On Mon, 22 Jul 2002, Martin J. Bligh wrote:
> > Was it purely Oracle which drove pte-highmem, or do you think
>
> I don't see you can get into pathalogical crap without heavy
> sharing of large amounts of data .... without sharing, you're at
> a fixed percentage of phys mem - with sharing, I can have more
> PTEs needed that I have phys mem.
... for which pte_highmem wouldn't fix the problem, either.
>From what I can see we really want/need 2 complementary
solutions to fix this problem:
1) large pages and/or shared page tables to reduce page
table overhead, which is a real "solution"
2) page table garbage collection, to reduce the amount
of (resident?) page tables when the shit hits the fan;
this is an "emergency" thing to have and we wouldn't
want to use it continously, but it might be important
to keep the box alive
Since I've heard that (1) is already in use by some people
I've started working on (2) the moment Linus asked me to put
the dentry/icache pages in the LRU ;)))
cheers,
Rik
-- Bravely reimplemented by the knights who say "NIH".http://www.surriel.com/ http://distro.conectiva.com/
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Jul 23 2002 - 22:00:38 EST