* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-09 16:46 Stephan Gerhold
0 siblings, 0 replies; 7+ messages in thread
From: Stephan Gerhold @ 2018-12-09 16:46 UTC (permalink / raw)
To: Hans de Goede; +Cc: Andy Shevchenko, Felipe Balbi, Greg Kroah-Hartman, USB
On Sun, Dec 09, 2018 at 04:28:07PM +0100, Hans de Goede wrote:
> Hi,
>
> On 09-12-18 16:07, Stephan Gerhold wrote:
> > On Sun, Dec 09, 2018 at 04:44:16PM +0200, Andy Shevchenko wrote:
> > > On Thu, Dec 6, 2018 at 8:49 PM Stephan Gerhold <stephan@gerhold.net> wrote:
> > > >
> > > > Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
> > > > the phy GPIOs") changed the code to claim the PHY GPIOs permanently
> > > > for Intel Baytrail devices.
> > > >
> > > > This causes issues when the actual PHY driver attempts to claim the
> > > > same GPIO descriptors. For example, tusb1210 now fails to probe with:
> > > >
> > > > tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
> > > >
> > > > dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
> > > > usually the PHY driver will then hold the GPIOs to turn off the
> > > > PHY when requested (e.g. during suspend).
> > > >
> > > > To fix the problem, this reverts the commit to restore the old
> > > > behavior to put the GPIOs immediately after usage.
> > > >
> > > > Link: https://www.spinics.net/lists/linux-usb/msg174681.html
> > > > Cc: stable@vger.kernel.org
> > > > Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
> > > > ---
> > > > drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
> > > > 1 file changed, 4 insertions(+), 4 deletions(-)
> > > >
> > > > diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
> > > > index 842795856bf4..fdc6e4e403e8 100644
> > > > --- a/drivers/usb/dwc3/dwc3-pci.c
> > > > +++ b/drivers/usb/dwc3/dwc3-pci.c
> > > > @@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
> > > > * put the gpio descriptors again here because the phy driver
> > > > * might want to grab them, too.
> > > > */
> > > > - gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
> > > > - GPIOD_OUT_LOW);
> > > > + gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
> > > > if (IS_ERR(gpio))
> > > > return PTR_ERR(gpio);
> > > >
> > > > gpiod_set_value_cansleep(gpio, 1);
> > > > + gpiod_put(gpio);
> > > >
> > > > - gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
> > > > - GPIOD_OUT_LOW);
> > > > + gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
> > > > if (IS_ERR(gpio))
> > > > return PTR_ERR(gpio);
> > > >
> > > > if (gpio) {
> > > > gpiod_set_value_cansleep(gpio, 1);
> > >
> > > > + gpiod_put(gpio);
> > > > usleep_range(10000, 11000);
> > >
> > > If something happens to GPIO line in between of these lines, the sleep
> > > might become obsolete. Shouldn't gpiod_put() be placed after?
> >
> > That's a good point, but I believe this would be more appropriately
> > fixed in a separate patch, since this is just an exact revert of
> > 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get the phy GPIOs")
> > (This is the way it was written when it was added to mainline 4 years
> > ago...)
> >
> > I can send a separate patch for this, or would you like to?
>
> Properly fixing this would require releasing *both* GPIOs *after* the
> ULPI vendor and product IDs have been read. Which will require adding
> some generic callback just for this to the generic non platform/pci
> specific dwc3 code. Which IMHO is not worth the trouble since in
> practice this is not really a problem.
Hmm, yeah, I was thinking about a specific situation where it would have
made a difference if both GPIOs were put after the delay (not after the
ULPI vendor/product IDs have been read). But now that I think about it
again, it does not make much sense... Never mind. (Maybe Andy has
another comment here..)
