[PATCH] Update arm,smmu.yaml bindings doc

[PATCH] Update arm,smmu.yaml bindings doc

[Sharat Masetty]

Patch [1] adds a "mem_iface_clock" in the clocks list for smmu device.
This patch updates the yaml doc file for smmu by adding the definition
for this new clock.

1: https://patchwork.freedesktop.org/patch/352718/

Sharat Masetty (1):
  dt-bindings: arm-smmu: update clocks and bindings for sc7180 SoC

 Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
 1 file changed, 3 insertions(+)

--
1.9.1

[PATCH] dt-bindings: arm-smmu: update the list of clocks

[Sharat Masetty]

This patch adds a clock definition needed for powering on the GPU TBUs
and the GPU TCU.

Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
---
 Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/Documentation/devicetree/bindings/iommu/arm,smmu.yaml b/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
index 6515dbe..235c0df 100644
--- a/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
+++ b/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
@@ -28,6 +28,7 @@ properties:
           - enum:
               - qcom,msm8996-smmu-v2
               - qcom,msm8998-smmu-v2
+              - qcom,sc7180-smmu-v2
               - qcom,sdm845-smmu-v2
           - const: qcom,smmu-v2

@@ -116,6 +117,7 @@ properties:
     items:
       - const: bus
       - const: iface
+      - const: mem_iface_clk

   clocks:
     items:
@@ -123,6 +125,7 @@ properties:
           smmu ptw
       - description: interface clock required to access smmu's registers
           through the TCU's programming interface.
+      - description: core clock required for the GPU SMMU TBUs and the GPU TCU

   power-domains:
     maxItems: 1
--
1.9.1

Re: [PATCH] dt-bindings: arm-smmu: update the list of clocks

[Rob Herring]

On Thu, 20 Feb 2020 13:42:22 +0530, Sharat Masetty wrote:
> This patch adds a clock definition needed for powering on the GPU TBUs
> and the GPU TCU.
> 
> Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
> ---
>  Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
>  1 file changed, 3 insertions(+)
> 

My bot found errors running 'make dt_binding_check' on your patch:

Documentation/devicetree/bindings/display/simple-framebuffer.example.dts:21.16-37.11: Warning (chosen_node_is_root): /example-0/chosen: chosen node must be at root node
/builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clock-names: ['bus', 'iface'] is too short
/builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clocks: [[4294967295, 123], [4294967295, 124]] is too short

See https://patchwork.ozlabs.org/patch/1241297
Please check and re-submit.

Re: [PATCH] dt-bindings: arm-smmu: update the list of clocks

[Sharat Masetty]

On 2/21/2020 2:05 AM, Rob Herring wrote:
> On Thu, 20 Feb 2020 13:42:22 +0530, Sharat Masetty wrote:
>> This patch adds a clock definition needed for powering on the GPU TBUs
>> and the GPU TCU.
>>
>> Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
>> ---
>>   Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
>>   1 file changed, 3 insertions(+)
>>
> My bot found errors running 'make dt_binding_check' on your patch:
>
> Documentation/devicetree/bindings/display/simple-framebuffer.example.dts:21.16-37.11: Warning (chosen_node_is_root): /example-0/chosen: chosen node must be at root node
> /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clock-names: ['bus', 'iface'] is too short
> /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clocks: [[4294967295, 123], [4294967295, 124]] is too short
>
> See https://patchwork.ozlabs.org/patch/1241297
> Please check and re-submit.
Hi Rob, These issues seem to be from the original code and not related 
to my patch. Are these going to be blocking errors?

Re: [PATCH] dt-bindings: arm-smmu: update the list of clocks

[Rob Herring]

On Wed, Feb 26, 2020 at 5:17 AM Sharat Masetty <smasetty@codeaurora.org> wrote:
>
>
> On 2/21/2020 2:05 AM, Rob Herring wrote:
> > On Thu, 20 Feb 2020 13:42:22 +0530, Sharat Masetty wrote:
> >> This patch adds a clock definition needed for powering on the GPU TBUs
> >> and the GPU TCU.
> >>
> >> Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
> >> ---
> >>   Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
> >>   1 file changed, 3 insertions(+)
> >>
> > My bot found errors running 'make dt_binding_check' on your patch:
> >
> > Documentation/devicetree/bindings/display/simple-framebuffer.example.dts:21.16-37.11: Warning (chosen_node_is_root): /example-0/chosen: chosen node must be at root node
> > /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clock-names: ['bus', 'iface'] is too short
> > /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clocks: [[4294967295, 123], [4294967295, 124]] is too short
> >
> > See https://patchwork.ozlabs.org/patch/1241297
> > Please check and re-submit.
> Hi Rob, These issues seem to be from the original code and not related
> to my patch. Are these going to be blocking errors?

