[PATCH] D84153: AMDGPU: Relax restriction on folding immediates into physregs

[Matt Arsenault via Phabricator via llvm-commits]

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

I never completed the work on the patches referenced by
f8bf7d7f42f28fa18144091022236208e199f331, but this was intended to
avoid folding immediate writes into m0 which the coalescer doesn't
understand very well. Relax this to allow simple SGPR immediates to
fold directly into VGPR copies. This pattern shows up routinely in
current GlobalISel code since nothing is smart enough to emit VGPR
constants yet.


https://reviews.llvm.org/D84153

Files:
  llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
  llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.kernarg.segment.ptr.ll
  llvm/test/CodeGen/AMDGPU/GlobalISel/zextload.ll
  llvm/test/CodeGen/AMDGPU/fold-imm-copy.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D84153.279194.patch
Type: text/x-patch
Size: 7195 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200720/fb5464ad/attachment.bin>