[libcamera-devel] [PATCH 2/2] libcamera: formats: add missing RGBX8888 info

[Tomi Valkeinen]

On 19/05/2022 10:07, Laurent Pinchart wrote:
> Hi Tomi,
> 
> On Thu, May 19, 2022 at 09:52:44AM +0300, Tomi Valkeinen wrote:
>> On 19/05/2022 02:20, Laurent Pinchart wrote:
>>> On Wed, May 18, 2022 at 03:20:14PM +0300, Tomi Valkeinen wrote:
>>>> Add missing RGBX8888 PixelFormatInfo.
>>>>
>>>> Signed-off-by: Tomi Valkeinen <tomi.valkeinen at ideasonboard.com>
>>>> ---
>>>>    src/libcamera/formats.cpp | 13 +++++++++++++
>>>>    1 file changed, 13 insertions(+)
>>>>
>>>> diff --git a/src/libcamera/formats.cpp b/src/libcamera/formats.cpp
>>>> index 3e60ec7d..1c710541 100644
>>>> --- a/src/libcamera/formats.cpp
>>>> +++ b/src/libcamera/formats.cpp
>>>> @@ -231,6 +231,19 @@ const std::map<PixelFormat, PixelFormatInfo> pixelFormatInfo{
>>>>    		.pixelsPerGroup = 1,
>>>>    		.planes = {{ { 4, 1 }, { 0, 0 }, { 0, 0 } }},
>>>>    	} },
>>>> +	{ formats::RGBX8888, {
>>>> +		.name = "RGBX8888",
>>>> +		.format = formats::RGBX8888,
>>>> +		.v4l2Formats = {
>>>> +			.single = V4L2PixelFormat(V4L2_PIX_FMT_RGBX32),
>>>
>>> I think this should be BGRX32. Furthermore, you need to also update the
>>
>> Thanks. Yes, it is obviously wrong, as there was another
>> V4L2_PIX_FMT_RGBX32 in the file already. I thought I checked that, but
>> apparently not...
>>
>> I have to say I don't get the V4L2 formats. Why is
>>
>> V4L2_PIX_FMT_XBGR32  BGRX-8-8-8-8
>>
>> so the X is not beside the R in the define. But
>>
>> V4L2_PIX_FMT_XRGB32  XRGB-8-8-8-8
>>
>> X is beside the R. Well, doesn't matter.
> 
> I think those are called historical mistakes. They tend to appear here
> and there. And again :-)

I guessed as much. Well, good that we don't make those. If only the 
other people would learn also.

  Tomi