2019-07-01 05:01:35

by Randy Dunlap

[permalink] [raw]
Subject: [PATCH] drm: bridge: DRM_SIL_SII8620 should depend on, not select INPUT

From: Randy Dunlap <[email protected]>

A single driver should not enable (select) an entire subsystem,
such as INPUT, so change the 'select' to "depends on".

Fixes: d6abe6df706c ("drm/bridge: sil_sii8620: do not have a dependency of RC_CORE")

Signed-off-by: Randy Dunlap <[email protected]>
Cc: Inki Dae <[email protected]>
Cc: Andrzej Hajda <[email protected]>
Cc: Laurent Pinchart <[email protected]>
Cc: [email protected]
---
Linus has written this a couple of times in the last 15 years or so,
but my search fu cannot find it. And there are a few drivers in the
kernel tree that do this, but we shouldn't be adding more that do so.

drivers/gpu/drm/bridge/Kconfig | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)

--- lnx-52-rc7.orig/drivers/gpu/drm/bridge/Kconfig
+++ lnx-52-rc7/drivers/gpu/drm/bridge/Kconfig
@@ -83,10 +83,9 @@ config DRM_PARADE_PS8622

config DRM_SIL_SII8620
tristate "Silicon Image SII8620 HDMI/MHL bridge"
- depends on OF
+ depends on OF && INPUT
select DRM_KMS_HELPER
imply EXTCON
- select INPUT
select RC_CORE
help
Silicon Image SII8620 HDMI/MHL bridge chip driver.



2019-07-01 09:24:09

by Andrzej Hajda

[permalink] [raw]
Subject: Re: [PATCH] drm: bridge: DRM_SIL_SII8620 should depend on, not select INPUT

On 01.07.2019 05:39, Randy Dunlap wrote:
> From: Randy Dunlap <[email protected]>
>
> A single driver should not enable (select) an entire subsystem,
> such as INPUT, so change the 'select' to "depends on".
>
> Fixes: d6abe6df706c ("drm/bridge: sil_sii8620: do not have a dependency of RC_CORE")
>
> Signed-off-by: Randy Dunlap <[email protected]>
> Cc: Inki Dae <[email protected]>
> Cc: Andrzej Hajda <[email protected]>
> Cc: Laurent Pinchart <[email protected]>
> Cc: [email protected]
> ---
> Linus has written this a couple of times in the last 15 years or so,
> but my search fu cannot find it. And there are a few drivers in the
> kernel tree that do this, but we shouldn't be adding more that do so.


The proper solution has been already posted [1], but it has not been
applied yet to input tree?

Randy are there chances your patchset will appear in ML in near future,
or should I merge your sii8620 patch alone?


[1]:
https://lore.kernel.org/lkml/[email protected]/


Regards

Andrzej



>
> drivers/gpu/drm/bridge/Kconfig | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> --- lnx-52-rc7.orig/drivers/gpu/drm/bridge/Kconfig
> +++ lnx-52-rc7/drivers/gpu/drm/bridge/Kconfig
> @@ -83,10 +83,9 @@ config DRM_PARADE_PS8622
>
> config DRM_SIL_SII8620
> tristate "Silicon Image SII8620 HDMI/MHL bridge"
> - depends on OF
> + depends on OF && INPUT
> select DRM_KMS_HELPER
> imply EXTCON
> - select INPUT
> select RC_CORE
> help
> Silicon Image SII8620 HDMI/MHL bridge chip driver.
>
>
>
>

2019-07-01 09:26:54

by Andrzej Hajda

[permalink] [raw]
Subject: Re: [PATCH] drm: bridge: DRM_SIL_SII8620 should depend on, not select INPUT

On 01.07.2019 11:23, Andrzej Hajda wrote:
> On 01.07.2019 05:39, Randy Dunlap wrote:
>> From: Randy Dunlap <[email protected]>
>>
>> A single driver should not enable (select) an entire subsystem,
>> such as INPUT, so change the 'select' to "depends on".
>>
>> Fixes: d6abe6df706c ("drm/bridge: sil_sii8620: do not have a dependency of RC_CORE")
>>
>> Signed-off-by: Randy Dunlap <[email protected]>
>> Cc: Inki Dae <[email protected]>
>> Cc: Andrzej Hajda <[email protected]>
>> Cc: Laurent Pinchart <[email protected]>
>> Cc: [email protected]
>> ---
>> Linus has written this a couple of times in the last 15 years or so,
>> but my search fu cannot find it. And there are a few drivers in the
>> kernel tree that do this, but we shouldn't be adding more that do so.
>
> The proper solution has been already posted [1], but it has not been
> applied yet to input tree?
>
> Randy are there chances your patchset will appear in ML in near future,
> or should I merge your sii8620 patch alone?


