tres@rap.ucar.EDU (Tres Hofmeister) writes:
> 	I'm noticing that pppd seems to have trouble with binary file
> transfers (certainly on my system, anyway... :).  Text transfers are
> fine, but binaries cause rcp(1) or ftp(1) to hang regularly, for
> 20-30 seconds at a time.  I'm running 1.0, with the base and comp
> stuff from ftp.eskimo.com built from the March 5th sources, on an
> SE/30.  Kernel is GENERIC #29 from the same place.
> 
> 	I'm assuming my modem configuration is fine (hardware flow
> control, appropriate cable, etc.), as I have no such trouble under
> MacOS using MacPPP.  Here's /etc/ppp/options:
> 
> tty00 38400 defaultroute passive crtscts netmask 255.255.255.0 connect "chat -f /etc/ppp/chatscript"
> 
> 	Any suggestions?  Thanks...

I've noticed something similar with a -current kernel built with the
OCELOT config from sources supped on March 11.  I'm also getting
console messages like this:

ser0: silo overflow, oflo = 0

This message apparently comes from the serial driver (the only
occurrence of 'silo' in /usr/src/sys/arch/mac68k/dev/ser.c), but I
haven't yet figured out exactly what the code is doing there.  It only
seems to happen when the serial port is active _and_ I'm moving the
mouse or typing.  Could the serial and adb drivers be stepping on each
others' toes somehow?

Hmm... I'm also running my serial (ppp) connection at 57600; it seemed
to work fine with older kernels.  Could that be causing the overflow
messages with the -current kernel?

- monroe
-------------------------------------------------------------------------
Monroe Williams    aka    monroe@cs.pdx.edu    aka    monroe@teleport.com