**** BEGIN LOGGING AT Fri Dec 09 02:59:58 2011 Dec 09 09:56:40 morning all Dec 09 10:24:11 morning Dec 09 10:25:41 morning Dec 09 10:40:58 morning all Dec 09 10:54:19 :-) Dec 09 11:14:52 Hi All, how would I setup yocto to spit out -src packages ? Dec 09 11:16:18 Been looking around the source, but I cant find any references on how to do it. Dec 09 11:20:24 sourcepkg.class seems to be the answer :) Dec 09 18:37:43 otavio: ping? Dec 09 18:38:15 incandescant: pong Dec 09 18:38:16 otavio: had you intended to submit ntp to oe-core? connman now DEPENDS it but we don't have it in oe-core Dec 09 18:38:49 incandescant: Ouch ... Dec 09 18:39:06 incandescant: it built here as I have meta-oe enabled Dec 09 18:39:23 otavio: yeah, I just realised that - sorry I did not catch that sooner Dec 09 18:39:37 * incandescant guessed he should only Ack patches he has built Dec 09 18:39:38 incandescant: I didn't catch it too Dec 09 18:39:54 incandescant: in fact I built, but did it on my tree Dec 09 18:39:59 incandescant: and my tree has meta-oe Dec 09 18:40:21 http://bugzilla.yoctoproject.org/show_bug.cgi?id=1817 Dec 09 18:40:29 RP__: should I add ntp or send a patch to remove ntp from depends? Dec 09 18:41:04 dvhart: thanks ;-) Dec 09 18:41:05 sgw: any opinion? Dec 09 18:41:25 I prefer to provide ntp Dec 09 18:41:29 otavio: I suspect add it Dec 09 18:41:34 OK Dec 09 18:41:41 will cook a patch for it Dec 09 18:42:57 what does connman use ntp for? Dec 09 18:43:06 how is it affected if it isn't there? Dec 09 18:43:25 dvhart: it has a pluging for it Dec 09 18:43:37 seems to me connman should rely on the system time, and it should be up to the distro if the system time is maintained via ntp Dec 09 18:44:04 dvhart: http://www.mail-archive.com/connman@connman.net/msg06788.html Dec 09 18:44:26 RP__, otavio: catching up and reading Dec 09 18:45:42 otavio, that doesn't really help explain WHY for me Dec 09 18:46:08 could be perfectly valid mind you, I was just asking b/c I didn't know Dec 09 18:46:29 dvhart: personally I don't use ntp plugin Dec 09 18:46:41 otavio: so it's untested? Dec 09 18:46:50 dvhart: but as it is modular, I think it is better to have it Dec 09 18:46:56 sgw: the ntp plugin yes Dec 09 18:47:15 I think it might be better to disable that functionality for now instead of adding a new depends. Dec 09 18:47:28 otavio: If nobody is using it I'm not sure we should have it Dec 09 18:47:36 we can easily add it if someone actually uses it Dec 09 18:47:42 RP__: right Dec 09 18:47:44 RP__: for me it is OK to drop for now Dec 09 18:48:24 RP__: cooking a patch for it then Dec 09 18:48:47 otavio: I think we kind of rushed this, otherwise I might have caught this. Dec 09 18:49:16 sgw: yes; it is holding my product building and I asked RP__ to get it in. Dec 09 18:49:20 sgw: my fault Dec 09 18:55:42 sgw: dvhart: incandescant: RP__: https://github.com/OSSystems/oe-core/commit/35343ba859919a927e5373ac7f39526dad301ca5 Dec 09 18:55:45 done Dec 09 18:56:31 IRC-Acked-By: dvhart Dec 09 18:56:50 typo in your commit log Dec 09 18:56:53 RP__: want me to send to ml? Dec 09 18:56:56 s/know/known/ Dec 09 18:56:59 dvhart: really? where Dec 09 18:57:00 grr Dec 09 18:57:04 I am not native Dec 09 18:57:05 otavio: please Dec 09 18:57:05 hold Dec 09 18:57:22 otavio: RP: seems right to me Dec 09 18:57:46 otavio, it's cool, that's what the review process is for :) Dec 09 18:57:55 * dvhart is native and still introduces all kinds of typos.... Dec 09 18:58:45 dvhart: do you close the bug once it is merged? Dec 09 18:58:58 RP__: sent Dec 09 18:59:16 otavio, no, we add the commit link to git.yoctoproject.org and mark it as FIXED Dec 09 18:59:41 dvhart: you do it, or I do it? Dec 09 19:00:22 otavio, you should move to fixed Dec 09 19:00:34 as the reported I will mark resolved after I confirm it works for me Dec 09 19:00:36 once merged Dec 09 19:00:46 gah, see - reported/reporter Dec 09 19:01:26 i.e. dev submits fix -> fix merged -> dev marks bug as fixed with link to git -> reporter verifies and marks bug appropriately Dec 09 19:01:49 so I mark as resolved/fixed? Dec 09 19:02:16 yes, once it's been merged Dec 09 19:02:56 I think this is documented somewhere, let me try and find it Dec 09 19:03:44 https://wiki.yoctoproject.org/wiki/Bugzilla_Configuration_and_Bug_Tracking Dec 09 19:03:45 ta da! Dec 09 