mbox series

[v4,00/11] A bunch of fix and optimization patches in spmi-pmic-arb.c

Message ID 1640071211-31462-1-git-send-email-quic_fenglinw@quicinc.com
Headers show
Series A bunch of fix and optimization patches in spmi-pmic-arb.c | expand

Message

Fenglin Wu Dec. 21, 2021, 7:19 a.m. UTC
Changes in v4:
  In [v4 02/11], separated spurious interrupt handling.
  In [v4 03/11], added Fixes tag for ("spmi: pmic-arb: do not ack and clear peripheral").
  In [v4 11/11], updated the binding to address few warnings in "make dtbs_check"

Changes in v3:
  Drop [v2 07/10] as this is no longer needed after this change:
		50fc4c8cd240 ("spmi: spmi-pmic-arb: fix irq_set_type race condition")
  In [v3 07/10], updated the author email to match with Signed-off-by.
  In [v3 10/10], added the binding change in this series, and addressed issues in "make dt_binding_check"

Changes in v2:
  In [v2 01/10], added code to handle spurious interrupt.
  In [v2 03/10], adressed minor comments to update the code logic.
  In [v2 04/10], minor update to detect spurious interrupt.
  In [v2 05/10], added Fixes tag.
  In [v2 07/10], added Fixes tag and updated commit text to explain the problem.
  In [v2 08/10], added binding change to make interrupt properties as optional.
  In [v2 09/10], updated to check presence of "interrupt-controller" property.

Abhijeet Dharmapurikar (1):
  spmi: pmic-arb: add a print in cleanup_irq

Ashay Jaiswal (1):
  spmi: pmic-arb: add support to dispatch interrupt based on IRQ status

David Collins (6):
  spmi: pmic-arb: check apid against limits before calling irq handler
  spmi: pmic-arb: correct duplicate APID to PPID mapping logic
  spmi: pmic-arb: block access for invalid PMIC arbiter v5 SPMI writes
  bindings: spmi: spmi-pmic-arb: mark interrupt properties as optional
  spmi: pmic-arb: make interrupt support optional
  spmi: pmic-arb: increase SPMI transaction timeout delay

Fenglin Wu (2):
  spmi: pmic-arb: handle spurious interrupt
  dt-bindings: convert qcom,spmi-pmic-arb binding to YAML format

Subbaraman Narayanamurthy (1):
  spmi: pmic-arb: do not ack and clear peripheral interrupts in
    cleanup_irq

 .../bindings/spmi/qcom,spmi-pmic-arb.txt           |  65 ---------
 .../bindings/spmi/qcom,spmi-pmic-arb.yaml          | 146 +++++++++++++++++++++
 drivers/spmi/spmi-pmic-arb.c                       | 136 +++++++++++++------
 3 files changed, 242 insertions(+), 105 deletions(-)
 delete mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt
 create mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml

Comments

Rob Herring (Arm) Dec. 21, 2021, 11:11 a.m. UTC | #1
On Tue, 21 Dec 2021 15:20:09 +0800, Fenglin Wu wrote:
> Convert the SPMI PMIC arbiter documentation to JSON/yaml.
> 
> Signed-off-by: Fenglin Wu <quic_fenglinw@quicinc.com>
> ---
>  .../bindings/spmi/qcom,spmi-pmic-arb.txt           |  67 ----------
>  .../bindings/spmi/qcom,spmi-pmic-arb.yaml          | 146 +++++++++++++++++++++
>  2 files changed, 146 insertions(+), 67 deletions(-)
>  delete mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt
>  create mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
> 

My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):

yamllint warnings/errors:

dtschema/dtc warnings/errors:
/builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.example.dt.yaml: spmi@fc4cf000: reg: [[4232900608, 4096], [4232884224, 4096], [4232880128, 4096]] is too long
	From schema: /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/spmi.yaml
/builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.example.dt.yaml: spmi@fc4cf000: reg: [[4232900608, 4096], [4232884224, 4096], [4232880128, 4096]] is too long
	From schema: /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml

doc reference errors (make refcheckdocs):
Documentation/devicetree/bindings/mfd/qcom,spmi-pmic.txt: Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt

