erh@nimenees.com (Eric Haszlakiewicz) writes:

>	It _does_ do the right thing.  Calling lseek saves that opaque
>cookie in the fp->fp_offset, which then (int vn_readdir) ends up in a
>uio_offset, which then gets used by tmpfs_dir_offset().
>	tmpfs directories are linked lists, so there _isn't_ a byte offset.

We don't need a byte offset. We do need cookies that are
32bit signed values and which are sign-extended for the 64bit
d_off field.

I had a look on 32bit NFS cookies (mount_nfs -X), these are
actually 30bit cookies (6bit hash + 24bit generation number)
and do not have problems with the sign.

-- 
-- 
                                Michael van Elst
Internet: mlelstv@serpens.de
                                "A potential Snark may lurk in every tree."