19:03:52 that's a lot of words, though Dec 09 19:04:36 incandescant, who's high maintenance now? Dec 09 19:04:38 ;) Dec 09 19:07:16 I'm not demanding it be read, I'm just pointing out its existence Dec 09 19:07:55 sgw: other day you were asking for eglibc patches Dec 09 19:08:13 sgw: they all are documented well in header and reflects the current status Dec 09 19:08:22 nothing has changed in there Dec 09 19:10:59 khem: what about the uclibc patches, since you work with those also, I thought I mentioned them also? Dec 09 19:11:20 uclibc ones I will take care of them one day Dec 09 19:11:28 and push them upstream Dec 09 19:11:35 khem: thanks Dec 09 19:13:33 there are about 400 patches I am holding on for uclibc Dec 09 19:13:45 there is one pesky bug thats lurking somehwere Dec 09 19:18:56 There's got to be a better way of doing this.... I'm moving some, but not all, of various recipe libraries from libdir to base_libdir (e.g, e2fsprogs)... Dec 09 19:19:30 I can mv the libxyz.so.* files into base_libdir easily enough, but do I really need to re-create the .so symlink (which should stay in libdir)? Dec 09 19:19:51 or is there a helper function to relocate libraries like this? Dec 09 19:20:32 The issue I'm running into is not knowing what the soname extension is when relinking the .so file Dec 09 19:20:36 zenlinux: look at meta-micro Dec 09 19:20:47 zenlinux: you can do it using prefix and exec_prefix Dec 09 19:20:59 otavio, looking... Dec 09 19:22:11 otavio, where is the meta-micro layer hosted? I'm not seeing it on the oe git repo Dec 09 19:23:24 zenlinux: openembedded git iirc Dec 09 19:23:41 dvhart: Is it possible for you to test a patch for connman build? Dec 09 19:23:43 ah, found it Dec 09 19:23:55 dvhart: here it builds but maybe I know how to fix it Dec 09 19:25:32 otavio, sure Dec 09 19:25:42 otavio, could you give me a more specific example? btw I'm not relocating *all* the libraries of e2fsprogs to base_libdir, just some of them known to be linked by binaries in base_bindir/base_sbindir Dec 09 19:25:44 otavio: zenlinux is trying something a little different Dec 09 19:26:10 zenlinux: I'm not sure meta-micro would help. A function for this kind of thing could be useful Dec 09 19:27:07 zenlinux: so no. This is not what I thought you were doing Dec 09 19:27:17 otavio, ok Dec 09 19:34:39 dvhart: https://github.com/OSSystems/oe-core/commit/695e24c9e389207412e0d39446f4ea1d95355ac2 Dec 09 19:36:32 dvhart: I know the comment is not right, i fix it after you confirm it builds Dec 09 19:38:48 hrm... can I download this as a patch directly from github.... not finding such a link Dec 09 19:40:37 otavio, building Dec 09 19:41:56 dvhart: about dosfstools Dec 09 19:42:05 dvhart: gplv3 are 3.0.x releases Dec 09 19:42:27 dvhart: what i did was to forward-port patches to 2.11 and remove 2.10 Dec 09 19:42:41 dvhart: and drop native specific recipe Dec 09 19:45:23 ok, so 2.11 includes the llseek patch and retains GPLV2 ? Dec 09 19:45:58 dvhart: yes Dec 09 19:48:39 dvhart: has connman built for you now? Dec 09 19:50:15 no, still seems to fail in the same way Dec 09 19:53:45 dvhart: https://github.com/OSSystems/oe-core/commit/835abb07c6bdc2d5c9e1c3e906819630c6dff023 Dec 09 19:53:53 dvhart: I added line 54 Dec 09 19:54:55 building Dec 09 19:57:34 that builds Dec 09 19:57:45 but what are we losing from this? Dec 09 20:02:54 dvhart: this was in meta-oe and since it built fine here I removed Dec 09 20:03:04 dvhart: it is a test that is not built Dec 09 20:03:11 dvhart: seems safe for now Dec 09 20:03:21 dvhart: i will finish the patch and send Dec 09 20:06:21 dvhart: sent to ml Dec 09 22:20:25 zeddii, ping Dec 09 22:20:29 zeddii, $ cat ../../features/rt/rt.scc Dec 09 22:20:29 patch rt-apply-patch-3.0.10-rt27.patch.patch Dec 09 22:20:46 the file only has one .patch prefix - this is bad right? Dec 09 23:57:26 dvhart: I should pull otavio's patch? Dec 09 23:58:03 RP__: I can confirm that without it I can't build, and with it I can Dec 09 23:58:35 RP__, it builds and solves the immediate problem Dec 09 23:58:46 RP__, but it feels like a hack Dec 09 23:58:58 ok, I'll add it on the grounds it makes things better rather than worse... Dec 09 23:58:58 RP__, I don't know what the repercussions might be Dec 09 23:59:06 * dvhart nods **** ENDING LOGGING AT Sat Dec 10 02:59:56 2011