| *** gitlab-br-bot has quit IRC | 04:27 | |
| *** gtristan has quit IRC | 05:45 | |
| *** gtristan has joined #baserock | 05:54 | |
| *** adds68 has joined #baserock | 06:47 | |
| *** jude_ has joined #baserock | 06:49 | |
| *** adds68 has quit IRC | 06:53 | |
| *** paulwaters_ has joined #baserock | 07:20 | |
| *** adds68 has joined #baserock | 07:48 | |
| *** adds68 has quit IRC | 08:03 | |
| *** adds68_ has joined #baserock | 08:03 | |
| *** adds68__ has joined #baserock | 08:40 | |
| *** adds68_ has quit IRC | 08:42 | |
| *** jonathanmaw has joined #baserock | 08:59 | |
| *** ssam2 has joined #baserock | 09:07 | |
| *** ChanServ sets mode: +v ssam2 | 09:07 | |
| ssam2 | pipeline for https://gitlab.com/baserock/definitions/merge_requests/59 succeeded! | 09:19 |
|---|---|---|
| ssam2 | could someone give it an approval ? | 09:23 |
| *** gitlab-br-bot has joined #baserock | 09:31 | |
| *** adds68__ has quit IRC | 09:32 | |
| gitlab-br-bot | definitions: merge request (sam/bootstrap-from-baserock-binaries->master: BuildStream builds should bootstrap from Baserock binaries, not the Freedesktop SDK) #58 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/58 | 10:05 |
| jjardon | pipeline for https://gitlab.com/baserock/definitions/merge_requests/64 succeeded! Can I get another review? | 10:22 |
| gitlab-br-bot | definitions: merge request (jjardon/no_master2->master: Do not use master of our build tools, but a specific version instead) #64 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/64 | 10:42 |
| *** adds68 has joined #baserock | 11:21 | |
| *** adds68 has joined #baserock | 11:23 | |
| *** adds68 has joined #baserock | 11:23 | |
| gitlab-br-bot | definitions: merge request (jjardon/no_master2->master: Do not use master of our build tools, but a specific version instead) #64 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/64 | 11:27 |
| jjardon | benbrown_: ssam2 sorry, can I get a reaproval ^? I had to do a manual rebase | 11:29 |
| gitlab-br-bot | definitions: merge request (jjardon/no_master2->master: Do not use master of our build tools, but a specific version instead) #64 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/64 | 11:30 |
| jjardon | ssam2: Do we still have the limitation to be a protected branch to be able to push to the bst cache server? | 11:32 |
| benbrown_ | jjardon: AFAIA, yes | 11:34 |
| ssam2 | yes | 11:37 |
| jjardon | ssam2: now that we have our own docker image; maybe we can put the SSH key there? | 11:43 |
| ssam2 | the image would have to be private | 11:43 |
| benbrown_ | jjardon: I definitely prefer how it is now to storing it in the docker image | 11:44 |
| jjardon | yeah | 11:45 |
| *** adds68 has quit IRC | 12:05 | |
| *** adds68 has joined #baserock | 12:32 | |
| *** adds68 has quit IRC | 12:59 | |
| *** adds68 has joined #baserock | 13:01 | |
| gitlab-br-bot | definitions: merge request (jjardon/no_master2->master: Do not use master of our build tools, but a specific version instead) #64 changed state ("merged"): https://gitlab.com/baserock/definitions/merge_requests/64 | 13:34 |
| *** toscalix has joined #baserock | 14:01 | |
| *** jude_ has quit IRC | 14:02 | |
| ssam2 | the compression of debug symbols has broken debug stripping in baserock | 14:26 |
| ssam2 | i guess the simplest thing is to override strip-commands in the baserock project.conf for now | 14:26 |
| ssam2 | i wonder why it doesn't work though ... perhaps we build stage2-binutils without libz | 14:26 |
| ssam2 | in fact, we definitely do that | 14:27 |
| gtristan | ssam2, has it ? | 14:27 |
| ssam2 | yeah | 14:28 |
| gtristan | ssam2, then I need to fix that, to be honest I assumed it would just work, but maybe it needs more fiddling than just adding the argument to the initial objcopy ? | 14:28 |
| ssam2 | https://paste.baserock.org/hupasozoce | 14:28 |
| ssam2 | and the result is no stripping is done at all | 14:28 |
| ssam2 | and the sysroot ends up at 1.2GB | 14:28 |
| ssam2 | maybe there's a way that objcopy could tell us whether it supports compression | 14:29 |
| ssam2 | or maybe we could add zlib to stage2 | 14:29 |
| gtristan | oh weird | 14:29 |
| gtristan | I dont get that error message at all | 14:29 |
| ssam2 | me neither | 14:29 |
| gtristan | you did get that error message, I mean, you received it | 14:30 |
| ssam2 | it's in the log | 14:30 |
| gtristan | I did not even receive it | 14:30 |
| ssam2 | oh, right | 14:30 |
| ssam2 | I thought you meant you don't understand it :-) | 14:30 |
| gtristan | yeah got that :) | 14:30 |
| gtristan | heh | 14:30 |
| ssam2 | did you do a baserock bootstrap to test this? I suspect it's due to the way we build binutils | 14:30 |
