**** BEGIN LOGGING AT Fri Jun 19 02:59:59 2015 Jun 19 06:15:43 Hi guys, this is Dieter (Kiermaier) and I am new at Arrow working as FAE for HIgh-End MPUs and System on Modules. Would be great to get in contact with the Yocto community! Jun 19 06:30:03 * LetoThe2nd hands dieter__ a cup of coffee, points over to the community couch "please have a seat while the rest of us is still waking up" Jun 19 06:32:17 Thank you Sir - will enjoy my coffee:) Jun 19 06:58:40 <_taw_> who calls check_defconfig in this file: http://git.yoctoproject.org/cgit/cgit.cgi/yocto-kernel-tools/tree/tools/createme Jun 19 06:59:19 <_taw_> I experiment whith this option: KBUILD_DEFCONFIG_KMACHINE ?= defconfig_file and it doesn't seem to work Jun 19 06:59:34 <_taw_> http://www.yoctoproject.org/docs/1.8/kernel-dev/kernel-dev.html#using-an-in-tree-defconfig-file Jun 19 09:12:21 hey Jun 19 09:15:26 <_4urele_> lakaye, ho Jun 19 09:37:53 i'm trying to get spdx working but it is not generated, i updated the SPDX_MANIFEST_DIR variable but it doesn't help, am i missing something ? Jun 19 09:50:22 when i run : bitbake target -c spdx it says "Task do_spdx does not exist for target target" Jun 19 09:50:28 help anyone ? :) Jun 19 09:56:16 <_4urele_> lakaye, if you want to build package spdx just do "bitbake spdx" Jun 19 09:57:32 <_4urele_> lakaye, if you want to execute a specific task for a package do : "bitbake spdx -c " (eg "bitbake busybox -c menuconfig") Jun 19 09:57:57 <_4urele_> lakaye, doest it fit your needs? Jun 19 09:58:15 actually, there is nothing that provides spdx on my side Jun 19 09:58:21 i'm probably missing something Jun 19 09:58:30 <_4urele_> so what is spdx? Jun 19 09:58:47 https://spdx.org/tools/community/yoctospdx Jun 19 09:59:25 <_4urele_> lakaye, sry I didn't knew this ;) Jun 19 10:03:54 <_4urele_> lakaye, you can maybe use "-c listtasks" Jun 19 10:05:16 is anything actually including the SPDX class in the build? you probably want something like INHERIT += "buildhistory" Jun 19 10:05:23 only, y'know, with SPDX Jun 19 10:06:34 OK, I added INHERIT += "spdx" to my local.conf and now have a bunch of failing do_spdx tasks Jun 19 10:07:22 lakaye: you will need to make sure some of the variables in meta/conf/licenses.conf are set appropriately Jun 19 10:07:48 i.e. point at a fossology instance, setting SPDX_MANIFEST_DIR, etc Jun 19 10:07:57 I guess it is. Jun 19 10:08:12 I already update the licenses.conf file Jun 19 10:08:27 and added the INHERIT ? Jun 19 10:09:54 I'll try, thanks Jun 19 10:10:18 sounds better, i've a lot of running tasks showing do_spdx Jun 19 10:10:26 thank you joshuagl :) Jun 19 10:13:28 np Jun 19 10:46:21 how can I use the buildhistory to see which files the SDK is supposed to generate? Jun 19 10:46:40 I do not get this "files-in-sdk.txt" Jun 19 10:46:58 and the generated SDK does not appear to be to the one that I want. Jun 19 11:19:40 joshuagl, RP: I'll be taking the autobuilder down in about 2 hours today. It'll be down for about 90 minutes. Jun 19 11:26:16 thanks for the notice pidge Jun 19 11:26:32 * joshuagl still has to figure out what's up with util-linux anyway Jun 19 11:42:32 bitbake -e -c populate_sdk myimage | grep ^BUILDHISTORY -> shows this: BUILDHISTORY_FEATURES="image package sdk" Jun 19 11:42:38 yet I cannot seem to find files-in-sdk.txt anywhere! Jun 19 11:42:42 I am using 1.6.3. Jun 19 12:55:53 joshuagl: broke fido-next again eh :/ Jun 19 12:56:05 annoying how the ab breaks stuff that works locally isn't it Jun 19 13:21:39 rburton: yeah, it's reproducing on my laptop though so that's something Jun 19 13:21:50 no idea what the fix is yet, but progress Jun 19 16:44:04 hello i had some questions about disabling sysvinit processes. i am overriding another recipe by using a .bbappend. is the intended way to disable this particular init to set INITSCRIPT_PARAMS=" " in the overriding recipe Jun 19 16:47:30 also, another init question -- are there tips on understanding