| *** jude has joined #buildstream | 08:12 | |
| *** tristan has quit IRC | 08:29 | |
| *** tlater has joined #buildstream | 08:30 | |
| *** jonathanmaw has joined #buildstream | 08:41 | |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 1 commit (last: _scheduler/job.py: Ignore SIGTSTP error for already exited process) https://gitlab.com/BuildStream/buildstream/commit/7bd590319df360d0aac84eff72f9dabd8b389533 | 09:22 |
|---|---|---|
| *** tristan has joined #buildstream | 09:27 | |
| juergbi | tristan: do we need to do something to get the automated gnome-modulesets conversion use latest jhbuild2bst? | 09:38 |
| tristan | juergbi, it selfupdates :) | 09:38 |
| juergbi | i don't see any recent updates, though | 09:39 |
| juergbi | nothing since the semi-manual triggered one yesterday afternoon | 09:39 |
| juergbi | also nothing in tracker | 09:39 |
| juergbi | *tracked | 09:39 |
| tristan | hmmm | 09:39 |
| tristan | juergbi, are you checking the gnome7 logs, or are you checking the commit history ? | 09:40 |
| tristan | looks like something got screwed there, lemme pop in and take a look... | 09:40 |
| tristan | crap this again | 09:41 |
| juergbi | i simply checked git commit history | 09:42 |
| tristan | so sometimes (and this happened on gnome builders as well), they all get blocked on a lock file and keep queueing | 09:42 |
| tristan | looks like one process never exited | 09:42 |
| tristan | So git history will not change if something didnt change | 09:42 |
| tristan | (empty commit) | 09:42 |
| tristan | But I suppose you were expecting a change :) | 09:43 |
| * tristan restarted it anyway | 09:43 | |
| juergbi | yes i was expecting changes :) | 09:43 |
| tristan | lets wait for the cron to trigger | 09:43 |
| tristan | last time I manually launched one, maybe that was the issue | 09:43 |
| juergbi | ok | 09:43 |
| tristan | that's gonna rebuild everything :) | 09:44 |
| tristan | should be 4 and a half minutes... | 09:45 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: element.py: Remove SandboxBwrap import) https://gitlab.com/BuildStream/buildstream/commit/66b2f1f086685f1cc51b0ad966f05189bcbb70d6 | 09:57 |
| juergbi | yes, it's fine now | 10:02 |
| juergbi | tristan: is "net-attempts" (default: 3) ok as name for the scheduler config option for network retries or do you have a better proposal? | 10:05 |
| juergbi | or shall we go for something like "net-retries" where 0 would be the minimum value as retry excludes the first attempt | 10:05 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: _artifactcache/__init__.py: Remove OSTreeCache import) https://gitlab.com/BuildStream/buildstream/commit/9402db7421903b00ad3b938f34f5a989656b3e7f | 10:05 |
| tristan | Ummm | 10:18 |
| tristan | I think I like 'retries' better, both for the inclusion of valid value 0, and also because it sounds like I would understand it better than 'attempts' | 10:19 |
| tristan | or understand it faster | 10:19 |
| tristan | also maybe 'network-retries' ? | 10:19 |
| tristan | no need to abbreviate that ? | 10:19 |
| juergbi | fine by me | 10:19 |
| gitlab-br-bot | push on buildstream@network-retries (by Jürg Billeter): 2 commits (last: _scheduler/job.py: Fix respawn) https://gitlab.com/BuildStream/buildstream/commit/855b63003ee15eb8cdd42f174a9193ee7b942ef5 | 10:59 |
| gitlab-br-bot | buildstream: merge request (network-retries->master: Network retries) #67 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/67 | 11:00 |
| ironfoot | <juergbi> rm refs/heads/'gnu-toolchain/core-linux-pam/af9a8fd2d234e497bed8f863b672768ef3b612be4ebee38f73b90c8198ea9377 | 11:01 |
| ironfoot | juergbi: I've done that | 11:01 |
| * ironfoot triggers failing pipeline https://gitlab.com/baserock/definitions/-/jobs/24906839 | 11:02 | |
| gitlab-br-bot | push on buildstream@jonathan/fix-dpkg-docs (by Jonathan Maw): 1 commit (last: Fix dpkg_build and dpkg_deploy documentation using wrong script names) https://gitlab.com/BuildStream/buildstream/commit/3f91671502de2a8efafa1d7e5a959a5744c6e82f | 11:04 |
