J. Bruce Fields wrote:
> On Fri, Aug 24, 2007 at 01:11:16PM -0400, Chuck Lever wrote:
>> The current mount.nfs implementation doesn't allow users to specify their
>> own addr= or clientaddr= option.
> 
> Don't we need clientaddr= in some cases to get the callback channel
> working?

The mount.nfs command adds a clientaddr option of its own before doing 
the syscall.  The point of the patch is to prevent passing in a 
user-created clientaddr=.

Does anyone know of a use case where an exposed clientaddr= is needed?

>> The new string-based interface does allow this, even though nfs(5)
>> does not document 'addr=' and specifically forbids adding
>> 'clientaddr='.
> 
> It says that it's currently ignored, which isn't the same as forbidding
> it.

nfs(5) is likely incorrect.  It's arguable what "currently ignored" 
means -- it probably means the mount command ignores it, if specified, 
and uses its own address anyway.