Using many modems

• Messages sorted by:[ date ][ thread ][ subject ][ author ]

Hi,

On Thu, Nov 05, 1998 at 08:02:18PM +0100, Oliver Sturm wrote:
> Well, I did find the logs. What I had already done here was to increase
> the ring count for one modem to two (the one on ttyS2). 

Well, from the log files, it sounds like a firmware bug in the modems,
but otherwise, it works nicely, doesn't it?

> ------------mgetty.ttyS1---------------------------------
> 07/30 14:39:52 yS1  mdm_send: 'AT+FCC=1,5,0,2,0,0,0,0' -> OK
> 07/30 14:39:53 yS1  waiting...
> 07/30 14:40:28 yS1  waiting for ``RING'' ** found **
> 07/30 14:40:28 yS1  send: ATA[0d]
> 07/30 14:40:28 yS1  waiting for ``CONNECT''
> 07/30 14:41:48 yS1  timeout in chat script, waiting for `CONNECT'
> 07/30 14:41:48 ##### failed timeout dev=ttyS1, pid=5966, caller='none',
> conn='', name=''

This one answers more slowly, and thus doesn't get the call (the ISDN
layer prevents it).  Obviously this confuses the I-Modem, and so it
hangs, not even returning NO CARRIER.

Get a more recent firmware version.

> -------------mgetty.ttyS2---------------------------------------
> 07/30 14:39:08 yS2  mgetty: experimental test release 1.1.14-Apr02
> 07/30 14:40:28 yS2  waiting for ``RING'' ** found **
> 07/30 14:40:28 yS2  waiting for ``RING''
> 07/30 14:40:36 yS2  found action string: ``+FCO_''

This modem is set to auto-answer, which is a very stoopid thing to do
(otherwise it wouldn't have answered).

> 07/30 14:40:36 yS2  start fax receiver...
[..]
> 07/30 14:41:13 ##### fax dev=ttyS2, pid=5965, caller='none', name='',
> id='_     +49 211 2232822_', +FHNG=000, pages=1, time=00:00:29

But nevertheless, the call is handled fine.  So what?

> -----------------------------------------------------
> 
> What happens, I interpret as follows:
> 
> 1) Both modems get a RING signal at the same time (14:40:28)
> 2) ttyS1 answers with ATA and waits for CONNECT (14:40:28), ttyS2 waits
> for second RING
> 3) ttyS2 gets +FCO out of the blue and starts receiving the incoming fax
> (14:40:36)
> 4) ttyS1 gets bored waiting for CONNECT and gives up (14:40:48)

Yes. So?

[..]
> Back in Juli I "fixed" the problem by commenting out a line in mgetty.c
> which handles the incoming +FCO while a modem is really waiting for
> additional RINGS. 

Now this is a very certain way to break things.  You set the modem to do
an auto-answer, and remove the part in mgetty that will cope with broken
modems? Gimme a break :-)

> This gave me no problems, apart from that I had to use
> different ring-counts for the two modems. (BTW, when using the same
> ring-count, one modem gets NO CARRIER alright, but the other gives up
> before coming to CONNECT ;()
> 
> What do you say?

Your modem on ttyS2 is setup wrongly, and the modem on ttyS1 has firmware
problems.

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert@greenie.muc.de
fax: +49-89-35655025                        gert.doering@physik.tu-muenchen.de