Vigneshwar Jayakumar 0d67b2edac

[AMDGPU] Add f32 cost model for exp/exp2/exp10 intrinsics (#185369) ...

exp2 maps to a single v_exp_f32 (quarter rate), while exp and exp10 need
additional full-rate costs for the base conversion. Also added cost based on 
non-afn and denorm mode.

This enables SLP vectorization of exp2 on targets with packed f32 ops.

2026-03-10 11:08:39 -05:00

..

abs.ll

…

add-sub.ll

…

addrspacecast.ll

…

aggregates.ll

…

arith-sminmax.ll

…

arith-ssat.ll

…

arith-uminmax.ll

…

arith-usat.ll

…

arithmetic_fence.ll

…

bit-ops.ll

…

canonicalize.ll

…

cast.ll

…

control-flow.ll

…

copysign.ll

…

div.ll

…

exp2.ll

[AMDGPU] Add f32 cost model for exp/exp2/exp10 intrinsics (#185369)

2026-03-10 11:08:39 -05:00

exp10.ll

[AMDGPU] Add f32 cost model for exp/exp2/exp10 intrinsics (#185369)

2026-03-10 11:08:39 -05:00

exp.ll

[AMDGPU] Add f32 cost model for exp/exp2/exp10 intrinsics (#185369)

2026-03-10 11:08:39 -05:00

extractelement.ll

…

fabs.ll

…

fadd.ll

…

fdiv.ll

IR: Promote "denormal-fp-math" to a first class attribute (#174293)

2026-02-05 13:31:26 +00:00

fma.ll

…

fmul.ll

…

fmuladd.ll

…

fneg.ll

…

fptosi.ll

…

fptoui.ll

…

frexp.ll

…

fround.ll

…

fsub.ll

…

fused_costs.ll

…

gep.ll

…

insertelement.ll

…

is_fpclass.ll

…

ldexp.ll

…

lit.local.cfg

…

load-to-trunc.ll

…

load.ll

…

log2.ll

…

log10.ll

…

log.ll

…

logicalop.ll

…

maximum.ll

…

maximumnum.ll

…

maxnum.ll

…

minimum.ll

…

minimumnum.ll

…

minnum.ll

…

mul.ll

…

ptrmask.ll

…

reduce-and.ll

…

reduce-or.ll

…

rem.ll

…

shifts.ll

…

shufflevector.ll

AMDGPU: Improve getShuffleCost accuracy for 8- and 16-bit shuffles (#168818)

2025-11-21 19:33:13 +00:00

special-argument-intrinsics.ll

…

sqrt.ll

…

store.ll

…