--LNKrWK8T5LDJo+fl
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Fri, Aug 05, 2005 at 03:26:25PM -0700, Bill Studenmund wrote:
> I think we send an ack each time userland reads something, in the hopes o=
f=20
> keeping the window as open as possible. That's kinda good. However it'd b=
e=20
> nice to meter them somehow.
>=20
> The problem is that to delay sending them means we need some sort of=20
> timeout.

Yes, we do; it's visible at net.inet.tcp.delack_ticks.

I suspect that yamt's issue is some other ack-generating event that's
not interacting with the delack mechanism - either not getting folded
into the pending delack, or not cancelling it as a result of sending
now.  I, too, would like a better answer :)

--
Dan.



--LNKrWK8T5LDJo+fl
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (NetBSD)

iD8DBQFC8++oEAVxvV4N66cRAtztAKDcW6sDjury+jgNQkJB6LpYC7pJwwCg5fz6
IIQcCeCYalYfDb+5WKq2X28=
=juFN
-----END PGP SIGNATURE-----

--LNKrWK8T5LDJo+fl--