Re: Linux TCP Fixing everyones problems? WAS(Re: TCP push sometimes missing under 2.2.5?

Geof Goodrum (ggoodrum@perigee.ncdc.noaa.gov)
Wed, 14 Apr 1999 08:40:29 -0400 (EDT)


On Wed, 14 Apr 1999, Chris Wedgwood wrote:

> On Wed, Apr 14, 1999 at 01:00:22AM +0100, Tim Fletcher wrote:
>
> > Can a kind of "compatablity mode" like the advanced router setup in
> > the config, were different fixes/workarounds are listed and can be
> > turned on or off. This means that the kernel has work arounds for
> > almost everything under the sun but most folks see only the slow
> > down that they have to due to local broken hardware on their
> > network. This means that Linux works and the true cause of the
> > problem is also highlighted.
>
> On the whole, I think this is a good idea; for some fixes this is
> certainly possible, but for others I think it would make the code
> really ugly with lots of #ifdef, #else, #endif bits all over the
> place.
>
> I guess it really depends on DaveM and Linus as to what is
> acceptable...

Realizing that this is a bit of a headache for the kernel maintainers, but
why isn't anyone else suggesting options in the kernel config? I believe
that the compatibility patches should be included in the kernel by
default, but could be deselected during a custom kernel config. The
patches could be bundled by platform, so ifdef SUN_PATCH, then all Sun
related patches get included.

Our site uses FTP Software's PC/TCP suite, and until the problems in that
suite were fixed, I always compiled in the compatibility option included
in the kernel.

I believe that separating the patches from the kernel source is a bad idea
as the kernel should maintain maximum compatability by default without
forcing users to search for patches, but I do agree with jamal that giving
the sysadmin the option to leave out unused compatibility code is
desirable. It also lets the kernel config documentation point out broken
implementations.

Geof Goodrum
Voice: +1-301-457-5100
Email: Geoffrey.P.Goodrum@noaa.gov

-
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/