j1939 - first steps

j1939 - first steps

[Oliver Hartkopp]

Hi Kurt,

following the idea from Marc (cloning linux-stable, check-out 3.7.10, adding
j1939 branch, applying the j1939-3.7 branch) i managed to get my linux-box
working with your stack.

Following the documentation, i did this (as root):

./ip link set vcan0 j1939 on
./ip addr show vcan0
./ip addr add dev vcan0 j1939 0x22
./ip addr show vcan0
6: vcan0: <NOARP,UP,LOWER_UP> mtu 72 qdisc noqueue state UNKNOWN
    link/can
    can-j1939 0x22 scope link

And this (as user):

./jsr -v vcan0:0x22
./jsr: bind(vcan0:0000000000000022), -99: Cannot assign requested address

./jsr -v 0x22
./jsr: bind(#34:-), -19: No such device


./jsr -v 22
./jsr: bind(#22:-), -19: No such device


Any idea?

Regards,
Oliver

Re: j1939 - first steps

[Kurt Van Dijck]

On Mon, Mar 18, 2013 at 09:13:33PM +0100, Oliver Hartkopp wrote:
> Hi Kurt,
> 
> following the idea from Marc (cloning linux-stable, check-out 3.7.10, adding
> j1939 branch, applying the j1939-3.7 branch) i managed to get my linux-box
> working with your stack.
> 
> Following the documentation, i did this (as root):
> 
> ./ip link set vcan0 j1939 on
> ./ip addr show vcan0
> ./ip addr add dev vcan0 j1939 0x22
> ./ip addr show vcan0
> 6: vcan0: <NOARP,UP,LOWER_UP> mtu 72 qdisc noqueue state UNKNOWN
>     link/can
>     can-j1939 0x22 scope link

static addressing, ok.
> 
> And this (as user):
> 
> ./jsr -v vcan0:0x22
you should use 'vcan0:,0x22'.
The format is: 'IFACE:<NAME>[,<ADDR>]'

Better proposals to combine dynamic & static addressing in strings?
> ./jsr: bind(vcan0:0000000000000022), -99: Cannot assign requested address
indeed. The 64bit name 00...0022 is tried.

Kind regards,
Kurt

Re: j1939 - first steps

[Oliver Hartkopp]

On 18.03.2013 22:41, Kurt Van Dijck wrote:

> On Mon, Mar 18, 2013 at 09:13:33PM +0100, Oliver Hartkopp wrote:


>> And this (as user):
>>
>> ./jsr -v vcan0:0x22
> you should use 'vcan0:,0x22'.
> The format is: 'IFACE:<NAME>[,<ADDR>]'
> 


./jsr -?
jsr: An SAE J1939 send/recv utility
Usage: jsr [OPTION...] SOURCE [DEST]

  -v, --verbose		Increase verbosity
  -p, --priority=VAL	J1939 priority (0..7, default 6)
  -S, --serialize	Strictly serialize outgoing packets
  -s, --size		Packet size, default autodetected

  SOURCE	[IFACE:][NAME|SA][,PGN]
  DEST			[NAME|SA]

Tells me something else ;-)

Btw. with your hint it works:

./jsr -v vcan0:,0x22
1122334455
112233
123123123123
112233445566778899AABBCCDDEEFF112233

leads to:
$ candump -ta any
 (1363674992.322459)  vcan0  18ECFF22   [8]  20 0B 00 02 FF 00 00 00
 (1363674992.372643)  vcan0  18EBFF22   [8]  01 31 31 32 32 33 33 34
 (1363674992.422713)  vcan0  18EBFF22   [5]  02 34 35 35 0A
 (1363674999.306921)  vcan0  1800FF22   [7]  31 31 32 32 33 33 0A
 (1363675021.549579)  vcan0  18ECFF22   [8]  20 0D 00 02 FF 00 00 00
 (1363675021.599657)  vcan0  18EBFF22   [8]  01 31 32 33 31 32 33 31
 (1363675021.649719)  vcan0  18EBFF22   [7]  02 32 33 31 32 33 0A
 (1363675049.932595)  vcan0  18ECFF22   [8]  20 25 00 06 FF 00 00 00
 (1363675049.982647)  vcan0  18EBFF22   [8]  01 31 31 32 32 33 33 34
 (1363675050.032709)  vcan0  18EBFF22   [8]  02 34 35 35 36 36 37 37
 (1363675050.082781)  vcan0  18EBFF22   [8]  03 38 38 39 39 41 41 42
 (1363675050.132851)  vcan0  18EBFF22   [8]  04 42 43 43 44 44 45 45
 (1363675050.182893)  vcan0  18EBFF22   [8]  05 46 46 31 31 32 32 33
 (1363675050.232954)  vcan0  18EBFF22   [3]  06 33 0A

First time to see j1939 on my linux box, yeah!

