Subject: Re: kern/31944 - Fix to reduce tmpfs memory usage
To: None <kern-bug-people@netbsd.org, gnats-admin@netbsd.org,>
From: Steven M. Bellovin <smb@cs.columbia.edu>
List: netbsd-bugs
Date: 04/14/2006 18:15:04
The following reply was made to PR kern/31944; it has been noted by GNATS.

From: "Steven M. Bellovin" <smb@cs.columbia.edu>
To: "Julio M. Merino Vidal" <jmmv84@gmail.com>
Cc: tech-kern@NetBSD.org, gnats-bugs@netbsd.org
Subject: Re: kern/31944 - Fix to reduce tmpfs memory usage
Date: Fri, 14 Apr 2006 14:13:35 -0400

 On Fri, 14 Apr 2006 19:51:14 +0200, "Julio M. Merino Vidal"
 <jmmv84@gmail.com> wrote:
 
 > Hello,
 > 
 > PR kern/31944 lists three issues that make tmpfs consume too much
 > memory.  The attached patch addresses the first of them.  It does the
 > following:
 > 
 > - Only keep the node identifier/generation number when deleting a node.
 >   The size of tmpfs_nid is much smaller than that of tmpfs_node.
 > - Avoid keeping nodes that were initialized but then discarded due to some
 >   error.  See the keepit parameter to the tmpfs_free_node function.
 > 
 > I have read the replies to that PR but the proposed solutions seem to be
 > "not right".  Also, one question that arises... can there be two live files in a
 > file system with the same node number?  If not, all those solutions do not
 > seem to address this...
 > 
 There's a lot of code that assumes that a <device,inode> pair uniquely
 identify files.  For example, in src/bin/pax/tables.h, there's this
 comment:
 
 /*
  * file hard link structure (hashed by dev/ino and chained) used to find the
  * hard links in a file system or with some archive formats (cpio)
  */
 
 
 		--Steven M. Bellovin, http://www.cs.columbia.edu/~smb