Add cu118 workflows

[ptrblck]

CC @atalman @malfet

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/90826

• 📄 Preview Python docs built from this PR
• 📄 Preview C++ docs built from this PR
• ❓ Need help or want to give feedback on the CI? Visit the bot commands wiki or our office hours

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 515c1ec:
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: ptrblck (0ea163a)

[ptrblck]

The MacOS build failures seem to come from an invalid usage of --jobs 0:

+ git submodule update --init --recursive --jobs 0
~/work/pytorch/pytorch/pytorch ~/work/pytorch/pytorch
BUG: run-command.c:1521: you must provide a non-zero number of processes!

while the Windows build issues seem to be real:

Path = C:\actions-runner\_work\pytorch\pytorch\builder\windows\internal\\..\temp_build\gpu_driver_dlls.zip
Type = zip
Physical Size = 5127591


Would you like to replace the existing file:
  Path:     C:\Windows\System32\nvcuda.dll
  Size:     2729776 bytes (2666 KiB)
  Modified: 2021-04-10 18:27:18
with the file from archive:
  Path:     nvcuda.dll
  Size:     17459864 bytes (17 MiB)
  Modified: 2019-11-14 22:47:12


Break signaled
? (Y)es / (N)o / (A)lways / (S)kip all / A(u)to rename all / (Q)uit? 
Archives with Errors: 1

Cleaning temp files

C:\actions-runner\_work\pytorch\pytorch\builder>goto set_cuda_env_vars 

C:\actions-runner\_work\pytorch\pytorch\builder>echo Setting up environment... 
Setting up environment...

C:\actions-runner\_work\pytorch\pytorch\builder>set "PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\bin;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\libnvvp;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0;C:\Windows\System32\OpenSSH;C:\Program Files\Amazon\cfn-bootstrap;C:\ProgramData\chocolatey\bin;C:\Program Files\Amazon\AWSCLIV2;C:\Program Files\Git\cmd;C:\Program Files\Git\mingw64\bin;C:\Program Files\Git\usr\bin;C:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit;C:\Users\runneruser\AppData\Local\Microsoft\WindowsApps" 

C:\actions-runner\_work\pytorch\pytorch\builder>set "CUDA_PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" 

C:\actions-runner\_work\pytorch\pytorch\builder>set "CUDA_PATH_V11_8=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" 

C:\actions-runner\_work\pytorch\pytorch\builder>set "NVTOOLSEXT_PATH=C:\Program Files\NVIDIA Corporation\NvToolsExt" 

C:\actions-runner\_work\pytorch\pytorch\builder>if errorlevel 1 exit /b 1 
Error: Process completed with exit code 1.

I'm currently unsure what's causing these issues and what the actual error is, so would need to dig a bit into it.

[atalman]

Windows issues where because of the windows 11.8 image was still not available on production. This issue should be resolved this morning. I am rerunning the tests now.

[ngimel]

If windows issues persist, we should start 11.8 runs and nightlies for linux only, it's been long enough.

[atalman]

@pytorchbot merge

[pytorch-bot]

This PR needs to be approved by an authorized maintainer before merge.

[atalman]

LGTM

[atalman]

@pytorchbot merge -f "Signal is green"

[pytorchmergebot]

Merge started

Your change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here