OMAP display kconfig options changing

OMAP display kconfig options changing

[Tomi Valkeinen]

Hi Tony,

I've just pushed a patch series to fbdev for-next branch which makes the
OMAP DRM and FB drivers independent of each other. This requires
changing the related Kconfig options.

What should I do with omap2plus_defconfig?

At the moment omap2plus_defconfig enables the omapfb driver and panels
as modules. Should I change it to do the same with the new code? Or
should I change it to use omapdrm, which is the better maintained and
developed driver (although no one has probably tried omapdrm on omap2)?

 Tomi

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20151218/582ec53b/attachment.sig>

OMAP display kconfig options changing

[Tony Lindgren]

Hi,

* Tomi Valkeinen <tomi.valkeinen@ti.com> [151218 00:55]:
> Hi Tony,
> 
> I've just pushed a patch series to fbdev for-next branch which makes the
> OMAP DRM and FB drivers independent of each other. This requires
> changing the related Kconfig options.

OK

> What should I do with omap2plus_defconfig?
> 
> At the moment omap2plus_defconfig enables the omapfb driver and panels
> as modules. Should I change it to do the same with the new code? Or
> should I change it to use omapdrm, which is the better maintained and
> developed driver (although no one has probably tried omapdrm on omap2)?

I'm all for changing omap2plus_defconfnig to use omapdrm.

Do you have a link to a page we can point people to for instructions
for what needs to be changed to get things working with omapdrm?

For omap2, I can try to peek into my rack for n800.

Regards,

Tony

OMAP display kconfig options changing

[Adam Ford]

Tomi,

I'd like to also test the LogicPD Torpedo which has been using the FB
drivers until this point. I just pushed some LCD info for inclusion in
4.5, but if I need to make changes, I'd like to do that sooner than
later.  Any insturctions would be helpful.  Will this impact people
who use the SGX drivers?


Thanks

adam

On Fri, Dec 18, 2015 at 7:53 AM, Tony Lindgren <tony@atomide.com> wrote:
> Hi,
>
> * Tomi Valkeinen <tomi.valkeinen@ti.com> [151218 00:55]:
>> Hi Tony,
>>
>> I've just pushed a patch series to fbdev for-next branch which makes the
>> OMAP DRM and FB drivers independent of each other. This requires
>> changing the related Kconfig options.
>
> OK
>
>> What should I do with omap2plus_defconfig?
>>
>> At the moment omap2plus_defconfig enables the omapfb driver and panels
>> as modules. Should I change it to do the same with the new code? Or
>> should I change it to use omapdrm, which is the better maintained and
>> developed driver (although no one has probably tried omapdrm on omap2)?
>
> I'm all for changing omap2plus_defconfnig to use omapdrm.
>
> Do you have a link to a page we can point people to for instructions
> for what needs to be changed to get things working with omapdrm?
>
> For omap2, I can try to peek into my rack for n800.
>
> Regards,
>
> Tony
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

OMAP display kconfig options changing

[Tomi Valkeinen]

On 18/12/15 16:03, Adam Ford wrote:
> Tomi,
> 
> I'd like to also test the LogicPD Torpedo which has been using the FB
> drivers until this point. I just pushed some LCD info for inclusion in
> 4.5, but if I need to make changes, I'd like to do that sooner than

The DT data is the same for omapfb and omapdrm, so no changes needed.

> later.  Any insturctions would be helpful.  Will this impact people
> who use the SGX drivers?

I'm no expert, but... Yes, I think so. If you were using SGX driver with
some additional omapfb patches, it won't work with omapdrm.

omapdrm + SGX works, but it's a bit tricky setup (need extra kernel
patches, specific userspace libs...). I hope we'll get the SGX side
sorted out during the first quarter, which should make omapdrm + SGX
much easier to get running.

