**** BEGIN LOGGING AT Mon Aug 10 02:59:59 2015 Aug 10 08:39:48 morning all Aug 10 08:47:35 bluelightning: hello Aug 10 08:53:07 hi Noor Aug 10 08:55:16 * Noor is circling arounf RP's new code for Pixbufcache :) Aug 10 09:21:33 aaahhh now I get it what is going on there Aug 10 09:21:41 I should have picked it early Aug 10 09:39:50 is it possible to use shell variables in a bb recipe i.e. Aug 10 09:40:12 do_install() { foo=bar; install ${foo} } Aug 10 09:40:32 as it seems to just print the literal, without expanding Aug 10 09:40:34 jackmitchell: sure, just use $foo not ${foo} Aug 10 09:41:54 yeah, that seems to do the same; it still uses the literal Aug 10 09:42:03 the error log shows Aug 10 09:42:05 install $foo Aug 10 09:46:49 jackmitchell: that's odd, because I just tried foo="hello"; echo $foo and I get "hello"... Aug 10 09:48:04 bluelightning: yeah, I think I've just realised what's happening Aug 10 09:48:46 bluelightning: the function was failing, and I was reading the output of the WARNING, which I was thinking was post-interpretation Aug 10 09:49:01 jackmitchell: oh, right... Aug 10 09:49:15 bluelightning: but thats the line of the bb file that was failing, not the line of the bash that failing Aug 10 09:50:41 example: WARNING: /home/jack/Work/build/openembedded/build/tmp-glibc/work/cortexa9hf-vfp-neon-oe-linux-gnueabi/firmware/gitAUTOINC+0912018eab-r0/temp/run.do_install.13687:1 exit 1 from Aug 10 09:50:41 install -D /home/jack/Work/build/openembedded/build/tmp-glibc/work/cortexa9hf-vfp-neon-oe-linux-gnueabi/firmware/gitAUTOINC+0912018eab-r0/git/file $firmwarelocation Aug 10 09:57:35 hi all, i'm trying to build a 32b in a 64b target and i'm checking the multlib support in yocto. besides https://wiki.yoctoproject.org/wiki/Multilib is there another place where to check for the actual status/example? also looking lib32-bash in OE. TIA Aug 10 10:04:48 sm0ketst: multilib works fine Aug 10 10:05:15 rburton: thanks for the status :) Aug 10 10:06:32 rburton: i'm trying to follow the wiki but I don't know if i need to do something more. I've also hit the MULTILIB_IMAGE_INSTALL variable googling and I'm almost lost .... still searching for example recipes Aug 10 10:09:11 I suspect the docs are more up to date https://www.yoctoproject.org/docs/1.8/dev-manual/dev-manual.html#combining-multiple-versions-library-files-into-one-image Aug 10 10:09:44 joshuagl: thanks for the link Aug 10 10:10:42 Using Dizzy here. Is it better to switch to 1.8 ?? may I stay with Dizzy? Aug 10 10:12:51 https://www.yoctoproject.org/docs/1.6/dev-manual/dev-manual.html#combining-multiple-versions-library-files-into-one-image ? :-) Aug 10 21:29:39 yocto is consistently failing to fetch a kernel repo properly. it claims a revision isnt in there, even though it is in there. Aug 10 21:30:01 does the branch= in the url match the branch that contains the SRCREV? Aug 10 21:30:04 now I fetched it manually, and added a .done file. is there anything else I should be aware of? Aug 10 21:30:04 if not, that'll fail Aug 10 21:30:22 yes, it matches Aug 10 21:30:48 what *exactly* is the error message? Aug 10 21:30:53 summary doesn't help in these situations Aug 10 21:32:50 I'll try to get more log output Aug 10 21:37:15 "WARNING: Failed to fetch URL git://github.com/beagleboard/linux.git;protocol=git;branch=3.14, attempting MIRRORS if available" and "DEBUG: Fetcher failure: Unable to find revision 14d84ae66d7c4c01ecbd3ab3165e371d7dd728e2 in branch 3.14 even from upstream" Aug 10 21:39:02 then it's not accessible from that branch, as i said before Aug 10 21:39:38 run git merge-base 14d84ae66d7c4c01ecbd3ab3165e371d7dd728e2 origin/3.14 in that git repo Aug 10 21:40:00 or git branch --all --contains 14d84ae66d7c4c01ecbd3ab3165e371d7dd728e2 Aug 10 21:40:52 I checked out that repo, and it does contain both the branch and the commit Aug 10 21:41:05 I'll try your commands Aug 10 21:42:18 no, i didn't say it doesn't contain the branch or th ecommit, i said that commit isn't included in the branch. not accessible from it. Aug 10 21:43:14 oh. okay. Aug 10 21:43:31 e.g. the commit was made on 4.0 not 3.14, or something. wrong branch for that commit Aug 10 21:43:58 generally either branch= or SRCREV will have to be changed to make sure they line up Aug 10 21:47:32 Looks like no branches contain that commit, only a tag. You might be able to set tag= in the uri, or branch=3.14.48-ti-r72 Aug 10 21:48:11 or change srcrev to something else Aug 10 21:49:13 yeah, weird. I wonder how these guys built this. they did use fido. Aug 10 21:49:46 either they used an old version taht didn't check this, or they could have rebased a branch so it no longer contains it when it used to Aug 10 21:49:48 afaict **** ENDING LOGGING AT Tue Aug 11 02:59:58 2015