Hi Sakari, On Thu, Sep 22, 2022 at 2:02 PM Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx> wrote: > > Hi Prabhakar, > > On Thu, Sep 22, 2022 at 01:53:49PM +0100, Lad, Prabhakar wrote: > > > > +int rzg2l_csi2_cmn_rstb_deassert(struct rzg2l_csi2 *csi2); > > > > +int rzg2l_csi2_dphy_setting(struct rzg2l_csi2 *csi2, bool on); > > > > +void rzg2l_csi2_mipi_link_setting(struct rzg2l_csi2 *csi2, bool on); > > > > > > Are these something that could be achieved using the standard interfaces, > > > as I believe the other drivers are doing? The pre_streamon and > > > post_streamon callbacks could be relevant for this. > > > > > Thanks for the pointer. I have now moved some code to pre_streamon () > > and rest to s_stream(). Is there any mandatory rule to have both > > implemented? (as I wont be needing post_streamoff(), nothing complains > > so for) > > In principle no. > OK. > But if you e.g. resume the device in pre_streamon, you'll need to suspend > it in post_streamon. > Agreed! Cheers, Prabhakar
