**** BEGIN LOGGING AT Fri Oct 11 02:59:58 2013 Oct 11 06:53:28 * kroon is on a flaky connection at the moment Oct 11 06:55:19 Is there a better way to include kernel-modules in an image other then using the recipe kernel-modules or selecting them individually? Oct 11 07:42:53 good morning Oct 11 07:55:15 if I just update SRVREV in a recipe (that pulls from git), should bitbake automagically understand that it needs to rerun do_fetch so that this new SRCREV is used? Oct 11 07:56:29 or do I explicitly have to state in the recipe that do_fetch depends on SRCREV? Oct 11 08:04:04 could it be that the newer source rev was already in the git clone in downloads/ and just not used before? Oct 11 08:04:22 erbo: SRCREV is used through PV. So what is important is to make sure that PV relies on SRCREV. Oct 11 08:05:04 and yes it is possible that the 'new' commit was there already. in which case, it would not fetch anything. however it should unpack a new build directory. Oct 11 08:05:17 ndec: that explains it, I just a bogus PV.. Oct 11 08:05:27 thanks Oct 11 08:05:44 it's indeed PV that bitbakes tracks for detecting 'new' stuff. Oct 11 08:53:20 morning all Oct 11 09:10:35 does OE support f2fs for images? Oct 11 09:13:26 dv_: I don't think so out of the box, but adding support shouldn't be too difficult Oct 11 09:14:26 hmmk Oct 11 09:14:37 dv_: we have a recipe for f2fs-tools in meta-oe; if you want to look into this have a look at meta/classes/image_types.bbclass to see what would need to be done Oct 11 09:15:03 should just be a matter of defining IMAGE_CMD_f2fs and maybe a few other variables Oct 11 10:55:28 hi all, hi bluelightning, hi mckoan Oct 11 11:10:16 broonie: what's up with linux-next ? Oct 11 11:10:48 broonie: stilll next-20130927 Oct 11 11:13:26 bluelightning, http://www.strava.com/activities/88295277?fb_source=708 Oct 11 11:14:00 Crofton|work: hi Oct 11 11:14:05 gm Oct 11 11:14:12 Crofton|work: impressive! Oct 11 11:14:18 heh Oct 11 11:14:23 not so bright, in the dak Oct 11 11:14:28 well I had a light Oct 11 11:14:45 some of the locals have a race .... Oct 11 11:15:02 wrt autobuilder, I was meaning it would be just great if we gather an host doing build for arch/machines not normally tested (think armv4 and armv5te) Oct 11 11:15:51 JaMa's atobuilder is already doing some Oct 11 11:16:21 armv5te is I think normally tested with qemuarm at least Oct 11 11:16:33 isn't yet moved to v7 ? Oct 11 11:16:47 no Oct 11 11:39:41 ant_work: sfr is on holiday, myself and tagr have been doing trees. Oct 11 11:40:33 but we're using a separate repo at git://gitorious.org/thierryreding/linux-next.git Oct 11 11:53:06 ok, I've already found that, thx. Will be merged soon then Oct 11 11:54:05 (I'm waiting for a fix probably in -rc5) Oct 11 11:54:35 at first is a bit hard with all that repos ;) Oct 11 11:54:58 btw, little progresses with the locomo-spi driver Oct 11 11:56:03 atm compiles but does not register, hope to clean it on weekend before refactor-it Oct 11 11:56:53 Is tools-profile a valid IMAGE_FEATURES option for standard OE? Oct 11 11:57:03 I'm trying to get perf to work Oct 11 12:08:46 hi! can someone please give me a hint what I have to add to my image recipe to get g_multi (usb gadget module)? Oct 11 12:09:33 exosyst: only if your recipe inherits core-image rather than image Oct 11 12:19:18 i don't suppose anyone has a system running uclibc they can run a five line C file on? Oct 11 12:35:28 bluelightning, Hmm, I'm using the anstrom scripts to create a systemd-image. How can I check? Oct 11 12:36:51 exosyst: what is your problemß Oct 11 12:37:03 exosyst: looks like systemd-image is in meta-angstrom; and looking at it it inherits image not core-image Oct 11 12:37:34 OK, so no tools-profile - fair enough :( Oct 11 12:42:38 what is perf packaged in? Oct 11 12:44:58 exosyst: not through IMAGE_FEATURES, however if you look at core-image.bbclass you can see PACKAGE_GROUP_tools-profile = "packagegroup-core-tools-profile" which