>
> Regards,
>
> Hans
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-09 15:28 Hans de Goede
0 siblings, 0 replies; 7+ messages in thread
From: Hans de Goede @ 2018-12-09 15:28 UTC (permalink / raw)
To: Stephan Gerhold, Andy Shevchenko; +Cc: Felipe Balbi, Greg Kroah-Hartman, USB
Hi,
On 09-12-18 16:07, Stephan Gerhold wrote:
> On Sun, Dec 09, 2018 at 04:44:16PM +0200, Andy Shevchenko wrote:
>> On Thu, Dec 6, 2018 at 8:49 PM Stephan Gerhold <stephan@gerhold.net> wrote:
>>>
>>> Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
>>> the phy GPIOs") changed the code to claim the PHY GPIOs permanently
>>> for Intel Baytrail devices.
>>>
>>> This causes issues when the actual PHY driver attempts to claim the
>>> same GPIO descriptors. For example, tusb1210 now fails to probe with:
>>>
>>> tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
>>>
>>> dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
>>> usually the PHY driver will then hold the GPIOs to turn off the
>>> PHY when requested (e.g. during suspend).
>>>
>>> To fix the problem, this reverts the commit to restore the old
>>> behavior to put the GPIOs immediately after usage.
>>>
>>> Link: https://www.spinics.net/lists/linux-usb/msg174681.html
>>> Cc: stable@vger.kernel.org
>>> Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
>>> ---
>>> drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
>>> 1 file changed, 4 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
>>> index 842795856bf4..fdc6e4e403e8 100644
>>> --- a/drivers/usb/dwc3/dwc3-pci.c
>>> +++ b/drivers/usb/dwc3/dwc3-pci.c
>>> @@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
>>> * put the gpio descriptors again here because the phy driver
>>> * might want to grab them, too.
>>> */
>>> - gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
>>> - GPIOD_OUT_LOW);
>>> + gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
>>> if (IS_ERR(gpio))
>>> return PTR_ERR(gpio);
>>>
>>> gpiod_set_value_cansleep(gpio, 1);
>>> + gpiod_put(gpio);
>>>
>>> - gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
>>> - GPIOD_OUT_LOW);
>>> + gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
>>> if (IS_ERR(gpio))
>>> return PTR_ERR(gpio);
>>>
>>> if (gpio) {
>>> gpiod_set_value_cansleep(gpio, 1);
>>
>>> + gpiod_put(gpio);
>>> usleep_range(10000, 11000);
>>
>> If something happens to GPIO line in between of these lines, the sleep
>> might become obsolete. Shouldn't gpiod_put() be placed after?
>
> That's a good point, but I believe this would be more appropriately
> fixed in a separate patch, since this is just an exact revert of
> 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get the phy GPIOs")
> (This is the way it was written when it was added to mainline 4 years
> ago...)
>
> I can send a separate patch for this, or would you like to?
Properly fixing this would require releasing *both* GPIOs *after* the
ULPI vendor and product IDs have been read. Which will require adding
some generic callback just for this to the generic non platform/pci
specific dwc3 code. Which IMHO is not worth the trouble since in
practice this is not really a problem.
Regards,
Hans
^ permalink raw reply [flat|nested] 7+ messages in thread
* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-09 15:07 Stephan Gerhold
0 siblings, 0 replies; 7+ messages in thread
From: Stephan Gerhold @ 2018-12-09 15:07 UTC (permalink / raw)
To: Andy Shevchenko; +Cc: Felipe Balbi, Hans de Goede, Greg Kroah-Hartman, USB
On Sun, Dec 09, 2018 at 04:44:16PM +0200, Andy Shevchenko wrote:
> On Thu, Dec 6, 2018 at 8:49 PM Stephan Gerhold <stephan@gerhold.net> wrote:
> >
> > Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
> > the phy GPIOs") changed the code to claim the PHY GPIOs permanently
> > for Intel Baytrail devices.
> >
> > This causes issues when the actual PHY driver attempts to claim the
> > same GPIO descriptors. For example, tusb1210 now fails to probe with:
> >
> > tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
> >
> > dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
> > usually the PHY driver will then hold the GPIOs to turn off the
> > PHY when requested (e.g. during suspend).
