mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Yin Li <yin.li@oss.qualcomm.com>
To: Rob Herring <robh@kernel.org>
Cc: "James Morse" <james.morse@arm.com>,
	"Shanker Donthineni" <sdonthineni@nvidia.com>,
	"Ben Horgan" <ben.horgan@arm.com>,
	"Krzysztof Kozlowski" <krzk+dt@kernel.org>,
	"Conor Dooley" <conor+dt@kernel.org>,
	"Catalin Marinas" <catalin.marinas@arm.com>,
	"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
	"Rafael J. Wysocki" <rafael@kernel.org>,
	"Danilo Krummrich" <dakr@kernel.org>,
	"Reinette Chatre" <reinette.chatre@intel.com>,
	"Fenghua Yu" <fenghuay@nvidia.com>,
	"Jonathan Cameron" <jic23@kernel.org>,
	"Bjorn Andersson" <andersson@kernel.org>,
	"Konrad Dybcio" <konradybcio@kernel.org>,
	"Gavin Shan" <gshan@redhat.com>,
	"Andre Przywara" <andre.przywara@arm.com>,
	"Drew Fustini" <fustini@kernel.org>,
	"Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>,
	"Shaopeng Tan" <tan.shaopeng@jp.fujitsu.com>,
	trilok.soni@oss.qualcomm.com, aiqun.yu@oss.qualcomm.com,
	ganapatrao.kulkarni@oss.qualcomm.com,
	"Srivathsa L Rao" <srivathsa.rao@oss.qualcomm.com>,
	"Huang Yiwei" <huang.yiwei@oss.qualcomm.com>,
	linux-arm-kernel@lists.infradead.org,
	linux-arm-msm@vger.kernel.org, linux-kernel@vger.kernel.org,
	driver-core@lists.linux.dev, devicetree@vger.kernel.org
Subject: Re: [PATCH v3 3/8] dt-bindings: arm: Add MPAM MSC binding
Date: Mon, 12 Oct 2026 18:24:06 +0800	[thread overview]
Message-ID: <51bc6576-7e52-4685-b7b2-a355c2b7e1ae@oss.qualcomm.com> (raw)
In-Reply-To: <2eddfb6d-c622-4d72-827e-aed685e72b9c@oss.qualcomm.com>



On 10/12/2026 3:05 PM, Yin Li wrote:
> 
> 
> On 10/9/2026 5:27 AM, Rob Herring wrote:
>> On Thu, Oct 08, 2026 at 02:41:12PM +0800, Yin Li wrote:
>>> From: Rob Herring <robh@kernel.org>
>>>
>>> The binding is designed around the assumption that an MSC will be a
>>> sub-block of something else such as a memory controller, cache 
>>> controller,
>>> or IOMMU. However, it's certainly possible a design does not have that
>>> association or has a mixture of both, so the binding illustrates how 
>>> we can
>>> support that with RIS child nodes.
>>>
>>> A key part of MPAM is we need to know about all of the MSCs in the 
>>> system
>>> before it can be enabled. This drives the need for the genericish
>>> 'arm,mpam-msc' compatible. Though we can't assume an MSC is accessible
>>> until a h/w specific driver potentially enables the h/w.
>>>
>>> Cc: James Morse <james.morse@arm.com>
>>> Signed-off-by: Rob Herring <robh@kernel.org>
>>> Signed-off-by: James Morse <james.morse@arm.com>
>>> [ Yin Li and Huang Yiwei: use oneOf for the compatible to allow both
>>>    standalone arm,mpam-msc and the two-item combination; fix schema 
>>> field
>>>    order, type constraints and yamllint spacing; fix example reg 
>>> addresses,
>>>    add arm,mpam-device phandle and remove TODO comment ]
>>> Co-developed-by: Huang Yiwei <huang.yiwei@oss.qualcomm.com>
>>> Signed-off-by: Huang Yiwei <huang.yiwei@oss.qualcomm.com>
>>> Signed-off-by: Yin Li <yin.li@oss.qualcomm.com>
>>> ---
>>>   .../devicetree/bindings/arm/arm,mpam-msc.yaml      | 202 ++++++++++ 
>>> +++++++++++
>>>   1 file changed, 202 insertions(+)
>>>
>>> diff --git a/Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml 
>>> b/Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml
>>> new file mode 100644
>>> index 000000000000..3c7b13a23ff2
>>> --- /dev/null
>>> +++ b/Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml
>>> @@ -0,0 +1,202 @@
>>> +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
>>> +%YAML 1.2
>>> +---
>>> +$id: http://devicetree.org/schemas/arm/arm,mpam-msc.yaml#
>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>> +
>>> +title: Arm Memory System Resource Partitioning and Monitoring (MPAM)
>>> +
>>> +maintainers:
>>> +  - Rob Herring <robh@kernel.org>
>>> +
>>> +description: |
>>> +  The Arm MPAM specification can be found here:
>>> +
>>> +  https://developer.arm.com/documentation/ddi0598/latest
>>> +
>>> +properties:
>>> +  compatible:
>>> +    oneOf:
>>> +      - const: arm,mpam-msc                   # Further details are 
>>> discoverable
>>> +      - items:
>>> +          - const: arm,mpam-memory-controller-msc
>>
>> Why do we need this?
>>
> 
> Hi Rob,
> 
> Thanks for the review.
> 
> The 'arm,mpam-memory-controller-msc' compatible is needed because,
> unlike cache MSCs, memory controller MSCs cannot rely on the parent node
> to identify the resource type.
> 
> Cache controllers universally carry a 'cache' compatible, so the driver
> can recognise any cache parent and derive the cache ID from standard
> properties. No equivalent convention exists for memory controllers, and
> the component ID comes from the MSC's own 'numa-node-id' rather than the
> parent, so there is no parent attribute to rely on.
> 

