[libcamera-devel] [PATCH v2 3/9] android: camera_device: Support multiple stream configurations

Kieran Bingham kieran.bingham at ideasonboard.com
Fri Jul 3 11:30:17 CEST 2020


Hi Niklas,

On 03/07/2020 00:22, Niklas Söderlund wrote:
> Hi Kieran,
> 
> Thanks for your work.
> 
> On 2020-07-02 22:36:48 +0100, Kieran Bingham wrote:
>> Create an initial Camera Configuration using an empty role set, and
>> populate the StreamConfigurations manually from each of the streams
>> given by the Android camera3_stream_configuration_t stream_list.
>>
>> Signed-off-by: Kieran Bingham <kieran.bingham at ideasonboard.com>
>> ---
>>  src/android/camera_device.cpp | 51 +++++++++++++++++------------------
>>  1 file changed, 25 insertions(+), 26 deletions(-)
>>
>> diff --git a/src/android/camera_device.cpp b/src/android/camera_device.cpp
>> index b9031ff0c2a4..03dcdd520794 100644
>> --- a/src/android/camera_device.cpp
>> +++ b/src/android/camera_device.cpp
>> @@ -942,6 +942,16 @@ PixelFormat CameraDevice::toPixelFormat(int format)
>>   */
>>  int CameraDevice::configureStreams(camera3_stream_configuration_t *stream_list)
>>  {
>> +	/*
>> +	 * Generate an empty configuration, and construct a StreamConfiguration
>> +	 * for each camera3_stream to add to it.
>> +	 */
>> +	config_ = camera_->generateConfiguration();
>> +	if (!config_) {
>> +		LOG(HAL, Error) << "Failed to generate camera configuration";
>> +		return -EINVAL;
>> +	}
>> +
>>  	for (unsigned int i = 0; i < stream_list->num_streams; ++i) {
>>  		camera3_stream_t *stream = stream_list->streams[i];
>>  
>> @@ -953,35 +963,18 @@ int CameraDevice::configureStreams(camera3_stream_configuration_t *stream_list)
>>  			       << ", height: " << stream->height
>>  			       << ", format: " << utils::hex(stream->format)
>>  			       << " (" << format.toString() << ")";
>> -	}
>>  
>> -	/* Only one stream is supported. */
>> -	if (stream_list->num_streams != 1) {
>> -		LOG(HAL, Error) << "Only one stream supported";
>> -		return -EINVAL;
>> -	}
>> -	camera3_stream_t *camera3Stream = stream_list->streams[0];
>> +		if (!format.isValid())
>> +			return -EINVAL;
>>  
>> -	/* Translate Android format code to libcamera pixel format. */
>> -	PixelFormat format = toPixelFormat(camera3Stream->format);
>> -	if (!format.isValid())
>> -		return -EINVAL;
>> +		StreamConfiguration streamConfiguration;
>>  
>> -	/*
>> -	 * Hardcode viewfinder role, replacing the generated configuration
>> -	 * parameters with the ones requested by the Android framework.
>> -	 */
>> -	StreamRoles roles = { StreamRole::Viewfinder };
>> -	config_ = camera_->generateConfiguration(roles);
>> -	if (!config_ || config_->empty()) {
>> -		LOG(HAL, Error) << "Failed to generate camera configuration";
>> -		return -EINVAL;
>> -	}
>> +		streamConfiguration.size.width = stream->width;
>> +		streamConfiguration.size.height = stream->height;
>> +		streamConfiguration.pixelFormat = format;
> 
> This is not related to this patch but blindly translating a HAL format 
> using a lookup table to a libcamera format will not work for RAW. I'm 
> just curious if it is sufficient for JPEG?
> 

No - it is not, but that's not tackled by this patch.

Lets consider this patch as supporting multiple streams of types that
are already supported ;-)

I would expect this patch to now support multiple YUV streams.


>>  
>> -	StreamConfiguration *streamConfiguration = &config_->at(0);
>> -	streamConfiguration->size.width = camera3Stream->width;
>> -	streamConfiguration->size.height = camera3Stream->height;
>> -	streamConfiguration->pixelFormat = format;
>> +		config_->addConfiguration(streamConfiguration);
>> +	}
>>  
>>  	switch (config_->validate()) {
>>  	case CameraConfiguration::Valid:
>> @@ -996,7 +989,13 @@ int CameraDevice::configureStreams(camera3_stream_configuration_t *stream_list)
>>  		return -EINVAL;
>>  	}
>>  
>> -	camera3Stream->max_buffers = streamConfiguration->bufferCount;
>> +	for (unsigned int i = 0; i < stream_list->num_streams; ++i) {
> 
> Small nit, would it make sens to iterate over config_ instead of 
> stream_list->num_streams? If we have a discrepancy we would end up with 
> less streams then requested but if we use ->at(i) an exception.

No... or at least - not in the /very/ near future.

There will very soon be an explicit discrepency between the number of
camera3 streams and the number of libcamera streams.

Adding a JPEG stream will not add a new stream, but we must still
support it.

Later I add a mapping which stores the correct libcamera index to
reference for this stream (for which multiple camera3 streams might
reference a single libcamera stream index).

If I iterate over config_ here, I would still need an index to map
/back/ to the camera3_stream index, so either way an explicit for loop
seems better here currently.



> 
> This is a small thing that poped out at me, with or without this 
> changed,
> 
> Reviewed-by: Niklas Söderlund <niklas.soderlund at ragnatech.se>

Thanks,

I hope the above makes you happier with this path...


> 
>> +		camera3_stream_t *stream = stream_list->streams[i];
>> +		StreamConfiguration &streamConfiguration = config_->at(i);
>> +
>> +		/* Use the bufferCount confirmed by the validation process. */
>> +		stream->max_buffers = streamConfiguration.bufferCount;
>> +	}
>>  
>>  	/*
>>  	 * Once the CameraConfiguration has been adjusted/validated
>> -- 
>> 2.25.1
>>
>> _______________________________________________
>> libcamera-devel mailing list
>> libcamera-devel at lists.libcamera.org
>> https://lists.libcamera.org/listinfo/libcamera-devel
> 

-- 
Regards
--
Kieran


More information about the libcamera-devel mailing list