> >
> > To fix the problem, this reverts the commit to restore the old
> > behavior to put the GPIOs immediately after usage.
> >
> > Link: https://www.spinics.net/lists/linux-usb/msg174681.html
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
> > ---
> > drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
> > 1 file changed, 4 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
> > index 842795856bf4..fdc6e4e403e8 100644
> > --- a/drivers/usb/dwc3/dwc3-pci.c
> > +++ b/drivers/usb/dwc3/dwc3-pci.c
> > @@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
> > * put the gpio descriptors again here because the phy driver
> > * might want to grab them, too.
> > */
> > - gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
> > - GPIOD_OUT_LOW);
> > + gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
> > if (IS_ERR(gpio))
> > return PTR_ERR(gpio);
> >
> > gpiod_set_value_cansleep(gpio, 1);
> > + gpiod_put(gpio);
> >
> > - gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
> > - GPIOD_OUT_LOW);
> > + gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
> > if (IS_ERR(gpio))
> > return PTR_ERR(gpio);
> >
> > if (gpio) {
> > gpiod_set_value_cansleep(gpio, 1);
>
> > + gpiod_put(gpio);
> > usleep_range(10000, 11000);
>
> If something happens to GPIO line in between of these lines, the sleep
> might become obsolete. Shouldn't gpiod_put() be placed after?
That's a good point, but I believe this would be more appropriately
fixed in a separate patch, since this is just an exact revert of
211f658b7b40 ("usb: dwc3: pci: Use devm functions to get the phy GPIOs")
(This is the way it was written when it was added to mainline 4 years
ago...)
I can send a separate patch for this, or would you like to?
>
> > }
> > }
>
>
> --
> With Best Regards,
> Andy Shevchenko
^ permalink raw reply [flat|nested] 7+ messages in thread
* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-09 14:53 Hans de Goede
0 siblings, 0 replies; 7+ messages in thread
From: Hans de Goede @ 2018-12-09 14:53 UTC (permalink / raw)
To: Andy Shevchenko, stephan; +Cc: Felipe Balbi, Greg Kroah-Hartman, USB
Hi,
On 09-12-18 15:44, Andy Shevchenko wrote:
> On Thu, Dec 6, 2018 at 8:49 PM Stephan Gerhold <stephan@gerhold.net> wrote:
>>
>> Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
>> the phy GPIOs") changed the code to claim the PHY GPIOs permanently
>> for Intel Baytrail devices.
>>
>> This causes issues when the actual PHY driver attempts to claim the
>> same GPIO descriptors. For example, tusb1210 now fails to probe with:
>>
>> tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
>>
>> dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
>> usually the PHY driver will then hold the GPIOs to turn off the
>> PHY when requested (e.g. during suspend).
>>
>> To fix the problem, this reverts the commit to restore the old
>> behavior to put the GPIOs immediately after usage.
>>
>> Link: https://www.spinics.net/lists/linux-usb/msg174681.html
>> Cc: stable@vger.kernel.org
>> Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
>> ---
>> drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
>> 1 file changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
>> index 842795856bf4..fdc6e4e403e8 100644
>> --- a/drivers/usb/dwc3/dwc3-pci.c
>> +++ b/drivers/usb/dwc3/dwc3-pci.c
>> @@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
>> * put the gpio descriptors again here because the phy driver
>> * might want to grab them, too.
>> */
>> - gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
>> - GPIOD_OUT_LOW);
>> + gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
>> if (IS_ERR(gpio))
>> return PTR_ERR(gpio);
>>
>> gpiod_set_value_cansleep(gpio, 1);
>> + gpiod_put(gpio);
>>
>> - gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
>> - GPIOD_OUT_LOW);
>> + gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
>> if (IS_ERR(gpio))
>> return PTR_ERR(gpio);
>>
>> if (gpio) {
>> gpiod_set_value_cansleep(gpio, 1);
>
>> + gpiod_put(gpio);
>> usleep_range(10000, 11000);
>
> If something happens to GPIO line in between of these lines, the sleep
> might become obsolete. Shouldn't gpiod_put() be placed after?
