**** BEGIN LOGGING AT Fri Oct 11 02:59:58 2013 Oct 11 06:44:05 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 07:42:55 good morning Oct 11 07:52:48 How can I disable -Werror flag in the linux-yocto-custom.bb recipe? Oct 11 08:00:50 hi guys, any idea how to populate the /boot directory in the image? I have uImage but still my dtb is missing Oct 11 08:03:59 n01, generating a BeagleBone Black image with angstrom has this Oct 11 08:04:03 n01, https://github.com/beagleboard/meta-beagleboard/tree/master/common-bsp/recipes-kernel/linux Oct 11 08:04:31 n01, I use the linux.inc in my own layer to do the same trick Oct 11 08:05:34 n01, I believe yocto has a similar method using meta/recipes-kernel/linux/linux-dtb.inc but I haven't tried/used it Oct 11 08:06:31 RagBal: honestly I wanted to avoid messing up with new inc files :) I thought yocto was able to do that Oct 11 08:06:44 but thanks for the pointer Oct 11 08:07:08 If it does, I'm interested as well =) Oct 11 08:38:33 RagBal: it seems that it is enough to add kernel-devicetree to IMAGE_INSTALL Oct 11 08:38:47 simple as that Oct 11 08:44:58 n01, hmm that would mean my linux.inc is pretty useless Oct 11 08:45:07 Since I only use it for that Oct 11 08:45:15 Will try it later, thnx! Oct 11 08:47:49 :) yw Oct 11 08:53:15 morning all Oct 11 09:59:09 Yocto hands-on labs at ELCE will be on VNC, amazing! Oct 11 09:59:44 just courious to see how good it will work ;-) Oct 11 10:14:43 hi i have used poky compiler.and used at91-3.13-soc kernel for at91sam9x5ek board while compiling the kernel it stops athttp://pastebin.com/vucFsZiy can you tell me what is that issues Oct 11 10:30:03 linu: is not clear what exactly you are doing Oct 11 10:34:21 mckoan, ya, i have been using kernel 3.13 version for at91sam9g35ek board and i have using yocto to cross compile the kernel,i have used default config which is given in kernel 3.13 for at91sam9g35,while compiling the kernel it stops and shows this error http://pastebin.com/vucFsZiy Oct 11 10:40:42 mckoan, arm-poky-linux-gnueabi-ld: no machine record defined make: *** [vmlinux] Error 1 i got the above error while compiling kernel 3.13, do you have any idea for this error Oct 11 10:56:52 rburton: hi ! trying to build core-image-weston with weston 1.3.0, I have the following error : Error: weston-examples not found in the base feeds (chiefriver x86_64 noarch any all), do you know what is weston-examples and if it is really needed ? Oct 11 11:11:54 elbc: its not needed, and presumably your new weston recipe doesn't install the examples. Oct 11 11:12:00 you can remove it from core-image-weston Oct 11 11:12:04 as the name says, its just exampes Oct 11 11:15:59 rburton: ok thanks, that's what I've done, I have successfully built my image but after booting, it is needed to start manually weston-lauch or /etc.init.d/weston script because systemd is not able to start the service with 1.3.0 version (it says that weston must be run from a virtual terminal), do you know how to proceed ? thanks again Oct 11 11:17:18 elbc: they've probably changed something in 1.3 Oct 11 11:17:39 elbc: probably best to do it Properly if you using systemd and create a user systemd session Oct 11 11:18:21 rburton: ok thanks I'll take a look at it Oct 11 11:19:20 elbc: git.yoctoproject.org has a poky-contrib repo, have a look at the ross/wayland branch Oct 11 11:19:27 that's very much work in progress! Oct 11 11:23:02 rburton: ok ! Oct 11 12:32:10 RP, I have a question about some code in oe-core/meta/recipies-core/initscripts/initscripts-1.0/checkroot.sh that you checked in 8 years ago.... Oct 11 12:32:55 RP: quick, hide! Oct 11 12:34:52 is there an option to automatically remove tmp/work folders, packages, etc. of older built recipes after building newer version? Oct 11 12:35:10 Net147: there is a file in scripts if I remember correctly