Failure due to missing (Exynos related) pinctrl patch

Failure due to missing (Exynos related) pinctrl patch

[Sachin Kamat]

Hi Linus, Kukjin,

Patch titled "pinctrl: exynos: change PINCTRL_EXYNOS option"
(linux-next commit Id: 7452b64d) which is present in linux-next is
missing in the mainline kernel. This patch is required along with the
patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
(mainline commit Id: e4a5da51) which has already made it into
mainline. Without the missing patch we get the following boot up
warnings and subsequent failures with dt boot on 4412 based board:

 WARNING: at drivers/gpio/gpio-samsung.c:3102 samsung_gpiolib_init+0x29c/0x2e8()
 Unknown SoC in gpio-samsung, no GPIOs added
 Modules linked in:
 [<c00142dc>] (unwind_backtrace+0x0/0xf8) from [<c00202bc>]
(warn_slowpath_common+0x4c/0x64)
 [<c00202bc>] (warn_slowpath_common+0x4c/0x64) from [<c0020368>]
(warn_slowpath_fmt+0x30/0x40)
 [<c0020368>] (warn_slowpath_fmt+0x30/0x40) from [<c028b924>]
(samsung_gpiolib_init+0x29c/0x2e8)
 [<c028b924>] (samsung_gpiolib_init+0x29c/0x2e8) from [<c000862c>]
(do_one_initcall+0x34/0x174)
 [<c000862c>] (do_one_initcall+0x34/0x174) from [<c0275960>]
(kernel_init_freeable+0xfc/0x1c8)
 [<c0275960>] (kernel_init_freeable+0xfc/0x1c8) from [<c01c0a5c>]
(kernel_init+0x8/0xe4)
 [<c01c0a5c>] (kernel_init+0x8/0xe4) from [<c000e158>] (ret_from_fork+0x14/0x3c)

-- 
With warm regards,
Sachin

RE: Failure due to missing (Exynos related) pinctrl patch

[Kukjin Kim]

Sachin Kamat wrote:
> 
> Hi Linus, Kukjin,
> 
> Patch titled "pinctrl: exynos: change PINCTRL_EXYNOS option"
> (linux-next commit Id: 7452b64d) which is present in linux-next is
> missing in the mainline kernel. This patch is required along with the
> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
> (mainline commit Id: e4a5da51) which has already made it into
> mainline. Without the missing patch we get the following boot up
> warnings and subsequent failures with dt boot on 4412 based board:
> 
>  WARNING: at drivers/gpio/gpio-samsung.c:3102
> samsung_gpiolib_init+0x29c/0x2e8()
>  Unknown SoC in gpio-samsung, no GPIOs added
>  Modules linked in:
>  [<c00142dc>] (unwind_backtrace+0x0/0xf8) from [<c00202bc>]
> (warn_slowpath_common+0x4c/0x64)
>  [<c00202bc>] (warn_slowpath_common+0x4c/0x64) from [<c0020368>]
> (warn_slowpath_fmt+0x30/0x40)
>  [<c0020368>] (warn_slowpath_fmt+0x30/0x40) from [<c028b924>]
> (samsung_gpiolib_init+0x29c/0x2e8)
>  [<c028b924>] (samsung_gpiolib_init+0x29c/0x2e8) from [<c000862c>]
> (do_one_initcall+0x34/0x174)
>  [<c000862c>] (do_one_initcall+0x34/0x174) from [<c0275960>]
> (kernel_init_freeable+0xfc/0x1c8)
>  [<c0275960>] (kernel_init_freeable+0xfc/0x1c8) from [<c01c0a5c>]
> (kernel_init+0x8/0xe4)
>  [<c01c0a5c>] (kernel_init+0x8/0xe4) from [<c000e158>]
> (ret_from_fork+0x14/0x3c)
> 
Sachin, thanks for your information.

Linus, can you check following again?

https://lkml.org/lkml/2013/1/18/461

Thanks.

- Kukjin

Re: Failure due to missing (Exynos related) pinctrl patch

[Linus Walleij]

On Tue, Jan 29, 2013 at 6:10 AM, Sachin Kamat <sachin.kamat@linaro.org> wrote:

> Patch titled "pinctrl: exynos: change PINCTRL_EXYNOS option"
> (linux-next commit Id: 7452b64d) which is present in linux-next is
> missing in the mainline kernel.

Do I haven't sent my pull request for the pinctrl fixes yet.

> This patch is required along with the
> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
> (mainline commit Id: e4a5da51) which has already made it into
> mainline. Without the missing patch we get the following boot up
> warnings and subsequent failures with dt boot on 4412 based board:

Hm I didn't realize there was a dependency between these
patches, oh well, better get the pull for the pinctrl tree finished
off then...

Yours,
Linus Walleij

Re: Failure due to missing (Exynos related) pinctrl patch

[Sachin Kamat]

>> This patch is required along with the
>> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
>> (mainline commit Id: e4a5da51) which has already made it into
>> mainline. Without the missing patch we get the following boot up
>> warnings and subsequent failures with dt boot on 4412 based board:
>
> Hm I didn't realize there was a dependency between these
> patches, oh well, better get the pull for the pinctrl tree finished
> off then...

Ideally those 2 patches should have been squashed together.

