Re: [RFC] changes to buffer.c (was Test12 ll_rw_block error)

[Chris Mason <mason@suse.com>]

On Friday, December 22, 2000 17:45:57 +0100 Daniel Phillips
<phillips@innominate.de> wrote:

[ flushing a page at a time in bdflush ]

> Um.  Why cater to the uncommon case of 1K blocks?  Just let
> bdflush/kupdated deal with them in the normal way - it's pretty
> efficient.  Only try to do the clustering optimization when buffer size
> matches memory page size.
> 

This isn't really an attempt at a clustering optimization.  The problem at
hand is that buffer cache buffers can be on relatively random pages.  So, a
page might have buffers that are very far apart, where one needs flushing
and the other doesn't.

In the blocksize == page size case, this won't happen, and we don't lose
any speed over the existing code.  In the blocksize < pagesize case, my new
code is slower, so my goal is to fix just that problem.

Real write clustering would be a different issue entirely, and is worth
doing ;-)

-chris

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/