Jiong Gong cb48f7855a [inductor cpu] fix uint8 add and sub (#113253) ... Fix https://github.com/pytorch/pytorch/issues/113016 and https://github.com/pytorch/pytorch/issues/113020 and https://github.com/pytorch/pytorch/issues/113141 and https://github.com/pytorch/pytorch/issues/113143 and https://github.com/pytorch/pytorch/issues/113144 Explicit typecast result of add/sub to uint8 (similar to how we fixed mul previously) to avoid implicit type promotion from C. Pull Request resolved: https://github.com/pytorch/pytorch/pull/113253 Approved by: https://github.com/lezcano, https://github.com/jansel		2023-11-10 04:06:42 +00:00
..
aoti_runtime	Support fp8 in AOTInductor + support optional<> in C ABI (#112527)	2023-11-08 22:56:48 +00:00
cuda	[Inductor CUTLASS backend] Epilogue fusion codegen (Step 1) (#110890)	2023-11-06 19:42:10 +00:00
__init__.py
common.py	[inductor] Memory planning (#112178)	2023-11-02 07:39:13 +00:00
cpp_prefix.h	S390x inductor support (#111367)	2023-10-20 19:38:46 +00:00
cpp.py	[inductor cpu] fix uint8 add and sub (#113253)	2023-11-10 04:06:42 +00:00
cuda_combined_scheduling.py	[Inductor CUTLASS backend] Epilogue fusion codegen (Step 1) (#110890)	2023-11-06 19:42:10 +00:00
memory_planning.py	[inductor] Memory planning (#112178)	2023-11-02 07:39:13 +00:00
triton_foreach.py	[Reland2] [inductor][BE] split triton_meta and inductor_meta (#112351)	2023-11-02 00:40:12 +00:00
triton_utils.py	[Inductor] Allow None values to be passed in as arguments to triton kernels (#113056)	2023-11-07 05:29:42 +00:00
triton.py	[inductor][easy] Fix fusion logging (#113308)	2023-11-09 03:19:39 +00:00
wrapper.py	[AOTI] Implement support for user defined kernels that use triton.autotune (#113229)	2023-11-10 01:39:00 +00:00