| Message ID | 20190722150302.29526-3-brgl@bgdev.pl (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | backlight: gpio: simplify the driver | expand |
pon., 22 lip 2019 o 18:06 Andy Shevchenko <andriy.shevchenko@linux.intel.com> napisał(a): > > On Mon, Jul 22, 2019 at 05:02:57PM +0200, Bartosz Golaszewski wrote: > > From: Bartosz Golaszewski <bgolaszewski@baylibre.com> > > > > Now that the last user of platform data (sh ecovec24) defines a proper > > GPIO lookup and sets the 'default-on' device property, we can drop the > > platform_data-specific GPIO handling and unify a big chunk of code. > > > > The only field used from the platform data is now the fbdev pointer. > > > -static int gpio_backlight_probe_dt(struct platform_device *pdev, > > - struct gpio_backlight *gbl) > > -{ > > - struct device *dev = &pdev->dev; > > - enum gpiod_flags flags; > > - int ret; > > - > > - gbl->def_value = device_property_read_bool(dev, "default-on"); > > - flags = gbl->def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW; > > - > > - gbl->gpiod = devm_gpiod_get(dev, NULL, flags); > > - if (IS_ERR(gbl->gpiod)) { > > - ret = PTR_ERR(gbl->gpiod); > > - > > - if (ret != -EPROBE_DEFER) { > > - dev_err(dev, > > - "Error: The gpios parameter is missing or invalid.\n"); > > - } > > - return ret; > > - } > > - > > - return 0; > > -} > > Why not leave this function (perhaps with different name)? > > -- > With Best Regards, > Andy Shevchenko > > Why would we do that if the entire probe() function is now less than 50 lines long? Also: it gets inlined by the compiler anyway. It doesn't make sense IMO. Bart
On Tue, Jul 23, 2019 at 08:28:00AM +0200, Bartosz Golaszewski wrote: > pon., 22 lip 2019 o 18:06 Andy Shevchenko > <andriy.shevchenko@linux.intel.com> napisał(a): > > > > On Mon, Jul 22, 2019 at 05:02:57PM +0200, Bartosz Golaszewski wrote: > > > From: Bartosz Golaszewski <bgolaszewski@baylibre.com> > > > > > > Now that the last user of platform data (sh ecovec24) defines a proper > > > GPIO lookup and sets the 'default-on' device property, we can drop the > > > platform_data-specific GPIO handling and unify a big chunk of code. > > > > > > The only field used from the platform data is now the fbdev pointer. > > > > > -static int gpio_backlight_probe_dt(struct platform_device *pdev, > > > - struct gpio_backlight *gbl) > > > -{ > > > - struct device *dev = &pdev->dev; > > > - enum gpiod_flags flags; > > > - int ret; > > > - > > > - gbl->def_value = device_property_read_bool(dev, "default-on"); > > > - flags = gbl->def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW; > > > - > > > - gbl->gpiod = devm_gpiod_get(dev, NULL, flags); > > > - if (IS_ERR(gbl->gpiod)) { > > > - ret = PTR_ERR(gbl->gpiod); > > > - > > > - if (ret != -EPROBE_DEFER) { > > > - dev_err(dev, > > > - "Error: The gpios parameter is missing or invalid.\n"); > > > - } > > > - return ret; > > > - } > > > - > > > - return 0; > > > -} > > > > Why not leave this function (perhaps with different name)? > > Why would we do that if the entire probe() function is now less than > 50 lines long? Also: it gets inlined by the compiler anyway. It > doesn't make sense IMO. I'm not against this, perhaps, dropping and moving can be split to two changes.
