| nanonyme | So did I check wrong or was https://github.com/apache/buildstream/pull/1570 the first PR after which all PR tests started failing? | 17:12 |
|---|---|---|
| nanonyme | Looks like it was merged with failing CI | 17:13 |
| nanonyme | I just created this empty PR against master, observing what happens https://github.com/apache/buildstream/pull/1593 | 17:22 |
| nanonyme | Last line seen at https://github.com/apache/buildstream/runs/5265739784?check_suite_focus=true tests/integration/artifact.py::test_cache_buildtrees PASSED [ 58%] | 17:37 |
| nanonyme | Seems to have broken after that. https://github.com/apache/buildstream/runs/5265739767?check_suite_focus=true is also broken but I didn't have tab open when it failed | 17:37 |
| nanonyme | https://github.com/apache/buildstream/runs/5265739802?check_suite_focus=true now also at tests/integration/artifact.py::test_cache_buildtrees PASSED [ 58%], not clear if it hanged yet | 17:42 |
| nanonyme | It looks probable though | 17:44 |
| nanonyme | So the first test that is not run based on console is test_preserve_environment | 17:45 |
| nanonyme | @abderrahim:gnome.org: do you think this is coincidence? | 17:54 |
| nanonyme | Hmm, it cannot actually be the composite work. I was accidentally using a lot older commit as base | 17:57 |
| nanonyme | juergbi tristan_ can you make PR with no content (git commit --allow-empty -m "Empty commit") from same repo, not fork? I want to see if it will fail same way or not | 17:58 |
| nanonyme | It still hangs https://github.com/apache/buildstream/runs/5267010113?check_suite_focus=true even when all parallelity is removed from the CI | 22:21 |
| nanonyme | At more or less same place | 22:22 |
| nanonyme | Next stop testing if that parallel fix didn't end up in a memory leak | 22:22 |
| nanonyme | I mean, not-parallel fix | 22:23 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!