Port-sparc archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: Magma bizarrity?
On Wed, Feb 25, 2009 at 04:01:51AM -0500, der Mouse wrote:
Thanks to a fellow Sun fan's generosity, I just acquired a Magma
16-port serial card. It's got a breakout box which mounts directly on
the back of the card, with 16 8P8C jacks on it. I've worked out the
pinout of ecah port, and it looks a little odd, but that's not a
problem since I'm going to be doing my own cabling anyway.
Magma pinouts:
http://www.magma.com/support/faqs/sbus/sun.htm
But there's a stranger surprise lurking. The ports on the breakout box
are arranged in a 4x4 grid and marked with the numbers 0 through 15 in
an obvious way:
0 1 2 3
4 5 6 7
8 9 10 11
12 13 14 15
But the arrangement of the ports in terms of minor device number,
rather than of the printing on the breakout box, is
14 5 0 10
13 4 8 9
12 7 3 11
15 6 1 2
This looks totally bizarre. I looked at the code, thinking maybe it
was some kind of race which port got which tty unit, but the code is
saner than that; port numbering is based on chip and channel numbers,
not something quasi-random like scheduling of kernel threads. (Also,
the arrangement does not change with a reboot, though that's a fairly
weak test.)
I assume you used MAKEDEV under netbsd to make the device special
files? How has that file changed since 1.4T and -current?
Cheers,
Jonathan.
Home |
Main Index |
Thread Index |
Old Index