Bernard Mentink wrote:
> >A PC is USELESS for any kind of bitbanging or the like
>
> Rubbish! The bottleneck is OpenOCD, A PC is more than capable to do
> all the work necessary to support "dumb (read simple)" interface chips
> like the FT2232,
I think you missed my point; even though the PC CPU is really fast,
there are so many busses between that CPU core and the JTAG signals it
drives, that really low latency is not really realistic.
I didn't miss your point. I just think that it doesn't matter
how many busses you have between CPU and JTAG, the only latency of
concern is the USB 1ms frame-rate, and that can be worked around very
easily, as I have done with running a USB->CANx2 at ~6000 frames/sec
each channel.
> I fully support improving OpenOCD on cheap adapters rather than
> "relying" on the purchase or building of an intelligent adapter ..
I think you missed my other point; since years and years it has been
quite possible to create intelligent adapters without high cost. But for
whatever reason, the only such product I've seen is Simon's Versaloon.
I'm glad it is there, because it is really nice!
Again, I didn't miss the point, I understand that in-expensive
adapters can be made. I just don't want to be forced to use certain
hardware ... opensource is all about choice ... I should be able to run
my FT2232 adapter at a reasonable speed, so if the aim of openocd is to
support a variety of adapters, then it should do them all at the best
speed possible .... For the FT2232D/LPCxxxx combo, that should be
256kB/sec, do you see my point?
There is also the problem of loading firmware on the device ...
Is openOCD going to do that transparently? I don't think so ....
Altera released their 8051 STAPL bytecode player with an open license
*in February 1999*. "We" are more than 10 years behind.
So??
I'm sorry that you thought I was advocating more expensive adapters,
that's pretty much the opposite of my point.
That wasn't my point, Sorry I didn't communicate that
effectively :-)
Cheers,
Bernie
_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development