[PATCH] D78772: [AMDGPU] Adopt GCNRegBankReassign for 16 bit subregs

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec created this revision.
rampitec added reviewers: arsenm, vpykhtin.
Herald added subscribers: kerbowa, hiraditya, t-tye, tpr, dstuttard, yaxunl, nhaehnle, wdng, jvesely, kzhuravl.
rampitec edited the summary of this revision.

It allows it not to crash and analyze 16 bit subregs if those
appear in the instructions. At the same time it does not attempt
to reassign these. It still can correctly identify register
banks to let larger registers to be reassigned.

More work will be needed here when real instructions will use
these registers and more tests as well.


https://reviews.llvm.org/D78772

Files:
  llvm/lib/Target/AMDGPU/GCNRegBankReassign.cpp
  llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
  llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp
  llvm/lib/Target/AMDGPU/SIRegisterInfo.h
  llvm/test/CodeGen/AMDGPU/regbank-reassign.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D78772.259756.patch
Type: text/x-patch
Size: 7969 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200423/9b887e8e/attachment.bin>