[PATCHv3 0/2] ARM: dts: socfpga: fix booting with SD/MMC

[PATCHv3 0/2] ARM: dts: socfpga: fix booting with SD/MMC

[dinguyen at opensource.altera.com]

From: Dinh Nguyen <dinguyen@opensource.altera.com>

The SOCFPGA dev kit was hanging during bootup on the SD/MMC driver loading.
The first patch fixes the booting and the 2nd patch adds a regulator node
for the SD/MMC driver to use.

v3 diff(s): Thanks to Doug Anderson, the real issue was that the GPIO bank used
by cd-gpios did not have status="okay" in the node. This cause the GPIO to not
get probed correctly.

In fact, prior to patch 3cf890fc42b "mmc: dw_mmc: Pass back errors from
mmc_of_parse()", the error was being masked, so it was not being detected.

v2 diff(s): Patch 2/2 ARM: dts: socfpga: Add a 3.3V fixed regulator node
        - Move the regulator nodes to their respective board dts file and
          correctly rename the regulator to match the schematic

Dinh Nguyen (2):
  ARM: dts: socfpga: Fix SD card detect
  ARM: dts: socfpga: Add a 3.3V fixed regulator node

 arch/arm/boot/dts/socfpga_arria5.dtsi         |  2 +-
 arch/arm/boot/dts/socfpga_arria5_socdk.dts    | 12 ++++++++++++
 arch/arm/boot/dts/socfpga_cyclone5_socdk.dts  | 15 +++++++++++++++
 arch/arm/boot/dts/socfpga_cyclone5_sockit.dts | 12 ++++++++++++
 4 files changed, 40 insertions(+), 1 deletion(-)

-- 
2.0.3

[PATCHv3 1/2] ARM: dts: socfpga: Fix SD card detect

[dinguyen at opensource.altera.com]

From: Dinh Nguyen <dinguyen@opensource.altera.com>

Without this patch, the booting the SOCFPGA platform would hang at the
SDMMC driver loading. The issue, debugged by Doug Anderson, turned out
to be that the GPIO bank used by the SD card-detect was not set to
status="okay".

Suggested-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
---
v3: Correctly degugged the issue to be a gpio node not having status="okay"
---
 arch/arm/boot/dts/socfpga_cyclone5_socdk.dts | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
index d7296a5..03a3745 100644
--- a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
+++ b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
@@ -37,6 +37,12 @@
 		 */
 		ethernet0 = &gmac1;
 	};
+
+	soc {
+		gpio at ff709000 {
+			status = "okay";
+		};
+	};
 };
 
 &gmac1 {
-- 
2.0.3

[PATCHv3 2/2] ARM: dts: socfpga: Add a 3.3V fixed regulator node

[dinguyen at opensource.altera.com]

From: Dinh Nguyen <dinguyen@opensource.altera.com>

Without the 3.3V regulator node, the SDMMC driver will give these warnings:

dw_mmc ff704000.dwmmc0: No vmmc regulator found
dw_mmc ff704000.dwmmc0: No vqmmc regulator found

This patch adds the regulator node, and points the SD/MMC to the regulator.

Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
Reviewed-by: Doug Anderson <dianders@chromium.org>
---
v3: Rename nodes to have schematic-name_regulator and remove "boot-on" and
    "always-on"
v2: Move the regulator nodes to their respective board dts file and
    correctly rename them to match the schematic
---
 arch/arm/boot/dts/socfpga_arria5.dtsi         |  2 +-
 arch/arm/boot/dts/socfpga_arria5_socdk.dts    | 12 ++++++++++++
 arch/arm/boot/dts/socfpga_cyclone5_socdk.dts  |  9 +++++++++
 arch/arm/boot/dts/socfpga_cyclone5_sockit.dts | 12 ++++++++++++
 4 files changed, 34 insertions(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/socfpga_arria5.dtsi b/arch/arm/boot/dts/socfpga_arria5.dtsi
index 03e8268..1907cc6 100644
--- a/arch/arm/boot/dts/socfpga_arria5.dtsi
+++ b/arch/arm/boot/dts/socfpga_arria5.dtsi
@@ -29,7 +29,7 @@
 			};
 		};
 
-		dwmmc0 at ff704000 {
+		mmc0: dwmmc0 at ff704000 {
 			num-slots = <1>;
 			broken-cd;
 			bus-width = <4>;
diff --git a/arch/arm/boot/dts/socfpga_arria5_socdk.dts b/arch/arm/boot/dts/socfpga_arria5_socdk.dts
index 27d551c..ccaf417 100644
--- a/arch/arm/boot/dts/socfpga_arria5_socdk.dts
+++ b/arch/arm/boot/dts/socfpga_arria5_socdk.dts
@@ -37,6 +37,13 @@
 		*/
 		ethernet0 = &gmac1;
 	};
+
+	regulator_3_3v: 3-3-v-regulator {
+		compatible = "regulator-fixed";
+		regulator-name = "3.3V";
+		regulator-min-microvolt = <3300000>;
+		regulator-max-microvolt = <3300000>;
+	};
 };
 
