| *** tristan has quit IRC | 00:34 | |
| *** tristan has joined #buildstream | 00:54 | |
| *** alexandrufazakas has joined #buildstream | 08:07 | |
| *** traveltissues has joined #buildstream | 08:07 | |
| gitlab-br-bot | danielsilverstone-ct opened issue #1089 (Reporting of unexpected deaths of worker processes could be better) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1089 | 08:29 |
|---|---|---|
| traveltissues | thanks | 08:29 |
| benschubert | valentind: have you had any chance of looking into https://gitlab.com/freedesktop-sdk/freedesktop-sdk/merge_requests/1560 ? :) | 08:34 |
| *** alexandrufazakas has quit IRC | 08:41 | |
| *** alexandrufazakas_ is now known as alexandrufazakas | 08:42 | |
| *** alexandrufazakas_ has joined #buildstream | 08:45 | |
| traveltissues | benschubert, are you guaranteeing that sourcefactory will never mutate the dict? | 08:46 |
| traveltissues | if that's well defined then it's fine | 08:47 |
| benschubert | traveltissues: that would be a pickle bug if that was the case AFAIK | 08:47 |
| traveltissues | ok, ty | 08:47 |
| gitlab-br-bot | traveltissues approved MR !1463 (aevri/pickle->master: Make ChildJobs and friends picklable) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1463 | 08:47 |
| traveltissues | everything else lgtm | 08:50 |
| valentind | benschubert, yes, it is coming. | 08:53 |
| valentind | But I need approval for merges on the docker image, so it takes some time because nobody looks at those | 08:53 |
| *** alexandrufazakas_ has joined #buildstream | 08:54 | |
| benschubert | oh I see, no worries :) thanks! | 08:55 |
| *** alexandrufazakas_ has quit IRC | 08:57 | |
| *** alexandrufazakas_ has joined #buildstream | 08:57 | |
| traveltissues | benschubert, did you have a chance to look at the benchmark for !1492 | 08:57 |
| gitlab-br-bot | MR !1492: Add _set_cache_keys_stability method https://gitlab.com/BuildStream/buildstream/merge_requests/1492 | 08:57 |
| *** bochecha has joined #buildstream | 08:59 | |
| benschubert | traveltissues: I fail to understand the performance gain, where were we doing too much work? | 08:59 |
| traveltissues | i'm not sure how significant that gain is or how precise the quick config is | 09:01 |
| traveltissues | i intended to demonstrate that it wouldn't have a significant negative impact | 09:02 |
| *** phoenix has joined #buildstream | 09:02 | |
| traveltissues | the change was really for maintainability reasons | 09:02 |
| traveltissues | well, for show there was actually a ~+0.5% time | 09:03 |
| benschubert | yeah, I would have expected a negative impact especially on py35 | 09:03 |
| *** alexandrufazakas has quit IRC | 09:04 | |
| *** alexandrufazakas has joined #buildstream | 09:06 | |
| benschubert | I'm slightly worried about this change since it adds more work on a performance critical part of the code (method call + more condition checks). So if it's for a short term (let's say 1 month) and we remove this method afterwards, that seem ok, otherwise I'm not sure | 09:06 |
| benschubert | benchmarks with 8 and 12 runners would also be a good thing for this one | 09:06 |
| *** alexandrufazakas_ has left #buildstream | 09:06 | |
| *** alexandrufazakas__ has joined #buildstream | 09:07 | |
| *** alexandrufazakas__ has left #buildstream | 09:07 | |
| benschubert | Also, we go back to the 'update_state' model, where we check for everything and it makes it hard to know what can actually change where :) | 09:07 |
| Kinnison | valentind: Is 1146 ready for merge? | 09:08 |
| gitlab-br-bot | AlexFazakas closed MR !1496 (alexfazakas/use-merge-trains->master: Use Gitlab merge trains) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1496 | 09:08 |
