Add cudnn related synchronization (pause and ping Meta)

CUDA_UPGRADE_GUIDE.MD

@@ -151,3 +151,4 @@ If you require to update CUDNN version for already existing CUDA version, please
 1. Builder PR: https://github.com/pytorch/builder/pull/1271. Important note: Builder PR and Pytorch PR need to be validated and landed togeather to avoid breakage of CI and nightly!
 2. Add new cudnn vesion to windows AMI: https://github.com/pytorch/test-infra/pull/1523. Rebuild and retest the AMI. Follow step 6  Generate new Windows AMI, test and deploy to canary and prod.
 3. Create PyTorch PR: https://github.com/pytorch/pytorch/pull/93086 and small wheel update PyTorch PR: https://github.com/pytorch/pytorch/pull/104757
+4. Note: in trying to get the pytorch PR (e.g. cudnn v8 upgrade to cudnn v9) to pass all tests, you may need to pause and contact Meta for help with 1) uploading docker tags to resolve issues like "the repository with name 'pytorch/pytorch-linux-focal-cuda12.1-cudnn9-py3-gcc9' does not exist in the registry with id '308535385114'". 2) uploading new dependency wheels (e.g. nvidia-cudnn-cu12) to resolve issues like "ERROR: Could not find a version that satisfies the requirement nvidia-cudnn-cu12==9.1.0.70; platform_system == "Linux" and platform_machine == "x86_64" (from torch) (from versions: 8.8.1.3, 8.9.2.26, 8.9.7.29)"