[comp.sys.att] Hardware Flow Control on tty

dave@westmark.WESTMARK.COM (Dave Levenson) (12/22/90)

I am trying to use a high speed V.32 modem on a tty port using AT&T
SysV/386r3.2.2.  The modem allows the DTE and DCE speeds to differ,
and then offers a choice of flow control methods.

The HARDWARE flow-control option causes the modem to assert CTS
until its buffer is nearly full, and then to drop CTS.  UNIX seems
to ignore the not-CTS state, and continues to send data. Is there
any way to get the UNIX tty device driver to recognize and react to
this signal?

The SOFTWARE flow-control option causes the modem to send X-OFF (^S)
to the the host when its buffer is nearly full, and then to send
X-ON (^Q) when it can again accept data.  In this mode, it is
incompatible with UUCP which wants to send binary data including
X-ON and X-OFF through the modem.
-- 
Dave Levenson			Internet: dave@westmark.com
Westmark, Inc.			UUCP: {uunet | rutgers | att}!westmark!dave
Warren, NJ, USA			AT&T Mail: !westmark!dave
[The Man in the Mooney]		Voice: 908 647 0900  Fax: 908 647 6857

dag@esleng.uucp (David A. Gilmour) (12/23/90)

In article <2173@westmark.WESTMARK.COM> dave@westmark.WESTMARK.COM (Dave Levenson) writes:
>I am trying to use a high speed V.32 modem on a tty port using AT&T
>SysV/386r3.2.2.  The modem allows the DTE and DCE speeds to differ,
>and then offers a choice of flow control methods.
>
>The HARDWARE flow-control option causes the modem to assert CTS
>until its buffer is nearly full, and then to drop CTS.  UNIX seems
>to ignore the not-CTS state, and continues to send data. Is there
>any way to get the UNIX tty device driver to recognize and react to
>this signal?
>

I had exactly the same problem with my ISC SVR3.2 system using a T2500
modem.  It turned out that RTS/CTS handshaking is *NOT* supported in the
ISC drivers.  This may be the problem with yours.

I fixed the problem by installing the FAS 2.07 drivers.  I don't know
if they will help you or not.  The source is available on the net, or
I can mail it to you.


-- 
__________________________________________________________________________
David A. Gilmour            |   
Excalibur Systems Limited   |   uunet!mitel!cunews!micor!esleng!dag
Kanata, Ontario, Canada     |

gil@limbic.ssdl.com (Gil Kloepfer Jr.) (12/23/90)

In article <2173@westmark.WESTMARK.COM> dave@westmark.WESTMARK.COM (Dave Levenson) writes:
>I am trying to use a high speed V.32 modem on a tty port using AT&T
>SysV/386r3.2.2.  The modem allows the DTE and DCE speeds to differ,
>and then offers a choice of flow control methods.

As of the last time I investigated this, the TTY driver on SVR3.2.2
didn't have HFC.  The only way to do this is to replace the TTY driver
with some other TTY device driver, such as FAS.

FAS is a public-domain TTY device driver which, as well as supporting
hardware flow control, also allows FIFO support for the NS16550 UART.
This is important when you're running a modem/terminal at high speeds
without losing characters.

There are some bugs in FAS, so you'd want to keep your eyes open for
new versions.  I'm using it on my 386 though, and it's worth it even
with the few bugs.

-- 
Gil Kloepfer, Jr.              gil@limbic.ssdl.com   ...!ames!limbic!gil 
Southwest Systems Development Labs (Div of ICUS)   Houston, Texas
"There are beautiful people I wish would have never opened their mouths,
because such ugliness oozes out."  Philosophy Prof. at NYIT

bill@unixland.uucp (Bill Heiser) (12/24/90)

In article <119@limbic.ssdl.com> gil@limbic.ssdl.com (Gil Kloepfer Jr.) writes:
>
>FAS is a public-domain TTY device driver which, as well as supporting
>hardware flow control, also allows FIFO support for the NS16550 UART.
>This is important when you're running a modem/terminal at high speeds
>without losing characters.

An interesting note here -- I haven't noticed any character loss running
a Trailblazer on my Esix system -- someone told me that the Esix asy
driver is a "clone" of the FAS driver.  The Esix driver is also 
reputed to support the NS16550 UART (although I've failed to get it 
to even work AT ALL with the WD version of that chip).  I have yet to
try the WD chips in the AST 4-port board I just installed.