There are no errors in this binding in mainline. You've added a 3rd
clock when all the existing users have exactly 2 clocks.

Rob

Re: [PATCH] dt-bindings: arm-smmu: update the list of clocks

[Sharat Masetty]

On 2/26/2020 8:03 PM, Rob Herring wrote:
> On Wed, Feb 26, 2020 at 5:17 AM Sharat Masetty <smasetty@codeaurora.org> wrote:
>>
>> On 2/21/2020 2:05 AM, Rob Herring wrote:
>>> On Thu, 20 Feb 2020 13:42:22 +0530, Sharat Masetty wrote:
>>>> This patch adds a clock definition needed for powering on the GPU TBUs
>>>> and the GPU TCU.
>>>>
>>>> Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
>>>> ---
>>>>    Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
>>>>    1 file changed, 3 insertions(+)
>>>>
>>> My bot found errors running 'make dt_binding_check' on your patch:
>>>
>>> Documentation/devicetree/bindings/display/simple-framebuffer.example.dts:21.16-37.11: Warning (chosen_node_is_root): /example-0/chosen: chosen node must be at root node
>>> /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clock-names: ['bus', 'iface'] is too short
>>> /builds/robherring/linux-dt-review/Documentation/devicetree/bindings/iommu/arm,smmu.example.dt.yaml: iommu@d00000: clocks: [[4294967295, 123], [4294967295, 124]] is too short
>>>
>>> See https://patchwork.ozlabs.org/patch/1241297
>>> Please check and re-submit.
>> Hi Rob, These issues seem to be from the original code and not related
>> to my patch. Are these going to be blocking errors?
> There are no errors in this binding in mainline. You've added a 3rd
> clock when all the existing users have exactly 2 clocks.

Rob,

Adding something like the following seems to be solving the bot errors, 
but I am not certain if this is the right way to address this issue. Can 
you please comment?

    clock-names:
+    minItems: 2
+    maxItems: 3
      items:
        - const: bus
        - const: iface
+      - const: mem_iface_clk

    clocks:
+    minItems: 2
+    maxItems: 3
      items:
        - description: bus clock required for downstream bus access and 
for the
            smmu ptw
        - description: interface clock required to access smmu's registers
            through the TCU's programming interface.
+      - description: core clock required for the GPU SMMU TBUs and the 
GPU TCU.

>
> Rob
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] dt-bindings: arm-smmu: update the list of clocks

[Matthias Kaehlcke]

On Thu, Feb 20, 2020 at 01:42:22PM +0530, Sharat Masetty wrote:
> This patch adds a clock definition needed for powering on the GPU TBUs
> and the GPU TCU.
> 
> Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
> ---
>  Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/iommu/arm,smmu.yaml b/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
> index 6515dbe..235c0df 100644
> --- a/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
> +++ b/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
> @@ -28,6 +28,7 @@ properties:
>            - enum:
>                - qcom,msm8996-smmu-v2
>                - qcom,msm8998-smmu-v2
> +              - qcom,sc7180-smmu-v2

The addition of the compatible string isn't (directly) related with $subject,
this should be done in a separate patch.

Re: [PATCH] dt-bindings: arm-smmu: update the list of clocks

[Robin Murphy]

[ /me fires off MAINTAINERS patch... ]

On 20/02/2020 9:02 pm, Matthias Kaehlcke wrote:
> On Thu, Feb 20, 2020 at 01:42:22PM +0530, Sharat Masetty wrote:
>> This patch adds a clock definition needed for powering on the GPU TBUs
>> and the GPU TCU.
>>
>> Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
>> ---
>>   Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 3 +++
>>   1 file changed, 3 insertions(+)
>>
>> diff --git a/Documentation/devicetree/bindings/iommu/arm,smmu.yaml b/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
>> index 6515dbe..235c0df 100644
>> --- a/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
>> +++ b/Documentation/devicetree/bindings/iommu/arm,smmu.yaml
>> @@ -28,6 +28,7 @@ properties:
>>             - enum:
>>                 - qcom,msm8996-smmu-v2
>>                 - qcom,msm8998-smmu-v2
>> +              - qcom,sc7180-smmu-v2
> 
> The addition of the compatible string isn't (directly) related with $subject,
> this should be done in a separate patch.

...or the patch should just be accurately described as "add support for 
new variant X, which requires an additional clock Y".

And speaking of which, can anyone clarify how "TCU and TBU core clock" 
manages to abbreviate to "mem_iface_clk"? I would have naively assumed 
something like "core" would be most logical.

Robin.