See https://patchwork.ozlabs.org/patch/1571409

This check can fail if there are any dependencies. The base for a patch
series is generally the most recent rc1.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit.
Rob Herring (Arm) Dec. 21, 2021, 2:42 p.m. UTC | #2
On Tue, Dec 21, 2021 at 03:20:09PM +0800, Fenglin Wu wrote:
> Convert the SPMI PMIC arbiter documentation to JSON/yaml.
> 
> Signed-off-by: Fenglin Wu <quic_fenglinw@quicinc.com>
> ---
>  .../bindings/spmi/qcom,spmi-pmic-arb.txt           |  67 ----------
>  .../bindings/spmi/qcom,spmi-pmic-arb.yaml          | 146 +++++++++++++++++++++
>  2 files changed, 146 insertions(+), 67 deletions(-)
>  delete mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt
>  create mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
> 

> diff --git a/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml b/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
> new file mode 100644
> index 0000000..df8cfb7
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
> @@ -0,0 +1,146 @@
> +# SPDX-License-Identifier: (GPL-2.0 OR BSD-2-Clause)
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/spmi/qcom,spmi-pmic-arb.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Qualcomm SPMI PMIC Arbiter
> +
> +maintainers:
> +  - Fenglin Wu <quic_fenglinw@quicinc.com>
> +  - Subbaraman Narayanamurthy <quic_subbaram@quicinc.com>
> +
> +description: |
> +  The SPMI PMIC Arbiter is found on Snapdragon chipsets. It is an SPMI
> +  controller with wrapping arbitration logic to allow for multiple
> +  on-chip devices to control a single SPMI master.
> +
> +  The PMIC Arbiter can also act as an interrupt controller, providing
> +  interrupts to slave devices.
> +
> +  See Documentation/devicetree/bindings/spmi/spmi.yaml for the generic
> +  SPMI controller binding requirements for child nodes.
> +
> +allOf:
> +  - $ref: spmi.yaml#
> +
> +properties:
> +  $nodename:
> +    pattern: "^spmi@.*"
> +
> +  compatible:
> +    const: qcom,spmi-pmic-arb
> +
> +  reg-names:
> +    $ref: /schemas/types.yaml#/definitions/string-array

reg-names already has a type defined.

> +    anyOf:
> +      - minItems: 3
> +      - maxItems: 3
> +      - enum: ["core", "intr", "cnfg"]
> +
> +      - minItems: 5
> +      - maxItems: 5
> +      - enum: ["core", "intr", "cnfg", "chnls", "obsrvr"]

I think you want something like this:

minItems: 3
items:
  - const: core
  - const: intr
  - const: cnfg
  - const: chnls
  - const: obsrvr


> +
> +  reg:
> +    minItems: 3
> +    maxItems: 5
> +    description: |
> +      Specifies base physical address and size of the registers in SPMI PMIC
> +      Arbiter HW module, with the following order.
> +        - SPMI PMIC arbiter core registers (core)
> +        - SPMI PMIC arbiter interrupt controller registers (intr)
> +        - SPMI PMIC arbiter configuration registers (cnfg)
> +        - SPMI PMIC arbiter tx-channel per virtual slave registers (chnls)
> +        - SPMI PMIC arbiter rx-channel per virtual slave registers (obsrvr).
> +      Register for "chnls" and "obsrvr" are only applicable for PMIC arbiter
> +      with HW version greater than V2.
> +
> +  "#address-cells":
> +    const: 2
> +
> +  "#size-cells":
> +    const: 0
> +
> +  interrupts:
> +    description: The summary interrupt for the PMIC Arb controller.
> +    maxItems: 1
> +
> +  interrupt-names:
> +    const: periph_irq
> +
> +  interrupt-controller: true
> +
> +  "#interrupt-cells":
> +    const: 4
> +    description: |
> +      Specifies the number of cells needed to encode any interrupt source.
> +      The 1st cell is the slave ID for the requested interrupt, its valid
> +      range is [0-15].
> +      The 2nd cell is the  peripheral ID for requested interrupt, its valid
> +      range is [0-255].
> +      The 3rd cell is the requested peripheral interrupt, its valid range
> +      is [0-7].
> +      The 4th cell is interrupt flags indicating level-sense information,
> +      as defined in dt-bindings/interrupt-controller/irq.h
> +
> +  qcom,ee:
> +    description: the active Execution Environment identifier
> +    $ref: /schemas/types.yaml#/definitions/uint32
> +    enum: [0, 1, 2, 3, 4, 5]
> +
> +  qcom,channel:
> +    description: which of the PMIC Arbiter provided channels to use for accesses
> +    $ref: /schemas/types.yaml#/definitions/uint32
> +    enum: [0, 1, 2, 3, 4, 5]
> +