-- 
With warm regards,
Sachin

RE: Failure due to missing (Exynos related) pinctrl patch

[Kukjin Kim]

Sachin Kamat wrote:
> 
> >> This patch is required along with the
> >> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
> >> (mainline commit Id: e4a5da51) which has already made it into
> >> mainline. Without the missing patch we get the following boot up
> >> warnings and subsequent failures with dt boot on 4412 based board:
> >
> > Hm I didn't realize there was a dependency between these
> > patches, oh well, better get the pull for the pinctrl tree finished
> > off then...
> 
> Ideally those 2 patches should have been squashed together.
> 
Err, Sachin, what are you talking about? As you said, one patch has been
merged into the mainline, so how does it can be squashed? In this case, just
applying the other patch into pinctrl-fixes is enough.

Thanks.

- Kukjin

Re: Failure due to missing (Exynos related) pinctrl patch

[Sachin Kamat]

On 30 January 2013 09:38, Kukjin Kim <kgene.kim@samsung.com> wrote:
> Sachin Kamat wrote:
>>
>> >> This patch is required along with the
>> >> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
>> >> (mainline commit Id: e4a5da51) which has already made it into
>> >> mainline. Without the missing patch we get the following boot up
>> >> warnings and subsequent failures with dt boot on 4412 based board:
>> >
>> > Hm I didn't realize there was a dependency between these
>> > patches, oh well, better get the pull for the pinctrl tree finished
>> > off then...
>>
>> Ideally those 2 patches should have been squashed together.
>>
> Err, Sachin, what are you talking about? As you said, one patch has been
> merged into the mainline, so how does it can be squashed? In this case, just
> applying the other patch into pinctrl-fixes is enough.

Yes you are right. It cannot be squashed now. I was referring to the
time of generation of these patches.

>
> Thanks.
>
> - Kukjin
>



-- 
With warm regards,
Sachin

Re: Failure due to missing (Exynos related) pinctrl patch

[Linus Walleij]

On Wed, Jan 30, 2013 at 5:08 AM, Kukjin Kim <kgene.kim@samsung.com> wrote:
> Sachin Kamat wrote:
>>
>> >> This patch is required along with the
>> >> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
>> >> (mainline commit Id: e4a5da51) which has already made it into
>> >> mainline. Without the missing patch we get the following boot up
>> >> warnings and subsequent failures with dt boot on 4412 based board:
>> >
>> > Hm I didn't realize there was a dependency between these
>> > patches, oh well, better get the pull for the pinctrl tree finished
>> > off then...
>>
>> Ideally those 2 patches should have been squashed together.
>>
> Err, Sachin, what are you talking about? As you said, one patch has been
> merged into the mainline, so how does it can be squashed? In this case, just
> applying the other patch into pinctrl-fixes is enough.

Torvalds has already pulled it in -> problem solved. :-)

Yours,
Linus Walleij

Re: Failure due to missing (Exynos related) pinctrl patch

[Sachin Kamat]

Hi Linus,

On 30 January 2013 15:18, Linus Walleij <linus.walleij@linaro.org> wrote:
> On Wed, Jan 30, 2013 at 5:08 AM, Kukjin Kim <kgene.kim@samsung.com> wrote:
>> Sachin Kamat wrote:
>>>
>>> >> This patch is required along with the
>>> >> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
>>> >> (mainline commit Id: e4a5da51) which has already made it into
>>> >> mainline. Without the missing patch we get the following boot up
>>> >> warnings and subsequent failures with dt boot on 4412 based board:
>>> >
>>> > Hm I didn't realize there was a dependency between these
>>> > patches, oh well, better get the pull for the pinctrl tree finished
>>> > off then...

>
> Torvalds has already pulled it in -> problem solved. :-)
>

I tested this on Linus T.'s latest tip (2e51b231) and the problem
still persists :(

You sent the following patch in your pull request to Linus T.:
Kukjin Kim (1):
      pinctrl: samsung: removing duplicated condition for PINCTRL_SAMSUNG

However, the missing dependent patch is:

"pinctrl: exynos: change PINCTRL_EXYNOS option"

-- 
With warm regards,
Sachin

Re: Failure due to missing (Exynos related) pinctrl patch

[Linus Walleij]

On Wed, Jan 30, 2013 at 11:15 AM, Sachin Kamat <sachin.kamat@linaro.org> wrote:
> On 30 January 2013 15:18, Linus Walleij <linus.walleij@linaro.org> wrote:
>> On Wed, Jan 30, 2013 at 5:08 AM, Kukjin Kim <kgene.kim@samsung.com> wrote:
>>> Sachin Kamat wrote:
>>>>
>>>> >> This patch is required along with the
>>>> >> patch "gpio: samsung: fix pinctrl condition for exynos and exynos5440"
>>>> >> (mainline commit Id: e4a5da51) which has already made it into
>>>> >> mainline. Without the missing patch we get the following boot up
>>>> >> warnings and subsequent failures with dt boot on 4412 based board:
>>>> >
>>>> > Hm I didn't realize there was a dependency between these
>>>> > patches, oh well, better get the pull for the pinctrl tree finished
>>>> > off then...
>
>>
>> Torvalds has already pulled it in -> problem solved. :-)
>>
>
> I tested this on Linus T.'s latest tip (2e51b231) and the problem
> still persists :(
>
> You sent the following patch in your pull request to Linus T.:
> Kukjin Kim (1):
>       pinctrl: samsung: removing duplicated condition for PINCTRL_SAMSUNG
>
> However, the missing dependent patch is:
>
> "pinctrl: exynos: change PINCTRL_EXYNOS option"

Aha how typically chaotical.

And that is in the middle of the devel branch that people are
basing their development on.

Not much to do, I cherry-picked that patch to the fixes branch
and hope that git will cope. (It usually does.)

Yours,
Linus Walleij