Skip to contentSkip to conversation

dt-bindings: pwm: imx: document i.MX compatibles

PATCH In review Krzysztof Kozlowski 4 participants 6 messages Sep 25, 2020 – Oct 12, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 dt-bindings: pwm: imx: document i.MX compatibles

Document all ARMv5, ARMv6 and ARMv7 i.MX compatibles to fix dtbs_check warnings like:

arch/arm/boot/dts/imx6dl-colibri-eval-v3.dt.yaml: pwm@2080000: compatible:0:

    'fsl,imx6q-pwm' is not one of ['fsl,imx8mm-pwm', 'fsl,imx8mn-pwm', 'fsl,imx8mp-pwm', 'fsl,imx8mq-pwm']
---
 Documentation/devicetree/bindings/pwm/imx-pwm.yaml | 11 +++++++++++
 1 file changed, 11 insertions(+)
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +11 -0
index 473863eb67e5..379d693889f6 100644
@@ -25,6 +25,17 @@ properties:
           - fsl,imx27-pwm
       - items:
           - enum:
+              - fsl,imx25-pwm
+              - fsl,imx31-pwm
+              - fsl,imx50-pwm
+              - fsl,imx51-pwm
+              - fsl,imx53-pwm
+              - fsl,imx6q-pwm
+              - fsl,imx6sl-pwm
+              - fsl,imx6sll-pwm
+              - fsl,imx6sx-pwm
+              - fsl,imx6ul-pwm
+              - fsl,imx7d-pwm
               - fsl,imx8mm-pwm
               - fsl,imx8mn-pwm
               - fsl,imx8mp-pwm
Signature · 1 line
2.17.1
#1 lore raw
Quoted 16 lines · Krzysztof Kozlowski

Document all ARMv5, ARMv6 and ARMv7 i.MX compatibles to fix dtbs_check warnings like:

arch/arm/boot/dts/imx6dl-colibri-eval-v3.dt.yaml: pwm@2080000: compatible:0:

    'fsl,imx6q-pwm' is not one of ['fsl,imx8mm-pwm', 'fsl,imx8mn-pwm', 'fsl,imx8mp-pwm', 'fsl,imx8mq-pwm']
---
 Documentation/devicetree/bindings/pwm/imx-pwm.yaml | 11 +++++++++++
 1 file changed, 11 insertions(+)
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +0 -0
index 473863eb67e5..379d693889f6 100644

Documentation/devicetree/bindings/pwm/imx-pwm.yaml properties: · line 28

Quoted 4 lines
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +1 -0
@@ -25,6 +25,17 @@ properties:
           - fsl,imx27-pwm
       - items:
           - enum:
+              - fsl,imx25-pwm

The driver actually used fsl,imx27-pwm to bind ...

Also since v5.1-rc1~38^2~17 the driver is split into pwm-imx27 and pwm-imx1. So maybe this file should be renamed to fsl,imx27-pwm.yaml? (And we need a volunteer to write fsl,imx1-pwm.yaml.)

Documentation/devicetree/bindings/pwm/imx-pwm.yaml properties: · line 41

Quoted 13 lines
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +10 -0
+              - fsl,imx31-pwm
+              - fsl,imx50-pwm
+              - fsl,imx51-pwm
+              - fsl,imx53-pwm
+              - fsl,imx6q-pwm
+              - fsl,imx6sl-pwm
+              - fsl,imx6sll-pwm
+              - fsl,imx6sx-pwm
+              - fsl,imx6ul-pwm
+              - fsl,imx7d-pwm
               - fsl,imx8mm-pwm
               - fsl,imx8mn-pwm
               - fsl,imx8mp-pwm

Best regards
Uwe

Signature · 2 lines
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |
#2 lore raw
Quoted 12 lines · Uwe Kleine-König
Quoted 11 lines · Krzysztof Kozlowski

Document all ARMv5, ARMv6 and ARMv7 i.MX compatibles to fix dtbs_check warnings like:

arch/arm/boot/dts/imx6dl-colibri-eval-v3.dt.yaml: pwm@2080000: compatible:0:

    'fsl,imx6q-pwm' is not one of ['fsl,imx8mm-pwm', 'fsl,imx8mn-pwm', 'fsl,imx8mp-pwm', 'fsl,imx8mq-pwm']
---
 Documentation/devicetree/bindings/pwm/imx-pwm.yaml | 11 +++++++++++
 1 file changed, 11 insertions(+)
Quoted 11 lines

Documentation/devicetree/bindings/pwm/imx-pwm.yaml properties: · line 28

Quoted 9 lines
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +1 -0
index 473863eb67e5..379d693889f6 100644
@@ -25,6 +25,17 @@ properties:
           - fsl,imx27-pwm
       - items:
           - enum:
+              - fsl,imx25-pwm

The driver actually used fsl,imx27-pwm to bind ...

