Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Ways to report trace when boot panics [Was NetBSD 7.0 i386 panic during boot]



2015-10-12 7:50 GMT+02:00 Robert Elz <kre%munnari.oz.au@localhost>:
> No, that makes no sense.   The whole point is that no-one has RS232
> on anything modern - forcing it to exist, just because it is what used
> to be used helps nothing.

This generalising statement is at least wrong in an industrial
environment for example. RS232 still going strong for many reasons. It
is quite noise immune due to its "ancient" voltage levels, USB would
miserably fail here, and it still has many applications where you want
to rely on a *simple* and established interface, where you can even
screw/drill strands together yourself and it works. There are even new
laptops with RS232 sold, although more in the professional series.

> The system that's to have it as its console would still need to have
> low level (polled, no interrupts) USB access, but would also need to
> implement everything needed to talk to such a device (in addition to
> talking to the controller) ...

And you won't have to do all that with a yet to be invented and
proprietary "NetBSD debug console" USB protocol?
What about an additional, yet to be written special application to
read the console data on the other side?

> What's more, that also needs extra hardware, which is also something to
> avoid (since it is likely to be used rarely, you'd never be able to
> find it when you need it, wouldn't have it with you when travelling, ...)
>
> Keep it simple.   That way it is likely to work, and be implementable.

Generalising again? Inventing a new protocol, requiring another NetBSD
machine on the other side, and USB *device* support for the target
doesn't sound "simple" or suitable for travelling to me, too...
With ucom(4), we already have code and a well-known interface, which I
would consider "simpler", at the moment.
And can you guarantee that your target machine will support the role
of USB *device* hardware-wise when connecting to another USB host, as
you suggested? Do you have a standard cable at home to connect a USB
host port to another USB host port? You will need extra hardware! For
RS232, there are just null modem cables (and I even think "USB null
modem cables" exist as ucom(4) presenting each host an "own" device
:), here we go again).

Felix


Home | Main Index | Thread Index | Old Index