Best regards,
Oliver

Re: j1939 - first steps

[Kurt Van Dijck]

On Tue, Mar 19, 2013 at 07:52:03AM +0100, Oliver Hartkopp wrote:
> On 18.03.2013 22:41, Kurt Van Dijck wrote:
> 
> > On Mon, Mar 18, 2013 at 09:13:33PM +0100, Oliver Hartkopp wrote:
> 
> 
> >> And this (as user):
> >>
> >> ./jsr -v vcan0:0x22
> > you should use 'vcan0:,0x22'.
> > The format is: 'IFACE:<NAME>[,<ADDR>]'
> > 
> 
> 
> ./jsr -?
> jsr: An SAE J1939 send/recv utility
> Usage: jsr [OPTION...] SOURCE [DEST]
> 
>   -v, --verbose		Increase verbosity
>   -p, --priority=VAL	J1939 priority (0..7, default 6)
>   -S, --serialize	Strictly serialize outgoing packets
>   -s, --size		Packet size, default autodetected
> 
>   SOURCE	[IFACE:][NAME|SA][,PGN]
>   DEST			[NAME|SA]
> 
> Tells me something else ;-)

I looked into the parser (can-utils/libj1939.c):
You should have omitted the 0x prefix.
jsr operates in hex (currently) by default.

I'll look for modifying this to be more intuitive.
> 
> Btw. with your hint it works:
> 
> ./jsr -v vcan0:,0x22
Yep, it is equal to './jsr -v vcan0'.
> 1122334455
> 112233
> 123123123123
> 112233445566778899AABBCCDDEEFF112233
> 
> leads to:
> $ candump -ta any
>  (1363674992.322459)  vcan0  18ECFF22   [8]  20 0B 00 02 FF 00 00 00
>  (1363674992.372643)  vcan0  18EBFF22   [8]  01 31 31 32 32 33 33 34
>  (1363674992.422713)  vcan0  18EBFF22   [5]  02 34 35 35 0A
>  (1363674999.306921)  vcan0  1800FF22   [7]  31 31 32 32 33 33 0A
>  (1363675021.549579)  vcan0  18ECFF22   [8]  20 0D 00 02 FF 00 00 00
>  (1363675021.599657)  vcan0  18EBFF22   [8]  01 31 32 33 31 32 33 31
>  (1363675021.649719)  vcan0  18EBFF22   [7]  02 32 33 31 32 33 0A
>  (1363675049.932595)  vcan0  18ECFF22   [8]  20 25 00 06 FF 00 00 00
>  (1363675049.982647)  vcan0  18EBFF22   [8]  01 31 31 32 32 33 33 34
>  (1363675050.032709)  vcan0  18EBFF22   [8]  02 34 35 35 36 36 37 37
>  (1363675050.082781)  vcan0  18EBFF22   [8]  03 38 38 39 39 41 41 42
>  (1363675050.132851)  vcan0  18EBFF22   [8]  04 42 43 43 44 44 45 45
>  (1363675050.182893)  vcan0  18EBFF22   [8]  05 46 46 31 31 32 32 33
>  (1363675050.232954)  vcan0  18EBFF22   [3]  06 33 0A
> 
> First time to see j1939 on my linux box, yeah!
congratulations.
Next step is to use flow control with transport protocol:

$ ./ip addr add dev vcan0 j1939 0x23
$ ./jsr -v vcan0:22,12345 23

And watch the can_id's & timestamps

Kind regards,
Kurt

Re: j1939 - first steps

[Kurt Van Dijck]

On Tue, Mar 19, 2013 at 07:52:03AM +0100, Oliver Hartkopp wrote:
> On 18.03.2013 22:41, Kurt Van Dijck wrote:
> 
> > On Mon, Mar 18, 2013 at 09:13:33PM +0100, Oliver Hartkopp wrote:
> 
> 
> >> And this (as user):
> >>
> >> ./jsr -v vcan0:0x22
> > you should use 'vcan0:,0x22'.
> > The format is: 'IFACE:<NAME>[,<ADDR>]'
> > 
> 
> 
> ./jsr -?
> jsr: An SAE J1939 send/recv utility
> Usage: jsr [OPTION...] SOURCE [DEST]
> 
>   -v, --verbose		Increase verbosity
>   -p, --priority=VAL	J1939 priority (0..7, default 6)
>   -S, --serialize	Strictly serialize outgoing packets
>   -s, --size		Packet size, default autodetected
> 
>   SOURCE	[IFACE:][NAME|SA][,PGN]
>   DEST			[NAME|SA]
> 
> Tells me something else ;-)

