Benjamin Kramer 0776f6e04d [LSV] Vectorize loads of vectors by turning it into a larger vector ...

Use shufflevector to do the subvector extracts. This allows a lot more
load merging on AMDGPU and also on NVPTX when <2 x half> is involved.

Differential Revision: https://reviews.llvm.org/D117219

2022-01-26 11:38:41 +01:00

..

4x2xhalf.ll

…

lit.local.cfg

…

merge-across-side-effects.ll

…

non-instr-bitcast.ll

…

propagate-invariance-metadata.ll

…