projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: b7587ca) | patch
AMDGPU: Use more accurate IR type for block handle
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Fri, 23 Dec 2022 20:52:41 +0000 (15:52 -0500)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Sat, 7 Jan 2023 02:23:28 +0000 (21:23 -0500)
commit47554a0c73ee703362d913d3a149f104819c8365
tree6e5fa2654698c51d6872dec2645e2c979fffa15btree | snapshot
parentb7587ca8377572ff45ce80aa070ac7a3aadde431commit | diff
AMDGPU: Use more accurate IR type for block handle

The device library uses this as a struct with a pointer sized integer
and 2 ints.
llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.cpp diff | blob | history
llvm/test/CodeGen/AMDGPU/enqueue-kernel.ll diff | blob | history
Domain: System / Toolchain;