**** BEGIN LOGGING AT Mon Jan 27 03:00:00 2014 Jan 27 09:42:46 morning all Jan 27 09:44:53 hey bluelightning Jan 27 09:45:00 hi koen Jan 27 11:45:25 morning all Jan 27 11:54:25 hi pb_ Jan 27 11:55:04 hi bluelightning Jan 27 14:00:10 has there been a change in OE warnings recently? Jan 27 14:00:16 all of a sudden I get this: "WARNING: gstreamer1.0-plugins-good-video4linux2-1.2.2 was registered as shlib provider for libgstvideo4linux2.so, changing it to gst-plugins-good-video4linux2-0.10.31 because it was built later" Jan 27 14:00:36 I was building images with both 1.0 and 0.10 plugins installed, and never got this warning Jan 27 14:02:37 dv_: yes, see package.bbclass commits Jan 27 14:02:48 ah Jan 27 14:03:46 hmm perhaps the gstreamer problem can be fixed by inserting -0.10 and -1.0 to the plugin filenames right before ".so" Jan 27 14:14:52 or marking them as private if you can confirm that nobody is supposed to link directly with plugins Jan 27 14:19:32 JaMa: it's not a plugin Jan 27 14:19:54 the gst-plugin-good/bad/ugly provide both shlibs and plugins Jan 27 14:20:28 e.g gst-plugin-v4l and gst-plugin-camera can link to libgstv4l to reuse bits Jan 27 16:56:53 is there a simple way to disable do_patch if I point S_pn to a local checkout Jan 27 16:58:15 heeen: I'd recommend using externalsrc.bbclass for that instead of setting S Jan 27 16:58:24 heeen: http://www.yoctoproject.org/docs/current/dev-manual/dev-manual.html#building-software-from-an-external-source Jan 27 16:59:34 the inherit line does not go into the recipe? Jan 27 17:04:40 heeen: depends which version you are using Jan 27 17:04:58 BB_VERSION = "1.18.0" Jan 27 17:05:30 heeen: right, so dylan/1.4 branch I guess - that means yes it needs to Jan 27 17:05:51 do_rootfs is painfully slow Jan 27 17:06:19 heeen: this is the 1.4 version: http://www.yoctoproject.org/docs/1.4/dev-manual/dev-manual.html#building-software-from-an-external-source Jan 27 17:10:19 Crofton|work: stop creating 4GB rootfs? Jan 27 17:11:39 :) Jan 27 17:12:39 Crofton|work: or replace that r/pi you try to build on with normal device ;d Jan 27 17:44:16 re: tune features, does armv7a make sense with callconvention-hard? Jan 27 17:44:39 as opposed to something like armv7a(t)hf Jan 27 17:53:58 Crofton|work: make sure you have pigz and/or pbzip/lbzip :) Jan 27 17:56:44 so compressing image is taking time then? Jan 27 17:56:54 thin I use gzip atm Jan 27 17:57:29 depends on your IMAGE_FSTYPES Jan 27 17:57:39 yeah Jan 27 17:57:46 but last time I've measured it it was spending a lot of time doing tar.bz2 and tar.gz Jan 27 18:06:33 denix needs to get the hamsters in the arago git server some speed. Jan 27 18:07:22 yeah, pigz certainly is a win for that stuff. we use it globally now. Jan 27 18:07:29 also "export GZIP=-1" Jan 27 18:37:48 so export GZIP=-1 will stop bitbake from using it and it will use pigz instead? Jan 27 18:40:12 * koen just replace /bin/gzip with pigz Jan 27 18:40:15 replaced* Jan 27 18:44:26 same here Jan 27 18:57:17 Does bitbake do some variable substitution in shell functions before executing them? I have a BOOT_SPACE?=8192 line in a bbclass. I then have two lines like these inside a shell function in the class: "BOOT_SPACE=123456 \n SPACE_NEEDED=$(expr ${BOOT_SPACE} + 4096)" The value of 8192 is used in the last line rather than 123456. Jan 27 18:57:43 sr105: anything in ${} is considered a bitbake variable Jan 27 18:57:53 ah... Jan 27 18:58:11 I have to unlearn muscle memory when doing shell methods in recipes Jan 27 18:58:18 didn't realize that since ${} is valid shell syntax, too Jan 27 18:58:47 exactly Jan 27 18:58:56 thx Jan 27 18:59:00 it makes it really easy to use bitbake vars Jan 27 18:59:13 but like I said, need to unlearn muscle memory :) Jan 27 19:00:58 I removed the braces and then changed the name of the variable to prevent accidents. Jan 27 20:21:45 why the heck would busybox's "make menuconfig" return 2 on failure when run from a devshell, but 0 on _failure_ when run via "bitbake -c menuconfig" ? Jan 27 20:30:09 oh Jan 27 20:30:53 should "$" be escaped in embedded python code in .bbclass files ? Jan 27 22:38:58 therre is some weirdness in the boost build Jan 27 22:39:06 need to figure out how to describe it :) **** ENDING LOGGING AT Tue Jan 28 02:59:59 2014