AX.25, NETROM and ROSE changes in 2.6.17
For those who haven't read my recent patches I'd like to point out that in the next kernel release, 2.6.17 the units for the AX.25 timers T1, T2 and T3 will change to use ms for all platforms away from the previous chaos of version, architecture, platform and configuration dependentant chaos. I've tried to explain things on
http://www.linux-ax25.org/wiki/Run_time_configurable_parameters
http://de.linux-ax25.org/wiki/Laufzeiteinstellungen
(1st page in English, 2nd in German)
This matters to any sysadmin that is modiyfing the values from their defaults.
Ralf
ax25-tools-0.0.8src.tgz
Hi,
I compiled OK a new kernel 2.6.16.14 with AX25 for my FEDORA CORE 4 and downloaded from FTP site ftp://f6bvp.org/pub
ax25-apps-0.0.6src.tgz
ax25-tools-0.0.8src.tgz
libax25-0.0.11src.tgz
I make and installed OK ax25-apps-0.0.6src.tgz and libax25-0.0.11src.tgz.
But with ax25-tools-0.0.8src.tgz have the following ERROR on make below at the end of the e-mail.
When I unpacked ax25-tools-0.0.8src.tgz, I see inside in the directory ax25-tools-0.008-2.6.tgz file. I tried tar xvfz ax25-tools-0.008-2.6.tgz and see nothing. What is that file? Is that a patch and how do I apply it?
If somebody can help me. Thanks.
73s JOSE / HP2AT
[root@fedora ax25-tools-0.0.8]# make
make all-recursive
make[1]: Entering directory `/home/hp2at/linux/ax25-tools-0.0.8' Making all in ax25
make[2]: Entering directory `/home/hp2at/linux/ax25-tools-0.0.8/ax25' gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
mheard.c
gcc -g -O2 -Wall -o mheard mheard.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
ax25d.c
gcc -g -O2 -Wall -o ax25d ax25d.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
axctl.c
gcc -g -O2 -Wall -o axctl axctl.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
axparms.c
gcc -g -O2 -Wall -o axparms axparms.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
axspawn.c
gcc -g -O2 -Wall -o axspawn axspawn.o -lax25 -lutil
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
beacon.c
gcc -g -O2 -Wall -o beacon beacon.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
bpqparms.c
gcc -g -O2 -Wall -o bpqparms bpqparms.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -cmheardd.cmheardd.c: In function âmainâ:
mheardd.c:272: warning: pointer targets in passing argument 1 of âftypeâ differ in signedness
mheardd.c:333: warning: case label value exceeds maximum value for type mheardd.c:339: warning: case label value exceeds maximum value for type mheardd.c:342: warning: case label value exceeds maximum value for type mheardd.c:345: warning: case label value exceeds maximum value for type mheardd.c:351: warning: case label value exceeds maximum value for type mheardd.c:354: warning: case label value exceeds maximum value for type mheardd.c:357: warning: case label value exceeds maximum value for type mheardd.c:360: warning: case label value exceeds maximum value for type gcc -g -O2 -Wall -o mheardd mheardd.o -lax25
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
rxecho.c
gcc -g -O2 -Wall -o rxecho rxecho.o -lax25
make[2]: Leaving directory `/home/hp2at/linux/ax25-tools-0.0.8/ax25' Making all in hdlcutil
make[2]: Entering directory `/home/hp2at/linux/ax25-tools-0.0.8/hdlcutil' Making all in fl
make[3]: Entering directory /home/hp2at/linux/ax25-tools-0.0.8/hdlcutil/fl' make[3]: Nothing to be done for all'.
make[3]: Leaving directory /home/hp2at/linux/ax25-tools-0.0.8/hdlcutil/fl' make[3]: Entering directory /home/hp2at/linux/ax25-tools-0.0.8/hdlcutil' gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
sethdlc.csethdlc.c: In function âmainâ:
sethdlc.c:540: warning: pointer targets in passing argument 1 of âdisplay_packetâ differ in signedness
gcc -DHAVE_CONFIG_H -I. -I. -I.. -DAX25_SYSCONFDIR=\""/usr/local/etc/ax25/"\"
-DAX25_LOCALSTATEDIR=\""/usr/local/var/ax25/"\" -g -O2 -Wall -c
hdrvcomm.c
In file included from hdrvcomm.c:43:/usr/include/asm/byteorder.h:6:2: warning: #warning using private kernel header; include <endian.h> instead!
