projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 71269a1) | patch
AMDGPU: Make signext/zeroext behave more sensibly over > i32
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Mon, 24 Feb 2020 20:43:18 +0000 (15:43 -0500)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Mon, 9 Mar 2020 19:56:10 +0000 (12:56 -0700)
commit156a1b59dfacacacb2c5ac81ee2db4b8e9d46279
tree89ab7c74cb492932520aacb933c5fa90924e4c0ftree | snapshot
parent71269a1f172cdad1cc0d7e2e6c94a7ece11ddb27commit | diff
AMDGPU: Make signext/zeroext behave more sensibly over > i32

Interpret these as extending to the next multiple of 32-bits. This had
no effect with i48 for example, which is really split into {i32, i16},
which should extend the high part.
llvm/lib/Target/AMDGPU/AMDGPUISelLowering.cpp diff | blob | history
llvm/lib/Target/AMDGPU/AMDGPUISelLowering.h diff | blob | history
llvm/test/CodeGen/AMDGPU/function-returns.ll diff | blob | history
Domain: System / Toolchain;