The note in c2adda27d202f ("video: backlight: Add of_find_backlight helper in backlight.c") says that gpio-backlight uses brightness as power state. Other backlight drivers do not, so limit this workaround to gpio-backlight. This fixes the case where e.g. pwm-backlight can perfectly well be set to brightness 0 on boot in DT, which without this patch leads to the display brightness to be max instead of off. Fixes: c2adda27d202f ("video: backlight: Add of_find_backlight helper in backlight.c") Signed-off-by: Marek Vasut <marex@denx.de> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Meghana Madhyastha <meghana.madhyastha@gmail.com> Cc: Noralf Trønnes <noralf@tronnes.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Thierry Reding <treding@nvidia.com> --- drivers/video/backlight/backlight.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/video/backlight/backlight.c b/drivers/video/backlight/backlight.c index 537fe1b376ad7..dfb66171dec41 100644 --- a/drivers/video/backlight/backlight.c +++ b/drivers/video/backlight/backlight.c @@ -676,6 +676,7 @@ EXPORT_SYMBOL(of_find_backlight_by_node); static struct backlight_device *of_find_backlight(struct device *dev) { struct backlight_device *bd = NULL; + bool is_gpio_backlight = false; struct device_node *np; if (!dev) @@ -685,6 +686,8 @@ static struct backlight_device *of_find_backlight(struct device *dev) np = of_parse_phandle(dev->of_node, "backlight", 0); if (np) { bd = of_find_backlight_by_node(np); + is_gpio_backlight = + of_device_is_compatible(np, "gpio-backlight"); of_node_put(np); if (!bd) return ERR_PTR(-EPROBE_DEFER); @@ -692,7 +695,7 @@ static struct backlight_device *of_find_backlight(struct device *dev) * Note: gpio_backlight uses brightness as * power state during probe */ - if (!bd->props.brightness) + if (is_gpio_backlight && !bd->props.brightness) bd->props.brightness = bd->props.max_brightness; } } -- 2.30.2
On Thu, Jul 08, 2021 at 11:10:58AM +0200, Marek Vasut wrote: > The note in c2adda27d202f ("video: backlight: Add of_find_backlight helper > in backlight.c") says that gpio-backlight uses brightness as power state. > Other backlight drivers do not, so limit this workaround to gpio-backlight. > > This fixes the case where e.g. pwm-backlight can perfectly well be set to > brightness 0 on boot in DT, which without this patch leads to the display > brightness to be max instead of off. > > Fixes: c2adda27d202f ("video: backlight: Add of_find_backlight helper in backlight.c") > Signed-off-by: Marek Vasut <marex@denx.de> > Cc: Daniel Thompson <daniel.thompson@linaro.org> > Cc: Meghana Madhyastha <meghana.madhyastha@gmail.com> > Cc: Noralf Trønnes <noralf@tronnes.org> > Cc: Sean Paul <seanpaul@chromium.org> > Cc: Thierry Reding <treding@nvidia.com> I have to admit that this patch really does makes it clear just how nasty the hack in of_find_backlight() currently is. Moreover I think it is also be obsolete. gpio-backlight power mode handling was pretty broken when this code was introduced. It was fixed in 2019 by ec665b756e6f ("backlight: gpio backlight: Correct initial power state handling") by trying to match the behaviour of PWM backlight. The new code always sets the brightness to 1 so I think we can just remove the hack from of_find_backlight() since I think it is unreachable. Daniel. > --- > drivers/video/backlight/backlight.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > diff --git a/drivers/video/backlight/backlight.c b/drivers/video/backlight/backlight.c > index 537fe1b376ad7..dfb66171dec41 100644 > --- a/drivers/video/backlight/backlight.c > +++ b/drivers/video/backlight/backlight.c > @@ -676,6 +676,7 @@ EXPORT_SYMBOL(of_find_backlight_by_node); > static struct backlight_device *of_find_backlight(struct device *dev) > { > struct backlight_device *bd = NULL; > + bool is_gpio_backlight = false; > struct device_node *np; > > if (!dev) > @@ -685,6 +686,8 @@ static struct backlight_device *of_find_backlight(struct device *dev) > np = of_parse_phandle(dev->of_node, "backlight", 0); > if (np) { > bd = of_find_backlight_by_node(np); > + is_gpio_backlight = > + of_device_is_compatible(np, "gpio-backlight"); > of_node_put(np); > if (!bd) > return ERR_PTR(-EPROBE_DEFER); > @@ -692,7 +695,7 @@ static struct backlight_device *of_find_backlight(struct device *dev) > * Note: gpio_backlight uses brightness as > * power state during probe > */ > - if (!bd->props.brightness) > + if (is_gpio_backlight && !bd->props.brightness) > bd->props.brightness = bd->props.max_brightness; > } > } > -- > 2.30.2
On 7/9/21 1:03 PM, Daniel Thompson wrote:
> On Thu, Jul 08, 2021 at 11:10:58AM +0200, Marek Vasut wrote:
>> The note in c2adda27d202f ("video: backlight: Add of_find_backlight helper
>> in backlight.c") says that gpio-backlight uses brightness as power state.
