From: Krzysztof Kozlowski <krzk@kernel.org>
To: chuan.liu@amlogic.com, Neil Armstrong <neil.armstrong@linaro.org>,
Michael Turquette <mturquette@baylibre.com>,
Stephen Boyd <sboyd@kernel.org>, Rob Herring <robh@kernel.org>,
Krzysztof Kozlowski <krzk+dt@kernel.org>,
Conor Dooley <conor+dt@kernel.org>
Cc: linux-amlogic@lists.infradead.org, linux-clk@vger.kernel.org,
devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 01/13] dt-bindings: clock: Add Amlogic A9 standardized model clock control units
Date: Mon, 9 Feb 2026 14:14:29 +0100 [thread overview]
Message-ID: <10caac08-1a1d-4fe9-8df8-e0b75c532e8f@kernel.org> (raw)
In-Reply-To: <20260209-a9_clock_driver-v1-1-a9198dc03d2a@amlogic.com>
On 09/02/2026 06:48, Chuan Liu via B4 Relay wrote:
> From: Chuan Liu <chuan.liu@amlogic.com>
>
> Add dt-binding documentation for standardized model clock control units
> in A9 SoC family.
>
> Signed-off-by: Chuan Liu <chuan.liu@amlogic.com>
> ---
> .../bindings/clock/amlogic,a9-model-ccu.yaml | 435 +++++++++++++++++++++
> 1 file changed, 435 insertions(+)
Brief review, you still have to read basic guidelines to not repeat the
basic mistakes.
>
> diff --git a/Documentation/devicetree/bindings/clock/amlogic,a9-model-ccu.yaml b/Documentation/devicetree/bindings/clock/amlogic,a9-model-ccu.yaml
> new file mode 100644
> index 000000000000..56c5cbe1b246
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/clock/amlogic,a9-model-ccu.yaml
> @@ -0,0 +1,435 @@
> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> +# Copyright (C) 2026 Amlogic, Inc. All rights reserved
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/clock/amlogic,a9-model-ccu.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Amlogic A9 Family Standardized Model Clock Control Unit
> +
> +maintainers:
> + - Chuan Liu <chuan.liu@amlogic.com>
> +
> +description:
> + The clock tree within the A9 is composed of numerous instances of these
> + standardized model CCU (Clock Control Units).
> +
> +properties:
> + compatible:
> + oneOf:
> + - items:
Drop
> + - const: amlogic,a9-composite-ccu
> + description: Supports clock source selection, frequency division, and
> + clock gating.
This is just one big enum. You can add comments if you insist...
> + - items:
> + - const: amlogic,a9-composite-ccu-mult
> + description: Some modules have multiple input clocks and contain
> + multiple composite-ccus internally.
> + - items:
> + - const: amlogic,a9-noglitch-ccu
> + description: Provides the same functionality as composite-ccu but
> + includes glitch suppression during frequency transitions.
> + - items:
> + - const: amlogic,a9-noglitch-ccu-mult
> + description: Some modules have multiple input clocks and contain
> + multiple noglitch-ccus internally.
> + - items:
> + - const: amlogic,a9-sysbus-ccu
> + description: Consists of multiple gating arrays, commonly used for
> + Amlogic's sys_clk and axi_clk.
> +
> + reg:
> + maxItems: 1
> +
> + clocks:
> + minItems: 1
> + maxItems: 16
> +
> + clock-names:
> + minItems: 1
> + maxItems: 16
> +
> + clock-output-names:
> + minItems: 1
> + # The sysbus-ccu of A9 supports up to 128 gates
> + maxItems: 128
> +
> + '#clock-cells':
> + description:
> + The clock controller of a module may contain one or more CCU(s). When a
> + clock controller has multiple CCUs, an index is required to specify a
> + particular CCU within the clock controller.
