[PATCH 06/21] ARM: OMAP: overo: use new display drivers

[Archit Taneja]

On Tuesday 30 July 2013 12:09 PM, Tomi Valkeinen wrote:
> On 30/07/13 09:21, Archit Taneja wrote:
>> Hi,
>>
>> On Friday 26 July 2013 12:38 PM, Tomi Valkeinen wrote:
>>> Use the new display drivers for OMAP3 Overo board.
>>>
>>> The new OMAP display drivers were merged for 3.11, and we can now change
>>> the board files to use the new ones and phase out the old ones.
>>>
>>> Note that the LCD add-on boards for lcd43 and lcd35 use the same GPIOs
>>> for the panels. This means that both panel devices cannot be probed at
>>> the same time.
>>>
>>> DT will handle this correctly, i.e. the DT data will contain the panel
>>> device only for the add-on board that is attached. However, for the
>>> board file we need a hackish solution: We parse the kernel boot command
>>> line, and see whether lcd43 or lcd35 is set as a default display, and
>>> add the given one. Or, if neither is given, default to lcd43.
>>>
>>
>> <snip>
>>
>>>    static struct omap_dss_board_info overo_dss_data = {
>>> -    .num_devices    = ARRAY_SIZE(overo_dss_devices),
>>> -    .devices    = overo_dss_devices,
>>> -    .default_device    = &overo_dvi_device,
>>> +    .default_display_name = "lcd43",
>>>    };
>>
>> The default display previously was the dvi device, if both lcd43 and
>> lcd35 are on add-on boards, then we should probably stick to dvi itself,
>> right? The hack won't work if dvi is the default device though.
>
> DVI is also on an add-on board, but it doesn't conflict with lcd43 or lcd35.
>
> The hack works fine even if DVI is the default device. In that case, it
> doesn't matter if lcd43 or lcd35 is added, because the user doesn't use
> them (as long as only one of them is added, because otherwise there'll
> be an error during probe).
>
> If DVI is the default device, we could actually skip adding both lcd43
> and lcd35. I just wanted to minimize the code in this hack, so I didn't
> do that.

Okay, thanks for the clarification, looks fine then.

Archit