[PATCH] D93337: AMDGPU/GlobalISel: Start cleaning up calling convention lowering

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUCallLowering.cpp:261
     const LLT RegTy = MRI.getType(ValVReg);
-    if (RegTy.getSizeInBytes() > Size)
-      Size = RegTy.getSizeInBytes();
-
-    assignValueToAddress(ValVReg, Addr, Size, MPO, VA);
+    MemSize = std::min(MemSize, (uint64_t)RegTy.getSizeInBytes());
+    assignValueToAddress(ValVReg, Addr, MemSize, MPO, VA);
----------------
cdevadas wrote:
> Why did you change the condition here?
Because the opposite case can happen. The legalized memory size can exceed the register type, which would form an illegal load


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D93337/new/

https://reviews.llvm.org/D93337