[PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Frank Li]

In driver, compatible string is silvaco,i3c-master instead of
silvaco,i3c-master-v1.

Signed-off-by: Frank Li <Frank.Li@nxp.com>
---
 Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml b/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
index 133855f11b4f..19e3f75eb45c 100644
--- a/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
+++ b/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
@@ -14,7 +14,7 @@ allOf:
 
 properties:
   compatible:
-    const: silvaco,i3c-master-v1
+    const: silvaco,i3c-master
 
   reg:
     maxItems: 1
@@ -49,7 +49,7 @@ unevaluatedProperties: false
 examples:
   - |
     i3c-master@a0000000 {
-        compatible = "silvaco,i3c-master-v1";
+        compatible = "silvaco,i3c-master";
         clocks = <&zynqmp_clk 71>, <&fclk>, <&sclk>;
         clock-names = "pclk", "fast_clk", "slow_clk";
         interrupt-parent = <&gic>;
-- 
2.34.1

[PATCH 2/2] arm64: dts: freescale: imx93: add i3c1 and i3c2

[Frank Li]

Add I3C1 and I3C2.

Signed-off-by: Frank Li <Frank.Li@nxp.com>
---
 arch/arm64/boot/dts/freescale/imx93.dtsi | 26 ++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/arch/arm64/boot/dts/freescale/imx93.dtsi b/arch/arm64/boot/dts/freescale/imx93.dtsi
index 6f85a05ee7e1..4d9ed0b32853 100644
--- a/arch/arm64/boot/dts/freescale/imx93.dtsi
+++ b/arch/arm64/boot/dts/freescale/imx93.dtsi
@@ -242,6 +242,19 @@ tpm2: pwm@44320000 {
 				status = "disabled";
 			};
 
+			i3c1: i3c-master@44330000 {
+				compatible = "silvaco,i3c-master";
+				reg = <0x44330000 0x10000>;
+				interrupts = <GIC_SPI 12 IRQ_TYPE_LEVEL_HIGH>;
+				#address-cells = <3>;
+				#size-cells = <0>;
+				clocks = <&clk IMX93_CLK_BUS_AON>,
+					 <&clk IMX93_CLK_I3C1_GATE>,
+					 <&clk IMX93_CLK_I3C1_SLOW>;
+				clock-names = "pclk", "fast_clk", "slow_clk";
+				status = "disabled";
+			};
+
 			lpi2c1: i2c@44340000 {
 				compatible = "fsl,imx93-lpi2c", "fsl,imx7ulp-lpi2c";
 				reg = <0x44340000 0x10000>;
@@ -496,6 +509,19 @@ tpm6: pwm@42510000 {
 				status = "disabled";
 			};
 
+			i3c2: i3c-master@42520000 {
+				compatible = "silvaco,i3c-master";
+				reg = <0x42520000 0x10000>;
+				interrupts = <GIC_SPI 61 IRQ_TYPE_LEVEL_HIGH>;
+				#address-cells = <3>;
+				#size-cells = <0>;
+				clocks = <&clk IMX93_CLK_BUS_WAKEUP>,
+					 <&clk IMX93_CLK_I3C2_GATE>,
+					 <&clk IMX93_CLK_I3C2_SLOW>;
+				clock-names = "pclk", "fast_clk", "slow_clk";
+				status = "disabled";
+			};
+
 			lpi2c3: i2c@42530000 {
 				compatible = "fsl,imx93-lpi2c", "fsl,imx7ulp-lpi2c";
 				reg = <0x42530000 0x10000>;
-- 
2.34.1

Re: [PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 1404 bytes --]

On Mon, Oct 16, 2023 at 11:24:49AM -0400, Frank Li wrote:
> In driver, compatible string is silvaco,i3c-master instead of
> silvaco,i3c-master-v1.

And what makes the driver right & the binding wrong? AFAICT, this is an
IP sold by silvaco & the -v1 suffix was explicitly requested during
review of the binding.

Thanks,
Conor.

> 
> Signed-off-by: Frank Li <Frank.Li@nxp.com>
> ---
>  Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml b/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
> index 133855f11b4f..19e3f75eb45c 100644
> --- a/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
> +++ b/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
> @@ -14,7 +14,7 @@ allOf:
>  
>  properties:
>    compatible:
> -    const: silvaco,i3c-master-v1
> +    const: silvaco,i3c-master
>  
>    reg:
>      maxItems: 1
> @@ -49,7 +49,7 @@ unevaluatedProperties: false
>  examples:
>    - |
>      i3c-master@a0000000 {
> -        compatible = "silvaco,i3c-master-v1";
> +        compatible = "silvaco,i3c-master";
>          clocks = <&zynqmp_clk 71>, <&fclk>, <&sclk>;
>          clock-names = "pclk", "fast_clk", "slow_clk";
>          interrupt-parent = <&gic>;
> -- 
> 2.34.1
> 

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Frank Li]

On Mon, Oct 16, 2023 at 06:30:11PM +0100, Conor Dooley wrote:
> On Mon, Oct 16, 2023 at 11:24:49AM -0400, Frank Li wrote:
> > In driver, compatible string is silvaco,i3c-master instead of
> > silvaco,i3c-master-v1.
> 
> And what makes the driver right & the binding wrong? AFAICT, this is an
> IP sold by silvaco & the -v1 suffix was explicitly requested during
> review of the binding.

