OMAPDSS: panel-sharp-ls037v7dw01: update to use gpiod

[ It's weird that I'm just now getting this warning from 2014... Oh well, looks legit. -dan ]

The patch 9522f9fe86f9: "OMAPDSS: panel-sharp-ls037v7dw01: update to use gpiod" from Apr 28, 2014, leads to the following static checker warning:

	drivers/gpu/drm/omapdrm/displays/panel-sharp-ls037v7dw01.c:213 sharp_ls_get_gpio()
	warn: 'gd' isn't an ERR_PTR

drivers/gpu/drm/omapdrm/displays/panel-sharp-ls037v7dw01.c
200 static int sharp_ls_get_gpio(struct device *dev, int gpio, unsigned long flags,
201 char *desc, struct gpio_desc **gpiod)

   202  {
   203          struct gpio_desc *gd;
   204          int r;
   205  
   206          *gpiod = NULL;
   207  
   208          r = devm_gpio_request_one(dev, gpio, flags, desc);
   209          if (r)
   210                  return r == -ENOENT ? 0 : r;
   211  
   212          gd = gpio_to_desc(gpio);
   213          if (IS_ERR(gd))
   214                  return PTR_ERR(gd) == -ENOENT ? 0 : PTR_ERR(gd);
                               ^^^^^^^^^^^^^^^^^^^^^^

gd can be an ERR_PTR if gpio_to_desc is defined out but it's never -ENOENT.

Hi,

Quoted 2 lines · Dan Carpenter

[ It's weird that I'm just now getting this warning from 2014... Oh well, looks legit. -dan ]

Sorry for the delay on this one, got distracted few times with other bugs to deal with. This seems like a valid warning yeah.

Tomi, do we really need two copies of the the same panels files in kernel?

For example:

$ find . -name panel-sharp-ls037v7dw01.c
./drivers/gpu/drm/omapdrm/displays/panel-sharp-ls037v7dw01.c
./drivers/video/fbdev/omap2/omapfb/displays/panel-sharp-ls037v7dw01.c

Quoted 24 lines

The patch 9522f9fe86f9: "OMAPDSS: panel-sharp-ls037v7dw01: update to use gpiod" from Apr 28, 2014, leads to the following static checker warning:

	drivers/gpu/drm/omapdrm/displays/panel-sharp-ls037v7dw01.c:213 sharp_ls_get_gpio()
	warn: 'gd' isn't an ERR_PTR

drivers/gpu/drm/omapdrm/displays/panel-sharp-ls037v7dw01.c
200 static int sharp_ls_get_gpio(struct device *dev, int gpio, unsigned long flags,
201 char *desc, struct gpio_desc **gpiod)

   202  {
   203          struct gpio_desc *gd;
   204          int r;
   205  
   206          *gpiod = NULL;
   207  
   208          r = devm_gpio_request_one(dev, gpio, flags, desc);
   209          if (r)
   210                  return r == -ENOENT ? 0 : r;
   211  
   212          gd = gpio_to_desc(gpio);
   213          if (IS_ERR(gd))
   214                  return PTR_ERR(gd) == -ENOENT ? 0 : PTR_ERR(gd);
                               ^^^^^^^^^^^^^^^^^^^^^^

Quoted 2 lines

gd can be an ERR_PTR if gpio_to_desc is defined out but it's never -ENOENT.

Seems like we can just remove the check for -ENOENT here.

Quoted 3 lines · Tony Lindgren

Tomi, do we really need two copies of the the same panels files in kernel?

Quoted 5 lines

For example:

$ find . -name panel-sharp-ls037v7dw01.c
./drivers/gpu/drm/omapdrm/displays/panel-sharp-ls037v7dw01.c
./drivers/video/fbdev/omap2/omapfb/displays/panel-sharp-ls037v7dw01.c

Strictly no, but I think this is the best way forward. Longer story:
http://comments.gmane.org/gmane.comp.video.dri.devel/143151

Quoted 1 line

Seems like we can just remove the check for -ENOENT here.

Yep. Writing a patch for this and the other panel-sharp-ls037v7dw01 error Dan reported is on my todo.