[PATCH] D23361: [OpenCL] AMDGCN: Fix size_t type

[Joey Gouly via cfe-commits]

joey added a subscriber: joey.

================
Comment at: lib/CodeGen/CodeGenModule.cpp:101
@@ -100,3 +100,3 @@
   DoubleTy = llvm::Type::getDoubleTy(LLVMContext);
   PointerWidthInBits = C.getTargetInfo().getPointerWidth(0);
   PointerAlignInBytes =
----------------
What if you create a new function in TargetInfo called getMaxPointerWidth(unsigned AddrSpace), and call that here? That would by default just call 'getPointerWidth', but in your AMDGPU TargetInfo you can override that.
That feels more generic.


https://reviews.llvm.org/D23361