Re: vm lock contention reduction

[Rik van Riel <riel@conectiva.com.br>]

On Fri, 5 Jul 2002, Linus Torvalds wrote:
> On Fri, 5 Jul 2002, Rik van Riel wrote:
> >
> > But it is, mmap() and anonymous memory don't trigger writeback.
>
> I don't think we can fix that, without going back to the approach of
> marking any writable memory as read-only and counting it at page-fault
> time.

I don't think we have to fix it, as long as the
shrink_caches/page_launder function is well
balanced and throttling is done intelligently.

> Wasn't it you who did that test-patch originally?

I don't remember who did it, so I suspect it wasn't me.
Could it be Ben ?

> There might be some way to avoid the page fault badness (the large page
> stuff will do this automatically, for example), which might make the
> "let's keep track of dirty mappings explicitly" approach acceptable
> again.

That's another approach, I guess it'll be worth looking
into both (and searching the web for what other people
have done before us with both ;))

regards,

Rik
-- 
Bravely reimplemented by the knights who say "NIH".

http://www.surriel.com/		http://distro.conectiva.com/

--
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-mm.org/