> +patternProperties:
> +  "@[0-9a-f]$":
> +    description: up to 16 child PMIC nodes
> +    type: object
> +
> +    properties:
> +      reg:
> +        items:
> +          - minItems: 1
> +            items:
> +              - minimum: 0
> +                maximum: 0xf
> +              - enum: [ 0 ]
> +                description:
> +                  0 means user ID address. 1 is reserved for group ID
> +                  address.
> +
> +    required:
> +      - reg

All this should be covered by spmi.yaml

> +
> +required:
> +  - compatible
> +  - reg-names
> +  - reg
> +  - "#address-cells"
> +  - "#size-cells"
> +  - qcom,ee
> +  - qcom,channel
> +
> +additionalProperties: false
> +
> +examples:
> +  - |
> +    spmi@fc4cf000 {
> +          compatible = "qcom,spmi-pmic-arb";
> +          reg-names = "core", "intr", "cnfg";
> +          reg = <0xfc4cf000 0x1000>,
> +                <0xfc4cb000 0x1000>,
> +                <0xfc4ca000 0x1000>;
> +          interrupt-names = "periph_irq";
> +          interrupts = <0 190 0>;
> +          interrupt-controller;
> +          #interrupt-cells = <4>;
> +
> +          qcom,ee = <0>;
> +          qcom,channel = <0>;
> +
> +          #address-cells = <2>;
> +          #size-cells = <0>;
> +    };
> -- 
> 2.7.4
> 
>
Fenglin Wu Dec. 21, 2021, 11:41 p.m. UTC | #3
On 2021/12/21 19:11, Rob Herring wrote:
> On Tue, 21 Dec 2021 15:20:09 +0800, Fenglin Wu wrote:
>> Convert the SPMI PMIC arbiter documentation to JSON/yaml.
>>
>> Signed-off-by: Fenglin Wu <quic_fenglinw@quicinc.com>
>> ---
>>   .../bindings/spmi/qcom,spmi-pmic-arb.txt           |  67 ----------
>>   .../bindings/spmi/qcom,spmi-pmic-arb.yaml          | 146 +++++++++++++++++++++
>>   2 files changed, 146 insertions(+), 67 deletions(-)
>>   delete mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt
>>   create mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
>>
> My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
> on your patch (DT_CHECKER_FLAGS is new in v5.13):
>
> yamllint warnings/errors:
>
> dtschema/dtc warnings/errors:
> /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.example.dt.yaml: spmi@fc4cf000: reg: [[4232900608, 4096], [4232884224, 4096], [4232880128, 4096]] is too long
> 	From schema: /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/spmi.yaml
> /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.example.dt.yaml: spmi@fc4cf000: reg: [[4232900608, 4096], [4232884224, 4096], [4232880128, 4096]] is too long
> 	From schema: /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
I re-based the change on the tip of spmi-next project which has this 
change included:
https://lore.kernel.org/all/20211119034613.32489-2-james.lo@mediatek.com/
With it, the constraint should be removed and this warning/error won't 
be seen.
> doc reference errors (make refcheckdocs):
> Documentation/devicetree/bindings/mfd/qcom,spmi-pmic.txt: Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt
>
> See https://patchwork.ozlabs.org/patch/1571409
>
> This check can fail if there are any dependencies. The base for a patch
> series is generally the most recent rc1.
>
> If you already ran 'make dt_binding_check' and didn't see the above
> error(s), then make sure 'yamllint' is installed and dt-schema is up to
> date:
>
> pip3 install dtschema --upgrade
>
> Please check and re-submit.
>
Fenglin Wu Dec. 22, 2021, 12:45 a.m. UTC | #4
resend with plain text


