>eh0   1500  <Link>      0.0.a4.0.45.b1       77732     0    89352  1281 60220

That seems to be rather a lot of collisions....

Ok there are various problems with the etherH driver. These have been getting
worse as I have improved the interrupt system. I have some more recent kernels
that I have not put online as the etherH driver is even worse.
The state of play is this ...

The current etherH driver is dying. From what Nut has told be, there is no
way that it can be fixed.
Nut has a new etherH driver rewritten from scratch. This driver is apparently
faster and a lot more reliable. However this driver will not be available to me
until the end of September as I understand it as that is the earliest that Nut
can get a copy to me.

So this will mean that nothing can really be done to fix the etherH problems
until then.

I will be releasing new kernels BUT if you use the etherH driver then don't use
them.
The problem with the recent kernels (~4530) is that if the etherH driver gets
upset during a DMA transfer it hangs the bus and thus the machine. I have found
that this happens frequently when transmitting packets and for development have
had to switch to an etherB.

More news when I get it.

Cheers,
				Mark

-- 
Mark Brinicombe				amb@physig.ph.kcl.ac.uk
Research Associate			http://www.ph.kcl.ac.uk/~amb/
Department of Physics			tel: 0171 873 2894
King's College London			fax: 0171 873 2716