 &gmac1 {
@@ -68,6 +75,11 @@
 	};
 };
 
+&mmc0 {
+	vmmc-supply = <&regulator_3_3v>;
+	vqmmc-supply = <&regulator_3_3v>;
+};
+
 &usb1 {
 	status = "okay";
 };
diff --git a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
index 03a3745..f962135 100644
--- a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
+++ b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
@@ -43,6 +43,13 @@
 			status = "okay";
 		};
 	};
+
+	regulator_3_3v: 3-3-v-regulator {
+		compatible = "regulator-fixed";
+		regulator-name = "3.3V";
+		regulator-min-microvolt = <3300000>;
+		regulator-max-microvolt = <3300000>;
+	};
 };
 
 &gmac1 {
@@ -76,6 +83,8 @@
 
 &mmc0 {
 	cd-gpios = <&gpio1 18 0>;
+	vmmc-supply = <&regulator_3_3v>;
+	vqmmc-supply = <&regulator_3_3v>;
 };
 
 &usb1 {
diff --git a/arch/arm/boot/dts/socfpga_cyclone5_sockit.dts b/arch/arm/boot/dts/socfpga_cyclone5_sockit.dts
index d26f155..16ea6f5 100644
--- a/arch/arm/boot/dts/socfpga_cyclone5_sockit.dts
+++ b/arch/arm/boot/dts/socfpga_cyclone5_sockit.dts
@@ -37,6 +37,13 @@
 		 */
 		ethernet0 = &gmac1;
 	};
+
+	regulator_3_3v: vcc3p3-regulator {
+		compatible = "regulator-fixed";
+		regulator-name = "VCC3P3";
+		regulator-min-microvolt = <3300000>;
+		regulator-max-microvolt = <3300000>;
+	};
 };
 
 &gmac1 {
@@ -53,6 +60,11 @@
 	rxc-skew-ps = <2000>;
 };
 
+&mmc0 {
+	vmmc-supply = <&regulator_3_3v>;
+	vqmmc-supply = <&regulator_3_3v>;
+};
+
 &usb1 {
 	status = "okay";
 };
-- 
2.0.3

[PATCHv3 1/2] ARM: dts: socfpga: Fix SD card detect

[Steffen Trumtrar]

Hi!

On Tue, Oct 21, 2014 at 11:15:10AM -0500, dinguyen at opensource.altera.com wrote:
> From: Dinh Nguyen <dinguyen@opensource.altera.com>
> 
> Without this patch, the booting the SOCFPGA platform would hang at the
> SDMMC driver loading. The issue, debugged by Doug Anderson, turned out
> to be that the GPIO bank used by the SD card-detect was not set to
> status="okay".
> 
> Suggested-by: Doug Anderson <dianders@chromium.org>
> Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
> ---
> v3: Correctly degugged the issue to be a gpio node not having status="okay"
> ---
>  arch/arm/boot/dts/socfpga_cyclone5_socdk.dts | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
> index d7296a5..03a3745 100644
> --- a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
> +++ b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
> @@ -37,6 +37,12 @@
>  		 */
>  		ethernet0 = &gmac1;
>  	};
> +
> +	soc {
> +		gpio at ff709000 {
> +			status = "okay";
> +		};
> +	};
>  };
>  

Looks weird. Don't you need to enable the GPIO bank instead:

	&gpio1 {
		status = "okay";
	}

???

Regards,
Steffen

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

[PATCHv3 1/2] ARM: dts: socfpga: Fix SD card detect

[Dinh Nguyen]

On 10/21/2014 12:10 PM, Steffen Trumtrar wrote:
> Hi!
> 
> On Tue, Oct 21, 2014 at 11:15:10AM -0500, dinguyen at opensource.altera.com wrote:
>> From: Dinh Nguyen <dinguyen@opensource.altera.com>
>>
>> Without this patch, the booting the SOCFPGA platform would hang at the
>> SDMMC driver loading. The issue, debugged by Doug Anderson, turned out
>> to be that the GPIO bank used by the SD card-detect was not set to
>> status="okay".
>>
>> Suggested-by: Doug Anderson <dianders@chromium.org>
>> Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
>> ---
>> v3: Correctly degugged the issue to be a gpio node not having status="okay"
>> ---
>>  arch/arm/boot/dts/socfpga_cyclone5_socdk.dts | 6 ++++++
>>  1 file changed, 6 insertions(+)
>>
>> diff --git a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>> index d7296a5..03a3745 100644
>> --- a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>> +++ b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>> @@ -37,6 +37,12 @@
>>  		 */
>>  		ethernet0 = &gmac1;
>>  	};
>> +
>> +	soc {
>> +		gpio at ff709000 {
>> +			status = "okay";
>> +		};
>> +	};
>>  };
>>  
> 
> Looks weird. Don't you need to enable the GPIO bank instead:
> 
> 	&gpio1 {
> 		status = "okay";
> 	}
> 