wt., 23 lip 2019 o 17:32 Andy Shevchenko <andriy.shevchenko@linux.intel.com> napisał(a): > > On Tue, Jul 23, 2019 at 08:28:00AM +0200, Bartosz Golaszewski wrote: > > pon., 22 lip 2019 o 18:06 Andy Shevchenko > > <andriy.shevchenko@linux.intel.com> napisał(a): > > > > > > On Mon, Jul 22, 2019 at 05:02:57PM +0200, Bartosz Golaszewski wrote: > > > > From: Bartosz Golaszewski <bgolaszewski@baylibre.com> > > > > > > > > Now that the last user of platform data (sh ecovec24) defines a proper > > > > GPIO lookup and sets the 'default-on' device property, we can drop the > > > > platform_data-specific GPIO handling and unify a big chunk of code. > > > > > > > > The only field used from the platform data is now the fbdev pointer. > > > > > > > -static int gpio_backlight_probe_dt(struct platform_device *pdev, > > > > - struct gpio_backlight *gbl) > > > > -{ > > > > - struct device *dev = &pdev->dev; > > > > - enum gpiod_flags flags; > > > > - int ret; > > > > - > > > > - gbl->def_value = device_property_read_bool(dev, "default-on"); > > > > - flags = gbl->def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW; > > > > - > > > > - gbl->gpiod = devm_gpiod_get(dev, NULL, flags); > > > > - if (IS_ERR(gbl->gpiod)) { > > > > - ret = PTR_ERR(gbl->gpiod); > > > > - > > > > - if (ret != -EPROBE_DEFER) { > > > > - dev_err(dev, > > > > - "Error: The gpios parameter is missing or invalid.\n"); > > > > - } > > > > - return ret; > > > > - } > > > > - > > > > - return 0; > > > > -} > > > > > > Why not leave this function (perhaps with different name)? > > > > Why would we do that if the entire probe() function is now less than > > 50 lines long? Also: it gets inlined by the compiler anyway. It > > doesn't make sense IMO. > > I'm not against this, perhaps, dropping and moving can be split to two changes. > This really is unnecessary - we can do it in a single patch alright. Bart
diff --git a/drivers/video/backlight/gpio_backlight.c b/drivers/video/backlight/gpio_backlight.c index e84f3087e29f..01262186fa1e 100644 --- a/drivers/video/backlight/gpio_backlight.c +++ b/drivers/video/backlight/gpio_backlight.c @@ -55,30 +55,6 @@ static const struct backlight_ops gpio_backlight_ops = { .check_fb = gpio_backlight_check_fb, }; -static int gpio_backlight_probe_dt(struct platform_device *pdev, - struct gpio_backlight *gbl) -{ - struct device *dev = &pdev->dev; - enum gpiod_flags flags; - int ret; - - gbl->def_value = device_property_read_bool(dev, "default-on"); - flags = gbl->def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW; - - gbl->gpiod = devm_gpiod_get(dev, NULL, flags); - if (IS_ERR(gbl->gpiod)) { - ret = PTR_ERR(gbl->gpiod); - - if (ret != -EPROBE_DEFER) { - dev_err(dev, - "Error: The gpios parameter is missing or invalid.\n"); - } - return ret; - } - - return 0; -} - static int gpio_backlight_probe(struct platform_device *pdev) { struct gpio_backlight_platform_data *pdata = @@ -86,6 +62,7 @@ static int gpio_backlight_probe(struct platform_device *pdev) struct backlight_properties props; struct backlight_device *bl; struct gpio_backlight *gbl; + enum gpiod_flags flags; int ret; gbl = devm_kzalloc(&pdev->dev, sizeof(*gbl), GFP_KERNEL); @@ -94,35 +71,20 @@ static int gpio_backlight_probe(struct platform_device *pdev) gbl->dev = &pdev->dev; - if (pdev->dev.fwnode) { - ret = gpio_backlight_probe_dt(pdev, gbl); - if (ret) - return ret; - } else if (pdata) { - /* - * Legacy platform data GPIO retrieveal. Do not expand - * the use of this code path, currently only used by one - * SH board. - */ - unsigned long flags = GPIOF_DIR_OUT; - + if (pdata) gbl->fbdev = pdata->fbdev; - gbl->def_value = pdata->def_value; - flags |= gbl->def_value ? GPIOF_INIT_HIGH : GPIOF_INIT_LOW; - - ret = devm_gpio_request_one(gbl->dev, pdata->gpio, flags, - pdata ? pdata->name : "backlight"); - if (ret < 0) { - dev_err(&pdev->dev, "unable to request GPIO\n"); - return ret; + + gbl->def_value = device_property_read_bool(&pdev->dev, "default-on"); + flags = gbl->def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW; + + gbl->gpiod = devm_gpiod_get(&pdev->dev, NULL, flags); + if (IS_ERR(gbl->gpiod)) { + ret = PTR_ERR(gbl->gpiod); + if (ret != -EPROBE_DEFER) { + dev_err(&pdev->dev, + "Error: The gpios parameter is missing or invalid.\n"); } - gbl->gpiod = gpio_to_desc(pdata->gpio); - if (!gbl->gpiod) - return -EINVAL; - } else { - dev_err(&pdev->dev, - "failed to find platform data or device tree node.\n"); - return -ENODEV; + return ret; } memset(&props, 0, sizeof(props));