[PATCH] D109008: [AMDGPU][NFC] Refactor AMDGPUCallingConv.td

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUCallingConv.td:147
+//===----------------------------------------------------------------------===//
+//     Compute Calling Conventions
+//===----------------------------------------------------------------------===//
----------------
It's not entirely accurate since amdgpu_gfx will use the same masks


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUCallingConv.td:230
+
+def CSR_AMDGPU_AllVGPRs : RegMask<
+  (sequence "VGPR%u", 0, 255)
----------------
Probably shouldn't have the CSR prefix if it's just all registers


================
Comment at: llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp:352-353
   case CallingConv::AMDGPU_Gfx:
     return MF->getSubtarget<GCNSubtarget>().hasGFX90AInsts()
-        ? CSR_AMDGPU_HighRegs_With_AGPRs_SaveList
-        : CSR_AMDGPU_HighRegs_SaveList;
+               ? CSR_AMDGPU_GFX90AInsts_SaveList
+               : CSR_AMDGPU_SaveList;
----------------
Changing the convention based on the subtarget seems problematic. Do we really need to do this?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D109008/new/

https://reviews.llvm.org/D109008