Linux/am4378: Spurious EDMA trigger from UART

Hi Patrick,
I am looking at it.

Steve K.

Sorry for the delay, but I am still looking into this.

Hi Steve,

Thanks for letting me know.  I'm not in a hurry, I was just surprised that the previous thread had been closed since it looks like a real hardware issue.  I hope that at some point, a fix or work around can be introduced into the linux drivers so that I can use stock drivers.

Regards,

Patrick

Hi Patrick,
I have to talk with someone over here who is not in office today.

Steve K.

Hi Steve,

Thank you for confirming the issue and the suspected cause (see previous thread ). 

Is there a work around or fix?

Regards,

Patrick

Hi Steve,

The work around for advisory 27 was already implemented in the stock kernel when I reported this problem.  Also from the previous thread, this appears to be a different issue and it occurs only on the USART instances and not on the UART instances.

The problem is probably caused by a spurious EDMA trigger rather than a spurious interrupt.

Please let me know if any new information becomes available.

Regards,

Patrick

As noted above I believe that advisory 27 addresses a different, but probably related, issue. The work around presented in advisory 27 is already handled by the stock driver (i.e., the spurious interrupt is acknowledged and cleared), but it does not solve the spurious DMA trigger.