The driver existed for the long time. DTS files already use
silvaco,i3c-master.

There are two options, one change doc, the other change dts and drivers.
I think change doc is easiest ways.

If there are v2 in future, we can added silvaco,i3c-master-v2 in future.

If everyone prefer change drivers and dts, I can work on new patch. Just
break back compatiblity.

Frank

> 
> Thanks,
> Conor.
> 
> > 
> > Signed-off-by: Frank Li <Frank.Li@nxp.com>
> > ---
> >  Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml b/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
> > index 133855f11b4f..19e3f75eb45c 100644
> > --- a/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
> > +++ b/Documentation/devicetree/bindings/i3c/silvaco,i3c-master.yaml
> > @@ -14,7 +14,7 @@ allOf:
> >  
> >  properties:
> >    compatible:
> > -    const: silvaco,i3c-master-v1
> > +    const: silvaco,i3c-master
> >  
> >    reg:
> >      maxItems: 1
> > @@ -49,7 +49,7 @@ unevaluatedProperties: false
> >  examples:
> >    - |
> >      i3c-master@a0000000 {
> > -        compatible = "silvaco,i3c-master-v1";
> > +        compatible = "silvaco,i3c-master";
> >          clocks = <&zynqmp_clk 71>, <&fclk>, <&sclk>;
> >          clock-names = "pclk", "fast_clk", "slow_clk";
> >          interrupt-parent = <&gic>;
> > -- 
> > 2.34.1
> > 

Re: [PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Krzysztof Kozlowski]

On 16/10/2023 21:32, Frank Li wrote:
> On Mon, Oct 16, 2023 at 06:30:11PM +0100, Conor Dooley wrote:
>> On Mon, Oct 16, 2023 at 11:24:49AM -0400, Frank Li wrote:
>>> In driver, compatible string is silvaco,i3c-master instead of
>>> silvaco,i3c-master-v1.
>>
>> And what makes the driver right & the binding wrong? AFAICT, this is an
>> IP sold by silvaco & the -v1 suffix was explicitly requested during
>> review of the binding.
> 
> The driver existed for the long time. DTS files already use
> silvaco,i3c-master.
> 
> There are two options, one change doc, the other change dts and drivers.
> I think change doc is easiest ways.
> 
> If there are v2 in future, we can added silvaco,i3c-master-v2 in future.
> 
> If everyone prefer change drivers and dts, I can work on new patch. Just
> break back compatiblity.

No, because Rob explicitly pointed out too generic name.
https://lore.kernel.org/all/20210111200821.GA2978830@robh.at.kernel.org/

Is it some way to avoid implementing feedback or how does it work?

Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Frank Li]

On Mon, Oct 16, 2023 at 09:57:18PM +0200, Krzysztof Kozlowski wrote:
> On 16/10/2023 21:32, Frank Li wrote:
> > On Mon, Oct 16, 2023 at 06:30:11PM +0100, Conor Dooley wrote:
> >> On Mon, Oct 16, 2023 at 11:24:49AM -0400, Frank Li wrote:
> >>> In driver, compatible string is silvaco,i3c-master instead of
> >>> silvaco,i3c-master-v1.
> >>
> >> And what makes the driver right & the binding wrong? AFAICT, this is an
> >> IP sold by silvaco & the -v1 suffix was explicitly requested during
> >> review of the binding.
> > 
> > The driver existed for the long time. DTS files already use
> > silvaco,i3c-master.
> > 
> > There are two options, one change doc, the other change dts and drivers.
> > I think change doc is easiest ways.
> > 
> > If there are v2 in future, we can added silvaco,i3c-master-v2 in future.
> > 
> > If everyone prefer change drivers and dts, I can work on new patch. Just
> > break back compatiblity.
> 
> No, because Rob explicitly pointed out too generic name.
> https://lore.kernel.org/all/20210111200821.GA2978830@robh.at.kernel.org/
> 
> Is it some way to avoid implementing feedback or how does it work?

Downstream use silvaco,i3c-master to match driver. Anyways, miss match
happened.

option 1: fix doc
option 2: fix driver and dts

what do you perfer? 