The actual reading of the ULPI product and vendor IDs happens a whole
bunch of functions after the dwc3_pci_quirks() function runs, so the
assumption is that nothing touches the GPIO until the phy driver
loads.
Also the current version of this patch is a straight revert of my patch,
IOW the put allowing potential mucking with the GPIO before the
IDs are read is a pre-existing problem.
Regards,
Hans
^ permalink raw reply [flat|nested] 7+ messages in thread
* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-09 14:44 Andy Shevchenko
0 siblings, 0 replies; 7+ messages in thread
From: Andy Shevchenko @ 2018-12-09 14:44 UTC (permalink / raw)
To: stephan; +Cc: Felipe Balbi, Hans de Goede, Greg Kroah-Hartman, USB
On Thu, Dec 6, 2018 at 8:49 PM Stephan Gerhold <stephan@gerhold.net> wrote:
>
> Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
> the phy GPIOs") changed the code to claim the PHY GPIOs permanently
> for Intel Baytrail devices.
>
> This causes issues when the actual PHY driver attempts to claim the
> same GPIO descriptors. For example, tusb1210 now fails to probe with:
>
> tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
>
> dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
> usually the PHY driver will then hold the GPIOs to turn off the
> PHY when requested (e.g. during suspend).
>
> To fix the problem, this reverts the commit to restore the old
> behavior to put the GPIOs immediately after usage.
>
> Link: https://www.spinics.net/lists/linux-usb/msg174681.html
> Cc: stable@vger.kernel.org
> Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
> ---
> drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
> index 842795856bf4..fdc6e4e403e8 100644
> --- a/drivers/usb/dwc3/dwc3-pci.c
> +++ b/drivers/usb/dwc3/dwc3-pci.c
> @@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
> * put the gpio descriptors again here because the phy driver
> * might want to grab them, too.
> */
> - gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
> - GPIOD_OUT_LOW);
> + gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
> if (IS_ERR(gpio))
> return PTR_ERR(gpio);
>
> gpiod_set_value_cansleep(gpio, 1);
> + gpiod_put(gpio);
>
> - gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
> - GPIOD_OUT_LOW);
> + gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
> if (IS_ERR(gpio))
> return PTR_ERR(gpio);
>
> if (gpio) {
> gpiod_set_value_cansleep(gpio, 1);
> + gpiod_put(gpio);
> usleep_range(10000, 11000);
If something happens to GPIO line in between of these lines, the sleep
might become obsolete. Shouldn't gpiod_put() be placed after?
> }
> }
^ permalink raw reply [flat|nested] 7+ messages in thread
* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-09 11:16 Hans de Goede
0 siblings, 0 replies; 7+ messages in thread
From: Hans de Goede @ 2018-12-09 11:16 UTC (permalink / raw)
To: Stephan Gerhold, Felipe Balbi; +Cc: Greg Kroah-Hartman, linux-usb
Hi,
Thank you.
On 06-12-18 19:42, Stephan Gerhold wrote:
> Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
> the phy GPIOs") changed the code to claim the PHY GPIOs permanently
> for Intel Baytrail devices.
>
> This causes issues when the actual PHY driver attempts to claim the
> same GPIO descriptors. For example, tusb1210 now fails to probe with:
>
> tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
>
> dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
> usually the PHY driver will then hold the GPIOs to turn off the
> PHY when requested (e.g. during suspend).
>
> To fix the problem, this reverts the commit to restore the old
> behavior to put the GPIOs immediately after usage.
>
> Link: https://www.spinics.net/lists/linux-usb/msg174681.html
> Cc: stable@vger.kernel.org
> Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Right, the original change was done because I did not realize
that the phy driver would take over managing the GPIOs (and on
my test device it doesn't due to an unrelated issue).