Hi Rob,

As a follow-up to my previous reply:
Whatever the outcome of the discussion on "arm_mpam: Add support for
memory controller MSC on DT platforms", we will keep the binding and
driver support consistent. If it's dropped, the 
'arm,mpam-memory-controller-msc' compatible, 'numa-node-id' property, 
and the memory controller example will be removed together.

Thanks,
Yin


> 'arm,mpam-memory-controller-msc' provides the explicit declaration that
> this MSC controls memory. It is used in "arm_mpam: Add support for
> memory controller MSC on DT platforms".
> 
> On 'oneOf': the original 'items' form requires both strings to be
> present, which is not appropriate for cache MSCs. Andre also noted this 
> in RFC v2, so we changed it to 'oneOf'.
> 
>>> +          - const: arm,mpam-msc
>>> +
>>> +  reg:
>>> +    maxItems: 1
>>> +    description: A memory region containing registers as defined in 
>>> the MPAM
>>> +      specification.
>>> +
>>> +  interrupts:
>>> +    minItems: 1
>>> +    items:
>>> +      - description: error (optional)
>>> +      - description: overflow (optional, only for monitoring)
>>> +
>>> +  interrupt-names:
>>> +    oneOf:
>>> +      - items:
>>> +          - enum: [error, overflow]
>>> +      - items:
>>> +          - const: error
>>> +          - const: overflow
>>
>> This can be expressed as:
>>
>>        minItems: 1
>>        items:
>>          - enum: [error, overflow]
>>          - const: overflow
>>
> 
> Agreed, will update.
> 
>>> +
>>> +  arm,not-ready-us:
>>> +    description: The maximum time in microseconds for monitoring 
>>> data to be
>>> +      accurate after a settings change. For more information, see the
>>> +      Not-Ready (NRDY) bit description in the MPAM specification.
>>> +
>>> +  numa-node-id:
>>> +    $ref: /schemas/types.yaml#/definitions/uint32
>>> +    description: NUMA node identifier for the memory resource this 
>>> MSC controls.
>>> +
>>> +  '#address-cells':
>>> +    const: 1
>>> +
>>> +  '#size-cells':
>>> +    const: 0
>>> +
>>> +patternProperties:
>>> +  '^ris@[0-9a-f]+$':
>>
>> Based on the 'reg' constraints, drop the '+'.
> 
> Agreed, will update.
> 
>>
>>> +    type: object
>>> +    additionalProperties: false
>>> +    description:
>>> +      RIS nodes for each resource instance in an MSC. These nodes 
>>> are required
>>> +      for each resource instance implementing known MPAM controls
>>> +
>>> +    properties:
>>> +      compatible:
>>> +        enum:
>>> +          - arm,mpam-cache
>>> +          # Memory bandwidth
>>> +          - arm,mpam-memory
>>> +
>>> +      reg:
>>> +        minimum: 0
>>> +        maximum: 0xf
>>> +
>>> +      cpus:
>>> +        $ref: /schemas/types.yaml#/definitions/phandle-array
>>
>> Drop. 'cpus' has a type already.
>>
> 
> Agreed, will update.
> 
>>> +        description:
>>> +          Phandle(s) to the CPU node(s) this RIS belongs to. By 
>>> default, the
>>> +          parent device's affinity is used.
>>> +
>>> +      arm,mpam-device:
>>> +        $ref: /schemas/types.yaml#/definitions/phandle
>>> +        description:
>>> +          By default, the MPAM enabled device associated with a RIS 
>>> is the MSC's
>>> +          parent node. It is possible for each RIS to be associated 
>>> with
>>> +          different devices in which case 'arm,mpam-device' should 
>>> be used.
>>
>> We should drop this until there's a use for it.
>>
> 
> On 'arm,mpam-device': we believe this property is structurally
> necessary. The MPAM specification (IHI0099B) states:
>      "Resource instance selection (RIS) allows an MPAM MSC to support
>      multiple resources, including multiple resources using the same
>      resource type or partitioning control. When RIS is implemented, an
>      MPAM MSC must implement independent resource controls and two or
