Joseph Huber dfcb8cb2a9 [OpenMP] Add pre sm_70 load hack back in (#138589) ...

Summary:
Different ordering modes aren't supported for an atomic load, so we just
do an add of zero as the same thing. It's less efficient, but it works.

Fixes https://github.com/llvm/llvm-project/issues/138560

2025-05-05 16:33:41 -05:00

..

Allocator.h

[OpenMP] Port the OpenMP device runtime to direct C++ compilation (#123673)

2025-02-05 08:18:52 -06:00

Configuration.h

…

Debug.h

…

DeviceTypes.h

[OpenMP] Replace use of target address space with <gpuintrin.h> local (#126119)

2025-02-09 10:25:25 -06:00

DeviceUtils.h

[OpenMP] Port the OpenMP device runtime to direct C++ compilation (#123673)

2025-02-05 08:18:52 -06:00

generated_microtask_cases.gen

…

Interface.h

…

LibC.h

…

Mapping.h

[OpenMP] Use 'gpuintrin.h' definitions for simple block identifiers (#131631)

2025-03-18 15:38:46 -05:00

Profiling.h

…

State.h

[OpenMP] Replace use of target address space with <gpuintrin.h> local (#126119)

2025-02-09 10:25:25 -06:00

Synchronization.h

[OpenMP] Add pre sm_70 load hack back in (#138589)

2025-05-05 16:33:41 -05:00

Workshare.h

[OpenMP] Port the OpenMP device runtime to direct C++ compilation (#123673)

2025-02-05 08:18:52 -06:00