David Green 03912a1de5
[GlobalISel] Translate scalar sequential vecreduce.fadd/fmul as fadd/fmul. (#153966)
A llvm.vector.reduce.fadd(float, <1 x float>) will be translated to
G_VECREDUCE_SEQ_FADD with two scalar operands, which is illegal
according to the verifier. This makes sure we generate a fadd/fmul
instead.
2025-08-18 14:59:44 +00:00
..
2025-03-06 09:08:21 -08:00