projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: def6158) | patch
clang/AMDGPU: Emit atomicrmw for atomic_inc/dec builtins
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Sat, 5 Nov 2022 20:07:49 +0000 (13:07 -0700)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Sat, 17 Jun 2023 00:18:50 +0000 (20:18 -0400)
commitb84721df636747a9955408a934e713192a9cfe21
treeded012d13e2e4e80cd9b587aa7d5ffde3949b1ddtree | snapshot
parentdef6158f957e77d67f74aa9c8bb342c4b64a159bcommit | diff
clang/AMDGPU: Emit atomicrmw for atomic_inc/dec builtins

This makes the scope and ordering arguments actually do something.
Also add some new OpenCL tests since the existing HIP tests didn't
cover address spaces.
clang/lib/CodeGen/CGBuiltin.cpp diff | blob | history
clang/test/CodeGenCXX/builtin-amdgcn-atomic-inc-dec.cpp diff | blob | history
clang/test/CodeGenOpenCL/builtins-amdgcn.cl diff | blob | history
Domain: System / Toolchain;