| gtristan | ssam2, So I suspect yeah it might have to do with zlib, but note that I changed the project defaults to split up the macros a bit | 14:31 |
| ssam2 | right | 14:31 |
| gtristan | so you can easily disable it for the bootstrap phase until it works, without having to rewrite a whole ton of strip commands | 14:31 |
| ssam2 | ohh, nice | 14:31 |
| gtristan | its something like objcopy-something-args | 14:31 |
| ssam2 | hmm, we're also using binutils from 2014 | 14:34 |
| ssam2 | still, one thing at a time... don't have time to fix every issue in baserock right now. i'll just disable compression | 14:35 |
| *** gtristan has quit IRC | 14:36 | |
| *** gtristan has joined #baserock | 14:39 | |
| *** adds68 has quit IRC | 14:48 | |
| *** adds68 has joined #baserock | 14:48 | |
| ssam2 | the stripping seems to work past gnu-toolchain.bst, so i think this is indeed just that zlib is not in stage2 | 14:52 |
| * ssam2 works around | 14:52 | |
| gtristan | ssam2, ok great to know that it works :) | 14:52 |
| gtristan | ssam2, any idea if the compression actually makes smaller debug files ? | 14:52 |
| ssam2 | not got an easy way to benchmark that | 14:53 |
| ssam2 | well, i guess i could strip something twice | 14:53 |
| ssam2 | let me try gcc | 14:53 |
| ssam2 | ah, here's the difference | 14:56 |
| ssam2 | cc1 debug symbols are 127MB when extracted normally, or 53MB when extracted with --compress-debug-symbols | 14:57 |
| ssam2 | so, yes :-) | 14:57 |
| ssam2 | on the flip side, objcopy takes 8 seconds vs. 0.3 seconds | 14:58 |
| ssam2 | so there's a time cost to the compression | 14:58 |
| *** adds68 has quit IRC | 15:00 | |
| gtristan | ssam2, sounds great to me thanks for trying it out ! | 15:14 |
| gtristan | :) | 15:14 |
| gtristan | overall, I think thats a desirable win, I guess some setups will want to disable it | 15:15 |
| gtristan | but for developers it should save some disk space | 15:16 |
| ssam2 | yeah. 8 seconds isn't too bad considering this was cc1, one of the biggest binaries we build | 15:16 |
| *** adds68 has joined #baserock | 15:59 | |
| *** toscalix has quit IRC | 16:29 | |
| *** adds68 has quit IRC | 16:30 | |
| gitlab-br-bot | definitions: merge request (sam/bootstrap-from-baserock-binaries->master: BuildStream builds should bootstrap from Baserock binaries, not the Freedesktop SDK) #58 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/58 | 16:51 |
| gitlab-br-bot | definitions: merge request (sam/bootstrap-from-baserock-binaries->master: BuildStream builds should bootstrap from Baserock binaries, not the Freedesktop SDK) #58 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/58 | 16:53 |
| ssam2 | alright! stage3-sysroot/x86_64 branch pushed to https://ostree.baserock.org/releases/ | 16:53 |
| ssam2 | bootstrap wise... we have liftoff | 16:54 |
| ssam2 | assuming https://gitlab.com/baserock/definitions/merge_requests/58 builds OK, which is a big assumption! | 16:54 |
| gtristan | ssam2, is that a circular build ? | 16:54 |
| gtristan | ssam2, boostrapping off of our bootstrap ? | 16:54 |
| ssam2 | there's a manual step, if that's what you mean | 16:54 |
| gtristan | yeah but, committed the result to an ostree, and then rebuilding off of that one ? | 16:55 |
| ssam2 | yes | 16:55 |
| gtristan | Nice ! | 16:55 |
| ssam2 | only x86_64 so far but.. baby steps | 16:55 |
| ssam2 | the sysroot is about 350MB; not perfect but reasonable | 16:55 |
| gtristan | It's the most important, from there if we have cross-bootstrap, even if only supported on x86_64, everything is possible :) | 16:56 |
| ssam2 | argh, i seem to have deleted the commits that update the README | 16:57 |
| gitlab-br-bot | definitions: merge request (sam/bootstrap-from-baserock-binaries->master: BuildStream builds should bootstrap from Baserock binaries, not the Freedesktop SDK) #58 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/58 | 16:58 |
| ssam2 | but, if that builds, then i just need to clean up some stuff on the infrastructure side tomorrow, and it's pretty done | 17:00 |
| *** jonathanmaw has quit IRC | 17:00 | |
| *** jonathanmaw has joined #baserock | 17:00 | |
| *** ssam2 has quit IRC | 17:10 | |
| *** adds68 has joined #baserock | 18:15 | |
| *** adds68 has quit IRC | 18:35 | |
| gitlab-br-bot | infrastructure: issue #1 ("Set up cache for GitLab CI runners") changed state ("opened") undefined | 18:44 |
| gitlab-br-bot | infrastructure: issue #1 ("Set up cache for GitLab CI runners") changed state ("opened") undefined | 19:24 |
| gitlab-br-bot | infrastructure: merge request (pedro/gitlab-bot-fixes-2->master: Route traffic to gitlab irc bot) #16 changed state ("opened"): https://gitlab.com/baserock/infrastructure/merge_requests/16 | 19:24 |
| *** gtristan has quit IRC | 22:45 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!