sound card modem on a laptop?

Hello,

I'm just about to begin playing with AX.25/Linux.. I'd like to get a soundcard modem running on a laptop (IBM Thinkpad), for packet and APRS use.

I'm fairly knowledgable with linux, but not with soundcards and AX.25..

Is anyone else doing this successfully?

Should I start with a 2.6 kernel, or is that too bleeding edge for now?

Thanks for any and all help.

- jim

p.s. Any good URLs I should visit to get my feet?

A comment about 2.6 kernels. Even though I haven't had a chance to try it myself yet (lazy me), I have been told that the MKISS driver in the 2.6 kernel doesn't compile.

The kernel mode soundmodem has been removed in 2.6 so you have to use the user mode soundmodem [1], and it's better anyway. However, the user mode soundmodem needs the MKISS driver to talk to the kernel. So if the driver really does not compile, you can't use 2.6...

/Tomi

[1] http://www.baycom.org/~tom/ham/soundmodem/

Tomi Manninen / OH2BNS / KP20ME04

I have made some patches a while ago to fix the scc and mkiss driver in 2.6.0 and 2.6.1. They haven't made it in mainline though.
In 2.6.1 you get another warning about mkiss using cli() I have made a patch to remove it but have yet to test it.

Jeroen

I have been experimenting Mandrake Linux 10.0 rc with kernel 2.6.0 and the last AX25 patches included into 2.6.1 kernel.

Mkiss driver is running well and there was no compilation problem here.

BTW compiling ax25-tools 0.0.8 gave me some warnings I was able to turn arround :
In ax25/axspawm.c
#include <sys/time.h> should be changed to
#include <time.h>

In netrom/net2kiss.c
#include <netinet/in.h> was missing
like in hdlcutil/hdrvcomm.c
#include <netinet/in.h> is needed for htons declaration and looking at other sources I think that
also in hdrvcomm.c
#include <net/ethernet.h> should be modified like this

#ifdef __GLIBC__
#include <net/ethernet.h>
#else
#include <linux/if_ether.h>
#endif

There is another warning about /netrom/kissnetd.c : 106 with sys_errlist is deprecated ; use strerror or sterror_r instead.

Although I am not familiar with diff I obtained a
diff of the ax25-tools.0.0.8 after I modified it.

73 de bernard, F6BVP

diff -ru ax25-tools-0.0.8/ax25/axspawn.c ax25-tools-0.0.8b/ax25/axspawn.c
@@ -124,7 +124,7 @@
 #include <paths.h>
 #include <errno.h>
 #include <syslog.h>
-#include <sys/time.h>
+#include <time.h>
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <sys/ioctl.h>

diff -ru ax25-tools-0.0.8/hdlcutil/hdrvcomm.c ax25-tools-0.0.8b/hdlcutil/hdrvcomm.c
@@ -48,7 +50,14 @@
 #include "hdrvcomm.h"
 #include "usersmdiag.h"
 #include <net/if_arp.h>
+
+#ifdef __GLIBC__
+#include <net/ethernet.h>
+#else
 #include <linux/if_ether.h>
+#endif
+#include <netinet/in.h>
+
 #include <linux/if_packet.h>
 #ifndef SOL_PACKET
 #define SOL_PACKET  263

diff -ru ax25-tools-0.0.8/kiss/net2kiss.c ax25-tools-0.0.8b/kiss/net2kiss.c
@@ -49,6 +49,8 @@
 #include <sys/socket.h>
 #include <net/if.h>
 
+#include <netinet/in.h>
+
 #ifdef __GLIBC__
 #include <net/ethernet.h>
 #else

It depends on how many warnings and debugging you have turned on, although it compiles fine there is a warning about cli() being used. It is buggy with that respect although very hard to trigger in normal use. What can be triggered easily is another bug in the reuse of the netdevice structure that happens when kissattach is killed and a new kissattach process is started...
I have fixes for both that are now being tested and it looks good sofar.

Jeroen