Ran into an interesting issue. We are running an Oracle Linux 6 server that has a directory mounted to a Windows NFS share. When we run a find command on that directory we got the following:

find: WARNING: Hard link count is wrong for ./: this may be a bug in your filesystem driver. Automatically turning on find's -noleaf option. Earlier results may have failed to include directories that should have been searched.
find: Filesystem loop detected; `./.@GMT-2023.02.02-18:00:01' has the same device number and inode as a directory which is 1 level higher in the filesystem hierarchy.
find: Filesystem loop detected; `./.@GMT-2023.02.02-13:00:01' has the same device number and inode as a directory which is 1 level higher in the filesystem hierarchy.
find: Filesystem loop detected; `./.@GMT-2023.02.01-18:00:00' has the same device number and inode as a directory which is 1 level higher in the filesystem hierarchy.
find: Filesystem loop detected; `./.@GMT-2023.02.01-13:00:01' has the same device number and inode as a directory which is 1 level higher in the filesystem hierarchy.

Doing an ls -la on the directory I see this at the top:

total 84024
drwxrwxrwx 2 jsmith 201 65536 Feb 3 07:00 .
drwxrwxrwx 2 jsmith 201 65536 Feb 3 07:00 .
dr-xr-xr-x. 74 root root 4096 Jan 17 13:27 ..
dr-xr-xr-x. 74 root root 4096 Jan 17 13:27 ..

I've there are quirky issues when using Windows NFS shares to Linux mounts but I've never run across this before. Anyone have any ideas how to correct this?

what kind of filesystem is it? How is it shared and mounted? You need to upgrade your OS to the latest available patch level.

Share is on a Windows 2008 R2 server using Services for Network File System (directory is also shared as a CIFS share for Windows users).
Mount is on an Oracle Linux 6.9 server using default NFS settings.
Also tried mounting on an Oracle Linux 8 server and seeing the same thing.
Subdirectories look fine.
Other NFS mounts on the same servers pointing to the sharing server look fine as well. It just seems to be this one.

So you see the same thing on two different hosts. That probably means a problem on the windows side. I would try umount it completely and run a filesystem check. And probably a reboot.

The Windows box is a production server so I'll need to set up a maintenance window.
In the meantime if someone has any other ideas feel free to chime in.
Thanks!