Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 11 Jan 1996 00:08:41 +0100 (MET)
From:      Pierre Beyssac <pb@fasterix.freenix.fr>
To:        dennis@etinc.com (dennis)
Cc:        nate@rocky.sri.MT.net, hackers@freebsd.org
Subject:   Re: pppd vs ijppp
Message-ID:  <199601102308.AAA26447@fasterix.frmug.fr.net>
In-Reply-To: <199601101744.MAA24543@etinc.com> from "dennis" at Jan 10, 96 12:44:12 pm

next in thread | previous in thread | raw e-mail | index | archive | help
dennis writes:
> ioctls were designed for setting options, and the options can still be
> largely managed
> in user space. Good data communications requires good reporting, and information
> availability is substantially better and more efficient in the kernel, not
> to mention the performance penalty of a user level implementation.

A kernel is something that needs to be as simple, small and efficient/reliable
as possible, that's why it generally contains only :

	1) stuff that *can't* be elsewhere because of the architecture of
	   the system.
	2) stuff that would not be efficient enough in user space.

PPP basically needs to (A) emit/receive network (IP or other) packets
and (B) encode/decode them. (A) *needs* to be in the kernel. (B) doesn't,
except for performance (when SLIP was first implemented, CPU resources
were scarce enough that this made a big difference even on personal
workstations. Not anymore).

Dial on demand and chat script features have nothing to do in the kernel,
because they don't need to be in and you generally don't care if you lose 2
milliseconds CPU for every 20+ seconds dialout procedure...$

> Improving kernel debugging is not difficult...we debug much more complicated
> things
> than ppp within the kernel space. The fact than pppd debugging is poor is
> not a good
> reason to adulterate the entire mechanism.

Looks like you've never debugged anything in kernel space :-)

Kernel-debugging space is *always* more difficult than user-space
debugging, because starting and stopping user processes is a lot
easier than starting and stopping kernels.

> Good datacomm should focus on the 99% and not the 1%. Imagine if you moved
> ethernet processing to user space to improve debugging?

That's exactly the point. Fancy features are the 1% so they belong in user
space.
-- 
Pierre Beyssac	    pb@fasterix.frmug.fr.net pb@fasterix.freenix.fr
{Free,Net,Open}BSD, Linux : il y a moins bien, mais c'est plus cher



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199601102308.AAA26447>