[comp.dcom.modems] Equinox hardware flow control

larry@nstar.rn.com (Larry Snyder) (02/14/91)

bill@astph.UUCP (Bill Dripps) writes:

>We are evaluating Equinox's 24 port Megaport. It looks like a good
>board. Unfortunately the key tech support person is on vacation this
>week. Sigh. 

>Anybody know how to turn on hardware flow control using ISC UNIX? The
>manual tells how for Xenix, but no hints for ISC. Did I miss it?

Hardware flow control for ISC Unix is not supported (so I've been
told) with the Equinox or stock ISC ASY driver.

-- 
   Larry Snyder, NSTAR Public Access Unix 219-289-0287 (HST/PEP/V.32/v.42bis)
                        regional UUCP mapping coordinator 
  {larry@nstar.rn.com, ..!uunet!nstar!larry, larry%nstar@iuvax

witr@rwwa.COM (Robert W. Withrow) (02/15/91)

>>Anybody know how to turn on hardware flow control using ISC UNIX?

Use the FAS driver, recently posted to one of the .sources groups.  I
can attest that it works well (at least on my os and system).
-- 
---
 Robert Withrow, R.W. Withrow Associates, Swampscott MA 01907 USA
 Tel: +1 617 598 4480, Fax: +1 617 598 4430, Uucp: witr@rwwa.COM

chip@chinacat.Unicom.COM (Chip Rosenthal) (02/16/91)

In article <1991Feb15.154233.19586@rwwa.COM>
	witr@rwwa.COM (Robert W. Withrow) writes:
>Use the FAS driver, recently posted to one of the .sources groups.  I
>can attest that it works well (at least on my os and system).

With an Equinox board???  Pretty twisted idea...
-- 
Chip Rosenthal  512-482-8260  |
Unicom Systems Development    |    I saw Elvis in my wtmp file.
<chip@chinacat.Unicom.COM>    |

scotto@ipars.UUCP (Scott O'Connell) (02/16/91)

In article <1991Feb14.134714.5407@nstar.rn.com> larry@nstar.rn.com (Larry Snyder) writes:
>bill@astph.UUCP (Bill Dripps) writes:
>
>>We are evaluating Equinox's 24 port Megaport. It looks like a good
>>board. Unfortunately the key tech support person is on vacation this
>>week. Sigh. 
>
>>Anybody know how to turn on hardware flow control using ISC UNIX? The
>>manual tells how for Xenix, but no hints for ISC. Did I miss it?

As I'm sure you're aware, the Equinox board does not support RTS/CTS flow
control without "stealing" the CD/DTR control signals from an adjacent
port.

I am running SCO Xenix 386 2.3.3 with the Megaport/24 and 2.0 drivers.  To
turn on hardware flow control with Xenix I use the following commands in
/etc/rc.d/5:

	megastty +fmc < /dev/ttyaa (rts/cts capability robbed from ttyab)
	megastty +fmc < /dev/ttyac (rts/cts capability robbed from ttyad)

Not knowing what version of the software you have this is all I can provide.

NOTE: I've had no problems with Equinox products after using them for
      more than 1 year.  Good stuff, IMHO.

john@zygot.ati.com (John Higdon) (02/18/91)

In article <1991Feb14.134714.5407@nstar.rn.com> larry@nstar.rn.com (Larry Snyder) writes:
>bill@astph.UUCP (Bill Dripps) writes:

>>Anybody know how to turn on hardware flow control using ISC UNIX? The
>>manual tells how for Xenix, but no hints for ISC. Did I miss it?

>Hardware flow control for ISC Unix is not supported (so I've been
>told) with the Equinox or stock ISC ASY driver.

According to the Equinox manual, flow control is not supported.
However, it goes on to say that flow control with a Trailblazer is not
required, since the ports have the ability to keep up with continuous
transmission at 38.4 K without it.

