On Mon, 1 Aug 2016 15:52:28 +0530
Madhavan Srinivasan [off-list ref] wrote:

On Monday 01 August 2016 02:21 PM, Benjamin Herrenschmidt wrote:

quoted

On Mon, 2016-08-01 at 18:07 +1000, Nicholas Piggin wrote:  

quoted

This will replay hardware_interrupt_common in the case we got a PMI
interrupt but no EE.

Should we just follow the normal pattern here, return 0xf00 for PMI,
and replay the same as the other cases?  

Agreed.  

OK. But can we handle that in the C itself.
With an additional check to see whether EE happended or not?
Because if PMI landed first, there will be no EE anyway?

I just don't know if it's worth making a special case for
it rather than handle it the same as the others?