[Xenomai-core] Re: RT-Socket-CAN bus error rate and latencies

[Wolfgang Grandegger <wg@domain.hid>]

Hartkopp, Oliver (K-GEFE/E) wrote:
> Even if i know, that my reply with our f*cking exchange mailserver will 
> break the thread ...
> 
> Jan Kiszka wrote:
>> I think Oliver's suggestions points in the right direction. But instead
>> of only coding a timer into the stack, I still vote for closing the loop
>> over the application:
>>
>> After the first error in a potential series, the related error frame is
>> queued, listeners are woken up, and BEI is disabled for now. Once some
>> listener read the error frame *and* decided to call into the stack for
>> further bus errors, BEI is enabled again.
>>
>>   
> IMO this is really a good idea! Even if it look just a bit different on 
> (non-RT) socketcan, the procedure could look like this:
> 
> 1. BEI occurs => disable BEI & lauch receiption of one error frame for 
> the user
> 2. Wait for user interaction
> 
> I would suggest to create an IOCTL, where you can define, if you only 
> want to receive one BE-frame (== default!) or every BE-frame (may flood 
> your system). After an occurance of a BEI the BE-frames are disabled (in 
> the default case) until the user decides to reenable it via the 
> suggested IOCTL.
> 
> Does this look acceptable to you?

No extra IOCTL please. I would place re-enabling of BEI in sendmsg.

> Btw. Currently the enabling of error frames (in socketcan) currently 
> does not depend on subscribed listeners (what is definitely a good idea!).

Why?

Wolfgang.