| valentind | Kinnison, I do not know. | 09:09 |
| Kinnison | valentind: and if so, should it go to the '1' branch? | 09:09 |
| valentind | Wait. | 09:09 |
| *** jonathanmaw has joined #buildstream | 09:10 | |
| valentind | Kinnison, I think that juergbi wanted to wait with that fix. | 09:10 |
| valentind | See #910 | 09:10 |
| gitlab-br-bot | Issue #910: Execution rights on an object can corrupt local cache https://gitlab.com/BuildStream/buildstream/issues/910 | 09:10 |
| valentind | We need to port it to 1.2 | 09:10 |
| valentind | But I will not do that now. | 09:11 |
| Kinnison | aah okay | 09:11 |
| juergbi | Kinnison: we'll need a solution on the casd side for this | 09:11 |
| Kinnison | Worth closing 1146 in the meantime? | 09:11 |
| juergbi | doesn't really make sense merging this now and soon casd support without that | 09:11 |
| Kinnison | juergbi: Presumably once buildstream moves to buildbox-only, that requirement is pushed entirely into casd/buildbox-run/buildbox-fuse ? | 09:12 |
| juergbi | yes | 09:12 |
| Kinnison | Also, given we're aiming at moving to all that, is it worth keeping !746 open or should we dump it? | 09:13 |
| gitlab-br-bot | MR !746: Some pycodestyle (PEP8) fixes https://gitlab.com/BuildStream/buildstream/merge_requests/746 | 09:13 |
| Kinnison | sorry !745 | 09:13 |
| gitlab-br-bot | MR !745: WIP: Depend on fusepy package instead internal copy https://gitlab.com/BuildStream/buildstream/merge_requests/745 | 09:13 |
| * Kinnison typo'd | 09:13 | |
| juergbi | it might still require a change on the buildstream side as the filesystem structure of the local cache is public API of casd and buildstream does some (read-only) direct access | 09:13 |
| juergbi | buildbox as sole sandboxing backend is a bit further off than casd support | 09:14 |
| juergbi | no strong opinion on whether to keep that MR open | 09:14 |
| Kinnison | "further off" is still 'before bst 2' ? | 09:14 |
| juergbi | yes, that's definitely the plan | 09:14 |
| Kinnison | so 745 either should be closed, or retargeted at the '1' branch? | 09:15 |
| juergbi | well, if this is to be merged to bst-1 before we remove fusepy from master, it's fine to merge it to master | 09:16 |
| Kinnison | fair | 09:16 |
| traveltissues | benschubert, fair points | 09:20 |
| traveltissues | my personal preference has always been to avoid 'deconstructed' methods | 09:21 |
| traveltissues | this whole area is probably subject to further change soon | 09:21 |
| traveltissues | i think jennis is going to be working on this | 09:21 |
| benschubert | traveltissues: I agree about deconstructed methods (with a delta for performance in python ;) ) but here, it is different state changes that can happen, and if some branches can _never_ be taken in most of the calls, I think it's worth splitting things out | 09:23 |
| gitlab-br-bot | jonathanmaw opened (was WIP) MR !1482 (jonathan/job-progress->master: Render progress information for loading and processing elements) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1482 | 09:26 |
| juergbi | traveltissues: I consider initialization of that variable to belong together with setting the cache key, so to me your variant doesn't really avoid 'deconstructed' methods | 09:28 |
| juergbi | might be very subjective, of course | 09:28 |
| gitlab-br-bot | beckyella16 opened (was WIP) MR !1497 (becky/artifact_access->master: Allow artifact subcommands to use artifact refs) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1497 | 09:29 |
| *** lachlan has joined #buildstream | 09:30 | |