Drop
> + oneOf:
> + - const: 0
> + description: Single clock output, no specifier needed
> + - const: 1
> + description: Multiple clocks, index selects specific output
Drop all descriptions. That's enum. Do not explain usu how DT works.
> +
> + amlogic,clock-max-frequency:
Drop property. So many wrong things here... First, start from basic
guidelines like talks or docs in kernel and understand the suffixes.
Second not a DT property.
> + description: |
> + Each clock's maximum output frequency is constrained during hardware
> + design to ensure proper timing requirements for the clock network. If the
> + clock frequency configured exceeds this design limit, it can lead to
> + abnormal behavior in modules relying on that clock and may even cause
> + cross-talk that affects other modules.
> +
> + In the driver, this property is parsed, and interface functions from the
Why would we care about driver?
This is binding, we talk about hardware.
> + CCF are called to enforce the clock's maximum frequency, preventing
> + potential issues caused by excessive clock frequency configurations.
> + $ref: /schemas/types.yaml#/definitions/uint32-array
> +
> + amlogic,reg-layout:
No, drop. Compatible defines it.
> + description:
> + These standardized model CCUs require register configuration for their
> + clock functions. This property node describes the register layout
> + parameters for each model's CCU.
> + $ref: /schemas/types.yaml#/definitions/uint32-matrix
> +
> +allOf:
> + - if:
> + properties:
> + compatible:
> + anyOf:
> + - contains:
> + const: amlogic,a9-composite-ccu
> + - contains:
> + const: amlogic,a9-noglitch-ccu
> + then:
> + properties:
> + '#clock-cells':
> + const: 0
> + clock-output-names:
> + minItems: 1
> + maxItems: 1
> + else:
> + properties:
> + '#clock-cells':
> + const: 1
> + clock-output-names:
> + minItems: 2
> +
> + - if:
> + properties:
> + compatible:
> + contains:
> + const: amlogic,a9-composite-ccu
> + then:
> + properties:
> + clocks:
> + minItems: 1
> + items:
> + - description: input clock source 0
> + - description: input clock source 1 (optional)
> + - description: input clock source 2 (optional)
> + - description: input clock source 3 (optional)
> + - description: input clock source 4 (optional)
> + - description: input clock source 5 (optional)
> + - description: input clock source 6 (optional)
> + - description: input clock source 7 (optional)
> + clock-names:
> + minItems: 1
> + items:
> + - const: clkin0
> + - const: clkin1
> + - const: clkin2
> + - const: clkin3
> + - const: clkin4
> + - const: clkin5
> + - const: clkin6
> + - const: clkin7
> + amlogic,reg-layout:
> + description: |
> + composite-ccu contains three register layout parameters:
> + * register offset
> + * bit offset
> + * divider effective bit width
> + required:
> + - amlogic,reg-layout
> + - clock-names
> +
> + - if:
> + properties:
> + compatible:
> + contains:
> + const: amlogic,a9-composite-ccu-mult
> + then:
> + properties:
> + clocks:
> + description:
> + Some clock controllers contain two composite-ccus (labeled
> + composite_a and composite_b). In certain controllers, composite_a
> + and composite_b share a common clock source, while in others they
> + have independent clock sources.
