Re: [PATCH 01/13] dt-bindings: interrupt-controller: create a header for RISC-V interrupts

[Anup Patel <anup@xxxxxxxxxxxxxx>]

On Tue, Nov 23, 2021 at 4:38 PM Heiko Stübner <heiko@xxxxxxxxx> wrote:
>
> Am Montag, 8. November 2021, 16:05:42 CET schrieb conor.dooley@xxxxxxxxxxxxx:
> > From: Ivan Griffin <ivan.griffin@xxxxxxxxxxxxx>
> >
> > Provide named identifiers for device tree for RISC-V interrupts.
> >
> > Licensed under GPL and MIT, as this file may be useful to any OS that
> > uses device tree.
> >
> > Signed-off-by: Ivan Griffin <ivan.griffin@xxxxxxxxxxxxx>
> > Signed-off-by: Conor Dooley <conor.dooley@xxxxxxxxxxxxx>
> > ---
> >  .../interrupt-controller/riscv-hart.h         | 19 +++++++++++++++++++
> >  1 file changed, 19 insertions(+)
> >  create mode 100644 include/dt-bindings/interrupt-controller/riscv-hart.h
> >
> > diff --git a/include/dt-bindings/interrupt-controller/riscv-hart.h b/include/dt-bindings/interrupt-controller/riscv-hart.h
> > new file mode 100644
> > index 000000000000..e1c32f6090ac
> > --- /dev/null
> > +++ b/include/dt-bindings/interrupt-controller/riscv-hart.h
> > @@ -0,0 +1,19 @@
> > +/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */
> > +/*
> > + * Copyright (C) 2021 Microchip Technology Inc.  All rights reserved.
> > + */
> > +
> > +#ifndef _DT_BINDINGS_INTERRUPT_CONTROLLER_RISCV_HART_H
> > +#define _DT_BINDINGS_INTERRUPT_CONTROLLER_RISCV_HART_H
> > +
> > +#define HART_INT_U_SOFT   0
> > +#define HART_INT_S_SOFT   1
> > +#define HART_INT_M_SOFT   3
> > +#define HART_INT_U_TIMER  4
> > +#define HART_INT_S_TIMER  5
> > +#define HART_INT_M_TIMER  7
> > +#define HART_INT_U_EXT    8
> > +#define HART_INT_S_EXT    9
> > +#define HART_INT_M_EXT    11
>
> (1) From checking clic doc [0] I see an additional
>         12   CLIC software interrupt
> defined.

Local IRQ #12 is for S-mode guest external interrupts as-per
the ratified H-extension specification.

I guess CLIC spec needs to be updated.

Regards,
Anup

>
> (2) The doc states that the ordering is a recommendation and
>         "not mandatory in all incarnations of the CLIC"
> Is that clarified somewhere else that this more than recommended?
>
> Thanks
> Heiko
>
>
> [0] https://github.com/riscv/riscv-fast-interrupt/blob/master/clic.adoc
>
> > +
> > +#endif /* _DT_BINDINGS_INTERRUPT_CONTROLLER_RISCV_HART_H */
> >
>
>
>
>