Yes, most of i.MX drivers use only few compatibles but DTSes and bindings use multiple of them. I was convinced during various talks that the specific compatibles (so "fsl,imx6q-pwm, fsl,imx27-pwm") are preferred than generic ones (so only "fsl,imx27-pwm"). NXP took it to the another level creating compatibles for absolutely every flavor of their CPU. And they mainlined it in DTSes...

The PWM is this crazy examples where, as you say, only two compatibles are actually used for binding but DTSes uses more.

Quoted 4 lines

Also since v5.1-rc1~38^2~17 the driver is split into pwm-imx27 and pwm-imx1. So maybe this file should be renamed to fsl,imx27-pwm.yaml? (And we need a volunteer to write fsl,imx1-pwm.yaml.)

I think there is no need, because the binding describes the hardware and could be supported by multiple drivers. I actually recently merged few bindings (i.MX 8 clocks, NXP PCA953x/max GPIO expanders).

Best regards,
Krzysztof

#3 lore raw
Quoted 32 lines · Krzysztof Kozlowski
Quoted 12 lines · Uwe Kleine-König
Quoted 11 lines · Krzysztof Kozlowski

Document all ARMv5, ARMv6 and ARMv7 i.MX compatibles to fix dtbs_check warnings like:

arch/arm/boot/dts/imx6dl-colibri-eval-v3.dt.yaml: pwm@2080000: compatible:0:

    'fsl,imx6q-pwm' is not one of ['fsl,imx8mm-pwm', 'fsl,imx8mn-pwm', 'fsl,imx8mp-pwm', 'fsl,imx8mq-pwm']
---
 Documentation/devicetree/bindings/pwm/imx-pwm.yaml | 11 +++++++++++
 1 file changed, 11 insertions(+)
Quoted 11 lines

Documentation/devicetree/bindings/pwm/imx-pwm.yaml properties: · line 28

Quoted 9 lines
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +1 -0
index 473863eb67e5..379d693889f6 100644
@@ -25,6 +25,17 @@ properties:
           - fsl,imx27-pwm
       - items:
           - enum:
+              - fsl,imx25-pwm

The driver actually used fsl,imx27-pwm to bind ...

Yes, most of i.MX drivers use only few compatibles but DTSes and bindings use multiple of them. I was convinced during various talks that the specific compatibles (so "fsl,imx6q-pwm, fsl,imx27-pwm") are preferred than generic ones (so only "fsl,imx27-pwm"). NXP took it to the another level creating compatibles for absolutely every flavor of their CPU. And they mainlined it in DTSes...

Quoted 2 lines

The PWM is this crazy examples where, as you say, only two compatibles are actually used for binding but DTSes uses more.

Yeah, these new compatible strings all seem to be used in the kernel, so we might as well document them.

That said, I did want to apply this patch, but that fails. Am I missing some other patch that you have sent out that touches this file? Actually it looks like this is because you've based this patch on linux-next, or perhaps the devicetree tree, because that contains commit d058717bdff4 ("dt-bindings: pwm: imx-pwm: Add i.MX 8M compatibles") from you that adds a couple more compatible strings. Probably best for Rob to pick this up, then:

Rob, here's a patchwork link for you if you need one:

	https://patchwork.ozlabs.org/project/devicetree-bindings/patch/[email protected]/

Although, looking at the devicetree-bindings instance version of that patch, I see that it's got a failing check attached (which looks like it can be ignored) and it's marked "Changes Requested", but no comments saying so.

Not sure if you want anything done here?

Thierry

#4 lore raw
Quoted 52 lines · Thierry Reding
Quoted 32 lines · Krzysztof Kozlowski
Quoted 12 lines · Uwe Kleine-König
Quoted 11 lines · Krzysztof Kozlowski

Document all ARMv5, ARMv6 and ARMv7 i.MX compatibles to fix dtbs_check warnings like:

arch/arm/boot/dts/imx6dl-colibri-eval-v3.dt.yaml: pwm@2080000: compatible:0:

    'fsl,imx6q-pwm' is not one of ['fsl,imx8mm-pwm', 'fsl,imx8mn-pwm', 'fsl,imx8mp-pwm', 'fsl,imx8mq-pwm']
---
 Documentation/devicetree/bindings/pwm/imx-pwm.yaml | 11 +++++++++++
 1 file changed, 11 insertions(+)
Quoted 11 lines

Documentation/devicetree/bindings/pwm/imx-pwm.yaml properties: · line 28

Quoted 9 lines
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +1 -0
index 473863eb67e5..379d693889f6 100644
@@ -25,6 +25,17 @@ properties:
           - fsl,imx27-pwm
       - items:
           - enum:
+              - fsl,imx25-pwm

The driver actually used fsl,imx27-pwm to bind ...

