Re: [PATCH v2 4/5] dt-bindings: phy-rockchip-inno-usb2: add otg mux irq quirk

[Frank Wang]

Hi Heiko,

On 2017/8/2 19:25, Heiko Stübner wrote:

Hi Frank,

Am Mittwoch, 2. August 2017, 15:53:09 CEST schrieb Frank Wang:

Add rockchip,otg_mux_irq_quirk property to support the otg-id/
otg-bvalid/linestate irqs that are multiplexed to one irq in
otg-port on some special Rockchip SoCs (e.g RV1108).

Signed-off-by: Frank Wang 
---
  Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt | 5 +
  1 file changed, 5 insertions(+)

diff --git
a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt index
ddf868a..d49572b 100644
--- a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
+++ b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
@@ -22,6 +22,8 @@ Optional properties:
 register files".
   - rockchip,companion_grf_quirk : when set driver will request
 "rockchip,usbgrf" phandle as one companion-grf.
+ - rockchip,otg_mux_irq_quirk: set if otg-id/otg-bvalid/linestate irqs
+are multiplexed to one irq in otg-port on some SoCs.

  Required nodes : a sub-node is required for each port the phy provides.
 The sub-node name is used to identify host or otg port,
@@ -36,6 +38,9 @@ Required properties (port (child) node):
* "otg-id" : for the otg id interrupt.
* "otg-bvalid" : for the otg vbus interrupt.
* "linestate" : for the host/otg linestate interrupt.
+   * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
+   linestate irqs to one irq. Should specify if
+   rockchip,otg_mux_irq_quirk property is set.

I'd think that should be auto-detectable in the driver without the quirk
property above?

I.e. check for the otg-mux interrupt first on probe and if can be found, you
know that this is the case, otherwise look for the regular interrupts.


Also I would write that down a bit differently, so that it doesn't confuse
people. Something similar to:

  - interrupt-names : a list which should
+Regular case:
* "otg-id" : for the otg id interrupt.
* "otg-bvalid" : for the otg vbus interrupt.
* "linestate" : for the host/otg linestate interrupt.
+Some socs use one interrupt with the obove muxed together, so for these
+   * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
+   linestate irqs to one irq.


Heiko



Thanks for your comments, I will fix it in next version.


BR.
Frank

Re: [PATCH v2 4/5] dt-bindings: phy-rockchip-inno-usb2: add otg mux irq quirk

[Frank Wang]

Hi Heiko,

On 2017/8/2 19:25, Heiko Stübner wrote:

Hi Frank,

Am Mittwoch, 2. August 2017, 15:53:09 CEST schrieb Frank Wang:

Add rockchip,otg_mux_irq_quirk property to support the otg-id/
otg-bvalid/linestate irqs that are multiplexed to one irq in
otg-port on some special Rockchip SoCs (e.g RV1108).

Signed-off-by: Frank Wang 
---
  Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt | 5 +
  1 file changed, 5 insertions(+)

diff --git
a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt index
ddf868a..d49572b 100644
--- a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
+++ b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
@@ -22,6 +22,8 @@ Optional properties:
 register files".
   - rockchip,companion_grf_quirk : when set driver will request
 "rockchip,usbgrf" phandle as one companion-grf.
+ - rockchip,otg_mux_irq_quirk: set if otg-id/otg-bvalid/linestate irqs
+are multiplexed to one irq in otg-port on some SoCs.

  Required nodes : a sub-node is required for each port the phy provides.
 The sub-node name is used to identify host or otg port,
@@ -36,6 +38,9 @@ Required properties (port (child) node):
* "otg-id" : for the otg id interrupt.
* "otg-bvalid" : for the otg vbus interrupt.
* "linestate" : for the host/otg linestate interrupt.
+   * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
+   linestate irqs to one irq. Should specify if
+   rockchip,otg_mux_irq_quirk property is set.

I'd think that should be auto-detectable in the driver without the quirk
property above?

I.e. check for the otg-mux interrupt first on probe and if can be found, you
know that this is the case, otherwise look for the regular interrupts.


Also I would write that down a bit differently, so that it doesn't confuse
people. Something similar to:

  - interrupt-names : a list which should
+Regular case:
* "otg-id" : for the otg id interrupt.
* "otg-bvalid" : for the otg vbus interrupt.
* "linestate" : for the host/otg linestate interrupt.
+Some socs use one interrupt with the obove muxed together, so for these
+   * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
+   linestate irqs to one irq.


Heiko



Thanks for your comments, I will fix it in next version.