> 
> Best regards,
> Krzysztof
> 

Re: [PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Krzysztof Kozlowski]

On 16/10/2023 22:10, Frank Li wrote:
> On Mon, Oct 16, 2023 at 09:57:18PM +0200, Krzysztof Kozlowski wrote:
>> On 16/10/2023 21:32, Frank Li wrote:
>>> On Mon, Oct 16, 2023 at 06:30:11PM +0100, Conor Dooley wrote:
>>>> On Mon, Oct 16, 2023 at 11:24:49AM -0400, Frank Li wrote:
>>>>> In driver, compatible string is silvaco,i3c-master instead of
>>>>> silvaco,i3c-master-v1.
>>>>
>>>> And what makes the driver right & the binding wrong? AFAICT, this is an
>>>> IP sold by silvaco & the -v1 suffix was explicitly requested during
>>>> review of the binding.
>>>
>>> The driver existed for the long time. DTS files already use
>>> silvaco,i3c-master.
>>>
>>> There are two options, one change doc, the other change dts and drivers.
>>> I think change doc is easiest ways.
>>>
>>> If there are v2 in future, we can added silvaco,i3c-master-v2 in future.
>>>
>>> If everyone prefer change drivers and dts, I can work on new patch. Just
>>> break back compatiblity.
>>
>> No, because Rob explicitly pointed out too generic name.
>> https://lore.kernel.org/all/20210111200821.GA2978830@robh.at.kernel.org/
>>
>> Is it some way to avoid implementing feedback or how does it work?
> 
> Downstream use silvaco,i3c-master to match driver. Anyways, miss match

Too bad for downstream.

> happened.
> 
> option 1: fix doc
> option 2: fix driver and dts
> 
> what do you perfer? 

Fix driver and dts, keep existing compatible as deprecated.

Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: i3c: Fix silvaco,i3c-master compatible string

[Miquel Raynal]

Hi Krzysztof,

krzysztof.kozlowski@linaro.org wrote on Mon, 16 Oct 2023 22:43:28 +0200:

> On 16/10/2023 22:10, Frank Li wrote:
> > On Mon, Oct 16, 2023 at 09:57:18PM +0200, Krzysztof Kozlowski wrote:  
> >> On 16/10/2023 21:32, Frank Li wrote:  
> >>> On Mon, Oct 16, 2023 at 06:30:11PM +0100, Conor Dooley wrote:  
> >>>> On Mon, Oct 16, 2023 at 11:24:49AM -0400, Frank Li wrote:  
> >>>>> In driver, compatible string is silvaco,i3c-master instead of
> >>>>> silvaco,i3c-master-v1.  
> >>>>
> >>>> And what makes the driver right & the binding wrong? AFAICT, this is an
> >>>> IP sold by silvaco & the -v1 suffix was explicitly requested during
> >>>> review of the binding.  
> >>>
> >>> The driver existed for the long time. DTS files already use
> >>> silvaco,i3c-master.
> >>>
> >>> There are two options, one change doc, the other change dts and drivers.
> >>> I think change doc is easiest ways.
> >>>
> >>> If there are v2 in future, we can added silvaco,i3c-master-v2 in future.
> >>>
> >>> If everyone prefer change drivers and dts, I can work on new patch. Just
> >>> break back compatiblity.  
> >>
> >> No, because Rob explicitly pointed out too generic name.
> >> https://lore.kernel.org/all/20210111200821.GA2978830@robh.at.kernel.org/
> >>
> >> Is it some way to avoid implementing feedback or how does it work?  

Not at all, it's my (faithful) fault, I did update the bindings upon
Rob request and apparently kept the previous string in the driver.

After thinking again I would drop silvaco,i3c-master entirely from the
driver, it was never intended to keep it.

Thanks,
Miquèl

Re: [PATCH 2/2] arm64: dts: freescale: imx93: add i3c1 and i3c2

[Rob Herring]

On Mon, Oct 16, 2023 at 11:24:50AM -0400, Frank Li wrote:
> Add I3C1 and I3C2.
> 
> Signed-off-by: Frank Li <Frank.Li@nxp.com>
> ---
>  arch/arm64/boot/dts/freescale/imx93.dtsi | 26 ++++++++++++++++++++++++
>  1 file changed, 26 insertions(+)
> 
> diff --git a/arch/arm64/boot/dts/freescale/imx93.dtsi b/arch/arm64/boot/dts/freescale/imx93.dtsi
> index 6f85a05ee7e1..4d9ed0b32853 100644
> --- a/arch/arm64/boot/dts/freescale/imx93.dtsi
> +++ b/arch/arm64/boot/dts/freescale/imx93.dtsi
> @@ -242,6 +242,19 @@ tpm2: pwm@44320000 {
>  				status = "disabled";
>  			};
>  
> +			i3c1: i3c-master@44330000 {
> +				compatible = "silvaco,i3c-master";

The real problem here is not whether we have "v1" or not, but you need 
an SoC specific compatible. Unless there's a public spec where we can 
know exactly how many resets, clocks, interrupts, etc.

Rob