In file included from hdrvcomm.c:57:/usr/include/netinet/in.h:354: error: syntax error before â(â token /usr/include/netinet/in.h:354: error: syntax error before â__u32â /usr/include/netinet/in.h:355: error: syntax error before â(â token /usr/include/netinet/in.h:355: error: syntax error before â__u16â /usr/include/netinet/in.h:357: error: syntax error before â(â token /usr/include/netinet/in.h:357: error: syntax error before â__u32â /usr/include/netinet/in.h:359: error: syntax error before â(â token /usr/include/netinet/in.h:359: error: syntax error before â__u16â hdrvcomm.c: In function âhdrvc_get_driver_nameâ:
hdrvcomm.c:678: warning: pointer targets in passing argument 2 of â__builtin_strncpyâ differ in signednesshdrvcomm.c: In function âhdrvc_get_mode_nameâ:
hdrvcomm.c:713: warning: pointer targets in passing argument 2 of â__builtin_strncpyâ differ in signedness
make[3]: *** [hdrvcomm.o] Error 1
make[3]: Leaving directory `/home/hp2at/linux/ax25-tools-0.0.8/hdlcutil' make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/home/hp2at/linux/ax25-tools-0.0.8/hdlcutil' make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/hp2at/linux/ax25-tools-0.0.8'
make: *** [all-recursive-am] Error 2
[root@fedora ax25-tools-0.0.8]#
-
To unsubscribe from this list: send the line "unsubscribe linux-hams" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: ax25-tools-0.0.8src.tgz
On Sat, May 06, 2006 at 06:16:40PM -0500, Jose Ng Lee wrote:
(Emily Postnews recommend learning about the difference between a fresh posting and a reply. Your posting has nothing to do with the my you've replied to)
Quoted 9 lines
I compiled OK a new kernel 2.6.16.14 with AX25 for my FEDORA CORE 4 and downloaded from FTP site ftp://f6bvp.org/pub
ax25-apps-0.0.6src.tgz
ax25-tools-0.0.8src.tgz
libax25-0.0.11src.tgzI make and installed OK ax25-apps-0.0.6src.tgz and libax25-0.0.11src.tgz.
Quoted 8 lines
But with ax25-tools-0.0.8src.tgz have the following ERROR on make below at the end of the e-mail.
When I unpacked ax25-tools-0.0.8src.tgz, I see inside in the directory ax25-tools-0.008-2.6.tgz file. I tried tar xvfz ax25-tools-0.008-2.6.tgz and see nothing. What is that file? Is that a patch and how do I apply it?
If somebody can help me. Thanks.
I've dumped a few packages I built for FC4 into
ftp://ftp.linux-mips.org/pub/linux/mips/people/ralf/ax25/
They used to work fine for me on FC4 but I switched to FC5 so there won't be any FC4 updates for these packages anymore.
Ralf
Quoted 9 lines · Ralf Baechle DL5RB wrote :
For those who haven't read my recent patches I'd like to point out that in the next kernel release, 2.6.17 the units for the AX.25 timers T1, T2 and T3 will change to use ms for all platforms away from the previous chaos of version, architecture, platform and configuration dependentant chaos. I've tried to explain things on
http://www.linux-ax25.org/wiki/Run_time_configurable_parameters
http://de.linux-ax25.org/wiki/Laufzeiteinstellungen
Quoted 6 lines
(1st page in English, 2nd in German)
This matters to any sysadmin that is modiyfing the values from their defaults.
Ralf
By the way, its a long time since I want to ask if there is any reason to keep the same name for different ax25.h include files.
-rw-r--r-- 1 root root 4877 mar 21 01:26 /usr/include/netax25/ax25.h [root@bernard tools]# ll /usr/src/linux/include/linux/ax25.h -rw-rw-rw- 1 root root 2752 mar 20 06:53 /usr/src/linux/include/linux/ax25.h [root@bernard tools]# ll /usr/src/linux/include/net/ax25.h -rw-rw-rw- 1 root root 14842 avr 30 11:27 /usr/src/linux/include/net/ax25.h
Somehow it may be confusing.
73 de Bernard, f6bvp
http://f6bvp.org
http://rose.fpac.free.fr/MINI-HOWTO/
http://rose.fpac.free.fr/MINI-HOWTO-FR/
Quoted 11 lines · Bernard Pidoux
By the way, its a long time since I want to ask if there is any reason to keep the same name for different
ax25.hinclude files.-rw-r--r-- 1 root root 4877 mar 21 01:26
/usr/include/netax25/ax25.h[root@bernard tools]# ll/usr/src/linux/include/linux/ax25.h-rw-rw-rw- 1 root root 2752 mar 20 06:53/usr/src/linux/include/linux/ax25.h
[root@bernard tools]# ll/usr/src/linux/include/net/ax25.h-rw-rw-rw- 1 root root 14842 avr 30 11:27/usr/src/linux/include/net/ax25.hSomehow it may be confusing.