| gitlab-br-bot | buildstream: merge request (jonathan/fix-dpkg-docs->master: Fix dpkg_build and dpkg_deploy documentation using wrong script names) #68 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/68 | 11:05 |
| gitlab-br-bot | buildstream: issue #30 ("Configurable retries for network related tasks") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/30 | 11:08 |
| gitlab-br-bot | buildstream: merge request (network-retries->master: Network retries) #67 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/67 | 11:08 |
| gitlab-br-bot | push on buildstream@master (by Tristan Van Berkom): 2 commits (last: _scheduler/job.py: Fix respawn) https://gitlab.com/BuildStream/buildstream/commit/855b63003ee15eb8cdd42f174a9193ee7b942ef5 | 11:08 |
| gitlab-br-bot | buildstream: Jürg Billeter deleted branch network-retries | 11:08 |
| tristan | ironfoot, looks like different failure now ? | 11:39 |
| tristan | xml-catalog | 11:39 |
| gitlab-br-bot | push on buildstream@pushreceive (by Jürg Billeter): 2 commits (last: _artifactcache/pushreceive.py: Add handshake after sending objects) https://gitlab.com/BuildStream/buildstream/commit/330615135053957b34a250ea919ecddeb67e82b3 | 11:40 |
| tristan | ironfoot, not sure what's up with that but suspect has something to do with missing DESTDIR in environment ? | 11:40 |
| gitlab-br-bot | buildstream: merge request (pushreceive->master: Artifact push robustness improvements) #69 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/69 | 11:41 |
| tristan | ironfoot, you will recall back when we started this conversion, we agreed it was more desirable to make the scripts mention DESTDIR="%{install-root}" explicitly rather than have an env var | 11:41 |
| tristan | I suspect that needs to be handled for the xml-catalog post-install case | 11:42 |
| ironfoot | hm.. so that is not happening automatically in defs2bst? | 11:44 |
| ironfoot | that may break lot of things | 11:45 |
| tristan | ironfoot, you were there :) | 11:47 |
| tristan | you know that code ! | 11:47 |
| ironfoot | :P | 11:47 |
| ironfoot | silly past-me | 11:48 |
| tristan | ironfoot, we talked about that, so yeah it would be more reliable to do "exactly what YBD did", but we considered this was a low amount of cases which required it and it would be better to require explicit make DESTDIR="%{install-root}" install across the board | 11:48 |
| tristan | ironfoot, the converter takes care of the majority of cases, though | 11:49 |
| ironfoot | so, is the ideal solution to explicitly put DESTDIR=$DESTDIR in all the places needed? like in xml-catalog post-install script? | 11:50 |
| tristan | ironfoot, DESTDIR="%{install-root}" ./post-install.sh | 11:51 |
| tristan | in this case | 11:51 |
| tristan | I dont think there are many | 11:51 |
| ironfoot | given that this is being auto-converted from definitions.git, i was thinking of fixing it in the .morph files | 11:51 |
| tristan | I see | 11:51 |
| tristan | ironfoot, you could alternatively add an exception in defs2bst | 11:52 |
| tristan | ironfoot, or checkout what defs2bst is doing first, and coax the morph file to be something it will convert well | 11:52 |
| tristan | ironfoot, I think there is a clause where we shove PREFIX and DESTDIR in the environment, your DESTDIR=$DESTDIR thing might trigger that | 11:52 |
| ironfoot | yeah, I'd have to check that first | 11:53 |
| ironfoot | unfortunately, little time :( | 11:53 |
| * ironfoot adds a note in the relevant issue | 12:00 | |
| gitlab-br-bot | push on buildstream@jonathan/fix-dpkg-docs (by Jonathan Maw): 1 commit (last: Fix dpkg_build and dpkg_deploy documentation using wrong script names) https://gitlab.com/BuildStream/buildstream/commit/e15ca6cf49bc3f8ef79338322299978620ef6f6d | 12:08 |
| gitlab-br-bot | buildstream: merge request (jonathan/fix-dpkg-docs->master: Fix dpkg_build and dpkg_deploy documentation using wrong script names) #68 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/68 | 12:08 |
