[PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Gary Thomas]

Previous versions of the kernel returned the logical value of
GPIO signals (i.e. 0 or 1).  This patch restores this behavior.

Signed-off-by: Gary Thomas <gary@mlbassoc.com>
---
 drivers/gpio/gpiolib.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 50c4922..45aa695 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -376,7 +376,7 @@ static ssize_t gpio_value_show(struct device *dev,
 	if (!test_bit(FLAG_EXPORT, &desc->flags))
 		status = -EIO;
 	else
-		status = sprintf(buf, "%d\n", gpiod_get_value_cansleep(desc));
+		status = sprintf(buf, "%d\n", !!gpiod_get_value_cansleep(desc));
 
 	mutex_unlock(&sysfs_lock);
 	return status;
-- 
1.9.1

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Otavio Salvador]

On Wed, May 20, 2015 at 12:54 PM, Gary Thomas <gary@mlbassoc.com> wrote:
> Previous versions of the kernel returned the logical value of
> GPIO signals (i.e. 0 or 1).  This patch restores this behavior.
>
> Signed-off-by: Gary Thomas <gary@mlbassoc.com>

I am not sure this is a correct fix. I tried to find out the commit
which changed the behaviour without success.

Fabio, is it possible for you to check how mainline is behaving? and
more important does it need any fix?

-- 
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Gary Thomas]

On 2015-05-20 10:34, Otavio Salvador wrote:
> On Wed, May 20, 2015 at 12:54 PM, Gary Thomas <gary@mlbassoc.com> wrote:
>> Previous versions of the kernel returned the logical value of
>> GPIO signals (i.e. 0 or 1).  This patch restores this behavior.
>>
>> Signed-off-by: Gary Thomas <gary@mlbassoc.com>
>
> I am not sure this is a correct fix. I tried to find out the commit
> which changed the behaviour without success.
>
> Fabio, is it possible for you to check how mainline is behaving? and
> more important does it need any fix?
>

If by mainline you mean 4.1+, the code is very different and looks much
more like the 3.10 version than the FSL 3.14.

My change simply matches the behaviour (and indeed uses the same construct)
as the 3.10 kernel.

Also, as long as I can recall, reading a GPIO pin has returned 0/1,
except for a small period during 2.6 where it was broken and quickly
repaired.

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Otavio Salvador]

On Wed, May 20, 2015 at 1:40 PM, Gary Thomas <gary@mlbassoc.com> wrote:
> On 2015-05-20 10:34, Otavio Salvador wrote:
>>
>> On Wed, May 20, 2015 at 12:54 PM, Gary Thomas <gary@mlbassoc.com> wrote:
>>>
>>> Previous versions of the kernel returned the logical value of
>>> GPIO signals (i.e. 0 or 1).  This patch restores this behavior.
>>>
>>> Signed-off-by: Gary Thomas <gary@mlbassoc.com>
>>
>>
>> I am not sure this is a correct fix. I tried to find out the commit
>> which changed the behaviour without success.
>>
>> Fabio, is it possible for you to check how mainline is behaving? and
>> more important does it need any fix?
>>
>
> If by mainline you mean 4.1+, the code is very different and looks much
> more like the 3.10 version than the FSL 3.14.
>
> My change simply matches the behaviour (and indeed uses the same construct)
> as the 3.10 kernel.
>
> Also, as long as I can recall, reading a GPIO pin has returned 0/1,
> except for a small period during 2.6 where it was broken and quickly
> repaired.

If there is a patch which fixes this upstream I would prefer to
cherry-pick this one.

Do you know which change has restored the behaviour?

-- 
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Eric Bénard]

Le Wed, 20 May 2015 13:34:24 -0300,
Otavio Salvador <otavio@ossystems.com.br> a écrit :

> On Wed, May 20, 2015 at 12:54 PM, Gary Thomas <gary@mlbassoc.com> wrote:
> > Previous versions of the kernel returned the logical value of
> > GPIO signals (i.e. 0 or 1).  This patch restores this behavior.
> >
> > Signed-off-by: Gary Thomas <gary@mlbassoc.com>
> 
> I am not sure this is a correct fix. I tried to find out the commit
> which changed the behaviour without success.
> 
> Fabio, is it possible for you to check how mainline is behaving? and
> more important does it need any fix?
> 
fixed mainline by :
25b35da gpio: generic: clamp retured value to [0,1]
which is not in 3.4.28

Eric

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Gary Thomas]

