Re: OMAP DSS: panel-dpi and enable gpios

[Tomi Valkeinen <tomi.valkeinen@ti.com>]

[-- Attachment #1: Type: text/plain, Size: 3458 bytes --]

On 16/05/14 00:02, Joachim Eastwood wrote:
> On 15 May 2014 15:18, Tomi Valkeinen <tomi.valkeinen@ti.com> wrote:
>> On 12/05/14 20:58, Joachim Eastwood wrote:
>>> Hello Tomi,
>>>
>>> There seems to be a mismatch between your panel-dpi code and DT docs.
>>>
>>> The docs state that enable-gpios is optinal but in panel-dpi.c you
>>> have the following code
>>> gpio = devm_gpiod_get(&pdev->dev, "enable");
>>> if (IS_ERR(gpio)) {
>>>     dev_err(&pdev->dev, "failed to parse enable gpio\n");
>>>     return PTR_ERR(gpio);
>>> } else {
>>>     gpiod_direction_output(gpio, 0);
>>>     ddata->enable_gpio = gpio;
>>> }
>>>
>>> Making probing fail on my DT since I don't use enable-gpios with
>>
>> Would this work? Only compile tested.
>>
>> diff --git a/drivers/video/fbdev/omap2/displays-new/panel-dpi.c b/drivers/video/fbdev/omap2/displays-new/panel-dpi.c
>> index dca6b10d1157..2ac38eaa4c8f 100644
>> --- a/drivers/video/fbdev/omap2/displays-new/panel-dpi.c
>> +++ b/drivers/video/fbdev/omap2/displays-new/panel-dpi.c
>> @@ -210,14 +210,19 @@ static int panel_dpi_probe_of(struct platform_device *pdev)
>>         struct gpio_desc *gpio;
>>
>>         gpio = devm_gpiod_get(&pdev->dev, "enable");
>> +
>>         if (IS_ERR(gpio)) {
>> -               dev_err(&pdev->dev, "failed to parse enable gpio\n");
>> -               return PTR_ERR(gpio);
>> +               r = PTR_ERR(gpio);
>> +               if (r == -EPROBE_DEFER || r != -ENOENT)
>> +                       return r;
>> +               else
>> +                       gpio = NULL;
>>         } else {
>>                 gpiod_direction_output(gpio, 0);
>> -               ddata->enable_gpio = gpio;
>>         }
>>
>> +       ddata->enable_gpio = gpio;
>> +
>>         ddata->backlight_gpio = -ENOENT;
>>
>>         r = of_get_display_timing(node, "panel-timing", &timing);
> 
> Seems to do the trick here.
> 
> Display is showing Tux's on boot up again :)

The check above was a bit too complex. Here's an updated patch.

From f48b44ca73e29b2328e7852d9beb06b161bb1cb4 Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Thu, 15 May 2014 16:19:44 +0300
Subject: [PATCH] OMAPDSS: panel-dpi: enable-gpio

The enable gpio should be optional, but the driver returns an error if
it doesn't get the gpio.

So change the driver to accept -ENOENT error.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
---
 drivers/video/fbdev/omap2/displays-new/panel-dpi.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/video/fbdev/omap2/displays-new/panel-dpi.c b/drivers/video/fbdev/omap2/displays-new/panel-dpi.c
index dca6b10d1157..3636b61dc9b4 100644
--- a/drivers/video/fbdev/omap2/displays-new/panel-dpi.c
+++ b/drivers/video/fbdev/omap2/displays-new/panel-dpi.c
@@ -210,14 +210,18 @@ static int panel_dpi_probe_of(struct platform_device *pdev)
 	struct gpio_desc *gpio;
 
 	gpio = devm_gpiod_get(&pdev->dev, "enable");
+
 	if (IS_ERR(gpio)) {
-		dev_err(&pdev->dev, "failed to parse enable gpio\n");
-		return PTR_ERR(gpio);
+		if (PTR_ERR(gpio) != -ENOENT)
+			return PTR_ERR(gpio);
+		else
+			gpio = NULL;
 	} else {
 		gpiod_direction_output(gpio, 0);
-		ddata->enable_gpio = gpio;
 	}
 
+	ddata->enable_gpio = gpio;
+
 	ddata->backlight_gpio = -ENOENT;
 
 	r = of_get_display_timing(node, "panel-timing", &timing);
-- 
1.9.1





[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]