Re: [PATCH] workaround for lost dirty bits on x86 SMP

["Stephen C. Tweedie" <sct@redhat.com>]

Hi,

On Mon, Sep 11, 2000 at 09:36:35PM -0400, bcrl@redhat.com wrote:

> Fwiw, with the patch, running a make -j bzImage on a 4 way box does not
> seem to have made a difference.

Of course it won't, because you aren't testing the new behaviour!
Anonymous pages are always dirty, and shared mmaped pages in
MAP_PRIVATE regions are always clean.  The only place where you need
to track the dirty bit dynamically is when you use shared writeable
mmaps --- can you measure a performance change there?

Cheers,
 Stephen
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux.eu.org/Linux-MM/