02:48 <marc2377-2> Test

03:03 <marc2377> test

05:40 <DavidHeidelberg> eric_engestrom: Heya. I'm seeing windows_msvc taking it's time.. while blocking the build-only jobs. Maybe win jobs could not block the transition to another stage?

05:41 <DavidHeidelberg> anyway, the job just got stuck, so most likely this is not occurring often

05:52 <eric_engestrom> the logic is "run the build jobs that the tests need first, and then the others", so by that logic the windows build used by windows tests should be included... but it was needed for shared runners, which windows isn't really, so perhaps we can remove it from `.build-for-tests-jobs` so that it still runs first but the build-only jobs don't wait on it anymore

05:52 <eric_engestrom> DavidHeidelberg: ^

05:55 <DavidHeidelberg> yeah, if I understand right, load on Windows runners doesn't involve cutting performance from Linux ones

05:55 <DavidHeidelberg> so it should be safe I hope

05:58 <eric_engestrom> yeah

05:58 <eric_engestrom> typing the commit now

06:06 <eric_engestrom> DavidHeidelberg: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31473

06:54 <daniels> we have much more than a single runner tho right? Microsoft added more …

09:06 <DragoonAethis> Looks like Patchwork is unstable over the last day or so (timeouts, 502s), could anyone take a look at it?

13:01 <eric_engestrom> daniels: this is about the number of potentially parallel build jobs per runner within a pipeline, and for windows there's only one build job

13:01 <eric_engestrom> whereas for linux we have a ton that have to be ordered somehow, and the current solution is the best I could come up with

13:03 <eric_engestrom> daniels, DavidHeidelberg: is there an issue with the a618 runners? they seem unable to pick up jobs and stay in "Waiting for job 15901830 to start." forever

13:04 <eric_engestrom> see https://gitlab.freedesktop.org/mesa/mesa/-/jobs/64527042 and the rest in the right column

13:06 <DavidHeidelberg> eric_engestrom: yes, https://lava.collabora.dev/scheduler/device_type/sc7180-trogdor-lazor-limozeen

13:06 <DavidHeidelberg> ping gallo[m]

14:12 <daniels> kicked them into healthcheck

14:14 <eric_engestrom> thanks!

14:17 <daniels> (we had some kind of spectacular network blowup)

15:53 <jrayhawk> i am restarting kemper for a kernel upgrade to see if i can stop NFS from hanging there, too