#
✕ nick:tellowkrinkle
#asahi-gpu · 2023-01-08 02:10
TellowKrinkleDoesn't robustness have to work with out of bounds reads too? That doesn't sound like it would protect against accidentally reading from another valid buffer that happened to be nearby
#asahi-gpu · 2023-01-07 21:21
TellowKrinkle(That's what Dolphin does in the MVK and Metal renderers)
#asahi-gpu · 2023-01-07 21:21
TellowKrinkleWell that's one way to do it
#asahi-gpu · 2023-01-07 21:18
TellowKrinkleIsn't lod bias part of GL 4.1? Do they just ignore it?
#asahi-gpu · 2023-01-05 19:42
TellowKrinkleAhh okay
#asahi-gpu · 2023-01-05 19:41
TellowKrinkleIf border color depends on the texture format, does that mean it's not part of the sampler? Since those are separate from textures in Metal
#asahi-gpu · 2022-12-29 17:51
TellowKrinkleBTW how do uniform registers get initialized? I noticed the Metal compiler seems to reference a lot of constants from them without loading them in a preamble shader, including things from constant address space buffers (with arbitrary constant offsets too), which can be bound to arbitrary GPU buffers. How does that work?
#asahi-gpu · 2022-12-20 04:22
TellowKrinkleSo far, it seems that if the input is smooth, doing anything other than just converting it to a uint removes the .TODO. But if the input is flat, I can do all sorts of math on it and it stays .TODO. Can't possibly imagine what such a flag would be useful for...
#asahi-gpu · 2022-12-20 04:04
TellowKrinklebuffer[uint(smooth.x) * 8 + 0] * buffer[uint(smooth.y) * 8 + 2] => device_load
#asahi-gpu · 2022-12-20 04:03
TellowKrinklebuffer[uint(smooth.x) * 8 + 0] * buffer[uint(smooth.x) * 8 + 2] => "device_load"
#asahi-gpu · 2022-12-20 04:02
TellowKrinklebuffer[uint(smooth.x)] * buffer[uint(smooth.y)] => "device_load.TODO"
#asahi-gpu · 2022-12-20 04:01
TellowKrinkleOkay I was able to get the TODO with smooth input
#asahi-gpu · 2022-12-20 03:59
TellowKrinklealyssa: Good to know
#asahi-gpu · 2022-12-20 03:55
TellowKrinkleOkay the ".TODO" seems to only be added if the input comes from a flat interpolated input?
#asahi-gpu · 2022-12-20 02:57
TellowKrinkleBTW on that fragment shader one, were you doing anything with side effects (e.g. writing to device memory)? I can only get the .TODO if I do that and I don't also add "[[early_fragment_tests]]"
#asahi-gpu · 2022-12-20 02:35
TellowKrinkle(Compute shader that loads 3 float4s, does an fma, and stores the result)
#asahi-gpu · 2022-12-20 02:35
TellowKrinkle 2c: 0551240e01c8f200 device_load 0, i32, quad, r10_r11_r12_r13, u2_u3, r9, unsigned, lsl 2
#asahi-gpu · 2022-12-20 02:35
TellowKrinkle 1c: 0501040e00c8f200 device_load 0, i32, quad, r0_r1_r2_r3, u2_u3, r0, unsigned, lsl 2
#asahi-gpu · 2022-12-20 02:35
TellowKrinkle 34: 3800 wait 0
#asahi-gpu · 2022-12-20 02:35
TellowKrinkle 24: 0529a40e00c8f200 device_load 0, i32, quad, r5_r6_r7_r8, u2_u3, r5, unsigned, lsl 2
#asahi-gpu · 2022-12-20 02:35
TellowKrinkleWeird, I'm getting this on my M1 (no .TODO):