I am using the Equinox Megaport without flow control for five
Trailblazers and ISC. It has absolutely no difficulty whatsoever.
-- 
        John Higdon         |   P. O. Box 7648   |   +1 408 723 1395
    john@zygot.ati.com      | San Jose, CA 95150 |       M o o !

karl@ddsw1.MCS.COM (Karl Denninger) (02/18/91)

In article <1991Feb14.134714.5407@nstar.rn.com> larry@nstar.rn.com (Larry Snyder) writes:
>bill@astph.UUCP (Bill Dripps) writes:
>
>>We are evaluating Equinox's 24 port Megaport. It looks like a good
>>board. Unfortunately the key tech support person is on vacation this
>>week. Sigh. 
>
>>Anybody know how to turn on hardware flow control using ISC UNIX? The
>>manual tells how for Xenix, but no hints for ISC. Did I miss it?
>
>Hardware flow control for ISC Unix is not supported (so I've been
>told) with the Equinox or stock ISC ASY driver.

Nonsense.

megamap +fmc </dev/ttyxx 

And it's on.  Full-time.  Until you reboot.

This only works on Megaport boards, not the stock ISC driver :-)

- 
Karl Denninger (karl@ddsw1.MCS.COM, <well-connected>!ddsw1!karl)
Public Access Data Line: [+1 708 808-7300], Voice: [+1 708 808-7200]
Macro Computer Solutions, Inc.   "Quality Solutions at a Fair Price"

todd@Quotron.COM (Todd_Booth) (02/19/91)

john@zygot.ati.com (John Higdon) writes:
>According to the Equinox manual, flow control is not supported...
>I am using the Equinox Megaport without flow control for five
>Trailblazers and ISC. It has absolutely no difficulty whatsoever.

I assume you have the modem drop the interface speed down to the link
speed.  Otherwise, when sending files _from_ ISC to a remote host (with
the link lower than the interface), your modem will drop many characters.

>        John Higdon         |   P. O. Box 7648   |   +1 408 723 1395
>    john@zygot.ati.com      | San Jose, CA 95150 |       M o o !
--
todd (booth) 

todd@quotron.com 213 302-4368

todd@gwinnett.UUCP (Todd Reese) (02/19/91)

In article <todd.666898719@premises1> todd@Quotron.COM (Todd_Booth) writes:
>john@zygot.ati.com (John Higdon) writes:
>>According to the Equinox manual, flow control is not supported...
>>I am using the Equinox Megaport without flow control for five
>>Trailblazers and ISC. It has absolutely no difficulty whatsoever.
>
>I assume you have the modem drop the interface speed down to the link
>speed.  Otherwise, when sending files _from_ ISC to a remote host (with
>the link lower than the interface), your modem will drop many characters.
>
>>        John Higdon         |   P. O. Box 7648   |   +1 408 723 1395
>>    john@zygot.ati.com      | San Jose, CA 95150 |       M o o !
>--
>todd (booth) 
>
>todd@quotron.com 213 302-4368



You know what happens when you assume!....

Trailblazers do not need flow control when using UUCP.  The modem has
UUCP -g protocol built in.  This is what makes it able to produce output
with PEP mode of 19200 baud. At a lower link speed (e.g. 2400 baud) the
UUCP from the host computer to the remote is handled directly.


---

					Todd F. Reese
					Gwinnett Computer Services
					Duluth, GA  30136
					emory!gwinnett!todd
					gatech!dscatl!gwinnett!todd

root@equinox.UUCP (Super user) (02/20/91)

In article <1991Feb14.134714.5407@nstar.rn.com> larry@nstar.rn.com (Larry Snyder) writes:
>bill@astph.UUCP (Bill Dripps) writes:
>
>>We are evaluating Equinox's 24 port Megaport. It looks like a good
>>board. Unfortunately the key tech support person is on vacation this
>>week. Sigh. 
>
>>Anybody know how to turn on hardware flow control using ISC UNIX? The
>>manual tells how for Xenix, but no hints for ISC. Did I miss it?
>
>Hardware flow control for ISC Unix is not supported (so I've been
>told) with the Equinox or stock ISC ASY driver.
>

I've checked into the ISC hardware flow control question and here is
what I've found out:

It's not quite true that ISC doesn't support hardware flow control since
the Dialer program and application software that turn hardware flow control
on work fine. It may be a question of semantics, but hardware flow control
is not supported "on the console or terminals" by ISC. For the most part,
xon/xoff works great. Because the megaport doesn't use UARTS, the megaport
can stop data transfer within two characters after receiving an XOFF--even
at 38.4K baud. That's significant, because most UART boards would not be able to
handle 38.4K baud with xon/xoff flow control because at that speed the 
buffer would overflow. Please remember, I'm talking about the need for
hardware flow with the use of terminals/printers -- not high speed modems.
For terminals/printers, use xon/xoff with MEGAPORT (why not?).

