[comp.sys.sun] Hardware flow control under SunOS4.0 question

khai%amara.uucp@mailgw.cc.umich.edu (S. Khai Mong) (02/16/89)

SunOS 4.0 supports flow control for the serial ports.  The question is how
do I set up my systems such that flow control is always on by default,
such as for outgoing uucp calls, incomming dialups.

The only way I seem to be able to set the hardware flow control is from a
program using ioctl calls.  I assume that the hardware bits are reset when
your program terminates, so that it is useless for my purpose.

--
Sao Khai Mong:   Applied Dynamics, 3800 Stone School Road, Ann Arbor, Mi48108
(313)973-1300 (uunet|sharkey)!amara!khai  khai%amara.uucp@mailgw.cc.umich.edu

joerg@sun.com (Joerg Schilling - H. Berthold AG Berlin) (03/02/89)

If you set up Hardware flowcontrol on the serial lines onboard (/dev/ttya
& /dev/ttyb) youwill have no luck. I do not know if this also applies to
other serial lines.

Fact for the ZILOG chip on the CPU borad is:
If CTS is set to AUTOENABLES (Hardware flowcontrol) the chip will not
allow you to talk to your modem if no carrier is present. This means that
you are no longer able to dial out ....... SORRY :-(

       J. Schilling 
        H. Berthold AG 
        Teltowkanalstr. 1-4 
        D 1000 Berlin 46 
        +49 30  7795 - 400 

joerg@berthold.DE

... tub!berthold!joerg 
... unido!berthold!joerg 
... sunmuc!berthold!joerg 
... sun!sunmuc!berthold!joerg 

guy@uunet.uu.net (Guy Harris) (03/14/89)

>If you set up Hardware flowcontrol on the serial lines onboard (/dev/ttya
>& /dev/ttyb) youwill have no luck. I do not know if this also applies to
>other serial lines.

It should apply to ALM-2 lines, since they use the same chip (although I
don't remember whether all ALM-2s ran all the interesting RS-232 lines to
the chip or not).  I don't think it applies to the ALM/ALM-1/Systech/MTI,
since that uses a different chip; of course, that chip *always* provides
hardware flow control and, like the Zilog chip, couples it to the "shut
off the receiver when DCD drops" feature.  Systech and the SunOS Systech
driver got around that, as I remember, by tying DCD high and using some
other line as DCD; this may mean that the problem doesn't apply to the
ALM-1.