Message-ID: <dfejhrchaxwze7gkipdgcx5byefv6hudi456na25yrts6viqvw@lfnv55dgvk32>
Date: Wed, 18 Jun 2025 21:55:24 +0200
From: Jorge Marques <gastmaier@...il.com>
To: Krzysztof Kozlowski <krzk@...nel.org>
Cc: Jorge Marques <jorge.marques@...log.com>, 
	Alexandre Belloni <alexandre.belloni@...tlin.com>, Frank Li <Frank.Li@....com>, Rob Herring <robh@...nel.org>, 
	Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>, 
	linux-i3c@...ts.infradead.org, devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 1/2] dt-bindings: i3c: Add adi-i3c-master

On Wed, Jun 18, 2025 at 05:45:22PM +0200, Krzysztof Kozlowski wrote:
> On 18/06/2025 14:15, Jorge Marques wrote:
> >>>
> >>> Signed-off-by: Jorge Marques <jorge.marques@...log.com>
> >>> ---
> >>>  .../devicetree/bindings/i3c/adi,i3c-master.yaml    | 63 ++++++++++++++++++++++
> >>>  MAINTAINERS                                        |  5 ++
> >>>  2 files changed, 68 insertions(+)
> >>>
> >>> diff --git a/Documentation/devicetree/bindings/i3c/adi,i3c-master.yaml b/Documentation/devicetree/bindings/i3c/adi,i3c-master.yaml
> >>> new file mode 100644
> >>> index 0000000000000000000000000000000000000000..718733bbb450c34c5d4924050cc6f85d8a80fe4b
> >>> --- /dev/null
> >>> +++ b/Documentation/devicetree/bindings/i3c/adi,i3c-master.yaml
> >>
> >> Filename based on the compatible, so adi,i3c-master-1.00.a.yaml
> >>
> > I agree, but I ended up following the pattern for the other adi,
> > bindings. I will move for v4. IMO the version suffix has no much use
> > since IP updates are handled in the driver.
> 
> Filename is not related to whether given ABI works with every device.
> Filename helps us to organize bindings and existing convention is that
> we want it to follow the compatible.
> 
Hi Krzysztof,

Understood.

> >>> @@ -0,0 +1,63 @@
> >>> +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
> >>> +%YAML 1.2
> >>> +---
> >>> +$id: http://devicetree.org/schemas/i3c/adi,i3c-master.yaml#
> >>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> >>> +
> >>> +title: Analog Devices I3C Controller
> >>> +
> >>> +description: |
> >>> +  FPGA-based I3C controller designed to interface with I3C and I2C peripherals,
> >>> +  implementing a subset of the I3C-basic specification.
> >>> +
> >>> +  https://analogdevicesinc.github.io/hdl/library/i3c_controller
> >>> +
> >>> +maintainers:
> >>> +  - Jorge Marques <jorge.marques@...log.com>
> >>> +
> >>> +properties:
> >>> +  compatible:
> >>> +    const: adi,i3c-master-1.00.a
> >>> +
> >>> +  reg:
> >>> +    maxItems: 1
> >>> +
> >>> +  clocks:
> >>> +    minItems: 1
> >>
> >> Why?
> >>
> > The IP core requires a clock, and the second is optional.
> 
> OK
> 
> > minItems sets the minimum number of required clocks and the maxItems is
> > inferred from the number of items.
> > 
> > On the IP core itself, one clock is required (axi), and if it is the
> > only provided, it means that the same clock for the AXI bus is used
> > also for the rest of the RTL logic.
> 
> Hm? What does it exactly mean - same clock? You mean one clock is routed
> to two pins? That's still two clocks. Or you mean that IP core will
> notice grounded clock input and do the routing inside?
> 

The routing is inside the IP core, and only one clock pin is used. In
fullness, since it is a FPGA-based IP Core, the number of input clock
pins are defined by the parameter [1] ASYNC_CLK that enables the
asynchronous i3c clock input pin. The devicetree then describes how
things are wired, if two clocks provided, it describes that both clock
inputs, axi and i3c, are wired to the IP Core, if only axi, then there
is no clock signal to the i3c input clock pin and axi clock drives the
whole IP.

[1] https://analogdevicesinc.github.io/hdl/library/i3c_controller/i3c_controller_host_interface.html#configuration-parameters

> > 
> > If a second clock is provided, i3c, it means it drives the RTL logic and is
> > asynchronous to the axi clock, which then just drives the register map logic.
> > For i3c specified nominal speeds, the RTL logic should run with a speed of
> > 100MHz. Some FPGAs, such as Altera CycloneV, have a default bus clock speed of
> > 50MHz. Changing the bus speed is possible, but affects timing and it may not be
> > possible from users to double the bus speed since it will affect timing of all
> > IP cores using the bus clock.
> >>> +    items:
> >>> +      - description: The AXI interconnect clock.
> >>> +      - description: The I3C controller clock.
> > I will update the descriptions to:
> > 
> >         - description: The AXI interconnect clock, drives the register map.
> >         - description: The I3C controller clock. AXI clock drives all logic if not provided.
> > 
> >>> +
> >>> +  clock-names:
> >>
> >> Not synced with clocks.
> >>
> > I will add `minItems: 1`.
> >>> +    items:
> >>> +      - const: axi
> >>> +      - const: i3c
> >>> +
> >>> +  interrupts:
> >>> +    maxItems: 1
> >>> +
> >>> +required:
> >>> +  - compatible
> >>> +  - reg
> >>> +  - clocks
> >>> +  - clock-names
> >>> +  - interrupts
> >>> +
> >>> +allOf:
> >>> +  - $ref: i3c.yaml#
> >>> +
> >>> +unevaluatedProperties: false
> >>> +
> >>> +examples:
> >>> +  - |
> >>> +    i3c@...00000 {
> >>> +        compatible = "adi,i3c-master";
> >>> +        reg = <0x44a00000 0x1000>;
> >>> +        interrupts = <0 56 4>;
> >>
> >> Use proper defines.
> >>
> > The following can added:
> > 
> >   #include <dt-bindings/interrupt-controller/irq.h>
> > 
> >   interrupts = <0 56 IRQ_TYPE_LEVEL_HIGH>;
> > 
> > Is there any other to be replaced?
> 
> Usually 0 has a meaning as well. Where is this used DTS snippet used (on
> which platform)?
> 

The example provided is used on AMD xilinx arm,cortex-a9-gic.
The IP core is tested on AMD Xilinx, Altera FPGAs (arm, microblaze, arm64).
Test/support to RiscV-based Lattice FPGAs should eventually also come.

In this example 0 means Shared Peripheral Interrupt, but has no header
file defining. I guess I could make it generic and set to

  interrupts = <3 IRQ_TYPE_LEVEL_HIGH>;

and let the user figure out for his target platform.
> Best regards,
> Krzysztof

Best regards,
Jorge

Powered by blists - more mailing lists