Re: Multithreaded TCP/IP stack

From: Stephen Torri (s.torri@lancaster.ac.uk)
Date: Thu Jul 13 2000 - 11:45:07 EST


Thanks to all who responded to this. I will relay the message on. As was
said by various people, misinformation can be turned out by the "spin
doctors" to paint the picture that is most flattering to their product.

The analogy of who is better in the operating system comes down to cars.
Some look truly decked out with the latest this and that to appear to be
quick. Yet when you put the pedal to the floor you realize its a pretty
car but a gutless wonder. The other side of the picture are cars that
appear normal. These are "sleepers". Secretly underneath the owners have
tricked out the engine. Other laugh at them until the see the back of the
car as it leaves them in the dust. Linux doesn't have the shinny
accessories but doesn't really need it. We don't even need to boast about
it because it does its own marketing by its performance. We just work at
this to make it work better than anything out there. If we are not better
in an area lets learn why and fix that. If we are better than lets
maintain that. So let the spin doctors make their product look nice while
we leaving them far behind.

Question answer. Thread terminated.

Stephen

On Thu, 13 Jul 2000, Alan Cox wrote:

> > A friend of mine is critical of linux for not having a multithreaded
> > TCP/IP stack. Is that true? If it is not then what kernel version
>
> The tx path and irq path have been multithreaded since 0.97 I believe.
> The receive path in userspace likewise. The receive stuff running in
> interrupt and post irq context is fully SMP multithreaded in 2.4test
>
>

-
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 : Sat Jul 15 2000 - 21:00:17 EST