[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-fsdevel
Subject:    Re: block bitmap readahead for ext2
From:       Andi Kleen <ak () suse ! de>
Date:       2001-01-24 16:13:11
[Download RAW message or body]

On Wed, Jan 24, 2001 at 11:00:41AM -0500, Jeff Darcy wrote:
> From: "Andi Kleen" <ak@suse.de>
> > Does this mean that unwriten extents are supported now in pagebuf?
> Otherwise
> > this ioctl would need to prezero the disks blocks to prevent old data
> > from being leaked, right?
> 
> I don't know specifically about the current Linux pagebuf code, but in
> general this is not necessarily the case.  If the interfaces are set up so
> that one can know when the whole block is about to be overwritten with new
> data, no zeroing is necessary for that block.

This is what XFS does on Irix according to the design documentation on the
XFS site.  It has special unwriten extents that on read return zeroes.
On write an unwriten extent is turned into a normal extent.
In the Linux port the VM support needed for that just seems not to be 
implemented yet. Pagebuf in this case is a XFS datastructure to describe 
multi page objects that are continuous on disk in memory -- normal linux page 
cache only knows about allocated individual pages and doesn't support
preallocation. 

-Andi
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org

[prev in list] [next in list] [prev in thread] [next in thread]