So reverting this is the right thing to do:
Acked-by: Hans de Goede <hdegoede@redhat.com>
Regards,
Hans
> ---
> drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
> index 842795856bf4..fdc6e4e403e8 100644
> --- a/drivers/usb/dwc3/dwc3-pci.c
> +++ b/drivers/usb/dwc3/dwc3-pci.c
> @@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
> * put the gpio descriptors again here because the phy driver
> * might want to grab them, too.
> */
> - gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
> - GPIOD_OUT_LOW);
> + gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
> if (IS_ERR(gpio))
> return PTR_ERR(gpio);
>
> gpiod_set_value_cansleep(gpio, 1);
> + gpiod_put(gpio);
>
> - gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
> - GPIOD_OUT_LOW);
> + gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
> if (IS_ERR(gpio))
> return PTR_ERR(gpio);
>
> if (gpio) {
> gpiod_set_value_cansleep(gpio, 1);
> + gpiod_put(gpio);
> usleep_range(10000, 11000);
> }
> }
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs"
@ 2018-12-06 18:42 Stephan Gerhold
0 siblings, 0 replies; 7+ messages in thread
From: Stephan Gerhold @ 2018-12-06 18:42 UTC (permalink / raw)
To: Felipe Balbi
Cc: Hans de Goede, Greg Kroah-Hartman, linux-usb, Stephan Gerhold
Commit 211f658b7b40 ("usb: dwc3: pci: Use devm functions to get
the phy GPIOs") changed the code to claim the PHY GPIOs permanently
for Intel Baytrail devices.
This causes issues when the actual PHY driver attempts to claim the
same GPIO descriptors. For example, tusb1210 now fails to probe with:
tusb1210: probe of dwc3.0.auto.ulpi failed with error -16 (EBUSY)
dwc3-pci needs to turn on the PHY once before dwc3 is loaded, but
usually the PHY driver will then hold the GPIOs to turn off the
PHY when requested (e.g. during suspend).
To fix the problem, this reverts the commit to restore the old
behavior to put the GPIOs immediately after usage.
Link: https://www.spinics.net/lists/linux-usb/msg174681.html
Cc: stable@vger.kernel.org
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
---
drivers/usb/dwc3/dwc3-pci.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c
index 842795856bf4..fdc6e4e403e8 100644
--- a/drivers/usb/dwc3/dwc3-pci.c
+++ b/drivers/usb/dwc3/dwc3-pci.c
@@ -170,20 +170,20 @@ static int dwc3_pci_quirks(struct dwc3_pci *dwc)
* put the gpio descriptors again here because the phy driver
* might want to grab them, too.
*/
- gpio = devm_gpiod_get_optional(&pdev->dev, "cs",
- GPIOD_OUT_LOW);
+ gpio = gpiod_get_optional(&pdev->dev, "cs", GPIOD_OUT_LOW);
if (IS_ERR(gpio))
return PTR_ERR(gpio);
gpiod_set_value_cansleep(gpio, 1);
+ gpiod_put(gpio);
- gpio = devm_gpiod_get_optional(&pdev->dev, "reset",
- GPIOD_OUT_LOW);
+ gpio = gpiod_get_optional(&pdev->dev, "reset", GPIOD_OUT_LOW);
if (IS_ERR(gpio))
return PTR_ERR(gpio);
if (gpio) {
gpiod_set_value_cansleep(gpio, 1);
+ gpiod_put(gpio);
usleep_range(10000, 11000);
}
}
^ permalink raw reply related [flat|nested] 7+ messages in thread
end of thread, other threads:[~2018-12-09 16:46 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-12-09 16:46 Revert "usb: dwc3: pci: Use devm functions to get the phy GPIOs" Stephan Gerhold
-- strict thread matches above, loose matches on Subject: below --
2018-12-09 15:28 Hans de Goede
2018-12-09 15:07 Stephan Gerhold
2018-12-09 14:53 Hans de Goede
2018-12-09 14:44 Andy Shevchenko
2018-12-09 11:16 Hans de Goede
2018-12-06 18:42 Stephan Gerhold
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.