Source-Changes-D archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: CVS commit: src/sys/kern

On Wed, Aug 10, 2011 at 07:22:06AM +0000, YAMAMOTO Takashi wrote:
 > > > why is this necessary?  can't we just let VOP_LOOKUP fail?
 > > 
 > > Not to fix PR 44658.
 > it's better to fix the vn_isunder check instead of avoiding running it.
 > IMO vn_isunder should return acutal error code (eg. ENOENT) rather
 > than just a boolean so that callers can decide what to do.

For some reason I thought that had already been done and found to be
inadequate. Not sure why I thought that, since it's obviously false.


 > > > the v_size == 0 check sounds wrong.  does it work for eg. nfs?
 > > 
 > > It apparently does break nullfs, so I've reverted it.
 > > 
 > > Is there any way to check this correctly/safely above the filesystem?
 > if "above the filesystem" means "without calling VOPs", i don't think
 > there's a way.

Yeah, that's what I'd thought, which is why checking the size looked
promising. Oh well.

David A. Holland

Home | Main Index | Thread Index | Old Index