Files
clang-p2996/llvm/lib/Target/AMDGPU/AMDGPUInline.cpp
Matt Arsenault 0ff901fba0 AMDGPU: Boost inline threshold with addrspacecasted alloca arguments
This was skipping GetUnderlyingObject for nonprivate addresses, but an
alloca could also be found through an addrspacecast if it's flat.

llvm-svn: 361649
2019-05-24 16:52:35 +00:00

7.4 KiB