[PATCH] D62310: AMDGPU: Correct maximum possible private allocation size

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
arsenm added reviewers: rampitec, nhaehnle, t-tye.
Herald added subscribers: arphaman, tpr, dstuttard, yaxunl, wdng, jvesely, kzhuravl.

We were assuming a much larger possible per-wave visible stack
allocation than is possible:

      

https://github.com/RadeonOpenCompute/ROCR-Runtime/blob/master/src/core/runtime/amd_gpu_agent.cpp#L70

Based on this, we can assume the high 15 bits of a frame index or sret
are 0. The frame index value is the per-lane offset, so the maximum
frame index value is MAX_WAVE_SCRATCH / wavesize.

      

Remove the corresponding subtarget feature and option that made
this configurable.


https://reviews.llvm.org/D62310

Files:
  lib/Target/AMDGPU/AMDGPU.td
  lib/Target/AMDGPU/AMDGPUSubtarget.cpp
  lib/Target/AMDGPU/AMDGPUSubtarget.h
  lib/Target/AMDGPU/SIISelLowering.cpp
  test/CodeGen/AMDGPU/frame-index-elimination.ll
  test/CodeGen/AMDGPU/function-returns.ll
  test/CodeGen/AMDGPU/huge-private-buffer.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D62310.200956.patch
Type: text/x-patch
Size: 9487 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190523/c9ef8163/attachment.bin>