Hi Geert,
On Wednesday, July 25, 2018, Geert Uytterhoeven wrote:
> > However, if "interrupt-names" is specified in DT, then the driver
> > determines what the interrupt are based on their names, not the order in
> which
> > they are listed.
> >
> > Correct?
>
> Correct.
One final note on this before I submit v2 of the patch.
I just coded up something that works and is more simple and I don't even
need "interrupt-names".
Basically by using your suggestion from the code review made everything
work.
On Friday, July 20, 2018, Geert Uytterhoeven wrote:
> > --- a/drivers/tty/serial/sh-sci.c
> > +++ b/drivers/tty/serial/sh-sci.c
> > @@ -65,6 +65,7 @@ enum {
> > SCIx_RXI_IRQ,
> > SCIx_TXI_IRQ,
> > SCIx_BRI_IRQ,
> > + SCIx_TEIDRI_IRQ,
>
> Why not separate enum values for TEI and DRI? According to the RZ/A2 docs,
> there are 6 separate interrupts, but they are multiplexed at the interrupt
> controller level.
enum {
SCIx_ERI_IRQ,
SCIx_RXI_IRQ,
SCIx_TXI_IRQ,
SCIx_BRI_IRQ,
+ SCIx_DRI_IRQ,
+ SCIx_TEI_IRQ,
SCIx_NR_IRQS,
SCIx_MUX_IRQ = SCIx_NR_IRQS, /* special case */
};
Listing the same interrupt ID number twice in the DT (because it is
muxed) is fine because the driver will check for that.
This seems to satisfy all the SCI/SCIF variants in all the SH and ARM
SoCs in the kernel today (DT and non-DT).
So as long as I describe the interrupt order in the DT Documentation,
all seems good.
Chris
