--=-9kXm8VncJ1yWw1yMzbVU
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Thu, 2004-11-25 at 11:12 +0200, Toomas Soome wrote:
> Volker Lendecke wrote:
> > How much does sendfile gain us at all? I've had to debug really nasty p=

roblems,
> > I scratched my head why samba simply stops responding, making XP reconn=

ect.
> > sendfile =3D no was the very simple answer after hours of staring at sn=

iffs,
> > trying to reproduce the errors. If we don't see a huge performance impr=

ovement
> > I'd VERY strongly argue that we dump sendfile completely. It is just wa=

y too
> > buggy in common kernels.

>=20
> linux is not the only os around.


That's actually the argument against sendfile(). If it were not used on
Linux, then it has better chance of being broken by Linux-bound
developers. (Unfortunately it does happen, even with the build farm).

Andrew Bartlett

--=20
Andrew Bartlett

--=-9kXm8VncJ1yWw1yMzbVU
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (GNU/Linux)

iD8DBQBBpcMDz4A8Wyi0NrsRAnA9AJ9mIngTS8BrANIaN84NZ0 Y5CZgZqwCcCZ0y
RRrIJV5UQkkwHzMGRXQq/jI=
=iDcg
-----END PGP SIGNATURE-----

--=-9kXm8VncJ1yWw1yMzbVU--