[PATCH] D74919: AMDGPU/GlobalISel: Manually RegBankSelect copies

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
arsenm added reviewers: qcolombet, nhaehnle, foad.
Herald added subscribers: kerbowa, hiraditya, t-tye, tpr, dstuttard, rovka, yaxunl, wdng, jvesely, kzhuravl.
Herald added a project: LLVM.

This was failng on any pre-assigned copy to the VCC bank.

      

This is something of a workaround for the default implementation in
getInstrMappingImpl, and how it treats copy-like operations in
general.

      

Copy-like operations are considered to only have one result register
bank, rather than separate banks for each source like a normal
instruction. To avoid potentially mishandling reg_sequence with
impossible operand combinations, the generic implementation errors on
impossible costs. If the bank was already assigned, is treated it
as-if it were an unsatisfiable REG_SEQUENCE mapping. We really don't
get any value from any of what getInstrMappingImpl tries to do for
copies, so just directly emit the simple mapping we really want.


https://reviews.llvm.org/D74919

Files:
  llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp
  llvm/test/CodeGen/AMDGPU/GlobalISel/regbankselect-copy.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D74919.245699.patch
Type: text/x-patch
Size: 6122 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200220/bdbcb01c/attachment.bin>