We don't need to expand to full shifts for the > 32-bit case. This
just switches to a sext_inreg of the high half.
Details
Details
Diff Detail
Diff Detail
Paths
| Differential D73414
AMDGPU/GlobalISel: Do a better job splitting 64-bit G_SEXT_INREG ClosedPublic Authored by arsenm on Jan 25 2020, 6:33 AM.
Details
Diff Detail
Revision Contents
Diff 240374 llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp
llvm/test/CodeGen/AMDGPU/GlobalISel/regbankselect-sext-inreg.mir
|