Bin Bao b0816e4714 [inductor] Fix AOTInductor output issues (#105773) ... Summary: This is a follow-up on https://github.com/pytorch/pytorch/pull/105496. There are several issues with the previous fix, 1) It explicitly does copy for every output at the end of the main function; 2) When an output is ReinterpretView, no as_strided was generated for it; 3) There can be duplicated buffer declarations. This PR fixes by making sure can_reuse behave consistently between two AOTIndcutor passes, and thus always generate the same set of kernels. It also adds handling of ReinterpretView. Differential Revision: [D47692214](https://our.internmc.facebook.com/intern/diff/D47692214) Pull Request resolved: https://github.com/pytorch/pytorch/pull/105773 Approved by: https://github.com/jansel		2023-07-24 01:58:49 +00:00
..
__init__.py
aot_inductor_interface.cpp	Add aot_inductor as a test backend for benchmarking (#105221)	2023-07-18 13:16:36 +00:00
common.py	ValueRange analysis for indirect indexing (#102611)	2023-07-14 13:43:05 +00:00
cpp_prefix.h	[inductor] Add single pass "var_unnormalized" reduction_type (#102486)	2023-07-08 20:48:29 +00:00
cpp.py	Revert "inductor: promote half/bfloat16 constant to float for cpu vectorization path (#105440)"	2023-07-20 03:56:44 +00:00
triton_foreach.py	Allow fusion of epilogue copies with upstream foreach ops (#104018)	2023-06-23 21:39:59 +00:00
triton_utils.py	Foreach kernel codegen in inductor (#99975)	2023-05-25 21:48:41 +00:00
triton.py	[Inductor] Provenance tracking for wrapper code (#105717)	2023-07-21 23:06:43 +00:00
wrapper.py	[inductor] Fix AOTInductor output issues (#105773)	2023-07-24 01:58:49 +00:00