Make PyTorch partially cross-compilable for Apple M1 #49701
Conversation
💊 CI failures summary and remediationsAs of commit d657a07 (more details on the Dr. CI page):
This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions to the (internal) Dr. CI Users group. This comment has been revised 11 times. |
There was a problem hiding this comment.
@malfet has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
Introduce changes similar to pytorch/cpuinfo#51 to detect that PyTorch is build on MacOS for different architecture Test Plan: Pull shibatch/sleef#376 and run `cmake -DCMAKE_OSX_ARCHITECTURES=x86_64 -DPYTHON_EXECUTABLE=/usr/bin/python3 -DUSE_NNPACK=NO -DUSE_XNNPACK=NO -DBUILD_TEST=YES .. -G Ninja; ninja basic`
malfet
force-pushed
the
malfet/update-cpuinfo
branch
from
5793a94
to
d657a07
Compare
There was a problem hiding this comment.
@malfet has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
| @@ -41,7 +41,7 @@ IF(CMAKE_SYSTEM_NAME MATCHES "Linux") | |||
| ENDIF (OMAP4_TRUE) | |||
|
|
|||
| ELSEIF(CMAKE_SYSTEM_NAME MATCHES "Darwin") | |||
| IF(CMAKE_SYSTEM_PROCESSOR STREQUAL "arm64") | |||
| IF(CMAKE_SYSTEM_PROCESSOR STREQUAL "arm64" AND NOT CMAKE_OSX_ARCHITECTURES STREQUAL "x86_64") | |||
There was a problem hiding this comment.
Do we have to worry about CMAKE_OSX_ARCHITECTURES equalling any other architecture or nah?
There was a problem hiding this comment.
In theory we should, but let's do it one step at a time: so far we can only use the system to cross-build it for another platform.
Summary: Update CPUINFO to include pytorch/cpuinfo#51 Update sleef to include shibatch/sleef#376 Modify aten/src/ATen/native/quantized/cpu/qnnpack/CMakeLists.txt to recognize CMAKE_OSX_ARCHITECTURES Pull Request resolved: pytorch#49701 Test Plan: `cmake -DCMAKE_OSX_ARCHITECTURES=x86_64 -DPYTHON_EXECUTABLE=/usr/bin/python3 -DUSE_XNNPACK=NO -DBUILD_TEST=YES .. -G Ninja; ninja basic` finishes successfully on Apple M1 Reviewed By: janeyx99 Differential Revision: D25669219 Pulled By: malfet fbshipit-source-id: 5ee36b64e3a7ac76448f2a300ac4993375a26de5
Update CPUINFO to include pytorch/cpuinfo#51
Update sleef to include shibatch/sleef#376
Modify aten/src/ATen/native/quantized/cpu/qnnpack/CMakeLists.txt to recognize CMAKE_OSX_ARCHITECTURES
Test Plan:
cmake -DCMAKE_OSX_ARCHITECTURES=x86_64 -DPYTHON_EXECUTABLE=/usr/bin/python3 -DUSE_XNNPACK=NO -DBUILD_TEST=YES .. -G Ninja; ninja basicfinishes successfully on Apple M1