Oliver, how about this change in the docs.
Would that have avoided this confusion (on both sides).

diff --git a/jacd.c b/jacd.c
index 1358918..088b72c 100644
--- a/jacd.c
+++ b/jacd.c
@@ -39,7 +39,8 @@ static const char help_msg[] =
 	"NAME is the 64bit nodename" "\n"
 	"\n"
 	"Example:" "\n"
-	"jacd -r 100,80-120 -c /tmp/1122334455667788.jacd 1122334455667788" "\n"
+	"	$ jacd -r 100,80-120 -c /tmp/0123456789abcdef.jacd"
+			" 0123456789abcdef" "\n"
 	;
 
 #ifdef _GNU_SOURCE
diff --git a/jspy.c b/jspy.c
index b48fea1..6b7a215 100644
--- a/jspy.c
+++ b/jspy.c
@@ -37,6 +37,10 @@ static const char help_msg[] =
 	"			(= receive traffic not for this ECU)" "\n"
 	"  -b, --block=SIZE	Use a receive buffer of SIZE (default 1024)" "\n"
 	"  -t, --time[=a|d|z|A]	Show time: (a)bsolute, (d)elta, (z)ero, (A)bsolute w date" "\n"
+	"\n"
+	"	NAME, SA & PGN are in hex, without 0x prefix.\n"
+	" Example\n:
+	"	$ jpy can0:,0ea00\n"
 	;
 
 #ifdef _GNU_SOURCE
diff --git a/jsr.c b/jsr.c
index 0303af0..aad6093 100644
--- a/jsr.c
+++ b/jsr.c
@@ -40,6 +40,10 @@ static const char help_msg[] =
 	"\n"
 	"  SOURCE	[IFACE:][NAME|SA][,PGN]" "\n"
 	"  DEST			[NAME|SA]" "\n"
+	"\n"
+	"	NAME, SA & PGN are in hex, without 0x prefix.\n"
+	"Example:\n"
+	"	$ jsr can0:0123456789abcdef,0da00 22\n"
 	;
 
 #ifdef _GNU_SOURCE

Re: j1939 - first steps

[Marc Kleine-Budde]

[-- Attachment #1: Type: text/plain, Size: 2100 bytes --]

On 03/20/2013 08:24 PM, Kurt Van Dijck wrote:
> On Tue, Mar 19, 2013 at 07:52:03AM +0100, Oliver Hartkopp wrote:
>> On 18.03.2013 22:41, Kurt Van Dijck wrote:
>>
>>> On Mon, Mar 18, 2013 at 09:13:33PM +0100, Oliver Hartkopp wrote:
>>
>>
>>>> And this (as user):
>>>>
>>>> ./jsr -v vcan0:0x22
>>> you should use 'vcan0:,0x22'.
>>> The format is: 'IFACE:<NAME>[,<ADDR>]'
>>>
>>
>>
>> ./jsr -?
>> jsr: An SAE J1939 send/recv utility
>> Usage: jsr [OPTION...] SOURCE [DEST]
>>
>>   -v, --verbose		Increase verbosity
>>   -p, --priority=VAL	J1939 priority (0..7, default 6)
>>   -S, --serialize	Strictly serialize outgoing packets
>>   -s, --size		Packet size, default autodetected
>>
>>   SOURCE	[IFACE:][NAME|SA][,PGN]
>>   DEST			[NAME|SA]
>>
>> Tells me something else ;-)
> 
> Oliver, how about this change in the docs.
> Would that have avoided this confusion (on both sides).
> 
> diff --git a/jacd.c b/jacd.c
> index 1358918..088b72c 100644
> --- a/jacd.c
> +++ b/jacd.c
> @@ -39,7 +39,8 @@ static const char help_msg[] =
>  	"NAME is the 64bit nodename" "\n"
>  	"\n"
>  	"Example:" "\n"
> -	"jacd -r 100,80-120 -c /tmp/1122334455667788.jacd 1122334455667788" "\n"
> +	"	$ jacd -r 100,80-120 -c /tmp/0123456789abcdef.jacd"
> +			" 0123456789abcdef" "\n"
>  	;
>  
>  #ifdef _GNU_SOURCE
> diff --git a/jspy.c b/jspy.c
> index b48fea1..6b7a215 100644
> --- a/jspy.c
> +++ b/jspy.c
> @@ -37,6 +37,10 @@ static const char help_msg[] =
>  	"			(= receive traffic not for this ECU)" "\n"
>  	"  -b, --block=SIZE	Use a receive buffer of SIZE (default 1024)" "\n"
>  	"  -t, --time[=a|d|z|A]	Show time: (a)bsolute, (d)elta, (z)ero, (A)bsolute w date" "\n"
> +	"\n"
> +	"	NAME, SA & PGN are in hex, without 0x prefix.\n"

What do you use to parse these numbers?

