Re: [reiserfs-dev] Re: [PATCH] sparc32: wrong type of nlink_t

[David S. Miller]

   From: Oleg Drokin <[e-mail]>
   Date: Thu, 5 Sep 2002 09:56:38 +0400
   
   On Thu, Sep 05, 2002 at 07:48:58AM +0200, Tomas Szepe wrote:
   > >    Does the internal reiserfs nlink value translate directly
   > >    to what stat() puts in st_nlink?
   > > It really doesn't matter.  Even if you have some huge value that can't
   > > be represented in st_nlink, you can report to the user that st_nlink
   > > is NLINK_MAX.
   > > This is one possible solution to this whole problem.
   >
   > And a pretty straightforward one, too. Convert the internal reiserfs
   > link stuff to an unsigned short, find NLINK_MAX using the code I posted
   
   Too bad it is 32bit nlink field in on disk format ;)
   
We're only talking about what the user is told is the
nlink value, not what you happen to compute and put/get
from disk.

Your nlink can be legitimately 128-bits if you want, it
still can be made to work :-)

   > last night (or maybe simply grab it from userspace includes) and add
   > a check to your stat() code to return NLINK_MAX if necessary.
   
   Ok, I think I will rework it to something sensible, because current code is
   somewhat a mess and corrupt correct nlink value on overflows. Hm.
   
Ok.

Franks a lot,
David S. Miller
[e-mail]