From: Yassine Oudjana <[email protected]>
Add the CAMSS power domain which is needed for the proper operation of
CAMSS, and add power-domain-names to ease fetching it as well as the other
power domains.
Signed-off-by: Yassine Oudjana <[email protected]>
Reviewed-by: Bryan O'Donoghue <[email protected]>
---
.../bindings/media/qcom,msm8996-camss.yaml | 13 ++++++++++++-
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml b/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml
index 8a10aa1cafc5..27c9a11f0df9 100644
--- a/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml
+++ b/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml
@@ -85,6 +85,13 @@ properties:
items:
- description: VFE0 GDSC - Video Front End, Global Distributed Switch Controller.
- description: VFE1 GDSC - Video Front End, Global Distributed Switch Controller.
+ - description: CAMSS GDSC - Camera Subsystem, Global Distributed Switch Controller.
+
+ power-domain-names:
+ items:
+ - const: vfe0
+ - const: vfe1
+ - const: camss
ports:
$ref: /schemas/graph.yaml#/properties/ports
@@ -209,6 +216,7 @@ required:
- interrupts
- iommus
- power-domains
+ - power-domain-names
- reg
- reg-names
- vdda-supply
@@ -326,7 +334,10 @@ examples:
<&vfe_smmu 3>;
power-domains = <&mmcc VFE0_GDSC>,
- <&mmcc VFE1_GDSC>;
+ <&mmcc VFE1_GDSC>,
+ <&mmcc CAMSS_GDSC>;
+
+ power-domain-names = "vfe0", "vfe1", "camss";
reg = <0x00a34000 0x1000>,
<0x00a00030 0x4>,
--
2.40.1
Yo Yassine,
On Fri, May 26, 2023 at 09:07:10PM +0300, Yassine Oudjana wrote:
> From: Yassine Oudjana <[email protected]>
>
> Add the CAMSS power domain which is needed for the proper operation of
> CAMSS, and add power-domain-names to ease fetching it as well as the other
> power domains.
>
> Signed-off-by: Yassine Oudjana <[email protected]>
> Reviewed-by: Bryan O'Donoghue <[email protected]>
> ---
> .../bindings/media/qcom,msm8996-camss.yaml | 13 ++++++++++++-
> 1 file changed, 12 insertions(+), 1 deletion(-)
>
> diff --git a/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml b/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml
> index 8a10aa1cafc5..27c9a11f0df9 100644
> --- a/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml
> +++ b/Documentation/devicetree/bindings/media/qcom,msm8996-camss.yaml
> @@ -85,6 +85,13 @@ properties:
> items:
> - description: VFE0 GDSC - Video Front End, Global Distributed Switch Controller.
> - description: VFE1 GDSC - Video Front End, Global Distributed Switch Controller.
> + - description: CAMSS GDSC - Camera Subsystem, Global Distributed Switch Controller.
> +
> + power-domain-names:
> + items:
> + - const: vfe0
> + - const: vfe1
> + - const: camss
>
> ports:
> $ref: /schemas/graph.yaml#/properties/ports
> @@ -209,6 +216,7 @@ required:
> - interrupts
> - iommus
> - power-domains
> + - power-domain-names
Why is this now required?
Thanks,
Conor.
> - reg
> - reg-names
> - vdda-supply
> @@ -326,7 +334,10 @@ examples:
> <&vfe_smmu 3>;
>
> power-domains = <&mmcc VFE0_GDSC>,
> - <&mmcc VFE1_GDSC>;
> + <&mmcc VFE1_GDSC>,
> + <&mmcc CAMSS_GDSC>;
> +
> + power-domain-names = "vfe0", "vfe1", "camss";
>
> reg = <0x00a34000 0x1000>,
> <0x00a00030 0x4>,
> --
> 2.40.1
>
On 26/05/2023 20:46, Conor Dooley wrote:
>> + - power-domain-names
> Why is this now required?
>
> Thanks,
> Conor.
>
Its an accurate description of the power/clock tree to have the top
power domain be switched on prior to the clocks that depend on it.
I think Yassine, you could probably include the majority of your
cover-letter text in this commit to explain this change a bit better.
---
bod
On Fri, May 26, 2023 at 09:05:47PM +0100, Bryan O'Donoghue wrote:
> On 26/05/2023 20:46, Conor Dooley wrote:
> > > + - power-domain-names
> > Why is this now required?
> >
> > Thanks,
> > Conor.
> >
>
> Its an accurate description of the power/clock tree to have the top power
> domain be switched on prior to the clocks that depend on it.
But what does that have to do with the *names* now being required?
> I think Yassine, you could probably include the majority of your
> cover-letter text in this commit to explain this change a bit better.
I think it would be good to have that regardless.
Cheers,
Conor.
> bod
^^^ I've been trying not to think about rugby since the weekend :(
On 26/05/2023 21:19, Conor Dooley wrote:
> On Fri, May 26, 2023 at 09:05:47PM +0100, Bryan O'Donoghue wrote:
>> On 26/05/2023 20:46, Conor Dooley wrote:
>>>> + - power-domain-names
>>> Why is this now required?
>>>
>>> Thanks,
>>> Conor.
>>>
>>
>> Its an accurate description of the power/clock tree to have the top power
>> domain be switched on prior to the clocks that depend on it.
>
> But what does that have to do with the *names* now being required?
oh the names
no toss that
>
>> I think Yassine, you could probably include the majority of your
>> cover-letter text in this commit to explain this change a bit better.
>
> I think it would be good to have that regardless.
>
> Cheers,
> Conor.
