Extend FSDP1 global clipping support for optimizers other than Shampoo

[wz337]

Summary: This diff is a followup of D73474285 and lets other dense optimizers take it the enable_global_grad_clip optim config. When enalbe_global_grad_clip=True and FSDP1 is used, it would calculate global gradient norm at a cost of extra communication.

Next steps:

1. global clipping more generic and work for FSDP2.

Differential Revision: D73969566

Privacy Context Container: L1235913

[facebook-github-bot]

This pull request was exported from Phabricator. Differential Revision: D73969566