On 2021/12/21 22:42, Rob Herring wrote:
> On Tue, Dec 21, 2021 at 03:20:09PM +0800, Fenglin Wu wrote:
>> Convert the SPMI PMIC arbiter documentation to JSON/yaml.
>>
>> Signed-off-by: Fenglin Wu <quic_fenglinw@quicinc.com>
>> ---
>>   .../bindings/spmi/qcom,spmi-pmic-arb.txt           |  67 ----------
>>   .../bindings/spmi/qcom,spmi-pmic-arb.yaml          | 146 +++++++++++++++++++++
>>   2 files changed, 146 insertions(+), 67 deletions(-)
>>   delete mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt
>>   create mode 100644 Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
>>
> 
>> diff --git a/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml b/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
>> new file mode 100644
>> index 0000000..df8cfb7
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.yaml
>> @@ -0,0 +1,146 @@
>> +# SPDX-License-Identifier: (GPL-2.0 OR BSD-2-Clause)
>> +%YAML 1.2
>> +---
>> +$id: http://devicetree.org/schemas/spmi/qcom,spmi-pmic-arb.yaml#
>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>> +
>> +title: Qualcomm SPMI PMIC Arbiter
>> +
>> +maintainers:
>> +  - Fenglin Wu <quic_fenglinw@quicinc.com>
>> +  - Subbaraman Narayanamurthy <quic_subbaram@quicinc.com>
>> +
>> +description: |
>> +  The SPMI PMIC Arbiter is found on Snapdragon chipsets. It is an SPMI
>> +  controller with wrapping arbitration logic to allow for multiple
>> +  on-chip devices to control a single SPMI master.
>> +
>> +  The PMIC Arbiter can also act as an interrupt controller, providing
>> +  interrupts to slave devices.
>> +
>> +  See Documentation/devicetree/bindings/spmi/spmi.yaml for the generic
>> +  SPMI controller binding requirements for child nodes.
>> +
>> +allOf:
>> +  - $ref: spmi.yaml#
>> +
>> +properties:
>> +  $nodename:
>> +    pattern: "^spmi@.*"
>> +
>> +  compatible:
>> +    const: qcom,spmi-pmic-arb
>> +
>> +  reg-names:
>> +    $ref: /schemas/types.yaml#/definitions/string-array
> 
> reg-names already has a type defined.
I understand there is a pattern property defined in dt-core.yaml and it 
defines ".*-names" as a "non-unique-string-array" type. But here, the 
strings in "reg-names" needs to be unique and it has to be ["core", 
"intr", "cnfg"] or ["core", "intr", "cnfg", "chnls", "obsrvr"] , that's 
why I redefined it as "string-array" type which requires each string to 
be unique. Otherwise, if any dtsi nodes define the "reg-name" as 
["core", "core", "core"] will not be caught as a fault.

> 
>> +    anyOf:
>> +      - minItems: 3
>> +      - maxItems: 3
>> +      - enum: ["core", "intr", "cnfg"]
>> +
>> +      - minItems: 5
>> +      - maxItems: 5
>> +      - enum: ["core", "intr", "cnfg", "chnls", "obsrvr"]
> 
> I think you want something like this:
> 
> minItems: 3
> items:
>    - const: core
>    - const: intr
>    - const: cnfg
>    - const: chnls
>    - const: obsrvr
> 
> 
As I said, the content for "reg-names" here only has two options , 
either ["core", "intr", "cnfg"] or ["core", "intr", "cnfg", "chnls", 
"obsrvr"]. In patch V3, I defined it as below and "make dtbs_check" 
threw out warnings because some of existing nodes defined "reg-names" 
with these strings are not having the same order as I defined here (I 
understood from the warnings that const items need to be followed 
strictly even in order wise, is this correct?), and I guess the order of 
the strings doesn't matter here and the schema here shouldn't have such 
limitation, so I updated it as the "array-string" type and specified the 
tuples can only be one of the strings defined in the enum. With this, 
the previous warning regarding "reg-names" in "make dtbs_check" are all 
fixed.

   reg-names:
     oneOf:
       - items:
           - const: core
           - const: intr
           - const: cnfg
       - items:
           - const: core
           - const: intr
           - const: cnfg
           - const: chnls
           - const: obsrvr


