> Alan Cox wrote:
> > I don't really want to see V34 in the kernel if at all possible. V34
> > isnt that real time. This is a test for the kernel I think. mlock()
> > and using the real time scheduling should be enough. If not then Ingo
> > and others have a target to polish up the scheduling patch.
>
> Well, my target is that V34 demod will go in user space, possibly with
> no real time requirement.
No real time requirements should certainly be the case.
> For modulation, it's possible that we could keep the hard stuff in user
> space, but with something in the kernel to insert idle frames should the
> Tx buffer underflow. I think that's better than dropping the line which
> _might_ happen with a cheap modem on the other end. Fabrice says V34
> modulation is not that complex. But a few things I've read give me a
> feeling that he might not have solved all the problems yet ;-) Things
> like frequency precompensation at the Tx end for example...
Its all relative - the modulator is very simple compared to the demodulator.
If you aim to have most output generated in user space, and some fudged
underflow fill in generated in the kernel, how will you make the echo
cancellor function reliably? Are you expecting the kernel to feed back its
output stream to the demodulator as it feeds the received stream? If the
output stream can underrun, its more likely to happen during the initial
training period, as the machine is working harder. Just letting the echo
cancellor hiccup during the transmit underrun might be a poor solution, as it
may screw up convergence. You could relegate the whole training process to
the kernel, but that puts more bloat in the kernel.
> > The other thing about V34 in user space is you ought to be able to use
> > it with any sound card ;)
>
> That is on my todo list :-)
Remember V34 is fairly fussy about the characteristics of the A/D and D/A
converters. A lot of sound cards may not function very well in this role.
Weird phase effects in some converters really screw up the dynamic
performance of the echo cancellor. The same phase effects don't have much
effect on the actual demod process, as this dynamically adapts well to such
effects. Still, I guess many sound cards will work OK.
Steve
-
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/