Re: [hylafax-users] Adaptive answer, MT5634ZBA-V92 (was Receive problems with MT2834ZDX modems)

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Bill Campbell wrote:

> Mar 21 15:18:03.18: [15400]: SESSION BEGIN 000005520 12062329186
> Mar 21 15:18:03.18: [15400]: HylaFAX (tm) Version 4.2.3
> Mar 21 15:18:03.18: [15400]: <-- [14:AT+FCLASS=1;A\r]
> Mar 21 15:18:11.56: [15400]: --> [3:FAX]
> Mar 21 15:18:11.56: [15400]: ANSWER: FAX CONNECTION  DEVICE '/dev/ttyS0'
> Mar 21 15:18:11.56: [15400]: RECV FAX: begin
> Mar 21 15:18:11.56: [15400]: <-- data [32]
> Mar 21 15:18:11.56: [15400]: <-- data [2]
> Mar 21 15:18:11.56: [15400]: --> [3:FAX]
> Mar 21 15:18:11.56: [15400]: DELAY 1500 ms
> Mar 21 15:18:13.06: [15400]: <-- [9:AT+FTH=3\r]
> Mar 21 15:18:13.06: [15400]: --> [6:^H ^H^H ^H]
> Mar 21 15:18:13.06: [15400]: DELAY 1500 ms
> Mar 21 15:18:14.56: [15400]: <-- [9:AT+FTH=3\r]
> Mar 21 15:18:14.56: [15400]: --> [5:ERROR]
> Mar 21 15:18:14.56: [15400]: RECV FAX: RSPREC error/got EOT
> Mar 21 15:18:14.56: [15400]: RECV FAX: end
> Mar 21 15:18:14.56: [15400]: SESSION END

Notice the double "FAX" response from the modem.  It says "FAX" twice at 
15:18:11.56.  So misbehavior by the modem is causing HylaFAX to trip 
over itself here.

I've seen Agere-chipsetted modems do this kind of thing before (after 
ATZ is my memory).

This may be fixed in a newer firmware (I don't know), but you're welcome 
to try the attached patch, also, which should keep HylaFAX from tripping 
up on it.

Thanks,

Lee.

--- hylafax.orig/faxd/FaxRecv.c++	2006-03-16 16:01:21.000000000 -0800
+++ hylafax/faxd/FaxRecv.c++	2006-03-21 15:42:37.840740128 -0800
@@ -64,6 +64,9 @@
     if (tif) {
 	recvPages = 0;			// total count of received pages
 	fileStart = pageStart = Sys::now();
+
+	modem->flushModemInput();
+
 	if (faxRecognized = modem->recvBegin(emsg)) {
 	    /*
 	     * If the system is busy then notifyRecvBegun may not return