Yes, most of i.MX drivers use only few compatibles but DTSes and bindings use multiple of them. I was convinced during various talks that the specific compatibles (so "fsl,imx6q-pwm, fsl,imx27-pwm") are preferred than generic ones (so only "fsl,imx27-pwm"). NXP took it to the another level creating compatibles for absolutely every flavor of their CPU. And they mainlined it in DTSes...

Quoted 2 lines

The PWM is this crazy examples where, as you say, only two compatibles are actually used for binding but DTSes uses more.

Yeah, these new compatible strings all seem to be used in the kernel, so we might as well document them.

That said, I did want to apply this patch, but that fails. Am I missing some other patch that you have sent out that touches this file? Actually it looks like this is because you've based this patch on linux-next, or perhaps the devicetree tree, because that contains commit d058717bdff4 ("dt-bindings: pwm: imx-pwm: Add i.MX 8M compatibles") from you that adds a couple more compatible strings. Probably best for Rob to pick this up, then:

Rob, here's a patchwork link for you if you need one:

Quoted 8 lines
        https://patchwork.ozlabs.org/project/devicetree-bindings/patch/[email protected]/

Although, looking at the devicetree-bindings instance version of that patch, I see that it's got a failing check attached (which looks like it can be ignored) and it's marked "Changes Requested", but no comments saying so.

Not sure if you want anything done here?

Thanks, I guess this will wait for the merge window to finish. It should then apply to your tree. I can resend in two weeks.

Best regards,
Krzysztof

maintainer
#5 lore raw
In reply to Thierry Reding
Quoted 49 lines · Thierry Reding
Quoted 32 lines · Krzysztof Kozlowski
Quoted 12 lines · Uwe Kleine-König
Quoted 11 lines · Krzysztof Kozlowski

Document all ARMv5, ARMv6 and ARMv7 i.MX compatibles to fix dtbs_check warnings like:

arch/arm/boot/dts/imx6dl-colibri-eval-v3.dt.yaml: pwm@2080000: compatible:0:

    'fsl,imx6q-pwm' is not one of ['fsl,imx8mm-pwm', 'fsl,imx8mn-pwm', 'fsl,imx8mp-pwm', 'fsl,imx8mq-pwm']
---
 Documentation/devicetree/bindings/pwm/imx-pwm.yaml | 11 +++++++++++
 1 file changed, 11 insertions(+)
Quoted 11 lines

Documentation/devicetree/bindings/pwm/imx-pwm.yaml properties: · line 28

Quoted 9 lines
Documentation/devicetree/bindings/pwm/imx-pwm.yaml +1 -0
index 473863eb67e5..379d693889f6 100644
@@ -25,6 +25,17 @@ properties:
           - fsl,imx27-pwm
       - items:
           - enum:
+              - fsl,imx25-pwm

The driver actually used fsl,imx27-pwm to bind ...

Yes, most of i.MX drivers use only few compatibles but DTSes and bindings use multiple of them. I was convinced during various talks that the specific compatibles (so "fsl,imx6q-pwm, fsl,imx27-pwm") are preferred than generic ones (so only "fsl,imx27-pwm"). NXP took it to the another level creating compatibles for absolutely every flavor of their CPU. And they mainlined it in DTSes...

Quoted 2 lines

The PWM is this crazy examples where, as you say, only two compatibles are actually used for binding but DTSes uses more.

Yeah, these new compatible strings all seem to be used in the kernel, so we might as well document them.

That said, I did want to apply this patch, but that fails. Am I missing some other patch that you have sent out that touches this file? Actually it looks like this is because you've based this patch on linux-next, or perhaps the devicetree tree, because that contains commit d058717bdff4 ("dt-bindings: pwm: imx-pwm: Add i.MX 8M compatibles") from you that adds a couple more compatible strings. Probably best for Rob to pick this up, then:

Quoted 8 lines

Rob, here's a patchwork link for you if you need one:

	https://patchwork.ozlabs.org/project/devicetree-bindings/patch/[email protected]/

Although, looking at the devicetree-bindings instance version of that patch, I see that it's got a failing check attached (which looks like it can be ignored) and it's marked "Changes Requested", but no comments saying so.

I was assuming a rename at least.

Quoted 1 line

Not sure if you want anything done here?

I guess not, so I've applied it.

Rob