| gitlab-br-bot | push on buildstream@master (by Jonathan Maw): 1 commit (last: Fix dpkg_build and dpkg_deploy documentation using wrong script names) https://gitlab.com/BuildStream/buildstream/commit/e15ca6cf49bc3f8ef79338322299978620ef6f6d | 12:24 |
| gitlab-br-bot | buildstream: merge request (jonathan/fix-dpkg-docs->master: Fix dpkg_build and dpkg_deploy documentation using wrong script names) #68 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/68 | 12:24 |
| gitlab-br-bot | buildstream: Jonathan Maw deleted branch jonathan/fix-dpkg-docs | 12:24 |
| gitlab-br-bot | push on buildstream@jonathan/dpkg_no_debiandir (by Jonathan Maw): 1 commit (last: dpkg_build.yaml: Remove useless debiandir from artifacts) https://gitlab.com/BuildStream/buildstream/commit/a56d9efba5a0095234ee5c26b7bb80c3eaa11958 | 13:48 |
| gitlab-br-bot | buildstream: merge request (jonathan/dpkg_no_debiandir->master: dpkg_build.yaml: Remove useless debiandir from artifacts) #70 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/70 | 13:54 |
| gitlab-br-bot | buildstream: merge request (jonathan/dpkg_no_debiandir->master: dpkg_build.yaml: Remove useless debiandir from artifacts) #70 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/70 | 14:04 |
| gitlab-br-bot | push on buildstream@master (by Tristan Van Berkom): 1 commit (last: dpkg_build.yaml: Remove useless debiandir from artifacts) https://gitlab.com/BuildStream/buildstream/commit/a56d9efba5a0095234ee5c26b7bb80c3eaa11958 | 14:04 |
| gitlab-br-bot | buildstream: Jonathan Maw deleted branch jonathan/dpkg_no_debiandir | 14:04 |
| gitlab-br-bot | push on buildstream@pushreceive (by Jürg Billeter): 2 commits (last: _artifactcache/pushreceive.py: Add handshake after sending objects) https://gitlab.com/BuildStream/buildstream/commit/f44da295a33f09a9294d6aa0b7782361cd7d288c | 14:10 |
| gitlab-br-bot | buildstream: merge request (pushreceive->master: Artifact push robustness improvements) #69 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/69 | 14:10 |
| *** tiagogomes has quit IRC | 14:10 | |
| gitlab-br-bot | buildstream: merge request (pushreceive->master: Artifact push robustness improvements) #69 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/69 | 14:25 |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 2 commits (last: _artifactcache/pushreceive.py: Add handshake after sending objects) https://gitlab.com/BuildStream/buildstream/commit/f44da295a33f09a9294d6aa0b7782361cd7d288c | 14:25 |
| gitlab-br-bot | buildstream: Jürg Billeter deleted branch pushreceive | 14:25 |
| tristan | ironfoot, can you upgrade buildstream on the baserock artifact server again ? | 14:53 |
| gitlab-br-bot | push on buildstream@master (by Tristan Van Berkom): 1 commit (last: Updated man pages.) https://gitlab.com/BuildStream/buildstream/commit/332fed488a8b8af995dbd346654b157b75846adf | 15:05 |
| ironfoot | tristan: should be done | 15:10 |
| tristan | ironfoot, great, there will be an error if old versions of buildstream try to push to that server now | 15:10 |
| tristan | so any builders, need to also be updated | 15:11 |
| ironfoot | gitlab jobs will use latest | 15:11 |
| tristan | yep | 15:12 |
| gitlab-br-bot | buildstream: issue #32 ("bst build does not handle elements with Consistency.INCONSISTENT") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/32 | 15:15 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: WIP: tarcache.py: Speed up tar cache extraction) https://gitlab.com/BuildStream/buildstream/commit/5892da9afcb200f6e12a8f25bce262b7c2cb9ae4 | 15:28 |
| *** tristan has quit IRC | 16:17 | |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: tarcache.py: Speed up tar cache extraction) https://gitlab.com/BuildStream/buildstream/commit/323679b3c8f8d4f240e6a58f4c378e815e04dd0f | 16:21 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: tarcache.py: Speed up tar cache extraction) https://gitlab.com/BuildStream/buildstream/commit/70e8b24e38f0b313b7fea7b77eac746aa25eb802 | 16:29 |
| *** tristan has joined #buildstream | 16:41 | |
| *** tlater has quit IRC | 17:09 | |
| *** jonathanmaw has quit IRC | 17:33 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!