| *** benschubert has quit IRC | 00:05 | |
| *** tristan has joined #buildstream | 05:18 | |
| *** ChanServ sets mode: +o tristan | 05:18 | |
| *** WSalmon has quit IRC | 05:44 | |
| *** WSalmon has joined #buildstream | 05:46 | |
| gitlab-br-bot | tristanvb closed issue #1240 (Commands to disable remote cache are being ignored / they do not work) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1240 | 06:19 |
|---|---|---|
| gitlab-br-bot | tristanvb closed issue #1396 (Add a way to force a local build) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1396 | 06:19 |
| tristan | Lets land this pronto: https://gitlab.com/BuildStream/website/-/merge_requests/144 | 07:04 |
| tristan | How do I run the src/buildstream/testing/_sourcetests/track_cross_junction.py tests ? | 07:27 |
| tristan | Maybe just tox -e py38-nocover -- --integration src/buildstream/testing/_sourcetests/track_cross_junction.py | 07:28 |
| * tristan tries | 07:28 | |
| tristan | nope | 07:29 |
| tristan | Thats some weird thing going on there which needs to be fixed iirc :-S | 07:29 |
| tristan | --plugins | 07:31 |
| tristan | yeah... /me can't wait for ben to set that straight... better lend a hand on this threading stuff so that can happen sooner... | 07:31 |
| tristan | And how do you selectively run one of the `--plugins` tests ? | 07:43 |
| *** toscalix has joined #buildstream | 07:56 | |
| *** toscalix has quit IRC | 08:35 | |
| *** santi has joined #buildstream | 08:47 | |
| tristan | It looks like we're not getting the START/SKIPPED messages one might expect for a task where the Queue.status() method returned QueueStatus.SKIP | 08:55 |
| juergbi | tristan: yes, for tasks that are skipped before even creating a job, there is no log message. for tasks that start a job and then set status as skipped, we get START+SKIPPED | 08:57 |
| juergbi | it can cut down noise but is somewhat inconsistent | 08:58 |
| tristan | juergbi, Would you be against adding a START/SKIPPED pair for that ? | 08:58 |
| juergbi | I'm not sure. it may make sense to discuss this as part of my WIP move of cache query into jobs | 08:58 |
| tristan | Hmmm | 08:59 |
| juergbi | we have to think about what logging is best there by default | 08:59 |
| tristan | Ok, I'll set that aside for now | 08:59 |
| tristan | Right, had part of that discussion in context of my other branch | 08:59 |
| tristan | with Ben, you probably noticed | 08:59 |
| juergbi | not sure, might not have read that in detail | 09:00 |
| juergbi | tristan: btw: !2073 is the filter fix with regression test I mentioned yesterday. is this ok for you? | 09:00 |
| gitlab-br-bot | MR !2073: filter.py: Combine integration commands in assemble() https://gitlab.com/BuildStream/buildstream/-/merge_requests/2073 | 09:01 |
| juergbi | (ignoring potential removal of dynamic public data or additional public data pass through for now) | 09:01 |
| tristan | juergbi, Quick review, don't understand the test; but I really gotta run straight away... | 09:09 |
| tristan | class starting in 20 | 09:09 |
| juergbi | ok, thanks | 09:09 |
| * tristan commented on mr... | 09:10 | |
| tristan | https://gitlab.com/BuildStream/buildstream/-/merge_requests/2072 is about ready now, should finally pass it's tests | 09:11 |
| tristan | and ensures consistent skip message behavior for sources which don't implement track() (like workspaces or local or patch sources), and adds a skip test | 09:12 |
| *** tristan has quit IRC | 09:16 | |
| *** tristan has joined #buildstream | 09:23 | |
| *** ChanServ sets mode: +o tristan | 09:23 | |
| *** tristan_ has joined #buildstream | 09:29 | |
| *** ChanServ sets mode: +o tristan_ | 09:29 | |
| *** tristan has quit IRC | 09:31 | |
| *** tomaz has joined #buildstream | 09:49 | |
| tomaz | people, a pipeline failed for me with: " OSError: [Errno 24] Too many open files: " | 09:50 |
| tomaz | this is something that buildstream should take care? the number of open files on the OS? | 09:51 |
| juergbi | tomaz: do you have a backtrace? is this with 1.x or master? | 10:01 |
| tomaz | juergbi: pipeline is running again, Ill paste the backtrace as soon as it fails. | 10:02 |
| tomaz | it's on 1.x | 10:02 |
| tomaz | freedesktopsdk is still on 1.x | 10:03 |
| juergbi | on 1.x we increase the process file limit as high as the system allows as pyfuse safehardlinks can need a lot of file descriptors | 10:03 |
| juergbi | if this still is insufficient, the system may have a low hard limit (it's configurable in Linux) | 10:04 |
| juergbi | there may be further improvements to this aspect that could be done in 1.x, however, master should no longer have this issue at all with buildbox-fuse | 10:05 |
| gitlab-br-bot | tristanvb approved MR !2072 (tristan/remove-bst-key-requires-stage->master: source.py: Remove BST_KEY_REQUIRES_STAGE) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/2072 | 11:11 |
| gitlab-br-bot | marge-bot123 merged MR !2072 (tristan/remove-bst-key-requires-stage->master: source.py: Remove BST_KEY_REQUIRES_STAGE) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/2072 | 11:11 |
| tristan_ | Are we okay to merge https://gitlab.com/BuildStream/buildstream/-/merge_requests/2070 without additional tests in advance of deciding on logging modes ? | 11:15 |
| tristan_ | I think so, but would like some review on the branch if possible | 11:15 |
| tristan_ | just me or the pipelines view on gitlab is totally borked | 11:51 |
| tristan_ | https://gitlab.com/BuildStream/buildstream/-/pipelines/193973435 | 11:51 |
| tristan_ | can't see the jobs correctly at the bottom | 11:51 |
| *** tristan_ has quit IRC | 12:06 | |
| *** tristan_ has joined #buildstream | 12:07 | |
| *** ChanServ sets mode: +o tristan_ | 12:07 | |
| gitlab-br-bot | juergbi approved MR !2073 (juerg/filter-pass-integration->master: filter.py: Combine integration commands in assemble()) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/2073 | 13:57 |
| gitlab-br-bot | marge-bot123 merged MR !2073 (juerg/filter-pass-integration->master: filter.py: Combine integration commands in assemble()) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/2073 | 13:57 |
| *** tristan_ has quit IRC | 14:11 | |
| *** tristan_ has joined #buildstream | 14:11 | |
| *** ChanServ sets mode: +o tristan_ | 14:11 | |
| gitlab-br-bot | juergbi opened MR !2074 (juerg/artifact-pull->master: Artifact pull code improvements) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/2074 | 14:58 |
| douglaswinship | tristan_: i'm having a lot of trouble with GitLab too. I found it could often get it to reset if I collapsed the sidebar, and then re-expanded it | 15:23 |
| *** santi has quit IRC | 17:38 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!