On 22.10.19 12:50, Jonathan Cameron wrote: > On Tue, 22 Oct 2019 10:43:09 +0200 > Martin Kepplinger <martink@xxxxxxxxx> wrote: > >> On 22.10.19 05:56, Andrey Smirnov wrote: >>> Current wording in the binding documentation doesn't make it 100% >>> clear that only one of "INT1" and "INT2" will ever be used by the >>> driver and that specifying both has no advantages. Re-word it to make >>> this aspect a bit more explicit. > > From a quick glance at the datasheet, it appears that the hardware > is capable of routing different interrupts to different pins, even > if the driver does not do so. > > CTRL_REG5 has 6 different bits to set whether particular blocks have > their interrupt routed to INT1 or INT2, so it appears to be possible > to send some each way. > > A binding is for the hardware not the driver, so if I'm right about this > the current text is correct and should be left alone. right, please ignore my "acked-by". Thanks a lot for taking a look. martin