No, this will not work as gpio1 is already a node inside gpio at ff709000.
The reason for this was that the Designware GPIO IP can support 1, 2 or
3 gpio ports of varying widths. [1][2]

Dinh

[1]
http://article.gmane.org/gmane.linux.drivers.devicetree/72771/match=dw+apb+gpio
[2] http://article.gmane.org/gmane.linux.kernel/1671576/match=dw+apb+gpio

[PATCHv3 1/2] ARM: dts: socfpga: Fix SD card detect

[Doug Anderson]

Dinh,

(without the HTML mail this time, sorry)

On Tue, Oct 21, 2014 at 11:48 AM, Dinh Nguyen
<dinguyen@opensource.altera.com> wrote:
> On 10/21/2014 12:10 PM, Steffen Trumtrar wrote:
>> Hi!
>>
>> On Tue, Oct 21, 2014 at 11:15:10AM -0500, dinguyen at opensource.altera.com wrote:
>>> From: Dinh Nguyen <dinguyen@opensource.altera.com>
>>>
>>> Without this patch, the booting the SOCFPGA platform would hang at the
>>> SDMMC driver loading. The issue, debugged by Doug Anderson, turned out
>>> to be that the GPIO bank used by the SD card-detect was not set to
>>> status="okay".
>>>
>>> Suggested-by: Doug Anderson <dianders@chromium.org>
>>> Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
>>> ---
>>> v3: Correctly degugged the issue to be a gpio node not having status="okay"
>>> ---
>>>  arch/arm/boot/dts/socfpga_cyclone5_socdk.dts | 6 ++++++
>>>  1 file changed, 6 insertions(+)
>>>
>>> diff --git a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>>> index d7296a5..03a3745 100644
>>> --- a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>>> +++ b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>>> @@ -37,6 +37,12 @@
>>>               */
>>>              ethernet0 = &gmac1;
>>>      };
>>> +
>>> +    soc {
>>> +            gpio at ff709000 {
>>> +                    status = "okay";
>>> +            };
>>> +    };

I haven't dug into the details you provided about the GPIO
infrastructure of socfpga, but it does seem like it might be nice to
some easier way to reference things in
"arch/arm/boot/dts/socfpga.dtsi".

Looking at other examples they call the "gpio at ff708000" "gpio0" and
the sub node something like "porta".  Given that this is the only
reference to the GPIO, it seems like now would be the time to change
it.

-Doug

[PATCHv3 1/2] ARM: dts: socfpga: Fix SD card detect

[Dinh Nguyen]

On 10/22/14, 10:45 AM, Doug Anderson wrote:
> Dinh,
> 
> (without the HTML mail this time, sorry)
> 
> On Tue, Oct 21, 2014 at 11:48 AM, Dinh Nguyen
> <dinguyen@opensource.altera.com> wrote:
>> On 10/21/2014 12:10 PM, Steffen Trumtrar wrote:
>>> Hi!
>>>
>>> On Tue, Oct 21, 2014 at 11:15:10AM -0500, dinguyen at opensource.altera.com wrote:
>>>> From: Dinh Nguyen <dinguyen@opensource.altera.com>
>>>>
>>>> Without this patch, the booting the SOCFPGA platform would hang at the
>>>> SDMMC driver loading. The issue, debugged by Doug Anderson, turned out
>>>> to be that the GPIO bank used by the SD card-detect was not set to
>>>> status="okay".
>>>>
>>>> Suggested-by: Doug Anderson <dianders@chromium.org>
>>>> Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
>>>> ---
>>>> v3: Correctly degugged the issue to be a gpio node not having status="okay"
>>>> ---
>>>>  arch/arm/boot/dts/socfpga_cyclone5_socdk.dts | 6 ++++++
>>>>  1 file changed, 6 insertions(+)
>>>>
>>>> diff --git a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>>>> index d7296a5..03a3745 100644
>>>> --- a/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>>>> +++ b/arch/arm/boot/dts/socfpga_cyclone5_socdk.dts
>>>> @@ -37,6 +37,12 @@
>>>>               */
>>>>              ethernet0 = &gmac1;
>>>>      };
>>>> +
>>>> +    soc {
>>>> +            gpio at ff709000 {
>>>> +                    status = "okay";
>>>> +            };
>>>> +    };
> 
> I haven't dug into the details you provided about the GPIO
> infrastructure of socfpga, but it does seem like it might be nice to
> some easier way to reference things in
> "arch/arm/boot/dts/socfpga.dtsi".
> 
> Looking at other examples they call the "gpio at ff708000" "gpio0" and
> the sub node something like "porta".  Given that this is the only
> reference to the GPIO, it seems like now would be the time to change
> it.

Yeah...I like your idea. Let met get a new patch for that...

Thanks,
Dinh