Re: Stability (2.2.14/15/16/17pre1)

From: Rik van Riel (riel@conectiva.com.br)
Date: Wed Jun 14 2000 - 13:44:44 EST


On Wed, 14 Jun 2000, Andrea Arcangeli wrote:
> On Wed, 14 Jun 2000, Marcelo Tosatti wrote:
>
> >2.2.15, mm/vmscan.c, kswapd thread:
> >
> > while (1) {
        [snip 2.2.16 kswapd loop]

> >What i'm missing here ?
>
> linux-2.2.15/mm/vmscan.c line 383 or linux-2.2.16/mm/vmscan.c line 390.

That's nice, but it doesn't seem to help. In theory you are right,
but since dozens of people complained about 2.2.15 VM performance
I guess things didn't exactly work out as expected...

> >mmap002 on 2.2.15 gets killed. mmap002 on 2.2.15 + 2.2.16's
> >thrasing heuristic runs fine. Try it.
>
> I bet mmap002 gets not killed in 2.2.16 because of the other
> stuff in vmscan.c (see flushcount addition). trashing_mem or not
> trashing_mem isn't a stability issue.

Indeed. It _did_ get killed, however, because try_to_free_pages()
in 2.2 can sometimes fail. By starting to call try_to_free_pages()
from applications _before_ memory gets critical the machine is
able to withstand far bigger memory loads without having to resort
to killing an application.

I've looked at trying to make try_to_free_pages() failsafe, but
there doesn't seem to be a way to achieve this without changing
too much of the way 2.2 VM works ;(

regards,

Rik

--
The Internet is not a network of computers. It is a network
of people. That is its real strength.

Wanna talk about the kernel? irc.openprojects.net / #kernelnewbies http://www.conectiva.com/ http://www.surriel.com/

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Thu Jun 15 2000 - 21:00:34 EST