| gitlab-br-bot | jjardon opened issue #1090 (overnigth tests are failing: 'CheckForbiddenElement' object has no attribute 'node_validate') on buildstream https://gitlab.com/BuildStream/buildstream/issues/1090 | 09:35 |
| jjardon | Hi! is this ^ a known breakage? | 09:36 |
| benschubert | jjardon: yes and there ishttps://gitlab.com/freedesktop-sdk/freedesktop-sdk/merge_requests/1560 for the fix | 09:36 |
| jjardon | ah, thanks benschubert ! | 09:36 |
| benschubert | waiting on a merge on the docker images for freedesktop ci | 09:36 |
| benschubert | valentind has more context | 09:36 |
| valentind | jjardon, thanks for merging that docker image fix. | 09:37 |
| benschubert | valentind: let me know when I can rebase my branch | 09:38 |
| jjardon | valentind: no problem, I guess we need that to merge https://gitlab.com/freedesktop-sdk/freedesktop-sdk/merge_requests/1560 ? | 09:38 |
| valentind | We need to rebase with current 18.08, then change DOCKER_REGISTRY to add /bst2 at the end, and put the correct hash for the docker images. | 09:39 |
| valentind | Plus benschubert's merge request. | 09:40 |
| jjardon | valentind: the docker changes will go in the buildstream CI, rigth? | 09:41 |
| valentind | When the images are published I can make an mr with that and benschubert's commit. | 09:41 |
| valentind | jjardon, What do you mean? | 09:41 |
| jjardon | AFAIK, the benschubert patch should be enough for the freedesktop-sdk part of the problem. If we need new buildstream images we should change the image used in the buildstream CI overnigth job | 09:43 |
| benschubert | jjardon: the problem is for installing bst2 for the freedesktop-sdk tests | 09:43 |
| benschubert | they are missing cython, a compiler, etc | 09:43 |
| jjardon | yeah | 09:44 |
| benschubert | for the freedesktop CI, not the buildstream one | 09:44 |
| benschubert | the buildstream images are fine | 09:44 |
| jjardon | so we need to fix https://gitlab.com/BuildStream/buildstream/blob/master/.gitlab-ci.yml#L220 | 09:45 |
| jjardon | oh ok | 09:45 |
| jjardon | let's merge that MR then (the freedesktop-sdk one)? valentind is anything blocking us to do that? | 09:46 |
| *** lachlan has quit IRC | 09:46 | |
| benschubert | jjardon: the tests for https://gitlab.com/freedesktop-sdk/freedesktop-sdk/merge_requests/1560 can't pass since the freedesktop-sdk CI image is unable of installing BuildStream, that was what valentind docker changes for the freedesktop CI where about | 09:47 |
| jjardon | benschubert: that's fine, the CI is not meant to pass as this is a special branch to be used by the buildstream CI, not the freedesktop-sdk CI | 09:48 |
| benschubert | oh ok, I did not know that, valentind ^ | 09:48 |
| jjardon | benschubert: feel free to push the fixes with [ci skip] on the commit message | 09:48 |
| benschubert | ok, let me update that | 09:49 |
| benschubert | jjardon: can you review and merge https://gitlab.com/freedesktop-sdk/freedesktop-sdk/merge_requests/1560 then? | 09:50 |
| jjardon | sure | 09:50 |
| *** lachlan has joined #buildstream | 09:53 | |
| benschubert | jjardon: cheers, merged. I'll continue my fixes on the overnight probably tonight :) | 09:54 |
| jjardon | benschubert: thanks a lot for the fix! I will open a MR to use that new HEAD | 09:55 |
| gitlab-br-bot | jjardon opened MR !1498 (jjardon/overnigth_yaml_bst2_chanches->master: .gitlab-ci.yml: Use new freedesktop-sdk with bst2 YAML API changes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1498 | 09:58 |
| jjardon | benschubert: when you have a second ^ | 09:58 |