 Tomi

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20151218/f2417bea/attachment-0001.sig>

OMAP display kconfig options changing

[Tomi Valkeinen]

On 18/12/15 15:53, Tony Lindgren wrote:
> Hi,
> 
> * Tomi Valkeinen <tomi.valkeinen@ti.com> [151218 00:55]:
>> Hi Tony,
>>
>> I've just pushed a patch series to fbdev for-next branch which makes the
>> OMAP DRM and FB drivers independent of each other. This requires
>> changing the related Kconfig options.
> 
> OK

Note that nothing has changed with omapfb and omapdrm as such. The only
(visible) change is the needed Kconfig options.

>> What should I do with omap2plus_defconfig?
>>
>> At the moment omap2plus_defconfig enables the omapfb driver and panels
>> as modules. Should I change it to do the same with the new code? Or
>> should I change it to use omapdrm, which is the better maintained and
>> developed driver (although no one has probably tried omapdrm on omap2)?
> 
> I'm all for changing omap2plus_defconfnig to use omapdrm.
> 
> Do you have a link to a page we can point people to for instructions
> for what needs to be changed to get things working with omapdrm?

Are you referring to Kconfig options? Or userspace?

But no, I don't have links to either.

For the Kconfig options, it's pretty similar to omapfb. After the
patches I'm pushing, one needs to enable DRM, DRM_OMAP, and then the
individual panel/encoder drivers (as for omapfb).

> For omap2, I can try to peek into my rack for n800.

There were a few fixes sent a few days ago which are needed for OMAP2/3.
I think both are now in Linus' tree.

 Tomi

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20151218/b81d6ce0/attachment-0001.sig>

OMAP display kconfig options changing

[Tony Lindgren]

* Tomi Valkeinen <tomi.valkeinen@ti.com> [151218 07:01]:
> 
> 
> On 18/12/15 15:53, Tony Lindgren wrote:
> > Hi,
> > 
> > * Tomi Valkeinen <tomi.valkeinen@ti.com> [151218 00:55]:
> >> Hi Tony,
> >>
> >> I've just pushed a patch series to fbdev for-next branch which makes the
> >> OMAP DRM and FB drivers independent of each other. This requires
> >> changing the related Kconfig options.
> > 
> > OK
> 
> Note that nothing has changed with omapfb and omapdrm as such. The only
> (visible) change is the needed Kconfig options.

OK

> >> What should I do with omap2plus_defconfig?
> >>
> >> At the moment omap2plus_defconfig enables the omapfb driver and panels
> >> as modules. Should I change it to do the same with the new code? Or
> >> should I change it to use omapdrm, which is the better maintained and
> >> developed driver (although no one has probably tried omapdrm on omap2)?
> > 
> > I'm all for changing omap2plus_defconfnig to use omapdrm.
> > 
> > Do you have a link to a page we can point people to for instructions
> > for what needs to be changed to get things working with omapdrm?
> 
> Are you referring to Kconfig options? Or userspace?

Well sounds like both need updating :)

> But no, I don't have links to either.

So change the Kconfig options and then deal with the flood
of angry emails? :) I think we need instructions..

> For the Kconfig options, it's pretty similar to omapfb. After the
> patches I'm pushing, one needs to enable DRM, DRM_OMAP, and then the
> individual panel/encoder drivers (as for omapfb).
> 
> > For omap2, I can try to peek into my rack for n800.
> 
> There were a few fixes sent a few days ago which are needed for OMAP2/3.
> I think both are now in Linus' tree.

OK I guess we have to wait your patches then to give it
a try.

Regards,

Tony

OMAP display kconfig options changing

[Tomi Valkeinen]

On 18/12/15 18:05, Tony Lindgren wrote:

>> Are you referring to Kconfig options? Or userspace?
> 
> Well sounds like both need updating :)

Depends... For simple fb apps, omapdrm works fine, and X with the
standard fbdev driver should work with omapdrm too.

But if the userspace is using any omapfb specific apps, then yes, update
is necessary.

>> But no, I don't have links to either.
> 
> So change the Kconfig options and then deal with the flood
> of angry emails? :) I think we need instructions..

Ok. I think it's then best that I just update the defconfig to enable
omapfb as modules, as it is currently.

