[PATCH 0/3] More small davinci device tree fixes

[PATCH 0/3] More small davinci device tree fixes

[David Lechner]

Here is another set of small davinci device tree fixes.

Tested working on LEGO MINDSTORMS EV3.

David Lechner (3):
  ARM: davinci: add spi0 to da850 device tree
  ARM: davinci: disable mdio and eth0 in da850.dtsi
  ARM: davinci: da850: There are 101 interrupts.

 arch/arm/boot/dts/da850-enbw-cmc.dts |  6 ++++++
 arch/arm/boot/dts/da850.dtsi         | 14 +++++++++++++-
 arch/arm/mach-davinci/da8xx-dt.c     |  1 +
 3 files changed, 20 insertions(+), 1 deletion(-)

-- 
1.9.1

[PATCH 1/3] ARM: DTS: da850: add node for spi0

[David Lechner]

Adds device definition for soc spi0 and also a aux data that is needed
for clock matching.

Signed-off-by: David Lechner <david@lechnology.com>
---
 arch/arm/boot/dts/da850.dtsi     | 10 ++++++++++
 arch/arm/mach-davinci/da8xx-dt.c |  1 +
 2 files changed, 11 insertions(+)

diff --git a/arch/arm/boot/dts/da850.dtsi b/arch/arm/boot/dts/da850.dtsi
index bbe7dd6..92b5f3c 100644
--- a/arch/arm/boot/dts/da850.dtsi
+++ b/arch/arm/boot/dts/da850.dtsi
@@ -295,6 +295,16 @@
 			reg = <0x308000 0x80>;
 			status = "disabled";
 		};
+		spi0: spi@41000 {
+			#address-cells = <1>;
+			#size-cells = <0>;
+			compatible = "ti,da830-spi";
+			reg = <0x41000 0x1000>;
+			num-cs = <6>;
+			ti,davinci-spi-intr-line = <1>;
+			interrupts = <20>;
+			status = "disabled";
+		};
 		spi1: spi@30e000 {
 			#address-cells = <1>;
 			#size-cells = <0>;
diff --git a/arch/arm/mach-davinci/da8xx-dt.c b/arch/arm/mach-davinci/da8xx-dt.c
index 64f3887..754f478 100644
--- a/arch/arm/mach-davinci/da8xx-dt.c
+++ b/arch/arm/mach-davinci/da8xx-dt.c
@@ -28,6 +28,7 @@ static struct of_dev_auxdata da850_auxdata_lookup[] __initdata = {
 	OF_DEV_AUXDATA("ti,da850-ecap", 0x01f06000, "ecap", NULL),
 	OF_DEV_AUXDATA("ti,da850-ecap", 0x01f07000, "ecap", NULL),
 	OF_DEV_AUXDATA("ti,da850-ecap", 0x01f08000, "ecap", NULL),
+	OF_DEV_AUXDATA("ti,da830-spi", 0x01c41000, "spi_davinci.0", NULL),
 	OF_DEV_AUXDATA("ti,da830-spi", 0x01f0e000, "spi_davinci.1", NULL),
 	OF_DEV_AUXDATA("ns16550a", 0x01c42000, "serial8250.0", NULL),
 	OF_DEV_AUXDATA("ns16550a", 0x01d0c000, "serial8250.1", NULL),
-- 
1.9.1

[PATCH 2/3] ARM: DTS: da850: disable mdio and eth0 by default

[David Lechner]

All other devices are disabled by default and not all boards will use
these devices, so these should be disabled too.

da850-evm.dtb already had status = "okay" for these devices.
da850-enbw-cmc.dts did not, so they were added.

Signed-off-by: David Lechner <david@lechnology.com>
---
 arch/arm/boot/dts/da850-enbw-cmc.dts | 6 ++++++
 arch/arm/boot/dts/da850.dtsi         | 2 ++
 2 files changed, 8 insertions(+)

diff --git a/arch/arm/boot/dts/da850-enbw-cmc.dts b/arch/arm/boot/dts/da850-enbw-cmc.dts
index 101d1a1..14dff3e 100644
--- a/arch/arm/boot/dts/da850-enbw-cmc.dts
+++ b/arch/arm/boot/dts/da850-enbw-cmc.dts
@@ -26,6 +26,12 @@
 		serial2: serial@10d000 {
 			status = "okay";
 		};
+		mdio: mdio@224000 {
+			status = "okay";
+		};
+		eth0: ethernet@220000 {
+			status = "okay";
+		};
 	};
 };
 
diff --git a/arch/arm/boot/dts/da850.dtsi b/arch/arm/boot/dts/da850.dtsi
index 92b5f3c..7f3fb98 100644
--- a/arch/arm/boot/dts/da850.dtsi
+++ b/arch/arm/boot/dts/da850.dtsi
@@ -322,6 +322,7 @@
 			#address-cells = <1>;
 			#size-cells = <0>;
 			reg = <0x224000 0x1000>;
+			status = "disabled";
 		};
 		eth0: ethernet@220000 {
 			compatible = "ti,davinci-dm6467-emac";
@@ -336,6 +337,7 @@
 					35
 					36
 					>;
+			status = "disabled";
 		};
 		usbphy: usbphy {
 			compatible = "ti,da830-usbphy";
-- 
1.9.1

