**** BEGIN LOGGING AT Wed Aug 22 03:00:00 2012 Aug 22 07:15:12 did anyone ever upgrade a glibc based image into a eglibc one ? Aug 22 07:37:50 Hi! .... Is is possible to blacklist a recipe in some way. The case is that meta-angstrom override oe-core's psplash_git.bb. For several reasons I don't want that. I could of cause jsut remove the file but I thing that is a bit too hacky. Aug 22 07:43:48 BBMASK should work Aug 22 07:45:45 mr_nice_: Where sould I write it? Aug 22 07:49:53 local.conf see local.conf example Aug 22 07:52:14 which package provides the config files under /etc/opkg/* ? Aug 22 07:53:32 nv Aug 22 07:53:48 mr_nice_: it seemns to work - thanks :) Aug 22 08:04:52 hi all Aug 22 08:05:07 holidays justy finished :( Aug 22 08:13:25 morning all Aug 22 08:55:17 hi, again! I have an error, while bitbaking monit, I guess, something wrong with system package: how I can reinstall it? Here is the log: http://pastebin.com/XfiUz0bJ Aug 22 09:39:42 bluelightning: I am building a root file system for collie, I will need your help in testing it out Aug 22 09:40:32 bluelightning: and kernel too Aug 22 09:40:39 khem: ok, sure thing Aug 22 09:40:54 I will try it on qemu as well since collie is now good on qemu master Aug 22 09:41:02 but h/w is still h/w Aug 22 09:41:28 indeed Aug 22 09:41:59 khem: I really appreciate you taking the time to help with this, thanks Aug 22 09:56:37 cool Aug 22 09:57:00 keep me informed too :) Aug 22 10:35:19 morning all Aug 22 10:37:17 morning pb_ Aug 22 10:39:35 hi pb_, florian Aug 22 10:40:07 hi bluelightning Aug 22 10:40:09 florian: btw, I noticed you only sent out your call for admins a while ago to the ltg mailing list - you might want to send it out to oe-devel as well Aug 22 10:42:44 bluelightning: that's something to think about... Aug 22 12:05:42 is it expected that eglibc build fails with 'cannot find -lgcc_eh'? Aug 22 12:14:40 ensc: FWIW, I just did an eglibc build an hour ago and did not get that failure Aug 22 12:19:41 bluelightning: a build from scratch or only egibc? Aug 22 12:19:46 I built from scratch Aug 22 12:20:26 it was a rebuild of almost everythin forced by signature change, but not from scratch Aug 22 12:20:33 I'll kick off a from-scratch build now Aug 22 12:20:51 I think it has something todo with the gcc-intermediate removal Aug 22 12:21:02 yes, it sounds like it Aug 22 12:31:04 JaMa: I don't suppose you have any more information for bug 2977 (broken error handling during parse)? Aug 22 12:42:29 Hi. What recipe provides a java plugin for firefox? Aug 22 13:01:15 ensc: eglibc seems to have compiled here just fine from scratch (for qemuarm) Aug 22 13:02:10 bluelightning: ok; I will post on ml Aug 22 13:34:28 ERROR: Function failed: Fetcher failure for URL: 'svn://www.eglibc.org/svn/branches/;module=eglibc-2_16;protocol=http'. Unable to fetch URL from any source. Aug 22 13:34:35 does someone has it too? Aug 22 14:10:57 Question: Aug 22 14:11:31 I am trying to port old recipes (machine and kernel) to oe-core. My kernel recipe is parsed but ignored. Can I ask bitbake to tell me why? Aug 22 14:18:32 fog: can you elaborate on "ignored"? Aug 22 14:19:10 Parsing of 824 .bb files complete (823 cached, 1 parsed). 1122 targets, 43 skipped, 0 masked, 0 errors. Aug 22 14:19:14 it is one of the skipped. Aug 22 14:19:30 (I know because if I touch it I get just 1 skipped after cache loading) Aug 22 14:20:01 fog: if you try to build it directly it should fail and tell you why it was skipped Aug 22 14:21:12 bluelightning: it fails but doesn't tell me why. I get a python traceback. Aug 22 14:21:26 fog: are you using -b? Aug 22 14:21:47 when it gives you the traceback Aug 22 14:21:51 bluelightning: yes: bitbake -b /home/fog/Workspaces/openembedded/oe-core/meta-ea31xx/recipes-kernel/linux/linux-ea314x_2.6.33.bb Aug 22 14:21:59 traceback that ends with: Aug 22 14:22:06 AttributeError: 'CoreRecipeInfo' object has no attribute 'task_deps' Aug 22 14:22:06 known issue I'm afraid Aug 22 14:22:21 try bitbake linux-ea314x instead Aug 22 14:22:38 assuming you have no other version recipe with the same PN Aug 22 14:22:48 ok. Aug 22 14:23:18 same error. maybe I made an error when porting the recipe. Aug 22 14:23:29 what, you get a traceback