We can look again at changing the default to omapdrm after we've tested
it on OMAP2 too, and we have a better idea what the user needs to take
care of.

 Tomi

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20151218/1ca25916/attachment-0001.sig>

OMAP display kconfig options changing

[Javier Martinez Canillas]

Hello Tomi,

On Fri, Dec 18, 2015 at 1:14 PM, Tomi Valkeinen <tomi.valkeinen@ti.com> wrote:
> On 18/12/15 18:05, Tony Lindgren wrote:
>
>>> Are you referring to Kconfig options? Or userspace?
>>
>> Well sounds like both need updating :)
>
> Depends... For simple fb apps, omapdrm works fine, and X with the
> standard fbdev driver should work with omapdrm too.
>

It does, I just tested omapdrm + X with the fbdev DDX and I was able
to start X and the xfce4 wm.

I just needed to enable DRM_FBDEV_EMULATION in omap2plus_defconfig to
have fbdev emulation over DRM.

> But if the userspace is using any omapfb specific apps, then yes, update
> is necessary.
>

Yes but these people should have a migration plan anyways since omapfb
(and fbdev in general) is going away right?

>>> But no, I don't have links to either.
>>
>> So change the Kconfig options and then deal with the flood
>> of angry emails? :) I think we need instructions..
>
> Ok. I think it's then best that I just update the defconfig to enable
> omapfb as modules, as it is currently.
>

In that case I think you should squash the defconfig changes with
commit ("70ba4e05531f omapfb/displays: change CONFIG_DISPLAY_* to
CONFIG_FB_OMAP2_*") to maintain bisectability.

Alternatively, the current panel and encoders Kconfig symbols could
remain for omapfb since that's the current ones used in
omap2plus_defconfig where omapfb is the default and have new Kconfig
symbols for omapdrm (i.e: CONFIG_DRM_OMAP_ENCODER_TFP410).

> We can look again at changing the default to omapdrm after we've tested
> it on OMAP2 too, and we have a better idea what the user needs to take
> care of.
>
>  Tomi
>

Best regards,
Javier

OMAP display kconfig options changing

[Tomi Valkeinen]

Hi Javier, Tony,

On 22/12/15 23:18, Javier Martinez Canillas wrote:

>> But if the userspace is using any omapfb specific apps, then yes, update
>> is necessary.
>>
> 
> Yes but these people should have a migration plan anyways since omapfb
> (and fbdev in general) is going away right?

Well, I would guess most people don't know/care about that.

>> Ok. I think it's then best that I just update the defconfig to enable
>> omapfb as modules, as it is currently.
>>
> 
> In that case I think you should squash the defconfig changes with
> commit ("70ba4e05531f omapfb/displays: change CONFIG_DISPLAY_* to
> CONFIG_FB_OMAP2_*") to maintain bisectability.

Here's the diff to change the defconfig to enable the same items as before:

diff --git a/arch/arm/configs/omap2plus_defconfig
b/arch/arm/configs/omap2plus_defconfig
index c5e1943e5427..b9581f1f0536 100644
--- a/arch/arm/configs/omap2plus_defconfig
+++ b/arch/arm/configs/omap2plus_defconfig
@@ -290,24 +290,23 @@ CONFIG_FB=y
 CONFIG_FIRMWARE_EDID=y
 CONFIG_FB_MODE_HELPERS=y
 CONFIG_FB_TILEBLITTING=y
-CONFIG_OMAP2_DSS=m
-CONFIG_OMAP5_DSS_HDMI=y
-CONFIG_OMAP2_DSS_SDI=y
-CONFIG_OMAP2_DSS_DSI=y
+CONFIG_FB_OMAP5_DSS_HDMI=y
+CONFIG_FB_OMAP2_DSS_SDI=y
+CONFIG_FB_OMAP2_DSS_DSI=y
 CONFIG_FB_OMAP2=m
-CONFIG_DISPLAY_ENCODER_TFP410=m
-CONFIG_DISPLAY_ENCODER_TPD12S015=m
-CONFIG_DISPLAY_CONNECTOR_DVI=m
-CONFIG_DISPLAY_CONNECTOR_HDMI=m
-CONFIG_DISPLAY_CONNECTOR_ANALOG_TV=m
-CONFIG_DISPLAY_PANEL_DPI=m
-CONFIG_DISPLAY_PANEL_DSI_CM=m
-CONFIG_DISPLAY_PANEL_SONY_ACX565AKM=m
-CONFIG_DISPLAY_PANEL_LGPHILIPS_LB035Q02=m
-CONFIG_DISPLAY_PANEL_SHARP_LS037V7DW01=m
-CONFIG_DISPLAY_PANEL_TPO_TD028TTEC1=m
-CONFIG_DISPLAY_PANEL_TPO_TD043MTEA1=m
-CONFIG_DISPLAY_PANEL_NEC_NL8048HL11=m
+CONFIG_FB_OMAP2_ENCODER_TFP410=m
+CONFIG_FB_OMAP2_ENCODER_TPD12S015=m
+CONFIG_FB_OMAP2_CONNECTOR_DVI=m
+CONFIG_FB_OMAP2_CONNECTOR_HDMI=m
+CONFIG_FB_OMAP2_CONNECTOR_ANALOG_TV=m
+CONFIG_FB_OMAP2_PANEL_DPI=m
+CONFIG_FB_OMAP2_PANEL_DSI_CM=m
+CONFIG_FB_OMAP2_PANEL_SONY_ACX565AKM=m
+CONFIG_FB_OMAP2_PANEL_LGPHILIPS_LB035Q02=m
+CONFIG_FB_OMAP2_PANEL_SHARP_LS037V7DW01=m
+CONFIG_FB_OMAP2_PANEL_TPO_TD028TTEC1=m
+CONFIG_FB_OMAP2_PANEL_TPO_TD043MTEA1=m
+CONFIG_FB_OMAP2_PANEL_NEC_NL8048HL11=m
 CONFIG_BACKLIGHT_LCD_SUPPORT=y
 CONFIG_LCD_CLASS_DEVICE=y
 CONFIG_LCD_PLATFORM=y

Tony, the diff above looks like something that's safe to merge via fbdev
tree. The context contains only display related configs. Is it ok if I
squash the above change to my series to keep bisectability?

> Alternatively, the current panel and encoders Kconfig symbols could
> remain for omapfb since that's the current ones used in
> omap2plus_defconfig where omapfb is the default and have new Kconfig
> symbols for omapdrm (i.e: CONFIG_DRM_OMAP_ENCODER_TFP410).

I want to change the config symbols, as the current ones are too generic
(e.g. CONFIG_DISPLAY_PANEL_DPI doesn't mention OMAP anywhere). I think
this is a good time to change them for omapfb. But this is probably a
good time to change them for omapdrm also, so I think I'll add that change.

 Tomi

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20151229/8458eaa2/attachment-0001.sig>

OMAP display kconfig options changing

[Javier Martinez Canillas]

Hello Tomi,

On Tue, Dec 29, 2015 at 7:23 AM, Tomi Valkeinen <tomi.valkeinen@ti.com> wrote:
> Hi Javier, Tony,
>
> On 22/12/15 23:18, Javier Martinez Canillas wrote:
>
>>> But if the userspace is using any omapfb specific apps, then yes, update
>>> is necessary.
>>>
>>
>> Yes but these people should have a migration plan anyways since omapfb
>> (and fbdev in general) is going away right?
>
> Well, I would guess most people don't know/care about that.
>

Fair enough.