>      more resources of the same type."
> 
> Once an MSC implements RIS, it must support ≥ 2 resources of the same
> type (a spec requirement). An MSC node has only one parent, so the
> association between each RIS and its corresponding device cannot be
> expressed through a shared parent alone. A per-RIS explicit association
> mechanism is therefore needed, and 'arm,mpam-device' is designed for
> exactly this purpose.
> 
> The property is used in the immediately following patch "arm_mpam: Add
> device tree support for MSC probing". If you have a different view on
> this design, we are happy to discuss further.
> 
> 
>>> +
>>> +    required:
>>> +      - compatible
>>> +      - reg
>>> +
>>> +required:
>>> +  - compatible
>>> +  - reg
>>> +
>>> +dependencies:
>>> +  interrupts: [interrupt-names]
>>> +
>>> +additionalProperties: false
>>> +
>>> +examples:
>>> +  - |
>>> +    L3: cache-controller@30000000 {
>>> +        compatible = "foo,a-l3-cache", "cache";
>>> +        cache-level = <3>;
>>> +        cache-unified;
>>> +
>>> +        ranges = <0x0 0x30000000 0x800000>;
>>> +        #address-cells = <1>;
>>> +        #size-cells = <1>;
>>> +
>>> +        msc@10000 {
>>> +            compatible = "arm,mpam-msc";
>>> +            reg = <0x10000 0x2000>;
>>> +            interrupts = <1>, <2>;
>>> +            interrupt-names = "error", "overflow";
>>> +            arm,not-ready-us = <1>;
>>> +            /* CPU affinity implied by parent cache node  */
>>> +        };
>>> +    };
>>> +
>>> +    mem: memory-controller@20000 {
>>> +        compatible = "foo,a-memory-controller";
>>> +        reg = <0x20000 0x1000>;
>>> +
>>> +        #address-cells = <1>;
>>> +        #size-cells = <1>;
>>> +        ranges;
>>> +
>>> +        msc@21000 {
>>> +            compatible = "arm,mpam-memory-controller-msc", 
>>> "arm,mpam-msc";
>>> +            reg = <0x21000 0x1000>;
>>> +            interrupts = <3>;
>>> +            interrupt-names = "error";
>>> +            arm,not-ready-us = <1>;
>>> +            numa-node-id = <1>;
>>> +        };
>>> +    };
>>> +
>>> +    iommu@40000 {
>>> +        reg = <0x40000 0x1000>;
>>> +
>>> +        ranges;
>>> +        #address-cells = <1>;
>>> +        #size-cells = <1>;
>>> +
>>> +        msc@41000 {
>>> +            compatible = "arm,mpam-msc";
>>> +            reg = <0x41000 0x1000>;
>>> +            interrupts = <5>, <6>;
>>> +            interrupt-names = "error", "overflow";
>>> +            arm,not-ready-us = <1>;
>>> +
>>> +            #address-cells = <1>;
>>> +            #size-cells = <0>;
>>> +
>>> +            ris@2 {
>>> +                compatible = "arm,mpam-cache";
>>> +                reg = <2>;
>>> +                arm,mpam-device = <&L3>;
>>> +            };
>>> +        };
>>> +    };
>>> +
>>> +    msc@80000 {
>>> +        compatible = "foo,a-standalone-msc";
>>> +        reg = <0x80000 0x1000>;
>>> +
>>> +        clocks = <&clks 123>;
>>> +
>>> +        ranges;
>>> +        #address-cells = <1>;
>>> +        #size-cells = <1>;
>>> +
>>> +        msc@10000 {
>>> +            compatible = "arm,mpam-msc";
>>> +            reg = <0x10000 0x2000>;
>>> +            interrupts = <7>;
>>> +            interrupt-names = "overflow";
>>> +            arm,not-ready-us = <1>;
>>> +
>>> +            #address-cells = <1>;
>>> +            #size-cells = <0>;
>>> +
>>> +            ris@0 {
>>> +                compatible = "arm,mpam-cache";
>>> +                reg = <0>;
>>> +                arm,mpam-device = <&L2_0>;
>>> +            };
>>> +
>>> +            ris@1 {
>>> +                compatible = "arm,mpam-memory";
>>> +                reg = <1>;
>>> +                arm,mpam-device = <&mem>;
>>> +            };
>>> +        };
>>> +    };
>>> +
>>> +...
>>>
>>> -- 
>>> 2.34.1
>>>
> 