>> +
>> +  reg:
>> +    minItems: 3
>> +    maxItems: 5
>> +    description: |
>> +      Specifies base physical address and size of the registers in SPMI PMIC
>> +      Arbiter HW module, with the following order.
>> +        - SPMI PMIC arbiter core registers (core)
>> +        - SPMI PMIC arbiter interrupt controller registers (intr)
>> +        - SPMI PMIC arbiter configuration registers (cnfg)
>> +        - SPMI PMIC arbiter tx-channel per virtual slave registers (chnls)
>> +        - SPMI PMIC arbiter rx-channel per virtual slave registers (obsrvr).
>> +      Register for "chnls" and "obsrvr" are only applicable for PMIC arbiter
>> +      with HW version greater than V2.
>> +
>> +  "#address-cells":
>> +    const: 2
>> +
>> +  "#size-cells":
>> +    const: 0
>> +
>> +  interrupts:
>> +    description: The summary interrupt for the PMIC Arb controller.
>> +    maxItems: 1
>> +
>> +  interrupt-names:
>> +    const: periph_irq
>> +
>> +  interrupt-controller: true
>> +
>> +  "#interrupt-cells":
>> +    const: 4
>> +    description: |
>> +      Specifies the number of cells needed to encode any interrupt source.
>> +      The 1st cell is the slave ID for the requested interrupt, its valid
>> +      range is [0-15].
>> +      The 2nd cell is the  peripheral ID for requested interrupt, its valid
>> +      range is [0-255].
>> +      The 3rd cell is the requested peripheral interrupt, its valid range
>> +      is [0-7].
>> +      The 4th cell is interrupt flags indicating level-sense information,
>> +      as defined in dt-bindings/interrupt-controller/irq.h
>> +
>> +  qcom,ee:
>> +    description: the active Execution Environment identifier
>> +    $ref: /schemas/types.yaml#/definitions/uint32
>> +    enum: [0, 1, 2, 3, 4, 5]
>> +
>> +  qcom,channel:
>> +    description: which of the PMIC Arbiter provided channels to use for accesses
>> +    $ref: /schemas/types.yaml#/definitions/uint32
>> +    enum: [0, 1, 2, 3, 4, 5]
>> +
> 
>> +patternProperties:
>> +  "@[0-9a-f]$":
>> +    description: up to 16 child PMIC nodes
>> +    type: object
>> +
>> +    properties:
>> +      reg:
>> +        items:
>> +          - minItems: 1
>> +            items:
>> +              - minimum: 0
>> +                maximum: 0xf
>> +              - enum: [ 0 ]
>> +                description:
>> +                  0 means user ID address. 1 is reserved for group ID
>> +                  address.
>> +
>> +    required:
>> +      - reg
> 
> All this should be covered by spmi.yaml
> 
>> +
>> +required:
>> +  - compatible
>> +  - reg-names
>> +  - reg
>> +  - "#address-cells"
>> +  - "#size-cells"
>> +  - qcom,ee
>> +  - qcom,channel
>> +
>> +additionalProperties: false
>> +
>> +examples:
>> +  - |
>> +    spmi@fc4cf000 {
>> +          compatible = "qcom,spmi-pmic-arb";
>> +          reg-names = "core", "intr", "cnfg";
>> +          reg = <0xfc4cf000 0x1000>,
>> +                <0xfc4cb000 0x1000>,
>> +                <0xfc4ca000 0x1000>;
>> +          interrupt-names = "periph_irq";
>> +          interrupts = <0 190 0>;
>> +          interrupt-controller;
>> +          #interrupt-cells = <4>;
>> +
>> +          qcom,ee = <0>;
>> +          qcom,channel = <0>;
>> +
>> +          #address-cells = <2>;
>> +          #size-cells = <0>;
>> +    };
>> -- 
>> 2.7.4
>>
>>
Fenglin Wu Dec. 22, 2021, 12:48 a.m. UTC | #5
On 2021/12/21 23:13, Rob Herring wrote:
> On Tue, 21 Dec 2021 15:20:06 +0800, Fenglin Wu wrote:
>> From: David Collins <collinsd@codeaurora.org>
>>
>> Mark all interrupt related properties as optional instead of
>> required.  Some boards do not required PMIC IRQ support and it
>> isn't needed to handle SPMI bus transactions, so specify it as
>> optional.
>>
>> Signed-off-by: David Collins <collinsd@codeaurora.org>
>> Signed-off-by: Fenglin Wu <quic_fenglinw@quicinc.com>
>> ---
>>   Documentation/devicetree/bindings/spmi/qcom,spmi-pmic-arb.txt | 2 ++
>>   1 file changed, 2 insertions(+)
>>
> 
> 
> Please add Acked-by/Reviewed-by tags when posting new versions. However,
> there's no need to repost patches *only* to add the tags. The upstream
> maintainer will do that for acks received on the version they apply.
> 
> If a tag was not added on purpose, please state why and what changed.
> 
I will remember to add the Acked-by/Reviewed-by tags next time when 
sending any update in this series. Thanks for the notice!