BR.
Frank

Re: [PATCH v2 4/5] dt-bindings: phy-rockchip-inno-usb2: add otg mux irq quirk

[Heiko Stübner]

Hi Frank,

Am Mittwoch, 2. August 2017, 15:53:09 CEST schrieb Frank Wang:
> Add rockchip,otg_mux_irq_quirk property to support the otg-id/
> otg-bvalid/linestate irqs that are multiplexed to one irq in
> otg-port on some special Rockchip SoCs (e.g RV1108).
> 
> Signed-off-by: Frank Wang 
> ---
>  Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt | 5 +
>  1 file changed, 5 insertions(+)
> 
> diff --git
> a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
> b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt index
> ddf868a..d49572b 100644
> --- a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
> +++ b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
> @@ -22,6 +22,8 @@ Optional properties:
>register files".
>   - rockchip,companion_grf_quirk : when set driver will request
>"rockchip,usbgrf" phandle as one companion-grf.
> + - rockchip,otg_mux_irq_quirk: set if otg-id/otg-bvalid/linestate irqs
> +  are multiplexed to one irq in otg-port on some SoCs.
> 
>  Required nodes : a sub-node is required for each port the phy provides.
>The sub-node name is used to identify host or otg port,
> @@ -36,6 +38,9 @@ Required properties (port (child) node):
>   * "otg-id" : for the otg id interrupt.
>   * "otg-bvalid" : for the otg vbus interrupt.
>   * "linestate" : for the host/otg linestate interrupt.
> + * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
> + linestate irqs to one irq. Should specify if
> + rockchip,otg_mux_irq_quirk property is set.

I'd think that should be auto-detectable in the driver without the quirk 
property above?

I.e. check for the otg-mux interrupt first on probe and if can be found, you 
know that this is the case, otherwise look for the regular interrupts.


Also I would write that down a bit differently, so that it doesn't confuse 
people. Something similar to:

 - interrupt-names : a list which should
+Regular case:
* "otg-id" : for the otg id interrupt.
* "otg-bvalid" : for the otg vbus interrupt.
* "linestate" : for the host/otg linestate interrupt.
+Some socs use one interrupt with the obove muxed together, so for these
+   * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
+   linestate irqs to one irq.


Heiko

Re: [PATCH v2 4/5] dt-bindings: phy-rockchip-inno-usb2: add otg mux irq quirk

[Heiko Stübner]

Hi Frank,

Am Mittwoch, 2. August 2017, 15:53:09 CEST schrieb Frank Wang:
> Add rockchip,otg_mux_irq_quirk property to support the otg-id/
> otg-bvalid/linestate irqs that are multiplexed to one irq in
> otg-port on some special Rockchip SoCs (e.g RV1108).
> 
> Signed-off-by: Frank Wang 
> ---
>  Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt | 5 +
>  1 file changed, 5 insertions(+)
> 
> diff --git
> a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
> b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt index
> ddf868a..d49572b 100644
> --- a/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
> +++ b/Documentation/devicetree/bindings/phy/phy-rockchip-inno-usb2.txt
> @@ -22,6 +22,8 @@ Optional properties:
>register files".
>   - rockchip,companion_grf_quirk : when set driver will request
>"rockchip,usbgrf" phandle as one companion-grf.
> + - rockchip,otg_mux_irq_quirk: set if otg-id/otg-bvalid/linestate irqs
> +  are multiplexed to one irq in otg-port on some SoCs.
> 
>  Required nodes : a sub-node is required for each port the phy provides.
>The sub-node name is used to identify host or otg port,
> @@ -36,6 +38,9 @@ Required properties (port (child) node):
>   * "otg-id" : for the otg id interrupt.
>   * "otg-bvalid" : for the otg vbus interrupt.
>   * "linestate" : for the host/otg linestate interrupt.
> + * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
> + linestate irqs to one irq. Should specify if
> + rockchip,otg_mux_irq_quirk property is set.

I'd think that should be auto-detectable in the driver without the quirk 
property above?

I.e. check for the otg-mux interrupt first on probe and if can be found, you 
know that this is the case, otherwise look for the regular interrupts.


Also I would write that down a bit differently, so that it doesn't confuse 
people. Something similar to:

 - interrupt-names : a list which should
+Regular case:
* "otg-id" : for the otg id interrupt.
* "otg-bvalid" : for the otg vbus interrupt.
* "linestate" : for the host/otg linestate interrupt.
+Some socs use one interrupt with the obove muxed together, so for these
+   * "otg-mux" : otg-port interrupt, which multiplex otg-id/otg-bvalid/
+   linestate irqs to one irq.


Heiko