Re: benchmarking squid on solaris/x86

From: Henrik Nordstrom <hno@dont-contact.us>
Date: Wed, 20 Mar 2002 23:10:41 +0100

On Wednesday 20 March 2002 20:41, Joe Cooper wrote:

> Sorry, Andres. I'm first! (I'm first for my medication, I'm first
> for everything!)
>
> Of course, I'll have to figure out the /dev/poll patch under Linux.
> Argh. I wish they would make up their mind about what to do about
> /dev/poll.

The RT signal approach also seems to getting into shape. The RT
scheme provides some quite interesting aspects even if a good
/dev/poll probably would have been more useful as it is a lot simpler
to make correct use of, and to debug..

To me it is a bit odd that Linux do have RT signal I/O, but not a
working /dev/poll. But I guess RT signal I/O was more fun to
implement..

Regards
Henrik
Received on Wed Mar 20 2002 - 15:12:37 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:14:52 MST