AW: Web GUI for can-utils (i.e. cangw, cansend, candump)

* AW: Web GUI for can-utils (i.e. cangw, cansend, candump)
@ 2013-06-10 16:30 Michael Luxen
  2013-06-10 18:22 ` Oliver Hartkopp
  0 siblings, 1 reply; 5+ messages in thread
From: Michael Luxen @ 2013-06-10 16:30 UTC (permalink / raw)
  To: Oliver Hartkopp; +Cc: linux-can Mailing List

Hello Oliver,

OK!
Thnak you for pointing me to the interestin toolchain with .kcd files (CANbabel translator), socketcand and Kayak.
Nevertheless I did some tests last weekend on my BeagleBone Black setup with DCAN0/DCAN1.

If I fire round about 37%@500kbit/s to DCAN0 I get a tast load of round about 45% of socketcand. In general it's possible for me to use Kayak on i.e. my Windows machine and connect to my BeagleBone Black via ethernet. But if I use the second CAN channel (DCAN1) or increase the busload of channel one (DCAN0) the taskload/CPUload increases unacceptable.

Did you have the same experience with socketcand and Kayak?
Why isn't it possible to connect directly to the can0/can1 socketcan interfaces without socketcand?

Regards
Mike

-----Ursprüngliche Nachricht-----
Von:	Oliver Hartkopp <socketcan@hartkopp.net>
Gesendet:	Sa 08.06.2013 14:31
Betreff:	Re: Web GUI for can-utils (i.e. cangw, cansend, candump)
An:	Michael Luxen <mluxen@gmx.net>; 
CC:	linux-can Mailing List <linux-can@vger.kernel.org>; 
> On 07.06.2013 22:10, Michael Luxen wrote:
> 
> > I'm looking for a easy solution to make the powerful cangw tool more 
> comfortable for endusers (dummy user).
> > 
> > Therefore I prefer to have a webfrontend / Web GUI to configure the cangw 
> rules with a webbrowser which is available on each PC/smartphone/etc
> > 
> > Is there already a configuration tool or best a Web GUI available which can 
> i.e.:
> > cangw
> > - Load two Vector .dbc files (one for can0 and one for can1)
> 
> I would suggest to start with kcd files instead of dbc files, which are Vector
> property.
> 
> E.g. The JAVA tool Kayak (http://kayak.2codeornot2code.org/) uses the kcd
> format (see: http://kayak.2codeornot2code.org/tutorial.html). There's also a
> kcd converter CANBabel at github: https://github.com/julietkilo/CANBabel
> 
> > - Select on one side (can0) the input CAN frame which should be routed from 
> can0 to can via cagw
> > - Select on the other side (can1) the input CAN fram whcih should be routed 
> from can1 to can0 via cangw
> > - Manipulate of CAN-frames with some logic operations +/-/*/div to use the 
> modification features of cangw
> 
> Today there are only AND/OR/XOR/SET defined for CAN frame modifications.
> 
> To modify signals on the fly is a much more complex attempt than doing some
> bit-wise operations. E.g. if you think about +/-/*/div operations: You need to
> provide the other operands, you need to extract and shift the data for
> processing and afterward inject it at the right place (shifting, adding the
> original information that was not modified, etc).
> 
> I wonder if this can/should be done in kernel context in a proper manner.
> 
> > - start cangw
> > 
> > cansend
> > - load a Vector .dbc file
> > - select a CAN-frame for transmitting
> > - select a CAN-signal to fillin a value to transmit
> > - manipulate the CAN-frames/CAN-signals content/values independent
> > - start cansend for CAN simulation
> > 
> > candump
> > - configure can interface to monitor/log
> > - start candump -l in background
> 
> I would suggest to preserve cansend and candump as-is.
> 
> Especially candump has become pretty complex by today to support all the
> CAN_RAW socket options and control information (from CAN FD to dropcounters &
> timestamps).
> 
> Probably some new tools like sigdump and sigsend (signal...) that make use of
> kcd files are a better option here.
> 
> But i don't know how this can be adapted to a web-based interface then.
> 
> Best regards,
> Oliver
> 
> 
> > 
> > 
> > Environment background:
> > I'm currently using the two internal CAN-Interfaces of the BeagleBone Black. 
> They are perfectly working for me with the latest Angstrom or Debian images.
> > 
> > Thanks for any hint or pointing me to the right place!
> > 
> > Regards
> > Mike
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-can" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > 
> 
>

^ permalink raw reply	[flat|nested] 5+ messages in thread