> + minItems: 1
> + items:
> + - description: composite_a/b's input clock source 0
> + - description: composite_a/b's input clock source 1 (optional)
> + - description: composite_a/b's input clock source 2 (optional)
> + - description: composite_a/b's input clock source 3 (optional)
> + - description: composite_a/b's input clock source 4 (optional)
> + - description: composite_a/b's input clock source 5 (optional)
> + - description: composite_a/b's input clock source 6 (optional)
> + - description: composite_a/b's input clock source 7 (optional)
> + - description: composite_b's input clock source 0 (optional)
> + - description: composite_b's input clock source 1 (optional)
> + - description: composite_b's input clock source 2 (optional)
> + - description: composite_b's input clock source 3 (optional)
> + - description: composite_b's input clock source 4 (optional)
> + - description: composite_b's input clock source 5 (optional)
> + - description: composite_b's input clock source 6 (optional)
> + - description: composite_b's input clock source 7 (optional)
> + clock-names:
> + minItems: 1
> + items:
> + - const: clkin0
> + - const: clkin1
> + - const: clkin2
> + - const: clkin3
> + - const: clkin4
> + - const: clkin5
> + - const: clkin6
> + - const: clkin7
> + - const: bclkin0
> + - const: bclkin1
> + - const: bclkin2
> + - const: bclkin3
> + - const: bclkin4
> + - const: bclkin5
> + - const: bclkin6
> + - const: bclkin7
> + amlogic,reg-layout:
> + description: |
> + composite-ccu contains three register layout parameters:
> + * register offset
> + * bit offset
> + * divider effective bit width
> + required:
> + - amlogic,reg-layout
> + - clock-names
> +
> + - if:
> + properties:
> + compatible:
> + contains:
> + const: amlogic,a9-noglitch-ccu
> + then:
> + properties:
> + clocks:
> + minItems: 1
> + items:
> + - description: input clock source 0
> + - description: input clock source 1 (optional)
> + - description: input clock source 2 (optional)
> + - description: input clock source 3 (optional)
> + - description: input clock source 4 (optional)
> + - description: input clock source 5 (optional)
> + - description: input clock source 6 (optional)
> + - description: input clock source 7 (optional)
> + clock-names:
> + minItems: 1
> + items:
> + - const: clkin0
> + - const: clkin1
> + - const: clkin2
> + - const: clkin3
> + - const: clkin4
> + - const: clkin5
> + - const: clkin6
> + - const: clkin7
> + required:
> + - clock-names
> +
> + - if:
> + properties:
> + compatible:
> + contains:
> + const: amlogic,a9-noglitch-ccu-mult
> + then:
> + properties:
> + clocks:
> + minItems: 1
Why?
> + items:
> + - description: input clock source 0
> + - description: input clock source 1 (optional)
Drop optional. Don't repeat constraints in free form text.
> + - description: input clock source 2 (optional)
> + - description: input clock source 3 (optional)
> + - description: input clock source 4 (optional)
> + - description: input clock source 5 (optional)
> + - description: input clock source 6 (optional)
> + - description: input clock source 7 (optional)
> + clock-names:
> + minItems: 1
> + items:
> + - const: clkin0
> + - const: clkin1
> + - const: clkin2
> + - const: clkin3
> + - const: clkin4
> + - const: clkin5
> + - const: clkin6
> + - const: clkin7
> + amlogic,reg-layout:
> + description: |
> + composite-ccu contains one register layout parameters:
> + * register offset
> + required:
> + - amlogic,reg-layout
> + - clock-names
> +
> + - if:
> + properties:
> + compatible:
> + contains:
> + const: amlogic,a9-sysbus-ccu
> + then:
> + properties:
> + clocks:
> + maxItems: 1
> + description: input clock of sysbus-ccu
list the items with description instead.
> + amlogic,reg-layout:
> + description: |
> + composite-ccu contains two register layout parameters:
> + * register offset
> + * bit offset
> + required:
> + - amlogic,reg-layout
This is huge and amount of ifs is clearly suggesting you combined way
too much into one file.
> +
> +required:
> + - compatible
> + - reg
> + - clocks
> + - clock-output-names
> + - "#clock-cells"
> +
> +additionalProperties: false
> +
> +examples:
> + - |
> + clk_dummy: clock-dummy {
> + compatible = "fixed-clock";
> + #clock-cells = <0>;
> + clock-frequency = <0>;
> + clock-output-names = "dummy";
> + status = "disabled";
> + };
Not relevant, drop entire node.