what happened during the build steps creating the init? i am trying to add an init for a dhcp-server (in dhcp package), but i see the init script does not have the header (provides/start/stop/etc) filled out correctly. in this case, i took the advice of this thread: http://comments.gmane.org/gmane.linux.embedded.yocto.general/10599 Jun 19 18:35:18 advice appreciated! Jun 19 19:35:08 crap, thought this was reproducible, so did a clean/rebuild, and now it won't happen again, and i destroyed the environment where it did Jun 19 19:35:08 oops Jun 19 19:38:24 is there any way to disable the checksums requirement for single packages? I am setting up a development layer/target and I have to get the code from a ftp server, not a svn/git/hg. If I have to update the checksums every time I want to build something quickly, I will go insane. Jun 19 19:44:15 heller_barde: I need to look it up exactly, but IIRC, using the 'closed' license also does that Jun 19 19:44:18 (ymmv, etc) Jun 19 19:46:10 rink_: if only. our software is already under the "closed" license. Sigh. If only we had sensible tooling... My kingdom for a git. Jun 19 19:46:19 but thanks for the pointer. Jun 19 19:47:12 CLOSED only affects license checksumming, not source checksumming from the fetch process Jun 19 19:47:25 kergoth: good to know. Jun 19 19:47:44 hmm I suppose that makes sense Jun 19 19:47:59 * rink_ thinks hard... there was a pretty useful way to doit Jun 19 19:48:10 we use it at work to drag internal repositories into the Yocto tree Jun 19 19:48:15 I'm near to writing some ridiculous python glue to automatically check the thing from the FTP into a local git and then pointing bitbake at that. ^^ Jun 19 19:48:26 but I'd rather not do that. Jun 19 19:48:35 heller_barde: ping me on monday, I can take a look at it Jun 19 19:49:04 so I've tried BB_STRICT_CHECKSUM and OE_UNSAFE_DOWNLOADS or something like that. Jun 19 19:49:13 rink_: alright. I'll try. Jun 19 19:49:48 when is monday _day_ for you? because I'm in UTC+1 Jun 19 19:50:01 CEST here, so that will match Jun 19 19:50:06 yeah, try setting BB_STRICT_CHECKSUM = "0" and also removing any md5sum/sha256sum flag from the recipe Jun 19 19:50:28 if there are checksums set, it'll still error about the mismatch even with strict checksums off, however Jun 19 19:50:33 kergoth I tried BB_STRICT_CHECKSUM="" for now because the documentation was talking about "undefining" it. Jun 19 19:50:35 afaict from a quick read of the appropriate code in bitbake Jun 19 19:50:43 it checks explicitly for a value of '1' Jun 19 19:50:46 so anything but 1 should work Jun 19 19:50:58 see line 580 of bitbake/lib/bb/fetch2/__init__.py Jun 19 19:51:22 hmm Jun 19 19:52:21 oddly, there's a member of the urldata (FetchData class) called ignore_checksums, which is obeyed, but nothing anywhere sets it to true Jun 19 19:55:15 I'm using version 1.7.1 and it didn't work with putting BB_STRICT_CHECKSUM="0" into the targets local.conf. I removed all SRC_URI[md5/sha256sum] entries from the recipe. Jun 19 19:55:24 thanks for all the effort though :D Jun 19 19:59:08 sorry, I meant bitbake version 1.24.0 Jun 19 20:18:17 kergoth: where can I set this BB_STRICT_CHECKSUM so it is actually set when it is checked? Because I dug into the code and it turns out that no matter where I set it to 0 so far, it is always 1 when it is checked. Jun 19 20:23:17 heller_barde: you'd want to set it in the recipe. if you set it in local.conf, it'll be overwritten by the set to 1 in default-distro.conf, which is included by defaultsetup.conf, which is included after local.conf Jun 19 20:27:32 kergoth: oooooh, in the recipe. I didn't really consider that. Thanks so much! Jun 19 20:27:35 I'll try right away Jun 19 20:27:56 (it's 10:30 pm now and I'm still at work. ack...) Jun 19 21:16:42 kergoth: thanks so much. it works now. I am relieved and going home now. good bye :) Jun 19 21:16:48 np **** ENDING LOGGING AT Sat Jun 20 02:59:58 2015