ChanServ changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard + Bifrost + Valhall - Logs https://oftc.irclog.whitequark.org/panfrost - I don't know anything about WSI. That's my story and I'm sticking to it.
Didgy has joined #panfrost
Didgy has quit [Quit: Konversation terminated!]
<alyssa>
Didgy (when they're back): plans are brewing for a Valhall-focused Vulkan driver in 2023
<alyssa>
but right now that's more of a dream than anything
<alyssa>
If you're not married to working on Vulkan frontend code specifically there's always the compiler
camus has joined #panfrost
camus1 has quit [Ping timeout: 480 seconds]
nsprr has quit []
jernej- has quit []
jernej has joined #panfrost
camus has quit [Remote host closed the connection]
rcf has quit [Ping timeout: 480 seconds]
rcf has joined #panfrost
kinkinkijkin has quit [Remote host closed the connection]
kinkinkijkin has joined #panfrost
rasterman has joined #panfrost
camus has joined #panfrost
nlhowell has joined #panfrost
floof58 has quit [Read error: No route to host]
floof58 has joined #panfrost
nlhowell has quit [Ping timeout: 480 seconds]
warpme_____ has joined #panfrost
atler is now known as Guest782
atler has joined #panfrost
Guest782 has quit [Ping timeout: 480 seconds]
paulk-bis has joined #panfrost
paulk has quit [Read error: Connection reset by peer]
megi has joined #panfrost
alyssa has quit [Quit: leaving]
alyssa has joined #panfrost
<alyssa>
q4a: copypaste fail, that should indeed be IADD.i32
<alyssa>
I'm also not convinced that sentence is right
<alyssa>
for the second q
<q4a>
Thanks. I'll add fix to my next MR about Valhall docs
<alyssa>
IADD_IMM.v2i16 should be used instead of IADD.v2i16 with a uniform
<alyssa>
What I'm trying to say in the second sentence is that IADD.v2i16 with an inline constant ("small constant") is preferrable to IADD_IMM.v2i16
<alyssa>
but I'm not actually sure that's right
<alyssa>
going back in time..
<alyssa>
MUX.i32 is (a & mask) | (b & ~mask)
<alyssa>
I guess for OpenCL you just swap the arguments
<alyssa>
jenneron[m]: LPAE is above my paygrade
<alyssa>
q4a: btw, the other "tricky" case with promoting constants to uniforms is for 64-bit constants (or i32vec2 constants, equivalently)
<alyssa>
which come up all the time for texturing
<alyssa>
e.g. look at shaders/glmark/1-4.shader_test
<alyssa>
this ends up compiling to IADD_IMM.i32 r2, 0x0, #0x20001800