[PATCH] D90979: [AMDGPU] Omit buffer resource with flat scratch.

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec updated this revision to Diff 303585.
rampitec added a comment.
Herald added a subscriber: qcolombet.

I have realized that SI_SPILL_* opcodes need a valid SReg_128, even if it is unused and will dropped after expansion. With this change ScratchRSrcReg is a $noreg which is not valid for the verifier.

I am using ttmp tuple as any reserved SReg_128 will work and at the end it will not be used. In general I could create a new unallocatable pseudo RC with SGPR_128 + PRIVATE_RSRC_REG just for that, but it looks like an overkill.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90979/new/

https://reviews.llvm.org/D90979

Files:
  llvm/lib/Target/AMDGPU/AMDGPUCallLowering.cpp
  llvm/lib/Target/AMDGPU/SIFrameLowering.cpp
  llvm/lib/Target/AMDGPU/SIISelLowering.cpp
  llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
  llvm/lib/Target/AMDGPU/SIMachineFunctionInfo.cpp
  llvm/test/CodeGen/AMDGPU/call-preserved-registers.ll
  llvm/test/CodeGen/AMDGPU/callee-frame-setup.ll
  llvm/test/CodeGen/AMDGPU/chain-hi-to-lo.ll
  llvm/test/CodeGen/AMDGPU/fast-unaligned-load-store.private.ll
  llvm/test/CodeGen/AMDGPU/flat-scratch.ll
  llvm/test/CodeGen/AMDGPU/load-lo16.ll
  llvm/test/CodeGen/AMDGPU/local-stack-alloc-block-sp-reference.ll
  llvm/test/CodeGen/AMDGPU/non-entry-alloca.ll
  llvm/test/CodeGen/AMDGPU/spill-offset-calculation.ll
  llvm/test/CodeGen/AMDGPU/stack-pointer-offset-relative-frameindex.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D90979.303585.patch
Type: text/x-patch
Size: 59965 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20201107/49251a3f/attachment.bin>