Re: Which is better at vm, and why? 2.2 or 2.4

Patrick McFarland (unknown@panax.com)
Sat, 13 Oct 2001 14:53:42 -0400


--vNrHrykRFvLVX6W3
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Also, I'd like to say about the documentation...

<quote>
Currently, these files are in /proc/sys/vm:
- bdflush
- buffermem
- freepages
- kswapd
- overcommit_memory
- page-cluster
- pagecache
- pagetable_cache
</quote>

but a simple ls of /proc/sys/vm reports:
bdflush kswapd overcommit_memory page-cluster pagetable_cache

Shouldnt the documentation be updated, seeing for the fact it was written i=
n the 2.2.10 days?

On 13-Oct-2001, Patrick McFarland wrote:
> Ill reiterate something here, im on a p133 with 16 megs. Yeah, the kind o=
f the crappy ide controller that eats cpu time to swap. (Enough so that my =
mouse pointer will freeze in X that its swapping so much. Swapping is the o=
nly thing ive found that can pull that off) Swapping the least ammount woul=
d be the best for a box like that.
>=20
> On 13-Oct-2001, Rik van Riel wrote:
> > On Sat, 13 Oct 2001, Patrick McFarland wrote:
> >=20
> > > Hmm, I see that as very bad. There should be a bunch of sysctls to do
> > > that easily.
> >=20
> > See /proc/sys/vm/* and the documentation ;)
> >=20
> > > Also, I heard that 2.4 (and I'm assuming 2.2 as well) swaps pages on a
> > > last-used-age basis, instead of either a number-of-times-used or a
> > > hybrid of the two. That kinda seems stupid,
> >=20
> > Don't worry since it's not true, at least the VM in the -ac
> > kernels _does_ use a hybrid of access recency and frequency
> > to determine page replacement.
> >=20
> > The -linus kernel, however only has LRU-like selection.
> >=20
> > At the moment the -linus kernel is faster than the -ac kernel
> > for some workloads. This may have something to do with better
> > clusterable IO ... when page replacement is less precise the
> > chance that IO is clusterable is probably larger due to the
> > way we scan.
> >=20
> > I plan to do more explicit IO clustering in -ac to try and
> > remedy this difference.
> >=20
> > regards,
> >=20
> > Rik
> > --=20
> > DMCA, SSSCA, W3C? Who cares? http://thefreeworld.net/ (volunteers ne=
eded)
> >=20
> > http://www.surriel.com/ http://distro.conectiva.com/
> >=20
>=20
> --=20
> Patrick "Diablo-D3" McFarland || unknown@panax.com

--=20
Patrick "Diablo-D3" McFarland || unknown@panax.com

--vNrHrykRFvLVX6W3
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)

iD8DBQE7yI208Gvouk7G1cURApQbAKC/S+jmYZ1uGjqgNwPFp5TH7a3wVQCfWb/0
0iACLOI7o0rY3zvq22RQsWQ=
=sG/R
-----END PGP SIGNATURE-----

--vNrHrykRFvLVX6W3--
-
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/