>> Other backlight drivers do not, so limit this workaround to gpio-backlight.
>>
>> This fixes the case where e.g. pwm-backlight can perfectly well be set to
>> brightness 0 on boot in DT, which without this patch leads to the display
>> brightness to be max instead of off.
>>
>> Fixes: c2adda27d202f ("video: backlight: Add of_find_backlight helper in backlight.c")
>> Signed-off-by: Marek Vasut <marex@denx.de>
>> Cc: Daniel Thompson <daniel.thompson@linaro.org>
>> Cc: Meghana Madhyastha <meghana.madhyastha@gmail.com>
>> Cc: Noralf Trønnes <noralf@tronnes.org>
>> Cc: Sean Paul <seanpaul@chromium.org>
>> Cc: Thierry Reding <treding@nvidia.com>
>
> I have to admit that this patch really does makes it clear just how
> nasty the hack in of_find_backlight() currently is.
>
> Moreover I think it is also be obsolete. gpio-backlight power mode
> handling was pretty broken when this code was introduced. It was fixed
> in 2019 by ec665b756e6f ("backlight: gpio backlight: Correct initial
> power state handling") by trying to match the behaviour of PWM
> backlight. The new code always sets the brightness to 1 so I think we
> can just remove the hack from of_find_backlight() since I think it is
> unreachable.
I assume by "new code" you mean the fixed gpio-backlight driver ?
Dropping the whole code after the Note: is fine by me.
Den 09.07.2021 15.19, skrev Marek Vasut:
> On 7/9/21 1:03 PM, Daniel Thompson wrote:
>> On Thu, Jul 08, 2021 at 11:10:58AM +0200, Marek Vasut wrote:
>>> The note in c2adda27d202f ("video: backlight: Add of_find_backlight
>>> helper
>>> in backlight.c") says that gpio-backlight uses brightness as power
>>> state.
>>> Other backlight drivers do not, so limit this workaround to
>>> gpio-backlight.
>>>
>>> This fixes the case where e.g. pwm-backlight can perfectly well be
>>> set to
>>> brightness 0 on boot in DT, which without this patch leads to the
>>> display
>>> brightness to be max instead of off.
>>>
>>> Fixes: c2adda27d202f ("video: backlight: Add of_find_backlight helper
>>> in backlight.c")
>>> Signed-off-by: Marek Vasut <marex@denx.de>
>>> Cc: Daniel Thompson <daniel.thompson@linaro.org>
>>> Cc: Meghana Madhyastha <meghana.madhyastha@gmail.com>
>>> Cc: Noralf Trønnes <noralf@tronnes.org>
>>> Cc: Sean Paul <seanpaul@chromium.org>
>>> Cc: Thierry Reding <treding@nvidia.com>
>>
>> I have to admit that this patch really does makes it clear just how
>> nasty the hack in of_find_backlight() currently is.
>>
>> Moreover I think it is also be obsolete. gpio-backlight power mode
>> handling was pretty broken when this code was introduced. It was fixed
>> in 2019 by ec665b756e6f ("backlight: gpio backlight: Correct initial
>> power state handling") by trying to match the behaviour of PWM
>> backlight. The new code always sets the brightness to 1 so I think we
>> can just remove the hack from of_find_backlight() since I think it is
>> unreachable.
>
> I assume by "new code" you mean the fixed gpio-backlight driver ?
>
> Dropping the whole code after the Note: is fine by me.
I haven't actually tested it, but when I saw the patch that Daniel
points to, it looked like this work around could be removed.
I haven't got time to tests it, but I will ack a patch that removes the
hack.
Noralf.