Ok, one reason why not is a problem we're working on with Bill Dripps.
WordPerfect 5.0 puts the wyse60 in PC Term mode (for keyboard compatibility
with DOS WordPerfect) and must use hardware flow control. To me, this 
sounds similar to the FoxBase 9600 baud limit that we came up with a fix
for (megastty +fb) to allow FoxBase to run at 38.4K baud with xon/xoff
handshaking (even though FoxBase disabled xon/xoff for its editor to
maintain DBASE III compatibility). We have a beta driver out with an
attempted fix for this WordPerfect problem. We're short-handed this week,
because two of our support guys are on installs, one high-level guru
who was working on these problems is on leave for a family emergency --
so the rest of us are getting support experience (which, I think, is
good -- since it really brings customer problems home to roost when you
have marketing types and kernel hacker types answering the phone.). These
problems will get solved!

wolf

john@zygot.ati.com (John Higdon) (02/23/91)

In article <todd.666898719@premises1> todd@Quotron.COM (Todd_Booth) writes:

>I assume you have the modem drop the interface speed down to the link
>speed.  Otherwise, when sending files _from_ ISC to a remote host (with
>the link lower than the interface), your modem will drop many characters.

Yes, this is the suggestion in the Megaport manual. S51 is set at 254,
which causes the interface to follow the connect rate and calls for
19200 baud when connected with PEP. In my particular application, the
modems are used virtually 100% at PEP anyway.
-- 
        John Higdon         |   P. O. Box 7648   |   +1 408 723 1395
    john@zygot.ati.com      | San Jose, CA 95150 |       M o o !

gandrews@netcom.COM (Greg Andrews) (02/24/91)

In article <30054@zygot.ati.com> john@zygot.ati.com (John Higdon) writes:
>
>S51 is set at 254, which causes the interface to follow the connect 
>rate and calls for 19200 baud when connected with PEP. 
>

Sorry to nitpick, John, but the S66 register determines whether the
interface is locked or follows the connect rate.  (S66=1 means locked, 
S66=0 means unlocked).  S51=254 means that the interface autobauds,
and chooses 19200 for PEP connections as you said.  (just wanted to
avoid any confusion about S51/S66...)

-- 
.-------------------------------------------.
| Greg Andrews      |   gandrews@netcom.COM |
`-------------------------------------------'

clewis@ferret.ocunix.on.ca (Chris Lewis) (02/24/91)

In article <1991Feb19.153604.13096@gwinnett.UUCP> todd@gwinnett.UUCP (Todd Reese) writes:
>Trailblazers do not need flow control when using UUCP.  The modem has
>UUCP -g protocol built in.  This is what makes it able to produce output
>with PEP mode of 19200 baud. At a lower link speed (e.g. 2400 baud) the
>UUCP from the host computer to the remote is handled directly.

Trailblazers *do* need to use flow control in most UUCP situations.  One situation
is where the modem is transmitting to "their" host, and the host can't
quite keep up to bursts.  If the host doesn't flow control, it'll lose characters.
This can often be bad enough to drop connections even with UUCP packetizing
and retransmissions of packets after errors.  (eg: if, from a quiescent state,
your host will frequently drop a character from a continuous 19200 baud input
burst, you're toast)

Even if the speed mismatches aren't bad enough to drop connections, the
packet retry rate can be sufficiently high to seriously degrade your
throughput.

Further, you can't use x-on/x-off flow control on a UUCP connection, because
x-on and x-off are part of the UUCP protocol.  Thus, hardware handshaking
is recommended to ensure that no characters are lost, and that as long as
the host can keep up, on average, with the modem, you'll get the highest
possible throughput.
-- 
Chris Lewis, Phone: (613) 832-0541, Internet: clewis@ferret.ocunix.on.ca
UUCP: uunet!mitel!cunews!latour!ecicrl!clewis; Ferret Mailing List:
(ferret-request@eci386); Psroff (not Adobe Transcript) enquiries:
psroff-request@eci386, current patchlevel is *7*.