The following reply was made to PR kern/35253; it has been noted by GNATS.

From: Elad Efrat <elad@NetBSD.org>
To: YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>
Cc: gnats-bugs@NetBSD.org
Subject: Re: kern/35253
Date: Tue, 26 Dec 2006 12:37:22 +0200

 YAMAMOTO Takashi wrote:
 >>>>> it's better to just remove fileassoc_table_add from API
 >>>>> and just do them internally, IMO.
 >>>> you mean, when a file entry is added, create the table if it's not
 >>>> already there?
 >>> yes, and resize hash appropriately according to the number of entries.
 >> on each addition? :)
 > 
 > appropriately. :)
 > 
 >> maybe we can add some "barriers", to intelligently resize the table...
 >>
 >> -e.
 > 
 > common ways are:
 > 
 > - rehash only when the difference between actual and ideal size is big
 >   enough.
 
 by ideal I guess you mean the "size it was last optimized for"?
 
 > - use a timer to avoid too frequent rehash.
 
 can that be done in the kernel safely?
 
 both sound good to me.
 
 -e.