**** BEGIN LOGGING AT Wed Jul 13 02:59:58 2016 Jul 13 09:23:05 that new bitbake progress stuff is pretty nice Jul 13 09:34:00 koen: yeah, its good Jul 13 10:13:39 hopefully the fetch progress patch gets merged since I believe that's still missing Jul 13 10:17:52 bluelightning: applied to mut :) Jul 13 10:18:01 thanks for remembering :) Jul 13 10:18:01 rburton: great :) Jul 13 10:18:40 I spent a lot of time poking at that patchset so I'm glad it's finally sorted :) Jul 13 10:19:10 can you verify that it all landed apart from the fetch bit? Jul 13 10:26:52 rburton: yep, just verified via rebase Jul 13 10:27:37 thanks Jul 13 10:28:14 just doing a test build of piglit so hopefully i get to see the git fetch progress now Jul 13 10:31:33 rburton: you did pick up the v3 of that patch right? Jul 13 10:31:59 hm Jul 13 10:32:11 I guess the answer is probably yes since it was you that found the issue with v2... Jul 13 10:32:30 well lets check :) Jul 13 10:32:54 yeah Jul 13 10:32:59 if it has --show-progress in it it's the old version Jul 13 10:40:41 I have seen several hardware vendors to use the google 'repo' tool for shipping their oe based bsps; is this recommended by oe/yocto? Jul 13 10:43:36 ensc|w: yocto doesn't really make a recommendation, but yes several do use repo Jul 13 10:44:05 awww had to answer the door and i missed the git fetch Jul 13 12:18:07 Hi there Jul 13 12:18:29 Is there a setting I can specify in local.conf to set prefix? Jul 13 12:18:41 I'd like to build everything under /opt Jul 13 12:19:07 I'm not trying to build a whole image, just one application and all it's dependencies (including libc) Jul 13 12:19:42 such that I can be certain it will run independent of the state of the rest of the filesystem Jul 13 12:21:30 I've considered setting EXTRA_OECONF_append_class-target = " --prefix=/opt/ostree " Jul 13 12:21:42 in local.conf Jul 13 12:21:58 WillManley: why not create a recipe for such application or building it externally after creating build_sdk? Jul 13 12:22:16 There's a whole bunch of dependencies Jul 13 12:22:33 including glib, libsoup, libc, etc. Jul 13 12:22:40 I want to bundle them all Jul 13 12:23:11 I was hoping to reuse openembedded's receipies to handle this Jul 13 12:23:38 WillManley: prefix=/opt and exec_prefix=/opt will do it Jul 13 12:24:18 oh, and you might want to set base_prefix too so stuff that would usually go into /lib will go into /opt Jul 13 12:24:36 WillManley: http://www.yoctoproject.org/docs/2.0/adt-manual/adt-manual.html#optionally-building-a-toolchain-installer Jul 13 12:31:31 Thanks, I'll give it a go Jul 13 12:32:13 Is there a way of setting prefix=/opt and exec_prefix=/opt just for class-target? Jul 13 12:32:31 in local.conf Jul 13 12:33:06 To save me from having to rebuild too many native packages? Jul 13 12:40:38 WillManley: just set the right prefix variables (see bitbake.conf) and you'll relocate everything. then create a bare rootfs with no kernel etc and you'll be sorted. Jul 13 12:44:38 i wonder if there's a good reason why exec_prefix doesn't default to ${prefix} Jul 13 12:57:44 rburton: Thanks Jul 13 13:36:33 zeromq isfialing on master Jul 13 13:36:35 http://pastebin.com/NJLy2K0q Jul 13 13:37:05 I am updating to 4.1.5 and still fails, but will fix by adding PACKAGECONFIG for libsodium and uses that instead of internal version Jul 13 15:07:27 Crofton|work: how does the recipe look like Jul 13 15:07:35 it seems its not finding its own headers Jul 13 15:08:00 which could mean it may not be generating Makefiles correctly Jul 13 16:17:09 khem, I want to say this is somethign with gcc6 changing header search Jul 13 16:17:23 oddly, the release notes for that release say they fixed something for 6 Jul 13 16:17:35 But I wonder if the tester builds against libsodium Jul 13 16:19:46 looked at commits and do not see anything about this Jul 13 16:21:39 hmm, Jul 13 16:25:15 recipe isn't "strange" Jul 13 16:25:42 I'm going to add the PACKAGECONFIG for libsodium, which need doing anyway Jul 13 16:25:49 and set default to work around this issue Jul 13 16:26:04 and include info in commit message Jul 13 18:29:50 damn it, need to change a commit message one back from HEAD Jul 13 18:31:14 Crofton|work: git rebase -i ? Jul 13 18:31:18 if it hasn't been pushed, just rebase -i Jul 13 18:31:18 yeah Jul 13 18:31:21 yeah Jul 13 18:31:27 something stupid is going on Jul 13 18:35:05 still wondering why it misses my rename on libsodium Jul 13 18:35:08 oh well Jul 13 18:39:35 Crofton|work and git, a love story in several acts Jul 13 18:39:43 :) Jul 13 18:39:43 lol Jul 13 18:39:52 pushing +branchname Jul 13 18:39:54 better than the alternatives Jul 13 18:40:07 WHo here would like to go back to monotone? Jul 13 18:40:13 Do not all speak at once. Jul 13 18:43:18 khem, git send-email -M -1 --to openembedded-devel@lists.openembedded.org --subject-prefix=meta-oe][PATCHv2 42da562f22f080bb5b0a34dd020b878e4f20572b Jul 13 18:43:31 not sure exactly what I did to break the rename Jul 13 19:06:10 I think put it in .gitconfig once for all Jul 13 19:08:07 yes, I just did that Jul 13 19:11:10 put what in gitconfig? Jul 13 19:13:36 [diff] Jul 13 19:13:36 renames = copies Jul 13 20:40:17 not sure if this would be useful for anyone else, but.. https://gist.github.com/kergoth/f6cb0e2bb3afd2560a2dd1ae46fc0cf5 Jul 13 20:40:41 doing a bunch of builds with minor changes for each one, and need control over cleanup Jul 13 20:40:42 heh Jul 13 21:01:57 system/bin/sh: ./i2cdump: not found Jul 13 21:02:09 did bitbake and downloaded this and getting this problem on my board. Jul 13 21:02:16 any pointers? Jul 13 21:15:25 peoliye: when you say "did bitbake" do you mean you actually added it to be installed into your image and then rebuilt and redeployed the image? Jul 13 21:16:40 bluelightning: i downloaded it using bitbake and pushed in on the device. Anyway just talked to my friend here and he said i need to enable software float for arm in my kernel. Let me rebuild my kernel and check again, will update you. Jul 13 21:23:25 peoliye: I'm still not sure what you've done exactly Jul 14 01:53:08 khem, are you working on backporting the two mips regression in musl 1.1.15? Jul 14 01:56:53 * armpit time to head home **** ENDING LOGGING AT Thu Jul 14 02:59:59 2016