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 263 bytes --]

Re: j1939 - first steps

[Kurt Van Dijck]

On Wed, Mar 20, 2013 at 09:44:18PM +0100, Marc Kleine-Budde wrote:
> On 03/20/2013 08:24 PM, Kurt Van Dijck wrote:
> > On Tue, Mar 19, 2013 at 07:52:03AM +0100, Oliver Hartkopp wrote:
> >> On 18.03.2013 22:41, Kurt Van Dijck wrote:
> >>
> >>> On Mon, Mar 18, 2013 at 09:13:33PM +0100, Oliver Hartkopp wrote:
> >>
> >>
> >>>> And this (as user):
> >>>>
> >>>> ./jsr -v vcan0:0x22
> >>> you should use 'vcan0:,0x22'.
> >>>
> >>
> >>
> >> ./jsr -?
> >> jsr: An SAE J1939 send/recv utility
> >> Usage: jsr [OPTION...] SOURCE [DEST]
> >>
> >>   -v, --verbose		Increase verbosity
> >>   -p, --priority=VAL	J1939 priority (0..7, default 6)
> >>   -S, --serialize	Strictly serialize outgoing packets
> >>   -s, --size		Packet size, default autodetected
> >>
> >>   SOURCE	[IFACE:][NAME|SA][,PGN]
> >>   DEST			[NAME|SA]
> >>
> >> Tells me something else ;-)
> > 
> > Oliver, how about this change in the docs.
> > Would that have avoided this confusion (on both sides).
> > 
> > diff --git a/jacd.c b/jacd.c
> > index 1358918..088b72c 100644
> > --- a/jacd.c
> > +++ b/jacd.c
> > @@ -39,7 +39,8 @@ static const char help_msg[] =
> >  	"NAME is the 64bit nodename" "\n"
> >  	"\n"
> >  	"Example:" "\n"
> > -	"jacd -r 100,80-120 -c /tmp/1122334455667788.jacd 1122334455667788" "\n"
> > +	"	$ jacd -r 100,80-120 -c /tmp/0123456789abcdef.jacd"
> > +			" 0123456789abcdef" "\n"
> >  	;
> >  
> >  #ifdef _GNU_SOURCE
> > diff --git a/jspy.c b/jspy.c
> > index b48fea1..6b7a215 100644
> > --- a/jspy.c
> > +++ b/jspy.c
> > @@ -37,6 +37,10 @@ static const char help_msg[] =
> >  	"			(= receive traffic not for this ECU)" "\n"
> >  	"  -b, --block=SIZE	Use a receive buffer of SIZE (default 1024)" "\n"
> >  	"  -t, --time[=a|d|z|A]	Show time: (a)bsolute, (d)elta, (z)ero, (A)bsolute w date" "\n"
> > +	"\n"
> > +	"	NAME, SA & PGN are in hex, without 0x prefix.\n"
> 
> What do you use to parse these numbers?
strtoul(str, &endp, 16);
		   ^^^^

The use is very similar to cansend, I believe, where a can_id of exactly 8 hex chars long
means 29bit, and shorter means 11bit.
Implicitely, everything is told to be in hex, without 0x.

And because this was not clear, Oliver got himself a 64bit nodename as strlen(0x22) > 2 :-)
> >>> The format is: 'IFACE:<NAME>[,<ADDR>]'
And my first attempt to help him, out of the head, was a 'legacy thought', not good either.
always check the code :-)

Kurt

Re: j1939 - first steps

[Oliver Hartkopp]

On 21.03.2013 06:53, Kurt Van Dijck wrote:

> On Wed, Mar 20, 2013 at 09:44:18PM +0100, Marc Kleine-Budde wrote:

>> What do you use to parse these numbers?
> strtoul(str, &endp, 16);
> 		   ^^^^
> 
> The use is very similar to cansend, I believe, where a can_id of exactly 8 hex chars long
> means 29bit, and shorter means 11bit.
> Implicitely, everything is told to be in hex, without 0x.
> 
> And because this was not clear, Oliver got himself a 64bit nodename as strlen(0x22) > 2 :-)


Yes that was my trap. I've seen the strlen but it didn't help me %-)

>>>>> The format is: 'IFACE:<NAME>[,<ADDR>]'
> And my first attempt to help him, out of the head, was a 'legacy thought', not good either.
> always check the code :-)


Regards,
Oliver