm off to CA to build this spectrograph! Feb 01 22:44:14 khem, when the patch will be in? Feb 01 22:44:21 no idea Feb 01 22:44:41 or may be it made in I havent tracked Feb 01 22:44:57 ok Feb 01 22:45:29 thanks a lot for the infos Feb 01 22:49:16 it didn't made it Feb 01 22:49:19 there was a ping **** ENDING LOGGING AT Thu Feb 02 02:59:57 2012