Sounds like the problem lies with the nfs lock daemon.
Are you sure it won't work unless you add ",udp"? udp is normally the default for NFS, so with or without -o udp should be the same. The active alternative is -o tcp, to change the default.

I also had problems with this.
I can mount my nfs shares on other systems either with plain nfs or nfs4. The debian server uses by default the portmap option "-i 127.0.0.1", but it doesn't seem to be a problem for the other clients.

On the N900 the are some problems though:
- The portmap daemon won't ever start because for some reason, the init script detects "pidof portmap" as portmap itself so it says "Already running." To get it to start I had to edit the lauch script commenting that line out.
- After that, mounting the share, with "-t nfs" gives a:
"mount: RPC: Unable to receive; errno = Connection refused"
- Editing the portmap conf on the server and removing the "-i 127.0.0.1" limitation makes the error message change to:
"mount: IP:/ failed, reason given by server: Permission denied"
- Trying to mount with NFSv4 always fails with:
"mount: mounting IP:/ on mountpoint failed: Invalid argument". Lack of kernel support?

Finally, the solution for me:
- hack the portmap init script to force it to launch the daemon
- remove the "-i 127.0.0.1" portmap option on the server side
- mount with IP:/path/from/server/root instead of IP:/path/from/share/root as I do on the rest of the computers

This is my first time playing with nfs, but as far as I've read on the Internet, this is some legacy behavior, isn't it?