**** BEGIN LOGGING AT Mon Nov 13 03:00:01 2017 Nov 13 07:34:03 good morning Nov 13 07:35:36 mckoan: agreed on the morning part. Nov 13 07:36:01 LetoThe2nd: :-) Nov 13 13:30:39 dv_: did you fix it? Nov 13 13:31:54 otavio: I don't yet know why this happens Nov 13 13:32:03 I wonder if anybody else saw the same problemn Nov 13 13:32:29 it seems to involve the quilt push/pop behavior Nov 13 13:32:50 and I'm concerned that this may be causing other silent errors, which is why I mention it here Nov 13 13:32:57 dv_: it is strage as I am using something similar with Go and it works Nov 13 13:33:14 dv_: see if the fetcher is broken; move to the root dir for the patch Nov 13 13:35:30 rburton: I sent the mesa stable bugfix Nov 13 13:35:36 rburton: important to go in Nov 13 13:36:00 rburton: fixes Android and Linux rendering issues Nov 13 13:36:32 otavio: indeed if I move the patch out of the subdir into the root dir, it works. I need to know why this is happening though. Nov 13 13:36:47 also, the patch is fetched. it is in the work dir Nov 13 13:37:03 it is just not applied. when I run patch -p1 --dry-run F0 -i , I can see that it cleanly applies. Nov 13 13:37:49 errr, I meant patch -p1 --dry-run -F0 -i Nov 13 13:38:27 I'll keep looking, but if this rings any bells for anybody here, I would welcome it Nov 13 13:50:58 dv_: this is new to rocko right? check the patch.bbclass changes since pyro Nov 13 13:51:58 hmm yes, good point Nov 13 14:45:31 Hey is there a way to enable symlinks for busybox apps in my image? Nov 13 15:50:06 * armpit 4.14 released Nov 13 15:50:43 * armpit time to update bsp Nov 13 15:52:48 armpit, to 4.1? Nov 13 15:53:31 kernel 4.14 is now mainline, no longer -rc Nov 13 16:31:22 I see that meta-oe is using the same major version of postgresql (9.4.x) for morty, pyro, rocko, master yet each are using a different minor release (which are only supposed to contain bug and security fixes). There are quite a few CVEs that have been fixed. Nov 13 16:32:17 I'm considering sending patches that upgrade morty, pyro, rocko, master to 9.4.15. Thoughts? Nov 13 16:33:23 I'd just rather not do the work if patches will be rejected due to some reason like "version has changed too much". Nov 13 16:35:17 if its *only* security fixes then you're in with a chance Nov 13 16:35:27 but i'm not a meta-oe maintainer so best to ask jama Nov 13 16:35:38 there's an argument here for a pyro-backports branch... Nov 13 16:36:58 rburton: It's not _just_ security fixes. "variety of fixes". So maybe that's pushing it? Nov 13 16:37:32 maybe Nov 13 16:40:58 If JaMa reappears I'll ask him Nov 13 17:03:30 JaMa: What are your thoughts on upgrading postgresql in morty, pyro, rocko, master all to use 9.4.15 to pick up CVE fixes? The minor versions contain a "variety of fixes". I'm not sure where the line is between pulling in legitimate fixes and pulling in too many changes. I wanted to ask before sending in any patches. Nov 13 17:09:49 georgem: it's question for armpit, but I think if everything is ok when it's merged to master that it could be backported to stable branches as well Nov 13 17:12:03 georgem, are postgresql dot release only bug fixes ? Nov 13 17:13:10 * armpit just had to look in up Nov 13 17:13:42 armpit, each dot release says whats in it. I checked back to the version in morty and all of them essentially say "variety of fixes" Nov 13 17:13:55 Example: https://www.postgresql.org/docs/9.4/static/release-9-4-15.html Nov 13 17:14:04 morty is 9.4.8 Nov 13 17:16:28 armpit: Yeah. I checked all of the dot releases from 9.4.15 back to 9.4.9 all they all say "variety of fixes" or "small fixes" on the summary line. Nov 13 17:17:37 that kind of information makes deciding if backporting to a stable branch easier Nov 13 17:18:54 please send a patch for master so we can start the process Nov 13 17:19:02 will do Nov 13 17:19:58 JaMa, I see a world build is