Re: [hylafax-users] Multiple "MODEM No carrier"

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Antonio,

Please try the attached patch.  The problem was that conf.t1Timer is in 
milliseconds and not seconds, and thus the timeout was 1000X too long.

Thanks,

Lee.
diff -Nru hylafax.orig/faxd/Class1.c++ hylafax/faxd/Class1.c++
--- hylafax.orig/faxd/Class1.c++	2007-10-17 17:40:24.000000000 -0700
+++ hylafax/faxd/Class1.c++	2007-10-18 11:40:35.461094816 -0700
@@ -1475,7 +1475,7 @@
 	     * modem response at that point, anyway.  So it is a good indicator for us.
 	     * So we simply repeat AT+FRH=3 in that case.
 	     */
-	} while (((u_int) Sys::now()-start < conf.t1Timer) && !gotframe && !wasTimeout() && 
+	} while (((u_int) Sys::now()-start < howmany(conf.t1Timer, 1000)) && !gotframe && !wasTimeout() && 
 	    ((conf.class1HasRHConnectBug && !frame.getLength() && lastResponse == AT_NOCARRIER && rhcnt++ < 30) ||
 	    (docrp && crpcnt++ < 3 && switchingPause(emsg, 3) && transmitFrame(dir|FCF_CRP))));	/* triple switchingPause to avoid sending CRP during TCF */
 	return (gotframe);
diff -Nru hylafax.orig/faxd/Class1Recv.c++ hylafax/faxd/Class1Recv.c++
--- hylafax.orig/faxd/Class1Recv.c++	2007-10-17 17:40:24.000000000 -0700
+++ hylafax/faxd/Class1Recv.c++	2007-10-18 11:31:12.034748640 -0700
@@ -292,7 +292,7 @@
 	 * time to make another attempt...
 	 */
 	if ((u_int) Sys::now()-start >= t1)
-	    break;
+	    return (false);
 	if (frame.getFCF() != FCF_CRP) {
 	    /*
 	     * Historically we waited "Class1TrainingRecovery" (1500 ms)