Re: RFC: COW for hugepages

[David Gibson]

On Wed, Apr 07, 2004 at 06:29:54PM +1000, Benjamin Herrenschmidt wrote:
> On Wed, 2004-04-07 at 17:53, Andrew Morton wrote:
> > David Gibson <david@xxxxxxxxxxxxxxxxxxxxx> wrote:
> > >
> > > Doing the COW for hugepages turns out not to be terribly difficult.
> > >  Is there any reason not to apply this patch?
> > 
> > Not much, except that it adds stuff to the kernel.
> > 
> > Does anyone actually have a real-world need for the feature?
> 
> Yup, porting some apps to use hugepages, when those apps
> rely on fork & cow semantics typically. Also, implicit use of
> hugepages (usually via a malloc override library).

I also have some experimental patches for putting ELF segments into
large pages (for HPC FORTRAN monsters with massive arrays in the BSS).
MAP_PRIVATE semantics (and hence COW) are a clear prerequisite for
that.

-- 
David Gibson			| For every complex problem there is a
david AT gibson.dropbear.id.au	| solution which is simple, neat and
				| wrong.
http://www.ozlabs.org/people/dgibson
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/