Re: Firewire/USB Card for Qube 2

--- Izumi Tsutsui <[email protected]> wrote:

> First, Qube2 is much slower than modern x86
> machines, and
> only "overmuscled" CPUs can saturate 100BaseTX
> network back in 1999.
> For example, the Tulip Ethernet chip can't transfer
> receive packets
> into unaligned DMA buffer and kernel has to copy all
> RX packets
> from DMA buffer to mbufs.
> 
> Second, implementation of network stack is too
> complicated and
> we can't see which part (CPU, devices, drivers, VM,
> protocol layer,
> checksumming, etc. etc) has bottleneck without
> proper benchmarks.
> 
> I agree current NetBSD/mips implementation is not so
> optimized
> for MIPS3 CPUS, though.
> ---
> Izumi Tsutsui
> 

While I agree with you about the DMA implementation (I
honestly don't know the codebase), I think the CPU
alone is capable of handling the traffic. After all it
is comparable to ARMs, used in all embedded appliances
in performance. It will be running little hot for the
moment, but it should be able to take the speed.

Most likely the DMA circuitry is to blame is what you
are saying, but I my guess is that the rest of the
circuitry was probably put together in a hurry.

Anyway the Qube is still a nice and practical solution
and I still love it. It is my Internet workhorse for
my home and regardless of the speed it is doing a
great job.

Thank you again for your hard work on it!