[PATCH 3/3] ARM: DTS: da850: There are 101 interrupts.

[David Lechner]

Fix off by one error in da850 device tree.

Signed-off-by: David Lechner <david@lechnology.com>
---
 arch/arm/boot/dts/da850.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/da850.dtsi b/arch/arm/boot/dts/da850.dtsi
index 7f3fb98..2c60751 100644
--- a/arch/arm/boot/dts/da850.dtsi
+++ b/arch/arm/boot/dts/da850.dtsi
@@ -19,7 +19,7 @@
 			compatible = "ti,cp-intc";
 			interrupt-controller;
 			#interrupt-cells = <1>;
-			ti,intc-size = <100>;
+			ti,intc-size = <101>;
 			reg = <0xfffee000 0x2000>;
 		};
 	};
-- 
1.9.1

Re: [PATCH 0/3] More small davinci device tree fixes

[Kevin Hilman]

David Lechner <david@lechnology.com> writes:

> Here is another set of small davinci device tree fixes.
>
> Tested working on LEGO MINDSTORMS EV3.

Also boot tested on da850-evm.

Tested-by: Kevin Hilman <khilman@baylibre.com>

Kevin

Re: [PATCH 1/3] ARM: DTS: da850: add node for spi0

[Sekhar Nori]

On Thursday 14 April 2016 04:00 AM, David Lechner wrote:
> Adds device definition for soc spi0 and also a aux data that is needed
> for clock matching.
> 
> Signed-off-by: David Lechner <david@lechnology.com>
> ---
>  arch/arm/boot/dts/da850.dtsi     | 10 ++++++++++
>  arch/arm/mach-davinci/da8xx-dt.c |  1 +
>  2 files changed, 11 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/da850.dtsi b/arch/arm/boot/dts/da850.dtsi
> index bbe7dd6..92b5f3c 100644
> --- a/arch/arm/boot/dts/da850.dtsi
> +++ b/arch/arm/boot/dts/da850.dtsi
> @@ -295,6 +295,16 @@
>  			reg = <0x308000 0x80>;
>  			status = "disabled";
>  		};
> +		spi0: spi@41000 {
> +			#address-cells = <1>;
> +			#size-cells = <0>;
> +			compatible = "ti,da830-spi";
> +			reg = <0x41000 0x1000>;
> +			num-cs = <6>;

This made me notice that num-cs is populated wrongly for spi1. It
actually has 8 chip selects. This is fine though.

Also, it will be nice to add pinctrl entries for spi0 like it is done
for spi1. You will need those anyway for using the interface.

> +			ti,davinci-spi-intr-line = <1>;
> +			interrupts = <20>;
> +			status = "disabled";
> +		};
>  		spi1: spi@30e000 {
>  			#address-cells = <1>;
>  			#size-cells = <0>;
> diff --git a/arch/arm/mach-davinci/da8xx-dt.c b/arch/arm/mach-davinci/da8xx-dt.c
> index 64f3887..754f478 100644
> --- a/arch/arm/mach-davinci/da8xx-dt.c
> +++ b/arch/arm/mach-davinci/da8xx-dt.c
> @@ -28,6 +28,7 @@ static struct of_dev_auxdata da850_auxdata_lookup[] __initdata = {
>  	OF_DEV_AUXDATA("ti,da850-ecap", 0x01f06000, "ecap", NULL),
>  	OF_DEV_AUXDATA("ti,da850-ecap", 0x01f07000, "ecap", NULL),
>  	OF_DEV_AUXDATA("ti,da850-ecap", 0x01f08000, "ecap", NULL),
> +	OF_DEV_AUXDATA("ti,da830-spi", 0x01c41000, "spi_davinci.0", NULL),
>  	OF_DEV_AUXDATA("ti,da830-spi", 0x01f0e000, "spi_davinci.1", NULL),
>  	OF_DEV_AUXDATA("ns16550a", 0x01c42000, "serial8250.0", NULL),
>  	OF_DEV_AUXDATA("ns16550a", 0x01d0c000, "serial8250.1", NULL),

I prefer DT updates are patches of their own and not combined with code
updates like this. Can you please split this up?

