On 9/30/18 12:45 AM, Vagrant Cascadian wrote:
(CC:ing Anatolij)
From: Vasily Khoruzhick anarsoul@gmail.com
If there's no sleep or reset GPIOs, video_bridge_set_active() returns -ENOENT. Don't fail in this case, since these GPIOs are optional.
Are really *both* optional? If yes, you should apply the same check I suggested in the last patch there as well, to also cover reset. Something like:
int video_bridge_set_active(struct udevice *dev, bool active) { int ret = 0;
/* The sleep GPIO is optional. */ if (uc_priv->sleep.dev) { ret = dm_gpio_set_value(&uc_priv->sleep, !active); if (ret) return ret; } if (uc_priv->reset.dev && active) { ret = dm_gpio_set_value(&uc_priv->reset, true); ... }
return ret; }
This avoids any attempts of cleverly ignoring errors, also allows you to get rid of this patch at all. Plus it's a more suitable approach for this generic driver file.
Cheers, Andre.
Signed-off-by: Vasily Khoruzhick anarsoul@gmail.com Signed-off-by: Vagrant Cascadian vagrant@debian.org
drivers/video/bridge/anx6345.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/video/bridge/anx6345.c b/drivers/video/bridge/anx6345.c index 3e3f0e2ce6..28cf2a9c2d 100644 --- a/drivers/video/bridge/anx6345.c +++ b/drivers/video/bridge/anx6345.c @@ -274,7 +274,7 @@ static int anx6345_enable(struct udevice *dev)
/* Deassert reset and enable power */ ret = video_bridge_set_active(dev, true);
- if (ret)
if (ret && ret != -ENOENT) return ret;
/* Reset */