| commit | a43d9af52c88025cc0d5747e8ab843df7a69fa03 | [log] [tgz] |
|---|---|---|
| author | SsnL <tongzhou.wang.1994@gmail.com> | Mon Jul 01 09:33:26 2019 -0700 |
| committer | Facebook Github Bot <facebook-github-bot@users.noreply.github.com> | Mon Jul 01 09:45:01 2019 -0700 |
| tree | d27ce91df0b5345e36eb453a8a05111928fb7be0 | |
| parent | 451c907a4748c49e41cf4ea09fb24daa7073cc65 [diff] |
Comment on why Windows build_pytorch.bat builds twice (#22363) Summary: I've noticed that Windows CI seems to build twice, e.g., https://ci.pytorch.org/jenkins/job/pytorch-builds/job/pytorch-win-ws2016-cuda9-cudnn7-py3-build/60304/console This adds a comment explaining why. Pull Request resolved: https://github.com/pytorch/pytorch/pull/22363 Differential Revision: D16073609 Pulled By: zou3519 fbshipit-source-id: ddb422b7c7e18cc436caff2c5838373a82f69429
diff --git a/.jenkins/pytorch/win-test-helpers/build_pytorch.bat b/.jenkins/pytorch/win-test-helpers/build_pytorch.bat index ccc87c9..77383d8 100644 --- a/.jenkins/pytorch/win-test-helpers/build_pytorch.bat +++ b/.jenkins/pytorch/win-test-helpers/build_pytorch.bat
@@ -69,6 +69,9 @@ set CMAKE_GENERATOR=Ninja +:: The following code will try to build PyTorch twice if USE_CUDA is neither 0 +:: nor 1. It is intended so that both builds can be folded into 1 CI run. + if not "%USE_CUDA%"=="1" ( if "%REBUILD%"=="" ( set NO_CUDA=1