Re: [PATCH] dt-bindings: serial: renesas,scif: Document R9A09G057 support

[Lad, Prabhakar]

Hi Geert,

On Thu, Mar 7, 2024 at 10:18 AM Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
>
> Hi Prabhakar,
>
> On Thu, Mar 7, 2024 at 11:09 AM Lad, Prabhakar
> <prabhakar.csengg@xxxxxxxxx> wrote:
> > On Wed, Mar 6, 2024 at 10:21 AM Lad, Prabhakar
> > <prabhakar.csengg@xxxxxxxxx> wrote:
> > > On Wed, Mar 6, 2024 at 10:15 AM Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
> > > > On Wed, Mar 6, 2024 at 11:06 AM Lad, Prabhakar
> > > > <prabhakar.csengg@xxxxxxxxx> wrote:
> > > > > On Wed, Mar 6, 2024 at 9:53 AM Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
> > > > > > On Tue, Mar 5, 2024 at 6:16 PM Prabhakar <prabhakar.csengg@xxxxxxxxx> wrote:
> > > > > > > From: Lad Prabhakar <prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx>
> > > > > > >
> > > > > > > Document support for the Serial Communication Interface with FIFO (SCIF)
> > > > > > > available in the Renesas RZ/V2H(P) (R9A09G057) SoC. The SCIF interface in
> > > > > > > the Renesas RZ/V2H(P) is similar to that available in the RZ/G2L
> > > > > > > (R9A07G044) SoC, with the only difference being that the RZ/V2H(P) SoC has
> > > > > > > three additional interrupts: one for Tx end/Rx ready and the other two for
> > > > > > > Rx and Tx buffer full, which are edge-triggered.
> > > > > > >
> > > > > > > No driver changes are required as generic compatible string
> > > > > > > "renesas,scif-r9a07g044" will be used as a fallback on RZ/V2H(P) SoC.
> > > > > >
> > > > > > If you declare SCIF on RZ/V2H compatible with SCIF on RZ/G2L, you
> > > > > > state that the current driver works fine (but perhaps suboptimal),
> > > > > > without adding support for the extra 3 interrupts?
> > > > > >
> > > > > Yes the current driver works without using the extra interrupts on the
> > > > > RZ/V2H. The extra interrupts on the RZ/V2H are just sort of duplicate
> > > > > ie
> > > > > - Transmit End/Data Ready interrupt , for which we we have two
> > > > > seperate interrupts already
> > > > > - Receive buffer full interrupt (EDGE trigger), for which we already
> > > > > have a Level triggered interrupt
> > > > > - Transmit buffer empty interrupt (EDGE trigger), for which we already
> > > > > have a Level triggered interrupt
> > > >
> > > > Thanks for the confirmation!
> > > >
> > > > > Are you suggesting to not fallback on RZ/G2L and instead make RZ/V2H
> > > > > an explicit one so that in future we handle these 3 extra interrupts?
> > > >
> > > > In light of the confirmation above, I am _not_ suggesting that.
> > > >
> > With the introduction of validation checks for interrupts, falling
> > back to "renesas,scif-r9a07g044" for RZ/V2H will be difficult for
> > validating interrupt count.
> >
> >   - if:
> >       properties:
> >         compatible:
> >           contains:
> >             enum:
> >               - renesas,scif-r7s9210
> >               - renesas,scif-r9a07g044
> >     then:
> >       properties:
> >         interrupts:
> >           minItems: 6
> >
> >         interrupt-names:
> >           minItems: 6
> >
> > With the above check RZ/V2H fall into this if block,
> >
> > Is there any way I can specify to match two compat strings?
>
> if r9a09g057 then ... else if r7s9210 || r9a07g044 then ...?
>
Thanks for the pointer (I was grepping for elif ;)).

Cheers,
Prabhakar