Hello again, Is anyone getting missed SPI TX / RX events at the end of an EDMA data transfer? Or have you found a solution to such a problem?
I was going through the McBSP EDMA code. Seems that they are requesting the EDMA channels at every data transfer request and then freeing it in the callback routine. Does anyone know why this approach has been taken? As I see it, this has a great performance impact. Which makes me wonder why do it this way. At first I used the McBSP EDMA approach in my SPI EDMA support. But later I tried shifting the EDMA request and freeing code into the SPI init and exit routines respectively. Only to find that EDMA hangs at the second TX transfer. Any body else noticed this? Regards, Uthappa
_______________________________________________ Davinci-linux-open-source mailing list [email protected] http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