>>> Ok. I think it's then best that I just update the defconfig to enable
>>> omapfb as modules, as it is currently.
>>>
>>
>> In that case I think you should squash the defconfig changes with
>> commit ("70ba4e05531f omapfb/displays: change CONFIG_DISPLAY_* to
>> CONFIG_FB_OMAP2_*") to maintain bisectability.
>
> Here's the diff to change the defconfig to enable the same items as before:
>
> diff --git a/arch/arm/configs/omap2plus_defconfig
> b/arch/arm/configs/omap2plus_defconfig
> index c5e1943e5427..b9581f1f0536 100644
> --- a/arch/arm/configs/omap2plus_defconfig
> +++ b/arch/arm/configs/omap2plus_defconfig
> @@ -290,24 +290,23 @@ CONFIG_FB=y
>  CONFIG_FIRMWARE_EDID=y
>  CONFIG_FB_MODE_HELPERS=y
>  CONFIG_FB_TILEBLITTING=y
> -CONFIG_OMAP2_DSS=m
> -CONFIG_OMAP5_DSS_HDMI=y
> -CONFIG_OMAP2_DSS_SDI=y
> -CONFIG_OMAP2_DSS_DSI=y
> +CONFIG_FB_OMAP5_DSS_HDMI=y
> +CONFIG_FB_OMAP2_DSS_SDI=y
> +CONFIG_FB_OMAP2_DSS_DSI=y
>  CONFIG_FB_OMAP2=m
> -CONFIG_DISPLAY_ENCODER_TFP410=m
> -CONFIG_DISPLAY_ENCODER_TPD12S015=m
> -CONFIG_DISPLAY_CONNECTOR_DVI=m
> -CONFIG_DISPLAY_CONNECTOR_HDMI=m
> -CONFIG_DISPLAY_CONNECTOR_ANALOG_TV=m
> -CONFIG_DISPLAY_PANEL_DPI=m
> -CONFIG_DISPLAY_PANEL_DSI_CM=m
> -CONFIG_DISPLAY_PANEL_SONY_ACX565AKM=m
> -CONFIG_DISPLAY_PANEL_LGPHILIPS_LB035Q02=m
> -CONFIG_DISPLAY_PANEL_SHARP_LS037V7DW01=m
> -CONFIG_DISPLAY_PANEL_TPO_TD028TTEC1=m
> -CONFIG_DISPLAY_PANEL_TPO_TD043MTEA1=m
> -CONFIG_DISPLAY_PANEL_NEC_NL8048HL11=m
> +CONFIG_FB_OMAP2_ENCODER_TFP410=m
> +CONFIG_FB_OMAP2_ENCODER_TPD12S015=m
> +CONFIG_FB_OMAP2_CONNECTOR_DVI=m
> +CONFIG_FB_OMAP2_CONNECTOR_HDMI=m
> +CONFIG_FB_OMAP2_CONNECTOR_ANALOG_TV=m
> +CONFIG_FB_OMAP2_PANEL_DPI=m
> +CONFIG_FB_OMAP2_PANEL_DSI_CM=m
> +CONFIG_FB_OMAP2_PANEL_SONY_ACX565AKM=m
> +CONFIG_FB_OMAP2_PANEL_LGPHILIPS_LB035Q02=m
> +CONFIG_FB_OMAP2_PANEL_SHARP_LS037V7DW01=m
> +CONFIG_FB_OMAP2_PANEL_TPO_TD028TTEC1=m
> +CONFIG_FB_OMAP2_PANEL_TPO_TD043MTEA1=m
> +CONFIG_FB_OMAP2_PANEL_NEC_NL8048HL11=m
>  CONFIG_BACKLIGHT_LCD_SUPPORT=y
>  CONFIG_LCD_CLASS_DEVICE=y
>  CONFIG_LCD_PLATFORM=y
>

Cool, that's what I changed to test omapfb as well. I can't test your
diff right now but looks correct to me.

