09:16Hazematman: Hey, is anyone here familar with minigbm? I'm trying to build it with support for amdgpu (roughly following the below link) but during the build it can't find defines for `__DRI_IMAGE_FORMAT_*`. Looking at the `dri_interface.h` header it says those are not exported anymore. Is building minigbm wth amdgpu support even supported anymore?
09:16Hazematman: https://www.collabora.com/news-and-blog/blog/2017/11/27/running-chromium-with-ozone-gbm-on-gnu-linux/
10:16daniels: Hazematman: https://gitlab.freedesktop.org/olv/hbm/-/tree/main is the replacement there
10:17Hazematman: Thanks daniels! I'll try using that instead
10:17daniels: np :)
13:05mlankhorst: tzimmermann: are you doing misc-fixes or is it mine now?
13:05tzimmermann: mlankhorst, i'd do the PR today
13:05tzimmermann: ok?
13:06mlankhorst: Ok. :)
13:48tzimmermann: airlied, sima, looks like last week's PR of drm-misc-fixes did not get merged. can you take care of it? drm-misc-fixes-2024-11-21
13:49sima: tzimmermann, airlied said earlier that he'll do a final -rc1 pr fri (australia tz) so should get included then
13:50tzimmermann: ok np
13:50sima: tzimmermann, I guess if you have more -fixes for -rc1 just do another incremental pull, and when airlied merges them in-order it should sort itself out
13:50sima: incremental pull as in only type up the summary for the new additional patches, not everything
13:50tzimmermann: ok
13:51sima: and maybe add a note to the pr mail that if airlied pulls that before the earlier one, he needs to include both summaries in the merge commit msg
13:51tzimmermann: sure
14:52emersion: soreau: i completely forgot your mesa patch ð«£
15:16mareko: jenatali: right, it doesn't fix it
15:20soreau: emersion: Apparently not :D
15:24jenatali: :(
15:25soreau: emersion: this can probably be closed https://gitlab.freedesktop.org/mesa/mesa/-/issues/12039
15:26emersion: it should get closed automatically once the MR is merged i think
15:56soreau: ð
16:47alyssa: mareko: I see i/o is getting close to passing ci (yay!!!), I guess now would be a good time for me to review st_nir_unlower_io_to_vars?
16:47alyssa: (at this point it seems pretty unlikely that the remaining bugs are specifically that pass)
17:09mareko: alyssa: up to you, workarounds for specific drivers might still be added there (there are already a few)
17:12alyssa: ah, alright
17:13alyssa: relatedly, I think with !31942, asahi goes vars -> lowered -> vars -> lowered
17:13alyssa: but other than that being rather silly, is there a benefit in reworking things to consume lowered directly and skip the new unlower pass?
17:14alyssa: (back when I wanted to in 2022 I couldn't because mesa/st choked on lowered I/O but you've since fixed that :) )
17:16mareko: alyssa: likely no benefit other than a cleanup
17:17alyssa: ack
17:37mareko: any idea why virgl-traces doesn't upload any results? https://gitlab.freedesktop.org/mesa/mesa/-/jobs/67380870
17:40mareko: eric_engestrom: ^^
19:05airlied: sima: fixes is in the fixes hole, where I just ignore it until rc1, as otherwise I get a messy backmerge