ncurses tests

Eric Norum eric at
Fri Jun 2 17:11:16 UTC 2000

Charles-Antoine Gauthier wrote:
> Is this the 16-bit baud rate problem that was reported before?
> FYI, in many console device drivers, the baud rate is represented as a
> signed 16 bit integer. You can't do more than 32767 bps. This should be
> fixed if it hasn't.

32-bit value in 68360 console driver.

But here's the problem that OUTWATER <vac4050 at>
has reported:
The initial open of a device sets the speed in the termios structure to
9600 baud -- see rtems_termios_open.  The smc1Initialize routine doesn't
use this (???) so the rate stays at the 115.2 kbaud (as OUTWATER has
configured) for `normal' operation.  When curses starts, it does an
ioctl to get the termios state, sets non-cooked mode and does an ioctl
to set the state -- and ends up changing the rate to 9600 baud.

Now, where's the bug?   Should the smc1Initialize abide by the termios
speed?  Probably.  But then how would anyone get a power-up serial line
rate of anything other than 9600 baud?  I'd say that there's some
configuration information missing on startup -- but where to put it?

Eric Norum                                 eric at
Canadian Light Source                      Phone: (306) 966-6308
University of Saskatchewan                 FAX:   (306) 966-6058
Saskatoon, Canada.

More information about the users mailing list