| gitlab-br-bot | BenjaminSchubert approved MR !1498 (jjardon/overnigth_yaml_bst2_chanches->master: .gitlab-ci.yml: Use new freedesktop-sdk with bst2 YAML API changes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1498 | 10:00 |
| benschubert | jjardon: lgtm thanks! | 10:00 |
| jjardon | thank you :) | 10:00 |
| benschubert | This will still not fix the CI, as we need to update sbt-experimental, but that's on my todo list and is now unblocked :D | 10:00 |
| jjardon | ah, let me change the MR description then | 10:03 |
| *** ssssam[m] has quit IRC | 10:04 | |
| *** skullone[m] has quit IRC | 10:04 | |
| *** nielsdg has quit IRC | 10:04 | |
| *** rafaelff[m] has quit IRC | 10:04 | |
| *** jjardon[m] has quit IRC | 10:04 | |
| *** connorshea[m] has quit IRC | 10:04 | |
| *** waltervargas[m] has quit IRC | 10:04 | |
| *** cgmcintyre[m] has quit IRC | 10:04 | |
| *** kailueke[m] has quit IRC | 10:04 | |
| *** mattiasb has quit IRC | 10:04 | |
| *** verdre[m]1 has quit IRC | 10:04 | |
| *** abderrahim[m] has quit IRC | 10:04 | |
| *** mrmcq2u[m] has quit IRC | 10:04 | |
| *** dbuch has quit IRC | 10:06 | |
| *** albfan[m] has quit IRC | 10:06 | |
| *** Trevinho[m] has quit IRC | 10:06 | |
| *** doras has quit IRC | 10:06 | |
| *** reuben640[m] has quit IRC | 10:06 | |
| *** tlater[m] has quit IRC | 10:06 | |
| *** Demos[m] has quit IRC | 10:06 | |
| *** asingh_[m] has quit IRC | 10:06 | |
| *** segfault3[m] has quit IRC | 10:07 | |
| *** dineshdb[m] has quit IRC | 10:07 | |
| *** pro[m] has quit IRC | 10:07 | |
| *** theawless[m] has quit IRC | 10:07 | |
| *** awacheux[m] has quit IRC | 10:07 | |
| *** lachlan has quit IRC | 10:18 | |
| *** lachlan has joined #buildstream | 10:22 | |
| *** cgmcintyre[m] has joined #buildstream | 10:23 | |
| gitlab-br-bot | jjardon merged MR !1498 (jjardon/overnigth_yaml_bst2_chanches->master: .gitlab-ci.yml: Use new freedesktop-sdk with bst2 YAML API changes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1498 | 10:30 |
| *** lachlan has quit IRC | 10:48 | |
| *** lachlan has joined #buildstream | 10:55 | |
| *** lachlan has quit IRC | 10:59 | |
| benschubert | jjardon: ^ poor marge, her job taken away from here ;) | 11:03 |
| benschubert | https://gitlab.com/BuildStream/buildstream/-/jobs/258191302 we seem to have runner problems again (ran 5 times, always docker not running) | 11:08 |
| paulsherwood | if marge is retiring, shouldn't there be a party? | 11:46 |
| *** bochecha has quit IRC | 11:48 | |
| gitlab-br-bot | LaurenceUrhegyi reopened issue #994 (Further remote execution testing) on buildstream https://gitlab.com/BuildStream/buildstream/issues/994 | 11:57 |
| laurence | Hang on, we're now not using marge at all? | 12:12 |
| laurence | That's not right, surely | 12:12 |
| benschubert | paulsherwood: we should still be using it | 12:27 |
| benschubert | Sorry, my comment was about don't forget to use marge, not the contrary | 12:27 |
| *** bochecha has joined #buildstream | 12:29 | |
| paulsherwood | benschubert: ack :) | 12:33 |
| benschubert | laurence: ^^ also :) | 12:34 |
| laurence | ok, thanks | 12:36 |
| laurence | :) | 12:36 |
| tpollard | could there be a situation in which an plugins full_name is not unique? | 12:49 |
| tpollard | afaict an element's full name will always be unique, and in terms of messaging in the UI I can't think of a situation in which another type of plugin wouldn't have a unique name | 12:56 |
| *** lachlan has joined #buildstream | 13:05 | |
| *** dftxbs3e has quit IRC | 13:06 | |
| gitlab-br-bot | BenjaminSchubert opened (was WIP) MR !1493 (bschubert/optimize-loader->master: Optimize loader) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1493 | 13:14 |
| *** lachlan has quit IRC | 13:19 | |
| *** phoenix has quit IRC | 13:21 | |
