Re: [PATCH v6 4/4] media: camss: sm8250: Pipeline starting and stopping for multiple virtual channels

[Bryan O'Donoghue <bryan.odonoghue@linaro.org>]

On 05/12/2022 16:49, Milen Mitkov (Consultant) wrote:
> 
> On 05/12/2022 18:43, Bryan O'Donoghue wrote:
>> On 05/12/2022 15:24, quic_mmitkov@quicinc.com wrote:
>>> media-ctl -v -d /dev/media0 -V '"imx577 
>>> '22-001a'":0[fmt:SRGGB10/3840x2160 field:none]'
>>
>> I really like the improved commit log, thank you for that.
>>
>> SRGGB10/3840x2160 drivers/media/i2c/imx412.c that's not a supported 
>> resolution.
>>
>> media-ctl -v -d /dev/media0 -V '"imx577 
>> '22-001a'":0[fmt:SRGGB10/4056x3040 field:none]'
>>
>> ?
>>
>> ---
>> bod
> 
> 
> Hi Bryan,
> 
> it's not a supported resolution with the current imx412/577 driver in 
> upstream. We tested with a different driver (with proprietary Sony 
> registers)
> 
> that uses this resolution that generates 2 multiplexed streams. It would 
> be impossible for pure upstream solution to test this driver with imx412 
> right now.
> 
> I couldn't have used the upstream supported resolution either, because 
> it's of a sensor mode with only 1 stream and that would create the 
> impression that anybody can just take this imx412 sensor and use the 
> upstream driver to test the solution, but this is in fact not possible 
> without register changes to the sensor driver itself.
> 
> Regards,
> 
> Milen
> 

OK fair enough, I think you should make a more explicit statement about 
the mapping of VC to RDI

Here's the real bit of salient information for whomever is trying to get 
VCs working in the future

media-ctl -l '"msm_csid0":1->"msm_vfe0_rdi0":0[1]'
media-ctl -l '"msm_csid0":2->"msm_vfe0_rdi1":0[1]'

The first VC msm_csid0":1 always maps to RDI0 the second VC always maps 
to RDI1

I think we should document that mapping clearly so that you don't have 
to read or understand the code to know

media-ctl -l '"msm_csid0":1->"msm_vfe0_rdi1":0[1]' is an invalid mapping 
and why that is.

---
bod