> +
> + apb {
> + #address-cells = <2>;
> + #size-cells = <2>;
> + clock-controller@350 {
> + compatible = "amlogic,a9-composite-ccu";
> + reg = <0x0 0x350 0x0 0x4>;
> + #clock-cells = <0>;
> + amlogic,clock-max-frequency = <50000000>;
> + amlogic,reg-layout = <0x0 0 7>;
> + clock-output-names = "sar_adc";
> + clocks = <&xtal_24m>,
> + <&scmi_clk 17>;
> + clock-names = "clkin0", "clkin1";
> + };
> +
> + clock-controller@290 {
> + compatible = "amlogic,a9-composite-ccu-mult";
> + reg = <0x0 0x290 0x0 0x8>;
> + #clock-cells = <1>;
> + amlogic,clock-max-frequency = <250000000>,
> + <250000000>,
> + <1200000000>;
> + amlogic,reg-layout = <0x0 0 7>,
> + <0x0 16 7>,
> + <0x4 0 7>;
> + clock-output-names = "sd_emmc_a",
> + "sd_emmc_b",
> + "sd_emmc_c";
> + clocks = <&xtal_24m>,
> + <&scmi_clk 6>,
> + <&scmi_clk 10>;
> + clock-names = "clkin0",
> + "clkin1",
> + "clkin2";
> + };
One or two are enough. Drop the rest.
>
Best regards,
Krzysztof
next prev parent reply other threads:[~2026-02-09 13:14 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-02-09 5:48 [PATCH 00/13] clk: amlogic: Introduce A9 PLL and CCU driver support Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 01/13] dt-bindings: clock: Add Amlogic A9 standardized model clock control units Chuan Liu via B4 Relay
2026-02-09 13:14 ` Krzysztof Kozlowski [this message]
2026-02-09 13:18 ` Krzysztof Kozlowski
2026-02-09 5:48 ` [PATCH 02/13] dt-bindings: clock: Add Amlogic A9 PLL controllers Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 03/13] dt-bindings: clock: Add Amlogic A9 misc clock control units Chuan Liu via B4 Relay
2026-02-09 13:15 ` Krzysztof Kozlowski
2026-02-09 5:48 ` [PATCH 04/13] clk: amlogic: Add basic clock driver Chuan Liu via B4 Relay
2026-02-09 13:17 ` Krzysztof Kozlowski
2026-02-09 5:48 ` [PATCH 05/13] clk: amlogic: Add composite " Chuan Liu via B4 Relay
2026-02-09 13:18 ` Krzysztof Kozlowski
2026-02-09 5:48 ` [PATCH 06/13] clk: amlogic: Add noglitch " Chuan Liu via B4 Relay
2026-02-09 21:51 ` Martin Blumenstingl
2026-02-09 5:48 ` [PATCH 07/13] clk: amlogic: Add duandiv " Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 08/13] clk: amlogic: Add PLL driver Chuan Liu via B4 Relay
2026-02-09 15:37 ` kernel test robot
2026-02-09 17:35 ` kernel test robot
2026-02-09 5:48 ` [PATCH 09/13] clk: amlogic: Add DT-based clock registration functions Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 10/13] clk: amlogic: Add A9 standardized model clock control units driver Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 11/13] clk: amlogic: Add A9 PLL controllers driver Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 12/13] clk: amlogic: Add A9 misc clock control units driver Chuan Liu via B4 Relay
2026-02-09 5:48 ` [PATCH 13/13] clk: amlogic: Add support for building as combined kernel module Chuan Liu via B4 Relay
2026-02-11 8:34 ` [PATCH 00/13] clk: amlogic: Introduce A9 PLL and CCU driver support Jerome Brunet
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=10caac08-1a1d-4fe9-8df8-e0b75c532e8f@kernel.org \
--to=krzk@kernel.org \
--cc=chuan.liu@amlogic.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=krzk+dt@kernel.org \
--cc=linux-amlogic@lists.infradead.org \
--cc=linux-clk@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mturquette@baylibre.com \
--cc=neil.armstrong@linaro.org \
--cc=robh@kernel.org \
--cc=sboyd@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox