[PATCH] D80364: [amdgpu] Teach load widening to handle non-DWORD aligned loads.

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added a comment.

LGTM in principle. We wanted to split CodeGenPrepare for a long time already. We also should drop widening from an early pass then.



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPULateCodeGenPrepare.cpp:118
+    return false;
+  auto Ty = LI.getType();
+  // Skip aggregate types.
----------------
Please use "auto *" as  tidy suggests.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPULateCodeGenPrepare.cpp:146
+  int64_t Offset = 0;
+  auto Base =
+      GetPointerBaseWithConstantOffset(LI.getPointerOperand(), Offset, *DL);
----------------
Also "auto *".


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPULateCodeGenPrepare.cpp:166
+  unsigned LdBits = DL->getTypeStoreSize(LI.getType()) * 8;
+  auto IntNTy = Type::getIntNTy(LI.getContext(), LdBits);
+
----------------
Same here and in another places.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D80364/new/

https://reviews.llvm.org/D80364