On 2015-05-20 10:49, Eric Bénard wrote:
> Le Wed, 20 May 2015 13:34:24 -0300,
> Otavio Salvador <otavio@ossystems.com.br> a écrit :
>
>> On Wed, May 20, 2015 at 12:54 PM, Gary Thomas <gary@mlbassoc.com> wrote:
>>> Previous versions of the kernel returned the logical value of
>>> GPIO signals (i.e. 0 or 1).  This patch restores this behavior.
>>>
>>> Signed-off-by: Gary Thomas <gary@mlbassoc.com>
>>
>> I am not sure this is a correct fix. I tried to find out the commit
>> which changed the behaviour without success.
>>
>> Fabio, is it possible for you to check how mainline is behaving? and
>> more important does it need any fix?
>>
> fixed mainline by :
> 25b35da gpio: generic: clamp retured value to [0,1]
> which is not in 3.4.28

True enough, but this patch won't apply to 3.14.28 as the
code is very different.

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Fabio Estevam]

On Wed, May 20, 2015 at 1:51 PM, Gary Thomas <gary@mlbassoc.com> wrote:

> True enough, but this patch won't apply to 3.14.28 as the
> code is very different.

Yes, this commit appeared in 3.15.

I suggest we go with Gary's patch for 3.14.

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Eric Bénard]

Le Wed, 20 May 2015 10:51:55 -0600,
Gary Thomas <gary@mlbassoc.com> a écrit :

> On 2015-05-20 10:49, Eric Bénard wrote:
> > Le Wed, 20 May 2015 13:34:24 -0300,
> > Otavio Salvador <otavio@ossystems.com.br> a écrit :
> >
> >> On Wed, May 20, 2015 at 12:54 PM, Gary Thomas <gary@mlbassoc.com> wrote:
> >>> Previous versions of the kernel returned the logical value of
> >>> GPIO signals (i.e. 0 or 1).  This patch restores this behavior.
> >>>
> >>> Signed-off-by: Gary Thomas <gary@mlbassoc.com>
> >>
> >> I am not sure this is a correct fix. I tried to find out the commit
> >> which changed the behaviour without success.
> >>
> >> Fabio, is it possible for you to check how mainline is behaving? and
> >> more important does it need any fix?
> >>
> > fixed mainline by :
> > 25b35da gpio: generic: clamp retured value to [0,1]
> > which is not in 3.4.28
> 
> True enough, but this patch won't apply to 3.14.28 as the
> code is very different.
> 
I think it should apply easily, as gpio-generic.c is very similar
between 3.14.28 and mainline :

3.14.28
https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142

commit 25b35da : 
https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb

Eric

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Fabio Estevam]

On Wed, May 20, 2015 at 1:57 PM, Eric Bénard <eric@eukrea.com> wrote:

> I think it should apply easily, as gpio-generic.c is very similar
> between 3.14.28 and mainline :
>
> 3.14.28
> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142
>
> commit 25b35da :
> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb

Yes, agreed. It is a one-line patch, so we can easily use the mainline fix.

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Gary Thomas]

On 2015-05-20 11:02, Fabio Estevam wrote:
> On Wed, May 20, 2015 at 1:57 PM, Eric Bénard <eric@eukrea.com> wrote:
>
>> I think it should apply easily, as gpio-generic.c is very similar
>> between 3.14.28 and mainline :
>>
>> 3.14.28
>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142
>>
>> commit 25b35da :
>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb
>
> Yes, agreed. It is a one-line patch, so we can easily use the mainline fix.
>

I concur - my change was local to just the sysfs code, but
this change is more correct (and already in mainline)

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Otavio Salvador]

On Wed, May 20, 2015 at 2:02 PM, Fabio Estevam <festevam@gmail.com> wrote:
> On Wed, May 20, 2015 at 1:57 PM, Eric Bénard <eric@eukrea.com> wrote:
>
>> I think it should apply easily, as gpio-generic.c is very similar
>> between 3.14.28 and mainline :
>>
>> 3.14.28
>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142
>>
>> commit 25b35da :
>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb
>
> Yes, agreed. It is a one-line patch, so we can easily use the mainline fix.

Yes. It applies fine in 3.14.

-- 
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Otavio Salvador]

On Wed, May 20, 2015 at 2:14 PM, Gary Thomas <gary@mlbassoc.com> wrote:
> On 2015-05-20 11:02, Fabio Estevam wrote:
>>
>> On Wed, May 20, 2015 at 1:57 PM, Eric Bénard <eric@eukrea.com> wrote:
>>
>>> I think it should apply easily, as gpio-generic.c is very similar
>>> between 3.14.28 and mainline :
>>>
>>> 3.14.28
>>>
>>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142
>>>
>>> commit 25b35da :
>>>
>>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb
>>
>>
>> Yes, agreed. It is a one-line patch, so we can easily use the mainline
>> fix.
>>
>
> I concur - my change was local to just the sysfs code, but
> this change is more correct (and already in mainline)

Can you generate a proper patch for it?

