06:05 <tomeu> MoeIcenowy: I will be getting soon a h64 for doing just that

07:26 <MoeIcenowy> tomeu: thanks

07:55 <tomeu> alyssa: the misnomers list is very cool, btw

13:54 <tomeu> alyssa: 18040afddada ("panfrost/midgard: Route new blending intrinsics") seems to break at least dEQP-GLES2.functional.shaders.random.all_features.fragment.0

14:22 <tomeu> alyssa: just found out that with the wallpaper jobs, I got this assertion in pandecode:

14:23 <tomeu> pandecode: ../mesa/src/gallium/drivers/panfrost/pandecode/decode.c:1544: pandecode_replay_vertex_tiler_postfix_pre: Assertion `stride_set == clamped_stride' failed.

14:23 <tomeu> if the wallpaper jobs aren't emitted, I don't get it and the decode is successful

14:23 <tomeu> any ideas?

14:25 <alyssa> tomeu: The joke ofc is that the list named "misnomers" is itself a misnomer :)

14:25 <alyssa> I don't remember that assert, let's see

14:26 <alyssa> ...Oh that's from code I wrote like last week, *blush*

14:26 <tomeu> looks recent, indeed :)

14:26 <tomeu> btw, this is on a veyron

14:26 <tomeu> maybe I should test on the nanopc as well

14:26 <alyssa> Oh! That changes things :)

14:27 <alyssa> tomeu: Basically what I was trying to say was "the field is u64-bit but stride is i32-bit, so assert to make sure we're not using the upper 32-bits"

14:28 <alyssa> But on Veyron the feild is u32, so that's likely bogus :)

14:28 <tomeu> yep, I have started a build for the nanopc, will check tomorrow

14:28 <alyssa> Nanopc is what hw?

14:28 <tomeu> rk3399

14:29 <alyssa> Gotcha

23:58 <hanetzer> blep