Messages in this thread | | | Date | Sat, 15 Nov 1997 03:26:31 +0100 | From | Martin von Loewis <> | Subject | Re: knfsd and system crashes |
| |
> I've currently got the third-level (when we have a parent dentry) search > working, but I'm a little unclear on the best strategy for implementing > the full volume search given device number and parent inode/inode > number. Can you offer some suggestions here?
PMFJI, I think the problem is the assumption that the volume has to be searched in the first place. Most file systems can open an inode very efficiently when given an inode number.
The Linux VFS currently offers the iget function to retrieve an inode when given an inode number. Unfortunately, this interface is not very reliable when it comes to error checking, like when the client gives a stale handle. Nevertheless, this is IMHO the route to go. You can either extend the read_inode operation to support better error reporting for deleted inodes, or introduce new interface.
Searching the entire volume is not an option, IMHO - at least not on file systems that have the clear notion of inode numbers.
Regards, Martin
| |