>
>> bod
> ^^^ I've been trying not to think about rugby since the weekend :(
Pockets O'Gara should have his paddy papers rescinded
On 26/05/2023 21:43, Konrad Dybcio wrote:
>
>
> On 26.05.2023 22:40, Bryan O'Donoghue wrote:
>> On 26/05/2023 21:36, Konrad Dybcio wrote:
>>>> oh the names
>>>>
>>>> no toss that
>>> this should be
>>>
>>> if:properties:compatible:blahblahmsm8996:then:required:power-domain-names
>>>
>>> Konrad
>>
>> Hmm, we don't depend on the names though.
> Check patch 3!
Hmm but we already count the number of power domains in
camss_configure_pd().
There's no logic in counting it twice using two different methods.
On 26.05.2023 22:40, Bryan O'Donoghue wrote:
> On 26/05/2023 21:36, Konrad Dybcio wrote:
>>> oh the names
>>>
>>> no toss that
>> this should be
>>
>> if:properties:compatible:blahblahmsm8996:then:required:power-domain-names
>>
>> Konrad
>
> Hmm, we don't depend on the names though.
Check patch 3!
Konrad
>
> ---
> bod
On 26.05.2023 22:21, Bryan O'Donoghue wrote:
> On 26/05/2023 21:19, Conor Dooley wrote:
>> On Fri, May 26, 2023 at 09:05:47PM +0100, Bryan O'Donoghue wrote:
>>> On 26/05/2023 20:46, Conor Dooley wrote:
>>>>> + - power-domain-names
>>>> Why is this now required?
>>>>
>>>> Thanks,
>>>> Conor.
>>>>
>>>
>>> Its an accurate description of the power/clock tree to have the top power
>>> domain be switched on prior to the clocks that depend on it.
>>
>> But what does that have to do with the *names* now being required?
>
> oh the names
>
> no toss that
this should be
if:properties:compatible:blahblahmsm8996:then:required:power-domain-names
Konrad
>
>>
>>> I think Yassine, you could probably include the majority of your
>>> cover-letter text in this commit to explain this change a bit better.
>>
>> I think it would be good to have that regardless.
>>
>> Cheers,
>> Conor.
>>
>>> bod
>> ^^^ I've been trying not to think about rugby since the weekend :(
>
> Pockets O'Gara should have his paddy papers rescinded
On 26/05/2023 21:36, Konrad Dybcio wrote:
>> oh the names
>>
>> no toss that
> this should be
>
> if:properties:compatible:blahblahmsm8996:then:required:power-domain-names
>
> Konrad
Hmm, we don't depend on the names though.
---
bod
On Fri, May 26 2023 at 10:36:32 PM +02:00:00, Konrad Dybcio
<[email protected]> wrote:
>
>
> On 26.05.2023 22:21, Bryan O'Donoghue wrote:
>> On 26/05/2023 21:19, Conor Dooley wrote:
>>> On Fri, May 26, 2023 at 09:05:47PM +0100, Bryan O'Donoghue wrote:
>>>> On 26/05/2023 20:46, Conor Dooley wrote:
>>>>>> + - power-domain-names
>>>>> Why is this now required?
>>>>>
>>>>> Thanks,
>>>>> Conor.
>>>>>
>>>>
>>>> Its an accurate description of the power/clock tree to have the
>>>> top power
>>>> domain be switched on prior to the clocks that depend on it.
>>>
>>> But what does that have to do with the *names* now being required?
>>
>> oh the names
>>
>> no toss that
> this should be
>
> if:properties:compatible:blahblahmsm8996:then:required:power-domain-names
The only compatible in this binding is qcom,msm8996-camss, so what
would this achieve?
>
> Konrad
>>
>>>
>>>> I think Yassine, you could probably include the majority of your
>>>> cover-letter text in this commit to explain this change a bit
>>>> better.
>>>
>>> I think it would be good to have that regardless.
>>>
>>> Cheers,
>>> Conor.
>>>
>>>> bod
>>> ^^^ I've been trying not to think about rugby since the weekend
>>> :(
>>
>> Pockets O'Gara should have his paddy papers rescinded
On 27.05.2023 08:05, Yassine Oudjana wrote:
>
> On Fri, May 26 2023 at 10:36:32 PM +02:00:00, Konrad Dybcio <[email protected]> wrote:
>>
>>
>> On 26.05.2023 22:21, Bryan O'Donoghue wrote:
>>> On 26/05/2023 21:19, Conor Dooley wrote:
>>>> On Fri, May 26, 2023 at 09:05:47PM +0100, Bryan O'Donoghue wrote:
>>>>> On 26/05/2023 20:46, Conor Dooley wrote:
>>>>>>> + - power-domain-names
>>>>>> Why is this now required?
>>>>>>
>>>>>> Thanks,
>>>>>> Conor.
>>>>>>
>>>>>
>>>>> Its an accurate description of the power/clock tree to have the top power
>>>>> domain be switched on prior to the clocks that depend on it.
>>>>
>>>> But what does that have to do with the *names* now being required?
>>>
>>> oh the names
>>>
>>> no toss that
>> this should be
>>
>> if:properties:compatible:blahblahmsm8996:then:required:power-domain-names
>
> The only compatible in this binding is qcom,msm8996-camss, so what would this achieve?
You're right - I didn't notice and assumed it was a camss-common one.
Konrad
>
>>
>> Konrad
>>>
>>>>
>>>>> I think Yassine, you could probably include the majority of your
>>>>> cover-letter text in this commit to explain this change a bit better.
>>>>
>>>> I think it would be good to have that regardless.
>>>>
>>>> Cheers,
>>>> Conor.
>>>>
>>>>> bod
>>>> ^^^ I've been trying not to think about rugby since the weekend :(
>>>
>>> Pockets O'Gara should have his paddy papers rescinded
>
>