Re: [PATCH 2.5.43-mm2] New shared page table patch

[Rik van Riel]

On 21 Oct 2002, Eric W. Biederman wrote:
> "Martin J. Bligh" <[e-mail]> writes:
>
> > > So why has no one written a pte_chain reaper?  It is perfectly sane
> > > to allocate a swap entry and move an entire pte_chain to the swap
> > > cache.
> >
> > I think the underlying subsystem does not easily allow for dynamic regeneration,
> > so it's non-trivial.
>
> We swap pages out all of the time in 2.4.x, and that is all I was
> suggesting swap out some but not all of the pages, on a very long
> pte_chain.  And swapping out a page is not terribly complex, unless
> something very drastic has changed.

Imagine a slightly larger than normal Oracle server.
Say 5000 processes with 1 GB of shared memory.

Just the page tables needed to map this memory would
take up 5 GB of RAM ... with shared page tables we
only need 1 MB of page tables.

The corresponding reduction in rmaps is a nice bonus,
but hardly any more dramatic than the page table
overhead.

In short, we really really want shared page tables.

regards,

Rik
-- 
Bravely reimplemented by the knights who say "NIH".
http://www.surriel.com/		http://distro.conectiva.com/
Current spamtrap:  <a href=mailto:"[e-mail]">[e-mail]</a>