19:09 anarsoul|2: enunes: I checked the shader for https://gitlab.freedesktop.org/mesa/mesa/-/issues/9323
19:09 anarsoul|2: it looks like it is heavily scalarized
19:11 anarsoul|2: and atan is lowered instead of using native instruction
19:15 anarsoul|2: looks like nir doesn't have support for native atan
19:25 anarsoul|2: and we really could use a copy propagation pass in ppir :)
19:26 anarsoul|2: It doesn't look good: mov.v1 $0 $1, stop
19:32 anarsoul|2: if I replace atan with a multiplication, instr number goes down to 36
19:32 anarsoul|2: from 67
19:39 anarsoul|2: and looks like we don't disassemble atan correctly :(
20:31 enunes: anarsoul|2: cool, I hadn't looked into the actual reason yet before posting there
20:32 enunes: so looks like at least atan can be added to nir and ppir for some improvement, maybe even relatively not too complicated work
22:20 anarsoul|2: enunes: yeah, shouldn't be too complicated