Thanks to everyone who applied. The OpenBSD approach to setting UFTDI baud
rates is definitely superior.
However, the root of my problem turned out to be Python. Even with the new
baud rate hardcoded in the UFTDI kernel module and manually added to termios.h,
Python was refusing to admit that it was a valid baud rate.
The issue is that Python (2.5.1) compiles its own termios interface module,
which builds a list of allowed baud rates from the defines in termios.h.
Python's termios.c does something like this:
include <termios.h>
termios_constants[] = {
{"B300",B300},
{"B1200",B1200},
{"B2400",B2400},
.
.
.
#ifdef B115200
{"B115200",B115200}
#endif
#ifdef B230400
{"B230400",B230400}
#endif
So of course my new buad rate never got added to the list. It's a fairly ugly
problem, because the valud baud rates are set in #defines in termios.h and
Python wants an array of them, and of course there's no way (that I know of) to
enumerate defines and get a list of those that start with "B" followed by
numbers (and, of course, for all I know there's some other BXXXXX define
somewhere that is not intended to indicate an allowed baud rate).
The real solution would be to use the OpenBSD UFTDI baud rate generator and
update Python's termios.c to avoid the list of valid baud rates and have it
just ask the serial port to set the requested rate and report back any error.
But that requires far more than my meager skills. I just added another
hardcoded #ifdef to Python's termios.c and it is all working now.
-Brooks
_______________________________________________
[email protected] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