-- 
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Gary Thomas]

On 2015-05-20 11:26, Otavio Salvador wrote:
> On Wed, May 20, 2015 at 2:14 PM, Gary Thomas <gary@mlbassoc.com> wrote:
>> On 2015-05-20 11:02, Fabio Estevam wrote:
>>>
>>> On Wed, May 20, 2015 at 1:57 PM, Eric Bénard <eric@eukrea.com> wrote:
>>>
>>>> I think it should apply easily, as gpio-generic.c is very similar
>>>> between 3.14.28 and mainline :
>>>>
>>>> 3.14.28
>>>>
>>>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142
>>>>
>>>> commit 25b35da :
>>>>
>>>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb
>>>
>>>
>>> Yes, agreed. It is a one-line patch, so we can easily use the mainline
>>> fix.
>>>
>>
>> I concur - my change was local to just the sysfs code, but
>> this change is more correct (and already in mainline)
>
> Can you generate a proper patch for it?
>

Why not just cherry-pick from 25b35da?  It then has all of the proper
signed-off-by, etc.

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Otavio Salvador]

On Wed, May 20, 2015 at 2:34 PM, Gary Thomas <gary@mlbassoc.com> wrote:
> On 2015-05-20 11:26, Otavio Salvador wrote:
>>
>> On Wed, May 20, 2015 at 2:14 PM, Gary Thomas <gary@mlbassoc.com> wrote:
>>>
>>> On 2015-05-20 11:02, Fabio Estevam wrote:
>>>>
>>>>
>>>> On Wed, May 20, 2015 at 1:57 PM, Eric Bénard <eric@eukrea.com> wrote:
>>>>
>>>>> I think it should apply easily, as gpio-generic.c is very similar
>>>>> between 3.14.28 and mainline :
>>>>>
>>>>> 3.14.28
>>>>>
>>>>>
>>>>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/gpio/gpio-generic.c?h=linux-3.14.y&id=v3.14.28#n142
>>>>>
>>>>> commit 25b35da :
>>>>>
>>>>>
>>>>> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/gpio/gpio-generic.c?id=25b35da7f4cce82271859f1b6eabd9f3bd41a2bb
>>>>
>>>>
>>>>
>>>> Yes, agreed. It is a one-line patch, so we can easily use the mainline
>>>> fix.
>>>>
>>>
>>> I concur - my change was local to just the sysfs code, but
>>> this change is more correct (and already in mainline)
>>
>>
>> Can you generate a proper patch for it?
>>
>
> Why not just cherry-pick from 25b35da?  It then has all of the proper
> signed-off-by, etc.

I cannot push at git.freescale.com so this is a problem. For
linux-fslc this is easy.

For now, I think a patch needs to be exported and applied on top of
the original Git tree. Fabio, can you ask if 25b35da can be applied at
Freescale's git?

-- 
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Fabio Estevam]

On Wed, May 20, 2015 at 2:36 PM, Otavio Salvador
<otavio@ossystems.com.br> wrote:

>> Why not just cherry-pick from 25b35da?  It then has all of the proper
>> signed-off-by, etc.
>
> I cannot push at git.freescale.com so this is a problem. For
> linux-fslc this is easy.
>
> For now, I think a patch needs to be exported and applied on top of
> the original Git tree. Fabio, can you ask if 25b35da can be applied at
> Freescale's git?

Lauren,

Could you ask the BSP team to apply 25b35da commit from mainline into 3.14?

Thanks

Re: [PATCH] kernel 3.14.28/gpiolib: Return GPIO logical value, not true value

[Otavio Salvador]

On Wed, May 20, 2015 at 2:49 PM, Fabio Estevam <festevam@gmail.com> wrote:
> On Wed, May 20, 2015 at 2:36 PM, Otavio Salvador
> <otavio@ossystems.com.br> wrote:
>
>>> Why not just cherry-pick from 25b35da?  It then has all of the proper
>>> signed-off-by, etc.
>>
>> I cannot push at git.freescale.com so this is a problem. For
>> linux-fslc this is easy.
>>
>> For now, I think a patch needs to be exported and applied on top of
>> the original Git tree. Fabio, can you ask if 25b35da can be applied at
>> Freescale's git?
>
> Lauren,
>
> Could you ask the BSP team to apply 25b35da commit from mainline into 3.14?

It is a shame but it has not been applied yet. I am cooking a forked
kernel for bugfixes like this and will send a patch for testing very
soon.

-- 
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750