| 19 Oct 2025 |
 ElvishJerricco | and the body can say "y'all deleted the code that used to do that" :P | 19:52:51 |
 dramforever | do you have any idea if we can work around this better than "just ignore the test for now"? | 19:54:24 |
| ElvishJerricco | looks like this bug report is the same problem: https://github.com/systemd/systemd/issues/39043
and is fixed by: https://github.com/systemd/systemd/pull/39071
Haven't looked at that code yet to make sure it would actually help us | 19:54:54 |
| ElvishJerricco | anyway we can workaround it by checking an actual dev node, not a static dev node, in this test | 19:55:12 |
| dramforever | but the bug is still there... | 19:56:02 |
| ElvishJerricco | lol the PR just adds ACL code back to logind | 19:56:07 |
| dramforever | do we just ignore that fact | 19:56:12 |
| ElvishJerricco | what do you mean? | 19:56:35 |
| ElvishJerricco | I think that fixes the bug? | 19:56:38 |
| ElvishJerricco | it's just not in 258 | 19:56:44 |
| ElvishJerricco | it's in 258.1 | 19:56:49 |
| dramforever | yeah but that rebuilds systemd right | 19:56:49 |
| ElvishJerricco | right, we can't do the patch this cycle | 19:56:59 |
| ElvishJerricco | so we'll do a workaround | 19:57:04 |
| dramforever | so if we merge staging-next we ship this bug to users | 19:57:11 |
| dramforever | and given the links this breaks real use cases kinda bad | 19:57:22 |
| dramforever | * and given the links in the pr thread this breaks real use cases kinda bad | 19:57:32 |
| ElvishJerricco | yea, I just don't think we can actually fix it, right? | 19:57:58 |
| ElvishJerricco | like we can fix the test by doing something dumb, but I don't think we can actually fix the bug without a systemd patch | 20:00:28 |
| dramforever | can we add a udev rule to fix this | 20:01:38 |
| ElvishJerricco | no because these aren't devices in /sys, right? | 20:01:54 |
| dramforever | oh, right, because static_node rules are basically magic | 20:02:43 |
| ElvishJerricco | ok yea, confirmed, the PR that's in 258.1 fixes it | 20:12:05 |
| ElvishJerricco | dramforever: do you agree that there's not a reasonable way to fix the underlying bug, other than to just update to 258.1? If so I'm just going to change the test to check a non-static dev node, like /dev/dri/card0, with a note to change it back after 258.1 | 20:13:07 |
| dramforever | when do we expect 258.1 to land? it's the next staging-next right? | 20:29:19 |
| dramforever | on the release schedule | 20:29:28 |
| dramforever | well basically, what is this line?
2025-11-01 master (Day before ZHF) Merge staging-next into master, prepare for ZHF. Begin first staging-next cycle | 20:31:10 |
| dramforever | is it the staging-next we're doing now? | 20:31:18 |
| dramforever | or the next one? | 20:32:19 |
| ElvishJerricco | yea | 20:34:39 |