then as well? Aug 22 14:23:33 yes. Aug 22 14:23:41 oh dear :( Aug 22 14:23:43 exact same traceback Aug 22 14:23:52 we should fix that as a matter of urgency Aug 22 14:24:06 thanks anyway. Aug 22 14:24:44 btw, should I do something special to generate a micro image (micro-meta) to test my machine layer? or just add it to bblayers? Aug 22 14:29:22 fog: well, I suspect micro-image is intended to be used with DISTRO = "micro", in which case yes you would want to add it to your bblayers.conf Aug 22 14:31:29 bluelightning: thanks. Aug 22 14:31:59 if you have problems with micro, feel free to ping pb_ ;) Aug 22 14:32:52 if you're just after testing your layer you should be able to use build core-image-minimal Aug 22 14:32:58 s/use// Aug 22 14:34:49 bluelightning: oh! I found the problem: I had COMPATIBLE_MACHINE = "ea31xxx" with an extra "x" Aug 22 14:35:08 ah right.. good catch Aug 22 14:35:29 now micro is lamenting missing recipes; I'll follow your advice and build core-image-minimal Aug 22 14:37:51 fog: which OE-Core branch are you using? Aug 22 14:37:59 bluelightning: master Aug 22 14:38:10 bluelightning: should I use a specific branch? Aug 22 14:38:17 no, just checking Aug 22 14:38:32 trying to reproduce the failure you mentioned here Aug 22 14:38:41 it's always best to try builds with just oe-core if you're making sure something works. The fewer the layers, the lower the possiblity of unrelated failures Aug 22 14:38:43 heh Aug 22 14:40:19 fog: so when you did "bitbake linux-ea314x" you definitely didn't specify -b ? Aug 22 14:40:26 I can reproduce it here but only with -b Aug 22 14:40:53 bluelightning: mm. let me look at the history. Aug 22 14:41:46 Ops. sorry, I fear I hed -b. I'll make the file wrong again and try without. Aug 22 14:42:19 ERROR: Nothing PROVIDES 'linux-ea314x' Aug 22 14:42:19 ERROR: linux-ea314x was skipped: incompatible with machine ea31xx (not in COMPATIBLE_MACHINE) Aug 22 14:42:21 bluelightning: :) Aug 22 14:42:38 great Aug 22 14:43:03 well, at least the failure is isolated to the area expected... Aug 22 14:43:24 -b is supposed to do the same thing, although personally I consider -b evil myself Aug 22 14:44:17 almost never used it except to test recipes without regenrating the cache. Aug 22 14:44:25 but now that I have an SSD, no problems. Aug 22 14:46:56 -b was only really useful for experts back when the parse times were even more crazy than they are now :) Aug 22 14:47:03 * kergoth can't remember teh last time he used it Aug 22 14:49:40 ok. 680 steps to go... bluelightning: thank you again, the conversion took just a couple fo hours. wonderful. Aug 22 14:49:48 fog: no worries Aug 22 15:50:58 another little question: it is possible to include an .inc file from a reciper from another layer? Aug 22 15:51:14 I need a specific version of uboot and I'd like to include u-boot.inb Aug 22 15:51:48 fog: yep, just use the path relative from the layer directory Aug 22 15:52:04 e.g. require recipes-core/somedir/somefile.inc Aug 22 15:53:54 bluelightning: and... thanks again! Aug 22 15:54:00 np Aug 22 15:59:05 FYI, I just sent a patch to fix the -b issue Aug 22 16:01:11 bluelightning: great. Aug 22 16:01:25 time to go home... thanks again and cu. Aug 22 16:01:30 take care Aug 22 16:27:28 bb all **** BEGIN LOGGING AT Wed Aug 22 17:34:23 2012 Aug 22 19:37:02 hi, a quick question: how do I setup a SRC_URI to download a git repo over https? Aug 22 19:37:16 git://bla.bla;protocol=https ? Aug 22 19:41:59 yep Aug 22 19:44:57 kergoth: thanks Aug 22 19:54:25 does ;patch=1 means something ? Aug 22 19:57:47 wakko: in the past it told the build to apply it as a patch, but nowadays the system will automatically do so for patch/diff files. so explicit is only needed for files named something different. also, the name of the parameter changed, its apply=yes to force it now Aug 22 19:58:03 short answer: just remove it, assuming yu're updating a recipe from oe classic to oe core Aug 22 20:00:22 ok Aug 22 20:03:10 is it expected that /lib/libgcc_s.so.1 is not stripped in the image? It makes a difference of 1 MB Aug 22 20:03:37 afaik no, that shouldnt' be the case Aug 22 20:03:44 khem would probably know best though **** ENDING LOGGING AT Thu Aug 23 02:59:59 2012