Nikita Shulga 9ca080db87 [MPS] Extend atomic operations to all int types (#158179) ...

That fixes `index_put(..., accumulate=True)` for all dtypes

int64 operation is not really atomic, but eventually consistent from the `index_put_accumulate` kernel point of view: i.e. by the end of the operation results in the global memory are indeed accumulation of the operands at given indices
Pull Request resolved: https://github.com/pytorch/pytorch/pull/158179
Approved by: https://github.com/dcci, https://github.com/Skylion007
ghstack dependencies: #158064, #158178

2025-07-14 04:25:05 +00:00

..

atomic.h

[MPS] Extend atomic operations to all int types (#158179)

2025-07-14 04:25:05 +00:00

common.h

[BE] Use simdgroup_size constexpr (#157751)

2025-07-08 03:46:20 +00:00

expm1f.h

[MPS] Fix binary builds (#155733)

2025-06-11 22:55:33 +00:00

indexing.h

[BE] fix typos in c10/ (#156078)

2025-06-18 10:24:44 +00:00

random.h

[BE] fix typos in c10/ (#156078)

2025-06-18 10:24:44 +00:00

reduction_utils.h

[BE] Use simdgroup_size constexpr (#157751)

2025-07-08 03:46:20 +00:00

special_math.h

Add isnan exit condition to special ops (#157464)

2025-07-05 04:19:50 +00:00

utils.h

[MPS] Implement scan metal kernels (#156100)

2025-06-17 17:44:22 +00:00