Em Thu, 20 Oct 2022 21:55:33 +0200 Hans de Goede <hdegoede@xxxxxxxxxx> escreveu: > Depending on which order userspace makes various v4l2 calls, the sensor > might still be powered down when set_fmt is called. > > What should really happen here is delay the writing of the mode-related > registers till streaming is started, but for now use the same quick fix > as the atomisp_ov2680 code and call power_up() from set_fmt() in > combination with keeping track of the power-state to avoid doing the > power-up sequence twice. Yes, that seems to be the right solution: to store the mode and use it during stream on. I was thinking on doing this at ov2680, but I opted not to. The rationale is that we need first to ensure that atomisp will use the standard API for the sensor drivers (this is currently not the case). As far as I recall, the try_fmt logic requires some things to come from the driver on a non-standard way, and for some, the device needs to be in power up state. It also uses it at atomisp/sensor's probe time. Regards, Mauro