One for building applications, one for the kernel itself.
73 de DL5RB op Ralf
--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21
different AX25.H files with same names
Quoted 8 lines · Ralf Baechle DL5RB wrote :
Quoted 6 lines · Bernard Pidoux
By the way, its a long time since I want to ask if there is any reason to keep the same name for different
ax25.hinclude files.
/usr/include/netax25/ax25.h/usr/src/linux/include/linux/ax25.h/usr/src/linux/include/net/ax25.h
Quoted 8 lines
Quoted 1 line
Somehow it may be confusing.
One for building applications, one for the kernel itself.
73 de DL5RB op Ralf
--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21
I understand that these files have different usage. This is why I think that they should have different names.
Here is the content of /usr/src/linux/include/linux/ax25.h
----------------
/*
* These are the public elements of the Linux kernel AX.25 code. A similar
* file netrom.h exists for the NET/ROM protocol.
*/
#ifndef AX25_KERNEL_H
#define AX25_KERNEL_H
----------------
So why not rename it ax25_kernel.h
73 de Bernard, f6bvp
http://f6bvp.org
http://rose.fpac.free.fr/MINI-HOWTO/
http://rose.fpac.free.fr/MINI-HOWTO-FR/
Re: different AX25.H files with same names
Quoted 16 lines · Bernard Pidoux
Quoted 15 lines · Ralf Baechle DL5RB wrote :
Quoted 8 lines · Bernard Pidoux
By the way, its a long time since I want to ask if there is any reason to keep the same name for different
ax25.hinclude files.
/usr/include/netax25/ax25.h/usr/src/linux/include/linux/ax25.h/usr/src/linux/include/net/ax25.hSomehow it may be confusing.
One for building applications, one for the kernel itself.
73 de DL5RB op Ralf
Quoted 7 lines
Quoted 2 lines
--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21I understand that these files have different usage. This is why I think that they should have different names.
So why not rename it
ax25_kernel.h
I just don't see a convcingint reason for it. If it aint broke don't fix it. Chances are renaming will break some software which against good advice is using the kernel header file anyway.
Ralf
Re: different AX25.H files with same names
If you really want to get confused, check out how many different versions of netrom.h are laying around.
/usr/include/linux/netrom.h/usr/include/netax25/kernel_netrom.h/usr/include/netrom/netrom.h/usr/src/linux/include/linux/netrom.h/usr/src/linux/include/net/netrom.h
The bottom line is you should never have the /usr/src/linux kernel includes in your build path if you are building applications.
Mike, W1NR
----- Original Message ----- From: "Ralf Baechle" <[email protected]>
To: "Bernard Pidoux" <[email protected]>
Cc: <[email protected]>
Sent: Thursday, May 18, 2006 10:49 AM
Subject: Re: different AX25.H files with same names
Quoted 36 lines · Bernard Pidoux
Quoted 17 lines
Quoted 15 lines · Ralf Baechle DL5RB wrote :
Quoted 8 lines · Bernard Pidoux
By the way, its a long time since I want to ask if there is any reason to keep the same name for different
ax25.hinclude files.
/usr/include/netax25/ax25.h/usr/src/linux/include/linux/ax25.h/usr/src/linux/include/net/ax25.hSomehow it may be confusing.
One for building applications, one for the kernel itself.
73 de DL5RB op Ralf
Quoted 7 lines
Quoted 2 lines
--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21I understand that these files have different usage. This is why I think that they should have different names.
So why not rename it
ax25_kernel.hI just don't see a convcingint reason for it. If it aint broke don't fix it. Chances are renaming will break some software which against good advice is using the kernel header file anyway.
Ralf
-
To unsubscribe from this list: send the line "unsubscribe linux-hams" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 9 lines · Ralf Baechle DL5RB wrote :
For those who haven't read my recent patches I'd like to point out that in the next kernel release, 2.6.17 the units for the AX.25 timers T1, T2 and T3 will change to use ms for all platforms away from the previous chaos of version, architecture, platform and configuration dependentant chaos. I've tried to explain things on
http://www.linux-ax25.org/wiki/Run_time_configurable_parameters
http://de.linux-ax25.org/wiki/Laufzeiteinstellungen
Quoted 7 lines
(1st page in English, 2nd in German)
This matters to any sysadmin that is modiyfing the values from their defaults.
