**** BEGIN LOGGING AT Wed Dec 23 02:59:58 2015 Dec 23 06:10:25 Hi I have made a yocto layer now the source code which the recipes in that layer builds already has the license Dec 23 06:10:41 I am wondering if adding sperate LICENSE file in the layer is really required? Dec 23 10:00:23 good morning Dec 23 15:12:49 Happy holidays, can you use a do_ script to not install telnet in busybox? Or would I be better off just removing telnet binary from rootfs/usr/bin/telnet and rootfs/usr/lib/opkg/alternatives/telnet with a ROOTFS_POSTPROCESS_COMMAND Dec 23 16:13:37 Has anyone customized busybox before in yocto? Dec 23 16:14:29 no, never in the past 11 years has anyone ever thought about customizing busybox before Dec 23 16:15:25 Kergoth - maybe the best way would be to just have my ROOTFS_POSTPROCESS_COMMAND remove telnet! 11 years tells me something!!! Dec 23 16:15:36 That was sarcasm Dec 23 16:15:43 changing teh busybox configuration is trivial Dec 23 16:15:49 kergoth: hold it together, it's almost christmas Dec 23 16:15:56 rburton: true Dec 23 16:16:08 mrk377: write a config fragment to disable telnet, trivials Dec 23 16:16:26 bitbake -c menuconfig busybox; bitbake -c diffconfig busybox; is a nice way to generate the fragment Dec 23 16:16:30 <3 that Dec 23 16:16:43 would be nice to make a recipetool command to do that and then shove the generated fragment into a layer Dec 23 16:16:50 been thinking about making that one of these days Dec 23 16:16:55 kergoth - Thanks for answer with little to no sarcasm!! lol Dec 23 16:19:02 Kergoth - I'm rolling an image that will be scanned according to Security Technical Instruction Guide (STIG) items, and they don't want telnet. These are Information Assurance things that certain customers are concerned about. Dec 23 16:19:58 Thanks again! Dec 23 16:21:01 yocto-layer create yourlayer; bitbake-layers add-layer meta-yourlayer; echo '# CONFIG_TELNET is not set' >no_telnet.cfg; recipetool appendsrcfile -W meta-yourlayer busybox no_telnet.cfg Dec 23 16:21:32 Thanks for the XMAS present! Dec 23 16:22:11 maybe i should make the meta-mentor kernel_set_defconfig, kernel_add_fragments, and kernel_set_configs recipetool subcommands more generic Dec 23 16:22:12 heh Dec 23 16:23:15 Yeah, maybe throw in two consecutive examples of different packages (kernel and busybox) Dec 23 16:33:21 Kergoth - this was very helpful. Thank you for the time mate! Dec 23 16:33:26 no problem Dec 23 16:38:25 Hmm, wonder if exec_build_env_command from devtool should go into scriptutils Dec 23 16:40:31 find . -name exec_build_env_command # This returns nothing... Is that the filename? Dec 23 17:08:42 * rburton goes back to testing his death-to-la-files class Dec 23 17:12:31 rburton: nice. are you killing them for the packages, or also from the sysroots? Dec 23 17:12:51 i have https://github.com/MentorEmbedded/meta-mentor/blob/master/meta-mel/classes/libtool-garbage.bbclass which is based on a class from pb, but that's packages only Dec 23 17:18:36 kergoth: EVERYWHERE Dec 23 17:18:40 yay Dec 23 17:19:15 my class is a do_install[postunc] Dec 23 17:19:27 erm, postfunc. Dec 23 17:20:25 hopefully none of the builds are relying on .la dependency_libraries, but should be easy enough to fix with pkg-config usage in those cases anyway Dec 23 17:20:37 soon find out Dec 23 17:20:53 oh look you cursed it Dec 23 17:21:01 openjade-native just broke Dec 23 17:21:19 | make[2]: *** No rule to make target '/data/poky-master/tmp/sysroots/x86_64-linux/usr/lib/libosp.la', needed by 'openjade'. Stop. Dec 23 17:21:20 hm Dec 23 17:21:27 i think that needs a re-clean Dec 23 17:23:46 meh, I still need to try getting the shallow git support submitted for review again, before we get too far along in the 2.1 cycle Dec 23 17:24:15 * rburton is getting annoyed at meta-fsl-* Dec 23 17:24:30 it now fails to build: openssl, valgrind, u-boot-qoriq Dec 23 17:28:39 oh, right, i was going to add BB_GIT_SHALLOW_EXTRA_REFS in case folks have extra refs they want to keep above and beyond the ones the SRCREVs are in, e.g. for gitpkgv Dec 23 17:30:52 * rburton shakes fist at openjade-native Dec 23 17:31:39 right that's it, time to sit down on the sofa and watch a bit more Fringe. sure fire way to get the kids to arrive home within seconds! Dec 23 17:31:55 hehe Dec 23 17:39:08 joshuagl: you will stay here, will you not? :) Dec 23 17:42:03 * nerdboy trying if a neo build is possible Dec 23 17:42:30 rburton: hopefully i will hit those meta-fsl issues... Dec 23 18:19:37 yay, i can "leverage" a fork with neo stuff Dec 23 20:26:16 IOError: [Errno 2] No such file or directory: '/proc/24177/io' <= wtf? Dec 23 20:28:35 possibly related to the recent buildstats changes? *shrug* Dec 23 20:47:31 can i turn it off? Dec 23 20:50:01 all i get is generic traceback and a bunch of errors... Dec 23 21:23:24 i thought you had to opt in to buildstats Dec 23 21:23:39 (local.conf) Dec 23 22:53:14 rburton: apparently not in fsl default local.conf... Dec 23 22:54:46 actually it's in every default USER_CLASSES i have Dec 23 22:55:03 so i'm not sure what "opt in" means Dec 23 23:02:54 even if i set the base dir it still barfs on the wrong /proc/xxxxx/io values Dec 23 23:04:36 seems to work as expected if i remove buildstats from USER_CLASSES completely Dec 23 23:04:44 never had to do that before... Dec 23 23:08:11 https://bpaste.net/raw/d53fe42969fb Dec 23 23:17:10 aha! Dec 23 23:18:02 missing a kernel option on the build host Dec 23 23:22:50 just on the machine i'm using too Dec 23 23:39:38 * nerdboy rebuilding kernel with TASK_IO accounting stuff enabled Dec 24 00:28:37 is that dependency documented anywhere? Dec 24 00:29:44 my guess is it was unintentional Dec 24 00:33:11 it worked fine not long ago on that machine Dec 24 00:33:29 other machines already had it enabled in the kernel Dec 24 00:41:10 see '[OE-core] [PATCH 5/7] buildstats: Improve to add getrusage data and corrected IO stats' Dec 24 01:25:02 "corrected" being the keyword... Dec 24 01:25:50 building again with updated kernel and default config, so that was the culprit Dec 24 01:26:59 MACHINE = "udooneo" <= need to add meta-uav next if this works... **** ENDING LOGGING AT Thu Dec 24 02:59:59 2015