| *** tiagogomes has left #buildstream | 13:24 | |
| *** tiagogomes has joined #buildstream | 13:24 | |
| Kinnison | jonathanmaw: When we last discussed the state of 1482 you told me that there were problems re: junctioned loading -- are they fixed in the current state of things? | 13:31 |
| jonathanmaw | Kinnison: probably not fixed, in the sense that while it seems like the results are correct, I haven't traced through the code to give a plausible explanation of why that is | 13:31 |
| Kinnison | jonathanmaw: Hmm okay, will you open new MRs for that and/or file issues should 1482 be merged? | 13:41 |
| jonathanmaw | Kinnison: will do | 13:44 |
| Kinnison | jonathanmaw: excellent | 13:44 |
| Kinnison | jonathanmaw: First two skims look pretty OK for 1482, I'll review for nits in a bit, but I wanted to try running it first, what should I run to give me the best taste for what it does? | 13:45 |
| jonathanmaw | Kinnison: it works best when used to look at large projects | 13:46 |
| jonathanmaw | so if you've got an extant huge debian-like project, then `bst show` in that will probably give a clear idea | 13:47 |
| Kinnison | okay cool, I have a debian project lying around | 13:47 |
| Kinnison | I see no ticker like activity during "loading elements" is that expected? | 13:49 |
| Kinnison | But I do get the many blank lines subtasks processed message | 13:49 |
| Kinnison | Okay so I see what it shows, it's a bit jarring | 13:50 |
| *** cgmcintyre[m] has quit IRC | 13:51 | |
| Kinnison | Somehow the success message doesn't render the same as other messages | 13:51 |
| Kinnison | is that deliberate? | 13:51 |
| Kinnison | jonathanmaw: ^^ | 13:54 |
| jonathanmaw | Kinnison: I'm not sure what you mean. The success messages should now print a detail section describing the number of subtasks processed | 13:56 |
| Kinnison | perhaps it's that the detail sections look jarring then to me | 13:56 |
| Kinnison | if that's "normal" for details then OK | 13:56 |
| Kinnison | :D | 13:56 |
| *** lachlan has joined #buildstream | 14:01 | |
| *** lachlan has quit IRC | 14:06 | |
| *** connorshea[m] has joined #buildstream | 14:26 | |
| *** asingh_[m] has joined #buildstream | 14:26 | |
| *** awacheux[m] has joined #buildstream | 14:26 | |
| *** abderrahim[m] has joined #buildstream | 14:26 | |
| *** albfan[m] has joined #buildstream | 14:26 | |
| *** Demos[m] has joined #buildstream | 14:26 | |
| *** doras has joined #buildstream | 14:26 | |
| *** dbuch has joined #buildstream | 14:26 | |
| *** Trevinho[m] has joined #buildstream | 14:26 | |
| *** dineshdb[m] has joined #buildstream | 14:26 | |
| *** gimpnet-irc[m] has joined #buildstream | 14:26 | |
| *** jjardon[m] has joined #buildstream | 14:26 | |
| *** kailueke[m] has joined #buildstream | 14:26 | |
| *** pro[m] has joined #buildstream | 14:26 | |
| *** m_22[m] has joined #buildstream | 14:26 | |
| *** rafaelff[m] has joined #buildstream | 14:26 | |
| *** theawless[m] has joined #buildstream | 14:26 | |
| *** ssssam[m] has joined #buildstream | 14:26 | |
| *** segfault3[m] has joined #buildstream | 14:26 | |
| *** waltervargas[m] has joined #buildstream | 14:26 | |
| *** mattiasb has joined #buildstream | 14:26 | |
| *** krichter[m] has joined #buildstream | 14:26 | |
| *** tlater[m] has joined #buildstream | 14:26 | |
| *** nielsdg has joined #buildstream | 14:26 | |
| *** skullone[m] has joined #buildstream | 14:26 | |
| *** mrmcq2u[m] has joined #buildstream | 14:26 | |
| *** tchaik[m] has joined #buildstream | 14:27 | |
| *** reuben640[m] has joined #buildstream | 14:27 | |
| *** verdre[m] has joined #buildstream | 14:27 | |