-- 
Thx and BRs,
Yin


  reply	other threads:[~2026-10-12 10:24 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-08  6:41 [PATCH v3 0/8] arm-mpam: Add basic device tree support for resctrl Yin Li
2026-10-08  6:41 ` [PATCH v3 1/8] arm_mpam: Fix the RIS index range check in mpam_ris_create_locked Yin Li
2026-10-08  6:41 ` [PATCH v3 2/8] arm_mpam: Fix MSC MMIO window size off-by-one with resource_size() Yin Li
2026-10-08  6:41 ` [PATCH v3 3/8] dt-bindings: arm: Add MPAM MSC binding Yin Li
2026-10-08 21:27   ` Rob Herring
2026-10-12  7:05     ` Yin Li
2026-10-12 10:24       ` Yin Li [this message]
2026-10-08  6:41 ` [PATCH v3 4/8] cacheinfo: Expose the code to generate a cache-id from a device_node Yin Li
2026-10-08  6:41 ` [PATCH v3 5/8] arm_mpam: Add device tree support for MSC probing Yin Li
2026-10-08  6:41 ` [PATCH v3 6/8] arm_mpam: Add support for memory controller MSC on DT platforms Yin Li
2026-10-08 21:29   ` Rob Herring
2026-10-12 10:11     ` Yin Li
2026-10-08  6:41 ` [PATCH v3 7/8] arm_mpam: Support MSC accessibility derivation from RIS nodes Yin Li
2026-10-08  6:41 ` [PATCH DNM v3 8/8] arm64: dts: qcom: kaanapali: Add MPAM MSC nodes for the L2 caches Yin Li
2026-10-08 21:00   ` Rob Herring
2026-10-12 10:48     ` Yin Li
2026-10-12 11:27       ` Ben Horgan

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=51bc6576-7e52-4685-b7b2-a355c2b7e1ae@oss.qualcomm.com \
    --to=yin.li@oss.qualcomm.com \
    --cc=aiqun.yu@oss.qualcomm.com \
    --cc=andersson@kernel.org \
    --cc=andre.przywara@arm.com \
    --cc=ben.horgan@arm.com \
    --cc=catalin.marinas@arm.com \
    --cc=conor+dt@kernel.org \
    --cc=dakr@kernel.org \
    --cc=devicetree@vger.kernel.org \
    --cc=driver-core@lists.linux.dev \
    --cc=fenghuay@nvidia.com \
    --cc=fustini@kernel.org \
    --cc=ganapatrao.kulkarni@oss.qualcomm.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=gshan@redhat.com \
    --cc=huang.yiwei@oss.qualcomm.com \
    --cc=ilpo.jarvinen@linux.intel.com \
    --cc=james.morse@arm.com \
    --cc=jic23@kernel.org \
    --cc=konradybcio@kernel.org \
    --cc=krzk+dt@kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rafael@kernel.org \
    --cc=reinette.chatre@intel.com \
    --cc=robh@kernel.org \
    --cc=sdonthineni@nvidia.com \
    --cc=srivathsa.rao@oss.qualcomm.com \
    --cc=tan.shaopeng@jp.fujitsu.com \
    --cc=trilok.soni@oss.qualcomm.com \
    /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

all inboxes | Powered by JetHome®