#asahi-gpu · 2022-10-28 04:34
phireupstream linux, or upstream rust-on-linux?#asahi-gpu · 2022-10-28 04:33
phirefair#asahi-gpu · 2022-10-27 22:02
phireWould allow you pass it around (and even compare) as if it was a 64bit int, but then the compiler can quickly unpack it into it's components just by just treating it as two 32bit ints#asahi-gpu · 2022-10-27 21:59
phireLina: consider using a soft float working format of 64bit, sign in bit63, 8/16 bits of exponent at bit 32+, and 32bits of signficand in bits 0-31#asahi-gpu · 2022-10-27 21:54
phireThe preformance isn't even that bad (when you inline it and avoid unpacking/packing into the proper ieee754 format)#asahi-gpu · 2022-10-27 21:45
phireThough, I do really like the "let's just do soft float" solution. People are too scared of soft float #asahi-gpu · 2022-10-27 21:44
phirestd::sync:: atomic::compiler_fence() // for rust#asahi-gpu · 2022-10-27 21:44
phireasm volitile("" ::: "memory") // for c#asahi-gpu · 2022-10-27 21:44
phireSurely there is a way to (safely) prevent reordering; compiler Instruction reordering barrier?#asahi-gpu · 2022-10-27 21:44
phireI said this on Twitter but I'll repeat it here:#asahi-gpu · 2022-10-27 05:00
phirem1 max seems to just set all those accleration pointers as soon as the depth buffer is enabled#asahi-gpu · 2022-10-27 03:26
phire"no it's not?" actually, you might be right. opengl makes it near impossible to tell the unresolved MSAA buffer won't be used again later#asahi-gpu · 2022-10-27 03:23
phireI have no idea how smart apple's opengl implementation is, and I assume it's all very explict in metal#asahi-gpu · 2022-10-27 03:22
phireeven when not using the EXT_ms_rtt extention, that's still something a smart opengl implemenation can infer from the commands and optimise in. #asahi-gpu · 2022-10-27 03:13
phirecya alyssa#asahi-gpu · 2022-10-27 03:12
phiredepth-test and triangle edeges are per sample. color is duplicated over all samples (masked by depth/stencil/rasteriation)#asahi-gpu · 2022-10-27 03:11
phireyes, multiple depth samples#asahi-gpu · 2022-10-27 03:11
phirecould the ultra have forced MSAA on for some reason?#asahi-gpu · 2022-10-27 03:09
phireI just really doubt paranoia is the justification. diffrent buffer layouts (store also resolving MSAA being the best example) is a much better reason#asahi-gpu · 2022-10-27 03:06
phireif apple think Metal is a safety boundary....#asahi-gpu · 2022-10-27 03:05
phirewait, these are all supplied from userspace, so app has control over the partial anyway#asahi-gpu · 2022-10-27 03:05
phiresafe? that's very paranoid. but I guess#asahi-gpu · 2022-10-27 03:04
phireyes, but if you are doing that, might as well use the same buffer for both partial and store. It's only if the final store buffer has a diffrent layout for some reason that you need seperate buffers#asahi-gpu · 2022-10-27 03:03
phireoh... multisample resolve?#asahi-gpu · 2022-10-27 03:03
phirethat's what I was already wondering. though why is it so keen to avoid writing to the store buffer until the end?#asahi-gpu · 2022-10-27 02:54
phireand framebuffer compresion *only* cares about reducing memory bandwidth, not buffer size#asahi-gpu · 2022-10-27 02:53
phireah, makes sense. so you can fallback to uncompressed if compression fails to produce a smaller result#asahi-gpu · 2022-10-27 02:52
phireso that means the buffer has the same layout when compressed? #asahi-gpu · 2022-10-27 02:51
phirealso, this is a waste of memory, having seperate buffers for the TVB overflows and the final depth#asahi-gpu · 2022-10-27 02:51
phirehmmm.... weird#asahi-gpu · 2022-10-27 02:49
phireunless... what if there is a hardware bug loading non-compressed depth/stencil buffers, and this is a workaround?#asahi-gpu · 2022-10-27 02:49
phirebut hardware bug would be weirder. how would the PBE know the diffrence between a final store and a TVB flush store?#asahi-gpu · 2022-10-27 02:48
phireagx firmware bug? I'd be supprised if it was a hardware bug#asahi-gpu · 2022-10-27 02:45
phireseperate control over TVB overflow and final depthbuffer is a useful feature#asahi-gpu · 2022-10-27 02:44
phirealso, who is forcing compression on? the kernel driver, or the agx firmware?#asahi-gpu · 2022-10-27 02:43
phireum... wat. is this just something that was always broken?#asahi-gpu · 2022-10-27 02:42
phirelina: is that cute flag demo somewhere?#asahi-gpu · 2022-10-27 02:34
phireguess that's a task for tomorrows stream#asahi-gpu · 2022-10-27 02:32
phireI can probally check from macos if my m1 max is allocating those accleration structures#asahi-gpu · 2022-10-27 02:30
phireyeah, figured#asahi-gpu · 2022-10-27 02:30
phirelina: did you push your changes to m1n1 somewhere?#asahi-gpu · 2022-10-27 02:27
phirereally?#asahi-gpu · 2022-10-27 02:25
phirebut given the Pro, Max and Ultra all have the exact same gpu name (G13X), I suspect they share features#asahi-gpu · 2022-10-27 02:22
phirethough I understand apple has a habbit of just not enabling features that weren't ready in time, and never enabling them#asahi-gpu · 2022-10-27 02:22
phirehey LunaFoxgirlVT#asahi-gpu · 2022-10-27 02:21
phirethere were a few unks that flipped from 0 to 1#asahi-gpu · 2022-10-27 02:17
phirefinal output, and in-progress?#asahi-gpu · 2022-10-27 02:17
phireor there are two depth buffers?#asahi-gpu · 2022-10-27 02:09
phireoh, point. I should chuck this into the hypervisor later#asahi-gpu · 2022-10-27 01:34
phireZLS = Z load store?