Regards,
Sekhar

Re: [PATCH 2/3] ARM: DTS: da850: disable mdio and eth0 by default

[Sekhar Nori]

On Thursday 14 April 2016 04:00 AM, David Lechner wrote:
> All other devices are disabled by default and not all boards will use
> these devices, so these should be disabled too.
> 
> da850-evm.dtb already had status = "okay" for these devices.
> da850-enbw-cmc.dts did not, so they were added.
> 
> Signed-off-by: David Lechner <david@lechnology.com>

Looks good to me.

Thanks,
Sekhar

Re: [PATCH 3/3] ARM: DTS: da850: There are 101 interrupts.

[Sekhar Nori]

On Thursday 14 April 2016 04:00 AM, David Lechner wrote:
> Fix off by one error in da850 device tree.
> 
> Signed-off-by: David Lechner <david@lechnology.com>

Looks good to me.

Regards,
Sekhar

Re: [PATCH 1/3] ARM: DTS: da850: add node for spi0

[David Lechner]

On 04/15/2016 05:24 AM, Sekhar Nori wrote:

>
> This made me notice that num-cs is populated wrongly for spi1. It
> actually has 8 chip selects. This is fine though.

I might as well fix it since I have to make changes anyway. Don't 
remember how I came up with 6.

>
> Also, it will be nice to add pinctrl entries for spi0 like it is done
> for spi1. You will need those anyway for using the interface.

I omitted this on purpose. For my use case, I am using the SPI as 
write-only, so not using the SOMI pin, which is actually muxed as a GPIO 
for something else. So having a pinctl like spi1 is of no use to me. I 
figured if someone needs it, they can add it, otherwise it just is 
wasted space to me.
>
> I prefer DT updates are patches of their own and not combined with code
> updates like this. Can you please split this up?

Ack.

Re: [PATCH 1/3] ARM: DTS: da850: add node for spi0

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 395 bytes --]

On Fri, 15 Apr 2016 11:17:55 -0500, David Lechner said:

> I omitted this on purpose. For my use case, I am using the SPI as
> write-only,

So your SPI accesses are fire-and-forget, and nothing ever comes back?
Seems a very dangerous way to design the use case, with no feedback if
something suddenly goes pear-shaped...

Or do you have ways to verify the status via some method other than SPI?

[-- Attachment #2: Type: application/pgp-signature, Size: 848 bytes --]

Re: [PATCH 1/3] ARM: DTS: da850: add node for spi0

[David Lechner]

On 04/15/2016 03:16 PM, Valdis.Kletnieks@vt.edu wrote:
> On Fri, 15 Apr 2016 11:17:55 -0500, David Lechner said:
>
>> I omitted this on purpose. For my use case, I am using the SPI as
>> write-only,
>
> So your SPI accesses are fire-and-forget, and nothing ever comes back?

Yes.

> Seems a very dangerous way to design the use case, with no feedback if
> something suddenly goes pear-shaped...

You should tell Sitronix. This is how their display controllers work.

>
> Or do you have ways to verify the status via some method other than SPI?
>

Nope.



I'm working with LEGO MINDSTORMS EV3, a mass-produced robotics system. 
It is what it is. And I have a logic analyzer for when things go 
pear-shaped. ;-)

Re: [PATCH 1/3] ARM: DTS: da850: add node for spi0

[Sekhar Nori]

On Friday 15 April 2016 09:47 PM, David Lechner wrote:
> On 04/15/2016 05:24 AM, Sekhar Nori wrote:
> 
>>
>> This made me notice that num-cs is populated wrongly for spi1. It
>> actually has 8 chip selects. This is fine though.
> 
> I might as well fix it since I have to make changes anyway. Don't
> remember how I came up with 6.

In section 3.7.7 of datasheet, there are 6 possible chip selects listed
for SPI0 and 8 possible chipselects for SPI1.

If you are fixing SPI1, please make that a separate patch.

>> Also, it will be nice to add pinctrl entries for spi0 like it is done
>> for spi1. You will need those anyway for using the interface.
> 
> I omitted this on purpose. For my use case, I am using the SPI as
> write-only, so not using the SOMI pin, which is actually muxed as a GPIO
> for something else. So having a pinctl like spi1 is of no use to me. I
> figured if someone needs it, they can add it, otherwise it just is
> wasted space to me.

Alright, makes sense.

Regards,
Sekhar