Hi Wolfgang, Many thanks for your answer. I'll take a look into the sja1000 source code.
Regards, Sandro > -----Ursprüngliche Nachricht----- > Von: Wolfgang Grandegger [mailto:[email protected]] > Gesendet: Dienstag, 6. Juli 2010 19:17 > An: Sandro Anders | CarMediaLab > Cc: [email protected] > Betreff: Re: [Socketcan-users] local loopback timestamp > > Hi Sandro, > > On 07/06/2010 04:55 PM, Sandro Anders | CarMediaLab wrote: > > Hi, > > > > I have a question according the local loopback. At which time the > > timestamp for the sent messages is created (and the message is forwarded > > to the local clients)? Is it on calling write() inside the client code > > or when the driver calls netif_wake_queue() after successful > > transmission? > > Most drivers send the loop back message upstream when the TX done > interrupt is handled. Here is the code for the SJA1000: > > http://lxr.linux.no/#linux+v2.6.34.1/drivers/net/can/sja1000/sja1000.c#L50 > 5 > > netif_rx() ist called in can_get_echo_skb(). > > Unfortunately, a few CAN controllers do not trigger TX done interrupts > and therefor the message is sent upstream earlier. Some other > controllers, like the AT91 CAN, are even able to do the local loopback > in hardware. > > Wolfgang. _______________________________________________ Socketcan-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/socketcan-users