Best regards,
Javier

OMAP display kconfig options changing

[Tony Lindgren]

* Javier Martinez Canillas <javier@dowhile0.org> [151229 04:14]:
> On Tue, Dec 29, 2015 at 7:23 AM, Tomi Valkeinen <tomi.valkeinen@ti.com> wrote:
> >> In that case I think you should squash the defconfig changes with
> >> commit ("70ba4e05531f omapfb/displays: change CONFIG_DISPLAY_* to
> >> CONFIG_FB_OMAP2_*") to maintain bisectability.
> >
> > Here's the diff to change the defconfig to enable the same items as before:
> >
> > diff --git a/arch/arm/configs/omap2plus_defconfig
> > b/arch/arm/configs/omap2plus_defconfig
> > index c5e1943e5427..b9581f1f0536 100644
> > --- a/arch/arm/configs/omap2plus_defconfig
> > +++ b/arch/arm/configs/omap2plus_defconfig
> > @@ -290,24 +290,23 @@ CONFIG_FB=y
> >  CONFIG_FIRMWARE_EDID=y
> >  CONFIG_FB_MODE_HELPERS=y
> >  CONFIG_FB_TILEBLITTING=y
> > -CONFIG_OMAP2_DSS=m
> > -CONFIG_OMAP5_DSS_HDMI=y
> > -CONFIG_OMAP2_DSS_SDI=y
> > -CONFIG_OMAP2_DSS_DSI=y
> > +CONFIG_FB_OMAP5_DSS_HDMI=y
> > +CONFIG_FB_OMAP2_DSS_SDI=y
> > +CONFIG_FB_OMAP2_DSS_DSI=y
> >  CONFIG_FB_OMAP2=m
> > -CONFIG_DISPLAY_ENCODER_TFP410=m
> > -CONFIG_DISPLAY_ENCODER_TPD12S015=m
> > -CONFIG_DISPLAY_CONNECTOR_DVI=m
> > -CONFIG_DISPLAY_CONNECTOR_HDMI=m
> > -CONFIG_DISPLAY_CONNECTOR_ANALOG_TV=m
> > -CONFIG_DISPLAY_PANEL_DPI=m
> > -CONFIG_DISPLAY_PANEL_DSI_CM=m
> > -CONFIG_DISPLAY_PANEL_SONY_ACX565AKM=m
> > -CONFIG_DISPLAY_PANEL_LGPHILIPS_LB035Q02=m
> > -CONFIG_DISPLAY_PANEL_SHARP_LS037V7DW01=m
> > -CONFIG_DISPLAY_PANEL_TPO_TD028TTEC1=m
> > -CONFIG_DISPLAY_PANEL_TPO_TD043MTEA1=m
> > -CONFIG_DISPLAY_PANEL_NEC_NL8048HL11=m
> > +CONFIG_FB_OMAP2_ENCODER_TFP410=m
> > +CONFIG_FB_OMAP2_ENCODER_TPD12S015=m
> > +CONFIG_FB_OMAP2_CONNECTOR_DVI=m
> > +CONFIG_FB_OMAP2_CONNECTOR_HDMI=m
> > +CONFIG_FB_OMAP2_CONNECTOR_ANALOG_TV=m
> > +CONFIG_FB_OMAP2_PANEL_DPI=m
> > +CONFIG_FB_OMAP2_PANEL_DSI_CM=m
> > +CONFIG_FB_OMAP2_PANEL_SONY_ACX565AKM=m
> > +CONFIG_FB_OMAP2_PANEL_LGPHILIPS_LB035Q02=m
> > +CONFIG_FB_OMAP2_PANEL_SHARP_LS037V7DW01=m
> > +CONFIG_FB_OMAP2_PANEL_TPO_TD028TTEC1=m
> > +CONFIG_FB_OMAP2_PANEL_TPO_TD043MTEA1=m
> > +CONFIG_FB_OMAP2_PANEL_NEC_NL8048HL11=m
> >  CONFIG_BACKLIGHT_LCD_SUPPORT=y
> >  CONFIG_LCD_CLASS_DEVICE=y
> >  CONFIG_LCD_PLATFORM=y
> >
> 
> Cool, that's what I changed to test omapfb as well. I can't test your
> diff right now but looks correct to me.

Sorry Tomi, I only now noticed this. Yes please feel free to queue
with your other patches (or as a fix):

Acked-by: Tony Lindgren <tony@atomide.com>