[reiserfs-list] RE: [OpenAFS-devel] weird issue with reiserfs on 2.4.3 and openaf s

[Derek Atkins]

I don't know.  I suspect we'd have to create a hash table and
store the extra information key'd of the inode number.

-derek

Chris Mason <mason@suse.com> writes:

> On Wednesday, April 11, 2001 01:38:30 PM -0400 Derek Atkins
> <warlord@MIT.EDU> wrote:
> 
> > Ok, how about this..  Let's assume that in user-mode we bring the file
> > into the inode cache and then pass just the inode number down to the
> > kernel.  In the kernel, is there some (easy) way to notice that we're
> > using reiserfs?  
> 
> No clean way to check for reiserfs, but in the kernel, we can detect for
> "provides interface to get around screwy behaviour". 
> 
> > If so, perhaps when we initiate the cache file we can
> > check if it's reiserfs and, if so, open the inode (which is currently
> > in the cache), grab the extra data, save the extra data (somewhere),
> > and then release it.  Then, later, we can check the saved data and
> > notice that the file was reiserfs and behave specially.
> > 
> > I really don't like this solution.. It's a complete kludge and
> > special-case for reiserfs..  But I think it would solve our problem
> > until, as you say, we move to 64-bit inodes :)
> > 
> 
> This is what I had in mind.  Where would we store the extra stuff in AFS?
> 
> -chris
> 
> 
> 

-- 
       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord@MIT.EDU                        PGP key available