Bill


-- 
home:	...!{uunet,bloom-beacon,esegue}!world!unixland!bill
	bill@unixland.uucp    Public Access Unix - Esix SYSVR3
508-655-3848(12/24)  508-651-8723(12/24/96-HST)  508-651-8733(12/24/96-PEP-V32)
other:	heiser@world.std.com

src@scuzzy.in-berlin.de (Heiko Blume) (12/29/90)

gil@limbic.ssdl.com (Gil Kloepfer Jr.) writes:
>There are some bugs in FAS, so you'd want to keep your eyes open for
>new versions.

you bet :-) in fact i run fas-2.08beta, no more problems.
[don't ask me to send it, wait for the real release]
-- 
      Heiko Blume <-+-> src@scuzzy.in-berlin.de <-+-> (+49 30) 691 88 93
                    public source archive [HST V.42bis]:
        scuzzy Any ACU,f 38400 6919520 gin:--gin: nuucp sword: nuucp
                     uucp scuzzy!/src/README /your/home

larry@nstar.rn.com (Larry Snyder) (12/29/90)

src@scuzzy.in-berlin.de (Heiko Blume) writes:

>you bet :-) in fact i run fas-2.08beta, no more problems.
>[don't ask me to send it, wait for the real release]

will 2.08 support streams and run under Unix release 4.0?

-- 
       Larry Snyder, Northern Star Communications, Notre Dame, IN USA 
  {larry@nstar.rn.com, uunet!nstar!larry, larry%nstar@iuvax.cs.indiana.edu}
                     backbone usenet newsfeeds available
         Public Access Unix Site (219) 289-0282 (5 high speed lines)

gemini@geminix.in-berlin.de (Uwe Doering) (12/30/90)

larry@nstar.rn.com (Larry Snyder) writes:

>src@scuzzy.in-berlin.de (Heiko Blume) writes:
>
>>you bet :-) in fact i run fas-2.08beta, no more problems.
>>[don't ask me to send it, wait for the real release]
>
>will 2.08 support streams and run under Unix release 4.0?

Someone here in Germany offered me patches for SVR4. But until now I
couldn't contact the person who actually made the changes. Maybe he
is on vacation.

The problem is that I want to release FAS 2.08 not later than January 15th.
There are two nasty bugs in FAS 2.07 that could lead to hung ports and
immortal processes. Therefor, I have to roll out FAS 2.08 soon, before
even more people will be bitten by these bugs. If I haven't heard anything
about the SVR4 patches by then I will post FAS 2.08 without SVR4 support.

If anyone else has ported FAS 2.07 to SVR4, please let me know, and send
me the patches. I will do my best to adapt them to FAS 2.08, hopefully
before the above mentioned deadline.

     Uwe
-- 
Uwe Doering  |  INET : gemini@geminix.in-berlin.de
Berlin       |----------------------------------------------------------------
Germany      |  UUCP : ...!unido!fub!geminix.in-berlin.de!gemini

fischer@utower.tower.in-berlin.de (Axel Fischer) (12/30/90)

larry@nstar.rn.com (Larry Snyder) writes:

>src@scuzzy.in-berlin.de (Heiko Blume) writes:

>>you bet :-) in fact i run fas-2.08beta, no more problems.
>>[don't ask me to send it, wait for the real release]

>will 2.08 support streams and run under Unix release 4.0?

No, 'cause Uwe Doering (the author of fas) has currently no access to Sys V.4

-Axel
-- 
fischer@utower.tower.in-berlin.de  /  fischer@tmpmbx.UUCP
  Class of '93
                               |  That is not dead, which can eternal lie.
  Go fast - take chances.      |  Yet with strange aeons, even death may die.

gil@limbic.ssdl.com (Gil Kloepfer Jr.) (01/03/91)

In article <1990Dec23.212620.7094@unixland.uucp> bill@unixland.uucp (Bill Heiser) writes:
>reputed to support the NS16550 UART (although I've failed to get it 
>to even work AT ALL with the WD version of that chip).  I have yet to
>try the WD chips in the AST 4-port board I just installed.