| gitlab-br-bot | marge-bot123 merged MR !1463 (aevri/pickle->master: Make ChildJobs and friends picklable) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1463 | 14:34 |
| tpollard | mmm pickles | 14:35 |
| *** lachlan has joined #buildstream | 14:35 | |
| *** lachlan has quit IRC | 14:51 | |
| *** lachlan has joined #buildstream | 15:10 | |
| *** lachlan has quit IRC | 15:19 | |
| *** tristan has quit IRC | 15:22 | |
| *** tristan has joined #buildstream | 15:51 | |
| *** lachlan has joined #buildstream | 16:05 | |
| *** phoenix has joined #buildstream | 16:10 | |
| *** lachlan has quit IRC | 16:12 | |
| *** lachlan has joined #buildstream | 16:21 | |
| *** bochecha has quit IRC | 16:24 | |
| jjardon | apologies benschubert , completely forgot about marge | 16:26 |
| benschubert | No worries:) | 16:28 |
| *** lachlan has quit IRC | 16:42 | |
| *** traveltissues has quit IRC | 16:54 | |
| *** traveltissues has joined #buildstream | 16:56 | |
| *** tpollard has quit IRC | 16:58 | |
| *** lachlan has joined #buildstream | 17:20 | |
| *** lachlan has quit IRC | 17:22 | |
| *** WSalmon has quit IRC | 17:22 | |
| *** lachlan has joined #buildstream | 17:22 | |
| *** adds68 has quit IRC | 17:22 | |
| *** laurence has quit IRC | 17:22 | |
| *** alexandrufazakas has quit IRC | 17:22 | |
| *** Becky has quit IRC | 17:22 | |
| *** samkirkham has quit IRC | 17:22 | |
| *** ikerperez has quit IRC | 17:22 | |
| *** valentind has quit IRC | 17:23 | |
| *** johnward has quit IRC | 17:23 | |
| *** bethw has quit IRC | 17:23 | |
| *** WSalmon has joined #buildstream | 17:23 | |
| *** Becky has joined #buildstream | 17:23 | |
| *** ikerperez has joined #buildstream | 17:23 | |
| *** samkirkham has joined #buildstream | 17:23 | |
| *** valentind has joined #buildstream | 17:23 | |
| *** alexandrufazakas has joined #buildstream | 17:23 | |
| *** johnward has joined #buildstream | 17:23 | |
| *** adds68 has joined #buildstream | 17:23 | |
| *** bethw has joined #buildstream | 17:24 | |
| *** laurence has joined #buildstream | 17:27 | |
| *** adds68 has quit IRC | 17:28 | |
| *** ikerperez has quit IRC | 17:28 | |
| *** samkirkham has quit IRC | 17:28 | |
| *** Becky has quit IRC | 17:28 | |
| *** alexandrufazakas has quit IRC | 17:28 | |
| *** adds68 has joined #buildstream | 17:29 | |
| *** adds68 has quit IRC | 17:31 | |
| *** adds68 has joined #buildstream | 17:32 | |
| *** WSalmon has quit IRC | 17:33 | |
| *** bethw has quit IRC | 17:33 | |
| *** jennis has quit IRC | 17:33 | |
| *** laurence has quit IRC | 17:33 | |
| *** valentind has quit IRC | 17:33 | |
| *** johnward has quit IRC | 17:33 | |
| *** bethw has joined #buildstream | 17:33 | |
| *** jennis has joined #buildstream | 17:34 | |
| *** paulsherwood has quit IRC | 17:35 | |
| *** lachlan has quit IRC | 17:35 | |
| *** valentind has joined #buildstream | 17:37 | |
| *** johnward has joined #buildstream | 17:37 | |
| *** laurence has joined #buildstream | 17:37 | |
| *** WSalmon has joined #buildstream | 17:38 | |
| *** lachlan has joined #buildstream | 17:45 | |
| *** traveltissues has quit IRC | 17:48 | |
| *** jonathanmaw has quit IRC | 18:08 | |
| *** lachlan has quit IRC | 18:17 | |
| *** phoenix_ has joined #buildstream | 18:48 | |
| *** phoenix has quit IRC | 18:49 | |
| *** phoenix_ is now known as phoenix | 18:50 | |
| *** lachlan has joined #buildstream | 19:00 | |
| *** lachlan has quit IRC | 19:25 | |
| *** phoenix has quit IRC | 21:35 | |
| *** phoenix has joined #buildstream | 21:36 | |
| *** phoenix has quit IRC | 22:00 | |
| *** traveltissues has joined #buildstream | 23:50 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!