07:01tzimmermann: jani, ack on the latest dumb-buffer fix: https://lore.kernel.org/intel-gfx/20260602112842.252279-1-tzimmermann@suse.de/T/#m2fa9830cb1fd8801a7c0f1b5893eae956fbe39e1 ? dows that fix the CI?
07:40mgrz: mlankhorst: / tzimmermann: https://lore.kernel.org/dri-devel/20260522135520.1862848-1-michal.grzelak@intel.com got reviewed, can it be merged via drm-misc/drm-misc-next? Asking since I don't have commit rights to any of drm-* repositories
08:13mahkoh: Is there a way to figure out which submission/process the fence in `amdgpu 0000:03:00.0: Fence fallback timer expired on ring comp_1.0.0` belongs to?
08:51tzimmermann: mgrz, thanks for the fixes. I'll add them to drm-misc-next
13:10hakzsam: anholt: as you are the deqp-runner (rust) author, is there a way to know the list of tests that were executed ? like if deqp-runner stops in a middle of a run when a GPU hang happens
13:18glehmann: alyssa: I assume you assigned the wrong MR to marge?
13:19alyssa: glehmann: yes, I noticed a minute later but you were 18 seconds faster :p
13:20zmike: nothing gets past glehmann.
13:22alyssa: I just tweaked my tooling to make this stop happening, since twice in a month means a tooling failure
13:22alyssa: sorry about that
13:23zmike: everyone makes mistakes
13:23alyssa: (meanwhile if I don't use the tool I'll request review from Marge, so.... gitlab...)
13:30jannau: so Marge should just lgtm and assign the merge request to herself
13:30zmike: if only
13:34alyssa: Sashiko and Marge are dating
14:32MrCooper: karolherbst: mesa 8711a88d965f ("rusticl/util: add Traits to help with usage of CString") broke the src/gallium/frontends/rusticl/libmesa_rust_util_test build target here, known issue?
14:33karolherbst: huh.. it shouldn't, I even build with tests
14:33karolherbst: what's the error?
14:33karolherbst: also CI would have complaiend
14:34MrCooper: https://paste.debian.net/hidden/8cb8ff68
14:35MrCooper: wiping the build directory didn't help
14:36MrCooper: also tried switching rust_std to 2024, that breaks src/gallium/frontends/rusticl/librusticl_test.a though
14:38karolherbst: MrCooper: what's the version of rustc?
14:38MrCooper: 1.94.1
14:38MrCooper: does it require 1.95?
14:39karolherbst: no, 1.85 rather
14:39karolherbst: could be an issue only visible on newer rustc
14:40karolherbst: I'll check that out
14:41karolherbst: gfxstream doesn't build..
14:42karolherbst: ahh yeah https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41978
14:48karolherbst: MrCooper: yeah.. so it fails on stable and compiles on 1.85
14:51karolherbst: MrCooper: please test https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41998
14:56MrCooper: does the trick, thanks
15:53alyssa: robclark: did we get consensus on https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41801 ?
15:54alyssa: Maybe sensible to add the countof() polyfill and converting code I own, while leaving ARRAY_SIZE in and letting other parts of the tree migrate at their leisure?
15:55alyssa: or maybe just close the MR and revisit in 5 years since we both have actual work to do :p
16:02karolherbst: can we clean up our atomics next 🙃
20:01robclark: alyssa: you could leave a #define ARRAY_SIZE() countof()... or not, I can adapt