Odd RS232 problem

[othermark]

Mike Tancsa wrote:

> At 05:53 AM 23/05/2006, Daniel O'Connor wrote:
>>On Saturday 13 May 2006 22:00, Holger Kipp wrote:
>>
>> > If you encounter silo overflows, you might need to increase
>> > cp4ticks in sio.c, eg
>> > - cp4ticks = speed / 10 / hz * 4;
>> > + cp4ticks = speed / 10 / hz * 40;
>> > and/or you might want to change hz from 1000 back to 100.
>>
>>OK, I'll try it.
> 
> This fixes the overflows for me as well. I have a number of boxes out
> in the field running with this local modification (40 vs 4). Without
> it, I have problems running a PCI modem at speeds better than 9600
> without causing overflows.

Yes, I had a laptop that exhibited this same problem and this change
did fix it.  I believe this is described in the following PR:

http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/51982

I had tried to work on this patch to provide a bounded sysctl
for the multiplier, that way you could just set the hint on 
boot up in order to fix this problem.  With a little trial and
error you could find the lowest tick value that works for the interrupt
load.

One of these days I'll get back to figuring out all the places I need to add
code to get that hint to work correctly.

-- 
othermark
atkin901 at nospam dot yahoo dot com
(!wired)?(coffee++):(wired);