Ralf
-
I have just installed kernel 2.6.16.17 and patched it with your postings. I compiled the kernel with the 1000 Hz timing option.
I wanted to compare my runtime AX25 parameters with your tables in
http://www.linux-ax25.org/wiki/Run_time_configurable_parameters
Here is what is listed when printing /proc/sys/net/ax25/ax0
ll /proc/sys/net/ax25/ax0/
total 0
-rw-r--r-- 1 root root 0 mai 21 23:25 ax25_default_mode -rw-r--r-- 1 root root 0 mai 21 23:25 backoff_type -rw-r--r-- 1 root root 0 mai 21 23:25 connect_mode -rw-r--r-- 1 root root 0 mai 21 23:25 dama_slave_timeout -rw-r--r-- 1 root root 0 mai 21 23:25 extended_window_size -rw-r--r-- 1 root root 0 mai 21 23:25 idle_timeout -rw-r--r-- 1 root root 0 mai 21 23:25 ip_default_mode -rw-r--r-- 1 root root 0 mai 21 23:25 maximum_packet_length -rw-r--r-- 1 root root 0 mai 21 23:25 maximum_retry_count -rw-r--r-- 1 root root 0 mai 21 23:25 protocol -rw-r--r-- 1 root root 0 mai 21 23:25 standard_window_size -rw-r--r-- 1 root root 0 mai 21 23:25 t1_timeout -rw-r--r-- 1 root root 0 mai 21 23:25 t2_timeout -rw-r--r-- 1 root root 0 mai 21 23:25 t3_timeout
It appears that dama_slave_timeout and protocol parameters are not in the AX25 table.
ll /proc/sys/net/netrom/
total 0
-rw-r--r-- 1 root root 0 mai 21 23:36 default_path_quality -rw-r--r-- 1 root root 0 mai 21 23:36 link_fails_count -rw-r--r-- 1 root root 0 mai 21 23:36 network_ttl_initialiser -rw-r--r-- 1 root root 0 mai 21 23:36 obsolescence_count_initialiser -rw-r--r-- 1 root root 0 mai 21 23:36 reset -rw-r--r-- 1 root root 0 mai 21 23:36 routing_control -rw-r--r-- 1 root root 0 mai 21 23:36 transport_acknowledge_delay -rw-r--r-- 1 root root 0 mai 21 23:36 transport_busy_delay -rw-r--r-- 1 root root 0 mai 21 23:36 transport_maximum_tries -rw-r--r-- 1 root root 0 mai 21 23:36 transport_no_activity_timeout -rw-r--r-- 1 root root 0 mai 21 23:36 transport_requested_window_size -rw-r--r-- 1 root root 0 mai 21 23:36 transport_timeout
reset parameter is not listed. Not ajustable ?
Here is the listing of ROSE parameters :
ll /proc/sys/net/rose/
total 0
-rw-r--r-- 1 root root 0 mai 21 23:33 acknowledge_hold_back_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 call_request_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 clear_request_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 link_fail_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 maximum_virtual_circuits -rw-r--r-- 1 root root 0 mai 21 23:33 no_activity_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 reset_request_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 restart_request_timeout -rw-r--r-- 1 root root 0 mai 21 23:33 routing_control -rw-r--r-- 1 root root 0 mai 21 23:33 window_size
Again, parameter no_activity_timeout is not in the ROSE table.
Any explanations ?
73 de Bernard, f6bvp
http://f6bvp.org
http://rose.fpac.free.fr/MINI-HOWTO/
http://rose.fpac.free.fr/MINI-HOWTO-FR/
Quoted 5 lines · Bernard Pidoux
(quoted diff) +0
-2-rw-r--r-- 1 root root 0 mai 21 23:25 dama_slave_timeout -rw-r--r-- 1 root root 0 mai 21 23:25 protocolIt appears that
dama_slave_timeoutand protocol parameters are not in the AX25 table.
They are only there if CONFIG_AX25_DAMA_SLAVE has been selected.
Quoted 5 lines
ll
/proc/sys/net/netrom/-rw-r--r-- 1 root root 0 mai 21 23:36 reset
reset parameter is not listed. Not ajustable ?
It's a relativly new thing, the G8ZPT-style NET/ROM circuit resets. It's of course ajustable, the file has "rw" set.
Quoted 5 lines
ll
/proc/sys/net/rose/-rw-r--r-- 1 root root 0 mai 21 23:33
no_activity_timeoutAgain, parameter
no_activity_timeoutis not in the ROSE table.
I suppose this was just inherited from the old, less than perfect documentation.
73 de DL5RB op Ralf
--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21