#asahi-gpu · 2026-02-28 07:56
ffxoh, in the firefox source, ok#asahi-gpu · 2026-02-28 07:56
ffxough. ok. thanks :) i really have no idea about linux yet. can see the reason for that?#asahi-gpu · 2026-02-28 07:55
jannauwebrender_compositor seems to be blocked on linux in general: https://searchfox.org/firefox-main/source/widget/gtk/GfxInfo.cpp#1023#asahi-gpu · 2026-02-28 07:54
ffxcould be even that compositor is used for putting a font onto any kind of background, it is a very imprtant base component in the browser. if its running on cpu, the browser cant be very fluid for anything drawn in realtime#asahi-gpu · 2026-02-28 07:52
ffxwebgpu actually runs smooth in asahi firefox#asahi-gpu · 2026-02-28 07:52
ffxbut later i found out that its stuttering because of the compositor, not webgpu#asahi-gpu · 2026-02-28 07:52
ffxthat was me, too#asahi-gpu · 2026-02-28 07:48
jannauah, my fault. someone was also looking into webgpu#asahi-gpu · 2026-02-28 07:48
jannauyou can't really compare gpu userspace features to macos since it's will be a different implementation due to metal vs. opengl/vulkan#asahi-gpu · 2026-02-28 07:47
ffxWEBRENDER is related to the normal canvas element, afaik#asahi-gpu · 2026-02-28 07:46
ffxi dont think that this is a "new" feature#asahi-gpu · 2026-02-28 07:46
ffxit seems to run on cpu fallback then#asahi-gpu · 2026-02-28 07:46
ffxif you stack alpha layers or so#asahi-gpu · 2026-02-28 07:46
ffxno, its not about webgpu, its about the normal compositing of layers#asahi-gpu · 2026-02-28 07:45
jannauI think it's entirely possible that firefox' webgpu support isn't finished and that it is disabled for reasons#asahi-gpu · 2026-02-28 07:43
ffxits normally avaialble under macos. So that's most likely the reason for the stuttering#asahi-gpu · 2026-02-28 07:42
ffxi ve came now to a new info, using about:support, it tells me that WEBRENDER_COMPOSITOR is blocklisted by gfxInfo#asahi-gpu · 2026-02-28 07:41
ffxoh ok#asahi-gpu · 2026-02-28 07:40
jannauonly radeonsi, intel and zink implement EXT_semaphore(_fd)? in mesa#asahi-gpu · 2026-02-28 07:39
jannaucalling a implementation broken because it doesn't implement an extension seems a little over the top. before EXT_semaphore_fd it needs needs EXT_semaphore first.#asahi-gpu · 2026-02-28 07:11
ffxBut i could be wrong totally.#asahi-gpu · 2026-02-28 07:11
ffxthis is printed by firefox into console at start, and I would assume it means that the asahi gpu driver needs to implement "EXT_semaphore_fd", to make modern browsers capable of gpu/cpu syncing...#asahi-gpu · 2026-02-28 07:10
ffxEXT_semaphore_fd is not supported" indicates a failure in your system's graphics stack, specifically regarding how the browser or application communicates with your graphics driver to synchronize GPU tasks (using EXT_semaphore_fd).#asahi-gpu · 2026-02-28 07:10
ffxHi, I am trying to investigate the cause of canvas render stuttering in Firefox and all the browsers with asahi... I came down to this result:#asahi-gpu · 2026-02-27 07:47
ffxi guess i reported some graphical issue then at the wrong place? https://github.com/AsahiLinux/linux/issues/459#asahi-gpu · 2026-02-27 07:46
ffxoh hi#asahi-gpu · 2026-02-22 15:03
alyssathat's the classic way to force a particular register pressure of your choosing#asahi-gpu · 2026-02-22 15:02
alyssaint temp0, ...., tempN; for(..) { temp0 = temp1; temp1 = temp2; ... tempN-1 = tempN; tempN = temp0; } foo(tempN)#asahi-gpu · 2026-02-22 15:02
alyssaThat you can confirm by varying the number of registers used, which is easy with an unrollable loop#asahi-gpu · 2026-02-22 15:01
alyssaPotentially one of the instructions adapts "USC Registers" - in this case, dynamically allocating GPRs from the cache for the rest of the program to use#asahi-gpu · 2026-02-22 15:00
alyssaIf you wanted to confirm that hypothesis, compare against shaders that use threadgroup memory (and different amounts of it)#asahi-gpu · 2026-02-22 15:00
alyssaBut if I did have to make a guess, one of those instructions does the equivalent of the M1 "USC Shared" struct.#asahi-gpu · 2026-02-22 14:59
alyssaI wouldn't worry about it too much#asahi-gpu · 2026-02-22 14:57
alyssaLikely the unknown instructions relate to dynamic caching / setting up other bits of state that traditionally USC control words managed but is now shader managed#asahi-gpu · 2026-02-22 14:56
alyssaAlso this looks like a preamble shader, which on M1 are independent binaries from the main shader, not sure if they merged that in M3#asahi-gpu · 2026-02-22 14:55
alyssaThe first two moves are hardbaking the address of the data to plumb in, which those last 4 moves (from the loaded data into the uniform registers) are dealing with#asahi-gpu · 2026-02-22 14:55
alyssaPresumably that mov.todo is a tail call#asahi-gpu · 2026-02-22 14:54
alyssaBased on the snippet you linked, the solution is pretty simple: dynamically generate a prolog that uses shader instructions to setup the state that would previously be done with USC control words.#asahi-gpu · 2026-02-22 14:53
alyssaBut that presents a chicken & egg problem - how do you get kernel arguments and things like that into the shader if you don't have the sideband?#asahi-gpu · 2026-02-22 14:52
alyssauniform registers are the spiciest - Likely they still exist in the form of something like AMD's SGPRs but you can't use them the traditional m1 way because that part of the control stream is gone#asahi-gpu · 2026-02-22 14:52
alyssaNot sure what hapens to coeff registers#asahi-gpu · 2026-02-22 14:52
alyssaTexture/sampler state regs likely go away by improving hw bindless support#asahi-gpu · 2026-02-22 14:51
alyssaWith M3, all of that is gone due to dynamic caching.#asahi-gpu · 2026-02-22 14:51
alyssaThe driver puts the right control words and then the program has the environment accessible easily.#asahi-gpu · 2026-02-22 14:51
alyssauniform registers, texture state registers, sampler state registers, coefficient registers#asahi-gpu · 2026-02-22 14:50
alyssaOn M1, there are special control words to load special registers with data from memory#asahi-gpu · 2026-02-22 14:50
alyssaIf i had to take a guess --#asahi-gpu · 2026-02-22 14:50
alyssapac85: those shaders would be fine here I believe#asahi-gpu · 2026-02-22 14:47
alyssagood work#asahi-gpu · 2026-02-22 14:47
alyssaoliver: that sounds believable