**** BEGIN LOGGING AT Fri Aug 09 02:59:59 2013 Aug 09 04:19:07 are there solib install commands in classic? Aug 09 04:20:06 nm, found it... Aug 09 07:48:01 morning #oe! Aug 09 07:52:27 crap, is it morning already? Aug 09 07:53:01 morning all Aug 09 07:53:06 nerdboy: for some yes :) Aug 09 07:54:45 but... it's not supposed to be tomorrow yet.. Aug 09 07:58:04 nerdboy: no problem. Close the blinds on your windows, turn off the lights and focus on your computer screen Aug 09 07:58:23 ah, use date to set the time back Aug 09 08:00:07 i'd have to kill ntp... Aug 09 08:02:12 nerdboy: small price to pay to live in the past Aug 09 08:02:27 nerdboy: think about how long this never ending hacking night could be Aug 09 08:04:27 i think i've had enough for one day... i pushed the new stuff before the nightly build, so time for some lack of consciousness... Aug 09 09:42:05 Hello. Probably a trivial question, but i still can't find an answer for quite some time yet: how do I force to rebuild EVERYTHING from sources? bitbake -c clean does not force to clean/rebuild dependancies. Aug 09 09:44:58 kbart: rm -rf tmp :-) Aug 09 09:45:08 oh, and sstate-cache Aug 09 09:46:31 I guess there is a more proper way too, but I don know it Aug 09 09:48:44 that's pretty much the only way Aug 09 09:48:55 if you want to rebuild everything Aug 09 09:49:23 usually that is not needed however... Aug 09 09:52:32 Thank you both. I believe, in my case it is needed as I have changed kernel but I still get old kernel modules installed in my filesystem. Or is there other way around? Aug 09 09:53:39 bitbake -c cleansstate virtual/kernel ought to be enough Aug 09 09:59:01 hm, some cryptic error message after I have issued the suggested command: Aug 09 09:59:06 ERROR: OE-core's config sanity checker detected a potential misconfiguration. Either fix the cause of this error or at your own risk disable the checker (see sanity.conf). Following is the list of potential problems / advisories: Error, the PACKAGE_ARCHS variable does not contain TUNE_PKGARCH (armv5).Toolchain tunings invalid: Tuning 'cortexa8thf-neon' has no defined features, and cannot be used. ERROR: Execution of event h Aug 09 09:59:21 any ideas what could cause that? Aug 09 10:07:51 kbart: is PACKAGE_ARCHS being set somewhere in your configuration/layers you are using? Aug 09 10:12:19 I'm using example image recipe from meta-ti layer, just made some changes to use kernel from my repository instead of default one. Before that image was built fine, so I guess it should be defined somewhere.. Aug 09 10:14:02 hmm, sounds like something has gone a bit awry there then Aug 09 10:14:42 ah, nevermind, my mistake here Aug 09 10:15:02 oh, what was the issue? Aug 09 10:15:31 just copied and paste "bitbake -c cleansstate virtual/kernel" without adding "MACHINE=" at the beginning.. Aug 09 10:16:30 too much OE today, getting dizzy already..;x Aug 09 10:16:57 ah, ok Aug 09 10:17:24 you may find it easier to just set MACHINE in your conf/local.conf to save having to specify it every time :) Aug 09 10:20:01 In fact, I use two machines simultaneously - one as an example and another is modified (one I'm currently working on). It's easier to debug when you can compare to something that works.. Anyway, thanks for the suggestion, I'll keep that in my next time. Aug 09 10:20:45 *mind Aug 09 10:24:00 ah ok, fair enough Aug 09 12:21:34 Hi. Has anyone seen an error like Aug 09 12:21:35 /home/yocto/src/fsl-community-bsp/build/tmp/sysroots/i686-linux/usr/share/quilt/scripts/patchfns: line 943: [: : integer expression expected Aug 09 12:21:35 The working tree was created by an older version of quilt. Please run 'quilt upgrade'. Aug 09 12:21:38 lately? Aug 09 12:24:29 That happens when building qt 4.8.5-r0 Aug 09 12:31:20 Hmmmm. The first lines of /home/yocto/src/fsl-community-bsp/build/tmp/sysroots/i686-linux/usr/share/quilt/scripts/patchfns are Aug 09 12:31:21 # This file contains the common functions used in all quilt script Aug 09 12:31:21 # It is meant to be sourced by bash scripts. Aug 09 12:31:36 I wonder if it is really being sourced by _bash_. Aug 09 12:31:43 I bet that is the problem. Aug 09 12:37:07 And the system has Aug 09 12:37:07 $ ls -l /bin/sh Aug 09 12:37:07 lrwxrwxrwx 1 root root 4 Mar 1 2012 /bin/sh -> dash Aug 09 12:40:13 mario-goulart: if this is the case, we need to fix it. Please try with bash in system and confirm if it fixes the issue. Aug 09 12:40:22 alright Aug 09 12:40:39 mario-goulart: but if it does, the right fix is to patch this script to make it 'dash-compatible' Aug 09 12:41:35 ok Aug 09 12:41:53 mario-goulart: what's the content of .version in the patches/ directory? Aug 09 12:42:35 or add a 'set -x' in version_check() to get some more information Aug 09 12:44:00 same error with /bin/sh -> /bin/bash Aug 09 12:44:11 ensc|w: I'll try to find that out. Aug 09 12:44:21 mario-goulart: not patches/, but .pc/ Aug 09 12:45:48 ok Aug 09 12:46:40 ensc|w: and check if host has quilt installed or not Aug 09 12:46:51 oops Aug 09 12:46:54 mario-goulart: ^ Aug 09 12:48:21 otavio: it doesn't seem to be installed. "dpkg --get-selections | grep quilt" doesn't show anything and there is no quilt in $PATH Aug 09 12:48:33 mario-goulart: great Aug 09 12:49:13 mario-goulart: in theory you'd be using 0.6 Aug 09 12:49:20 mario-goulart: in theory you'd be using 0.60 in Poky Aug 09 12:52:21 ensc|w: pc/.version is empty. Aug 09 13:08:04 mario-goulart: some out-of-diskspace problem? or crash while patching? Aug 09 13:11:09 ensc|w: no. I still suspect the shell being dash is the problem. Now I have /bin/sh as a link to /bin/bash and I cleaned sstate for qt. Aug 09 13:11:26 I didn't get the qt error, but a different for for pulseaudio. :-/ Aug 09 13:12:46 OTOH, I just didn't get an error for qt because bitbake didn't attempt to build it. Aug 09 13:43:46 oh, time to prepare hambedded distribution layet Aug 09 13:43:48 layer* Aug 09 17:52:34 yo Aug 09 18:19:02 kergoth: is --as-needed buried somewhere in the oe build settings? i suppose it could be in the gcc specs... Aug 09 18:23:18 not sure actually, i'm a bit out of date on the internal toolchain handling Aug 09 18:24:39 who might might be more current on that topic? Aug 09 18:25:46 khem may be your best bet Aug 09 19:08:15 khem ping Aug 09 19:17:44 * mr_science dislikes workarounds that mask underlying problems Aug 09 19:18:09 s/dislikes/hates-with-a-passion/ Aug 09 19:19:52 heh Aug 09 19:21:08 i might need to get out my cluebat and smash something... Aug 09 19:22:48 now that i look, i see LDFLAGS_append getting abused for the same issue Aug 09 22:20:12 well, at least it was a good lunch day... Aug 10 01:41:00 yo **** ENDING LOGGING AT Sat Aug 10 02:59:58 2013