Oct 11 12:35:38 Net147: cleanup-workdir looks promising Oct 11 12:35:48 The line ``exec 0>&9 9>&-'' has the unindented consequence of causing stdout to be closed resulting in all the echo commands in the scripts being noops Oct 11 12:35:58 jackmitchell: that looks to be it Oct 11 12:36:19 I have 7 copies of kernel work directory and each is 4.9 GB... Oct 11 12:36:55 Net147: there's a inherit rm_work Oct 11 12:37:06 RP, trying to figure out if redirecting stdin is necessary or if it should be jsut ``exec 9<&-'' Oct 11 12:37:19 Net147: or there's a rm_old_work in meta-oe i think, which wipes out old work directories Oct 11 12:37:31 jchonig: could you point me at the change I made? :) Oct 11 12:37:45 ^4b46c1f openembedded/packages/initscripts/initscripts-1.0/checkroot.sh (Richard Purdie 2005-08-31 10:45:47 +0000 55) exec 0>&9 9>&- Oct 11 12:39:30 jchonig: That is the initial population of the poky repo! Oct 11 12:39:45 RP, OK, you are off the hook for that one.... Oct 11 12:40:47 running cleanup-workdir... Oct 11 12:40:48 RP, I guess the owner is the last one to touch the code doesn't apply for initial population... Oct 11 12:41:23 jchonig: http://git.openembedded.org/openembedded/commit/packages/initscripts/initscripts-1.0/checkroot.sh?id=c8e5702127e507e82e6f68a4b8c546803accea9d Oct 11 12:41:30 jchonig: I'm guessing that reference is even older Oct 11 12:42:00 something like rm_old_work in oe-core would be useful I think... Oct 11 12:42:40 jchonig: You're right though, closing stdout isn't intended Oct 11 12:43:16 RP Going to have to get seebs to review my change (exec 9<&-) It seems to work Oct 11 12:43:54 jchonig: You might be able to find further into in the bitkeeper repo... Oct 11 12:43:59 RP: sooner or later I'll also have a question about locomo_kbd_tweak-r2.patch. 2006 stuff :p Oct 11 12:44:23 ant_work: I never did much with locomo ;-) Oct 11 12:44:25 RP The problem I was debugging is read-only-rootfs dosn't work. It's because bootlogd starts and is writing to fs before checkroot.sh runs. Fix is to change checkroot.sh to S06 to run before bootlogd. Will send a patch today or Monday Oct 11 12:45:48 bluelightning: do we have a bk history repo anywhere in git? Oct 11 12:46:44 RP: if you mean before the history in the OE-Classic repo, not that I'm aware of Oct 11 12:51:32 bluelightning: I could have sworn I generated this at some point, must never have published it :( Oct 11 12:54:12 that is an archaeology dig there Oct 11 14:03:25 is there a command to clear out old sstate-cache data too? Oct 11 14:04:08 -c cleansstate Oct 11 14:04:29 preserving the current sstate data? Oct 11 14:07:59 Net147: have a look at sstate-cache-management.sh Oct 11 14:08:19 that script could probably could do with a nicer name Oct 11 14:14:14 talking about sstate-cache-management.sh I think that --stamps-dir param doesn't play well with latest rm_work Oct 11 14:16:03 rm_work removes some stamps for which sstate-archive is still valid, but sstate-cache-management.sh removes it as unused when --stamps-dir is used Oct 11 14:26:52 how would I append to a variable for x86 architecture only? Oct 11 14:28:41 _append_x86, _append_x86-64 it seems Oct 11 14:41:48 ERROR: Multiple .bb files are due to be built which each provide virtual/libgl (.../mesa_9.1.6.bb .../mesa-gl_9.1.6.bb). is that normal? Oct 11 15:30:48 I wish meta-fsl-arm used Freescale instead of the more generic 'Proprietary' license in its proprietary recipes Oct 11 15:48:19 kergoth: have you tried poking the maintainer(s)? Oct 11 15:55:31 not yet, minor annoyance, not a priority yet. its a good idea, though. best to be as specific as possible with licensing, after all Oct 11 15:55:38 * kergoth adds to todo Oct 11 21:02:16 halstead: ping Oct 11 21:02:29 Hi scottrif . Oct 11 21:03:00 halstead: going PM **** ENDING LOGGING AT Sat Oct 12 02:59:59 2013