Judy Scheltema and I have tried the WD 16550 chips in Steve Nuchia's 
async board, and they fail there as well.  It seems that the only ones
which work properly are the National chips.  Why?  I don't know.  As
I recall, the symptoms we were having were problems like seeing repeating
garbage popping-in all over the place, kind of like the internal FIFO
pointers getting munged.  I never tried WD's 16550 chips without FIFOs
enabled, but why should I?  The National parts are more expensive (about
twice the price of the WD version), but are worth it.
-- 
Gil Kloepfer, Jr.              gil@limbic.ssdl.com   ...!ames!limbic!gil 
Southwest Systems Development Labs (Div of ICUS)   Houston, Texas
"There are beautiful people I wish would have never opened their mouths,
because such ugliness oozes out."  Philosophy Prof. at NYIT

eric@egsner.cirr.com (Eric Schnoebelen) (01/04/91)

In article <129@limbic.ssdl.com> gil@limbic.ssdl.com (Gil Kloepfer Jr.) writes:
- In article <1990Dec23.212620.7094@unixland.uucp> bill@unixland.uucp
-	 (Bill Heiser) writes:
- >reputed to support the NS16550 UART (although I've failed to get it 
- >to even work AT ALL with the WD version of that chip).  I have yet to
- >try the WD chips in the AST 4-port board I just installed.
- 
- Judy Scheltema and I have tried the WD 16550 chips in Steve Nuchia's 
- async board, and they fail there as well.  It seems that the only ones
- which work properly are the National chips.  Why?  I don't know.

        Why?  As I recall, it seems that Western Digital got copies of
the 16550 dies, before all the fixes to the FIFO's got in.  National
Semi is selling 16550A parts which have several fixes for problems with
the FIFO's when they are enabled.  When the FIFO is not enabled, there
is little difference between the '550 and the '450, so the WD part would
work fine in that situation.

        I have two of the WD parts that I have never gotten to work (as
well the shouldn't) and they are just sitting in my junk box.  All (but
one) of the NS parts have worked flawlessly, and the one that didn't, I
had aquired second hand, and hadn't been treated terribly well.

-- 
Eric Schnoebelen		eric@cirr.com		 schnoebe@convex.com
	Real Programmers never work 9 to 5.  If any real programmers
	    are around at 9am it's because they were up all night.

grebus@nobozo.enet.dec.com (01/05/91)

In article <1990Dec23.212620.7094@unixland.uucp>, bill@unixland.uucp (Bill Heiser) writes:

> An interesting note here -- I haven't noticed any character loss running
> a Trailblazer on my Esix system -- someone told me that the Esix asy
> driver is a "clone" of the FAS driver.  The Esix driver is also 
> reputed to support the NS16550 UART (although I've failed to get it 
> to even work AT ALL with the WD version of that chip).  I have yet to
> try the WD chips in the AST 4-port board I just installed.

Interesting...I've also had problems getting 16550's to work, and the
chips were the WD version.
I had the same experience as Bill with the Esix asy driver.  I don't
believe FAS worked either.
The symptoms were a hung port.  Poking around with crash led me to
believe a xmit done
interrupt had gotten lost.  At the time I wrote it off to bugs in the
el-cheapo serial card.
Anyone else seen problems with the WD 16550's.

/gary
	Gary Grebus
	grebus@nobozo.enet.dec.com     (work)
	decvax!balrog!glg                        (home)
	"Speaking for me, not for my employer"

bill@unixland.uucp (Bill Heiser) (01/06/91)

In article <1991Jan4.220906.11895@e2big.mko.dec.com> grebus@nobozo.enet.dec.com () writes:
>
>Anyone else seen problems with the WD 16550's.
>

I got several letters from people who said they also had problems with
the WD chips.  Apparently, according to one person who seemed to know
about such things, the WD chip is a clone of the NAS16550 rather than
the NAS16550A.  As I understand it, there were (and are) bugs in that
chip that cause it to not work properly with FIFOs enabled -- this, of
course, defeats the purpose of the 16550 chip!



-- 
home:	...!{uunet,bloom-beacon,esegue}!world!unixland!bill
	bill@unixland.uucp    Public Access Unix - Esix SYSVR3
508-655-3848(12/24)  508-651-8723(12/24/96-HST)  508-651-8733(12/24/96-PEP-V32)
other:	heiser@world.std.com

blue@techunix.BITNET (Baruch Cochavy) (01/07/91)

About a somewhat related issue: what does it take, in terms of hardware
modification to convert a PC 2 serial ports to UNIX usable board ?
The way I see it, all it takes is proper shareing of the IRQs, which
are open collector (open drain) devices ....

        Thanks in advance,
        Baruch
--
Baruch Cochavy |   Disk space --           |  blue@techunix.BITNET
Technion       |      the final frontier ! |  blue@techunix.technion.AC.IL
Haifa 32000    |                           |  phone:
ISRAEL         |                           |          (972)-4-380337