Ups, I used wrong surname, I meant Ronald, added him input ML to cc.


Regards

Andrzej



>
>
> [1]:
> https://lore.kernel.org/lkml/[email protected]/
>
>
> Regards
>
> Andrzej
>
>
>
>> drivers/gpu/drm/bridge/Kconfig | 3 +--
>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>
>> --- lnx-52-rc7.orig/drivers/gpu/drm/bridge/Kconfig
>> +++ lnx-52-rc7/drivers/gpu/drm/bridge/Kconfig
>> @@ -83,10 +83,9 @@ config DRM_PARADE_PS8622
>>
>> config DRM_SIL_SII8620
>> tristate "Silicon Image SII8620 HDMI/MHL bridge"
>> - depends on OF
>> + depends on OF && INPUT
>> select DRM_KMS_HELPER
>> imply EXTCON
>> - select INPUT
>> select RC_CORE
>> help
>> Silicon Image SII8620 HDMI/MHL bridge chip driver.
>>
>>
>>
>>

Subject: Re: [PATCH] drm: bridge: DRM_SIL_SII8620 should depend on, not select INPUT


Hi Andrzej,

On Mon, Jul 01, 2019 at 11:26:25AM +0200, Andrzej Hajda wrote:
> On 01.07.2019 11:23, Andrzej Hajda wrote:
> > On 01.07.2019 05:39, Randy Dunlap wrote:
> >> From: Randy Dunlap <[email protected]>
> >>
> >> A single driver should not enable (select) an entire subsystem,
> >> such as INPUT, so change the 'select' to "depends on".
> >>
> >> Fixes: d6abe6df706c ("drm/bridge: sil_sii8620: do not have a dependency of RC_CORE")
> >>
> >> Signed-off-by: Randy Dunlap <[email protected]>
> >> Cc: Inki Dae <[email protected]>
> >> Cc: Andrzej Hajda <[email protected]>
> >> Cc: Laurent Pinchart <[email protected]>
> >> Cc: [email protected]
> >> ---
> >> Linus has written this a couple of times in the last 15 years or so,
> >> but my search fu cannot find it. And there are a few drivers in the
> >> kernel tree that do this, but we shouldn't be adding more that do so.
> >
> > The proper solution has been already posted [1], but it has not been
> > applied yet to input tree?
> >
> > Randy are there chances your patchset will appear in ML in near future,
> > or should I merge your sii8620 patch alone?
>
>
> Ups, I used wrong surname, I meant Ronald, added him input ML to cc.

Looks like the applespi patches got merged, but they were changed at
the last minute to remove the imply's that lead to the Kconfig
dependency on the changes here. And I'm guessing for that reason the
patch [1] was not sent with that pull request in the end.

In short it seems to me that it would be best now if you sent a pull
request for patch [1] and Arnd Bergmann's followup patch that was
discussed in the last few days [2]. Sorry for all the delay.


Cheers,

Ronald


> > [1]:
> > https://lore.kernel.org/lkml/[email protected]/

[2] https://lore.kernel.org/lkml/[email protected]/


> > Regards
> >
> > Andrzej
> >
> >
> >
> >> drivers/gpu/drm/bridge/Kconfig | 3 +--
> >> 1 file changed, 1 insertion(+), 2 deletions(-)
> >>
> >> --- lnx-52-rc7.orig/drivers/gpu/drm/bridge/Kconfig
> >> +++ lnx-52-rc7/drivers/gpu/drm/bridge/Kconfig
> >> @@ -83,10 +83,9 @@ config DRM_PARADE_PS8622
> >>
> >> config DRM_SIL_SII8620
> >> tristate "Silicon Image SII8620 HDMI/MHL bridge"
> >> - depends on OF
> >> + depends on OF && INPUT
> >> select DRM_KMS_HELPER
> >> imply EXTCON
> >> - select INPUT
> >> select RC_CORE
> >> help
> >> Silicon Image SII8620 HDMI/MHL bridge chip driver.
> >>
> >>
> >>
> >>
>