[4288] in linux-net channel archive
Tell me why I'm attracting flames?
daemon@ATHENA.MIT.EDU (Olaf Titz)
Tue Sep 3 22:42:13 1996
Cc: recipient.list.not.shown:;@vger.rutgers.edu
From: Olaf Titz <olaf@bigred.inka.de>
Date: 03 Sep 1996 23:28:03 +0200
To: ;@unlisted-recipients (no To-header on input)
Alan Cox <alan@cymru.net> wrote:
> Repeat for about the 10,000th time on linux-kernel. the route lookup code
> is atomic. kerneld route lookup has to fail like that, use diald to do
> dial on demand stuff. This isnt going to change and as many people as they
Perhaps someone can just explain to me where the race condition is so
that I can actually _understand_ the stuff, instead of simply throwing
flames at suggestions. I get a bit tired of that lately. :-(
(At least it is anything but obvious that the code is atomic, it has
explicit locking in it and I have tried carefully to find where it
occurs.)
The request-route stuff is extremely useful in that it can greatly
simplify the routing setup. Just more nice than fiddling with diald,
esp. when the latter would mean to stuff everything through a chain of
slip interfaces. Yes, "niceness" does matter.
Actually, it works for my setup without the patch, I just put a ping
before each critical connection request. A fully transparent routing
setup would be a big advantage for Linux in non-trivial networking
situations. I don't see the cause for bashing here. (Except perhaps
for the pro/con kerneld religious wars of the past which I had
killfiled, then.)
olaf
--
___ Olaf.Titz@inka.de or @{stud,informatik}.uni-karlsruhe.de ____
__ o <URL:http://www.inka.de/~bigred/> <IRC:praetorius>
__/<_ >> Just as long as the wheels keep on turning round
_)>(_)______________ I will live for the groove 'til the sun goes down << ____