means adding packagegroup-core-tools-profile to IMAGE_INSTALL would accomplish the same thing Oct 11 12:46:01 Crofton|work: the perf package is called... perf Oct 11 12:46:44 bluelightning, Crofton|work I wanted to know that yesterda Oct 11 12:46:47 too obvious Oct 11 12:46:59 y as the perf shipped with the image is to do with framebuffers ;-/ Oct 11 12:47:18 I'd sort of like to know what package that comes from Oct 11 12:48:07 although that is from the part of me that does not have enough time to spend on OE :) Oct 11 12:48:24 hi Oct 11 12:49:00 Crofton|work: exosyst: I would like to know that as well Oct 11 12:49:40 grepping for perf isn't helpful :) Oct 11 12:50:01 bluelightning, Crofton|work well it's not in my self-made systemd image - perhaps a remnant in an old recipe somewhere on the build server? Oct 11 12:50:08 I assume part of the issue is the default bbb image has loads of stuff Oct 11 12:50:23 the shipping image has loads of preinstalled stuff Oct 11 12:57:29 Nothing provides gdbserver? Really? That's not a default recipe? - How can I find where it went? Oct 11 12:58:23 part of gdb, iirc Oct 11 12:58:40 one recipe may produce several packages Oct 11 12:59:12 Oh really? I still wish there was a who-provides script or something but wouldn't know how to query it! Oct 11 12:59:13 PACKAGES =+ "gdbserver" Oct 11 12:59:15 says gdb Oct 11 12:59:31 hm, bb might have something like that Oct 11 13:01:09 whilst playing package pub quiz - where's the full version of ping? Is it now iputils-ping or similar? Oct 11 13:04:45 exosyst: I believe iputils-ping yes Oct 11 13:05:05 Tried and failed, maybe it's just iputils now Oct 11 13:05:56 yup - love moving targets :D Oct 11 13:06:16 which recipe is actually giving the default big BBB image? is it /setup-scripts/sources/meta-ti/recipes-misc/images/cloud9-gnome-image.bb Oct 11 13:07:05 Anguel, I believe it is Oct 11 13:12:25 exosyst: so cloud9-gnome-image.bb -> require cloud9-image.bb -> require ti-hw-bringup-image.bb -> include recipes-images/angstrom/systemd-image.bb -> inherit image Oct 11 13:12:37 easy :) Oct 11 13:13:15 I didn't realise it brought in ti-hw-bringup-image.. That's gotta be a mistake as that's some low level stuff in there (and it had a strange license last I looked) Oct 11 13:14:00 exosyst: but it looks like g_multi is only in this ti-hw-bringup-image, it is not in systemd-image Oct 11 13:46:39 exosyst: btw, dgbserver is installed by cloud9-image.bb Oct 11 13:47:04 Anguel, I saw cheers - I generated an opkg for it anyway so we're good :D Oct 11 13:47:21 :) Oct 11 13:47:23 Also, my little Package.gz script has saved me soooo much pain haha Oct 11 13:49:14 I'm sure there's a script (or maybe debians dpkg-scandependencies?) but I wrote opkg-scanpackages as a stop-gap Oct 11 13:54:27 OE General Assembly http://www.openembedded.org/wiki/Edinburgh,_2013 Oct 11 13:56:35 mckoan, thanks :) Oct 11 13:56:50 Crofton|work: yw Oct 11 13:58:33 mckoan: thx, added myself Oct 11 14:10:39 Is the GA due to the ELCE that's on too/ Oct 11 14:17:34 exosyst: that's the cause for the location at least Oct 11 15:01:46 morning Oct 11 15:03:22 gm chouimat Oct 11 20:04:28 how should I decide the latyer priority for a new layer? Oct 11 20:07:03 ? Oct 11 20:07:09 depends on the layer, and what you want it to be able to override Oct 11 20:07:29 lets say low priority Oct 11 20:07:40 the point is add recipes, not override stuff Oct 11 20:07:52 look at hte priorities of other layers, and decide appropriately based on that Oct 11 20:07:55 thats what i usually do Oct 11 20:08:30 I assume the lower the number, the lower the priority Oct 11 20:08:38 yep Oct 11 20:10:54 kergoth: hi Oct 11 23:43:49 * kergoth fires up a reviewboard vm appliance to play with it Oct 12 02:58:54 so what happened to SITEINFO_BITS ? all of a sudden i'm getting a non-integer argument error from a recipe with an "expr ..." in it... **** ENDING LOGGING AT Sat Oct 12 02:59:59 2013