running. what branch is it building ? Nov 13 17:20:00 armpit: btw the world builds finished during weekend, so I've triggered them again with newer oe-core Nov 13 17:20:08 your branch Nov 13 17:20:09 hehe Nov 13 17:20:11 in meta-oe Nov 13 17:20:12 k Nov 13 17:20:34 the report is generated, do you want me to send it to ML and update on wiki or will you? Nov 13 17:21:06 JaMa, is that a jenkins job too ? Nov 13 17:21:11 yes Nov 13 17:21:21 http://jenkins.nas-admin.org/view/OE/job/oe_world_workspace-parse-results/lastBuild/console Nov 13 17:21:32 I'll take a shot at it Nov 13 17:33:00 JaMa, I don't think the log on wiki was saved. i don't see http://logs.nslu2-linux.org/buildlogs/oe/world/sumo/log.report.20171111_042827.log Nov 13 17:47:18 armpit: checking, it was rsynced as a file http://logs.nslu2-linux.org/buildlogs/oe/world/sumo Nov 13 17:47:21 will fix it now Nov 13 17:48:06 http://logs.nslu2-linux.org/buildlogs/oe/world/sumo is the log file. there is not .log in that dir Nov 13 17:49:16 armpit: fixed Nov 13 17:49:30 at least for the main log Nov 13 17:49:45 the MACHINE logs are missing, trying to re-rsync them now Nov 13 17:54:15 other log rsynced as well now Nov 13 17:57:03 thaytan, Nov 13 17:57:09 opps.. thanks Nov 13 18:14:47 armpit: it looks like you included oe-devel ML twice in To Nov 13 18:15:10 To: "openembedded-devel@lists.openembedded.org >> OpenEmbedded Devel List" Nov 13 18:15:13 Nov 13 18:56:16 oh Nov 13 20:19:46 armpit: and do you plan to update the wiki as well? http://www.openembedded.org/wiki/Bitbake_World_Status_Sumo Nov 13 20:23:11 JaMa, is that a manual process? /me nothing in my notes on that issue Nov 13 20:25:04 ah, looks like a copy & paste ??? Nov 13 20:27:59 done Nov 13 20:30:50 armpit: yes, copy & paste, you should add it to the list (last line of the table) Nov 13 20:30:59 it looks like you've replaced whole page Nov 13 20:32:18 sounds like a good job for a script... ;) Nov 13 20:35:26 armpit: I've fixed it Nov 13 20:59:15 JaMa, thank Nov 13 21:52:14 devtool seems to be pretty bad about taking files added/appended to SRC_URI and adding them to the original location SRC_URI is set. Makes it pretty unusable for a lot of recipes and layer combinations. Nov 13 21:52:14 georgem: if you can give me a reproducer I can look into it Nov 13 21:53:47 bluelightning: I just used it on oe-core master to add a patch to the shadow recipe. When you do "devtool finish shadow meta" it adds everything that was in SRC_URI_append_class-target to SRC_URI. I've also seen similar happen if you have another layer that has a bbappend that adds to SRC_URI. Nov 13 21:55:25 georgem: right, so appends... I just added conditional append support though it still doesn't do anything special to stop the value being munged together, that's a separate bug Nov 13 21:57:34 I appreciate it's a tricky set of problems Nov 13 22:13:34 bluelightning: Seems like you might be able to crawl through the appends (bitbake -e seems to have this kind of information) and remove anything in those from the list of files added to SRC_URI. Nov 13 22:14:51 georgem: well, we can't remove those - they have to end up in the source tree and we want to enable updating them Nov 13 22:15:29 right. maybe I didn't phrase that correctly. Nov 13 22:16:39 I mean something like: Check if $file is already in an append to SRC_URI. If so, don't add it to SRC_URI. Nov 13 22:17:44 but yes. it needs to be generally more aware of the appends and when you make a modification even go as far as letting you tell it which append to put it in. Nov 13 22:17:57 I think we're on the same page Nov 13 22:23:21 georgem: right yes... it'll have to be smarter about how it modifies the value of SRC_URI Nov 13 22:23:44 yup Nov 13 22:23:44 luckily we have the in-built history information that we can use for that, it just has to take it into account **** ENDING LOGGING AT Tue Nov 14 03:00:01 2017