On Fri, Sep 24, 2021 at 01:48:51PM +0300, Jani Nikula wrote: > On Thu, 23 Sep 2021, Ville Syrjala <ville.syrjala@xxxxxxxxxxxxxxx> wrote: > > From: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > > > > Let's start to reject user modes whose refresh rate is > > nowhere near the actual fixed mode refresh rate we're > > going to use. ATM we're just flat out lying to the user. > > I eyeballed through this quickly, mostly seems reasonable, I'll try to > do detailed review later. > > One question though. I think we have bug reports [1][2] about panels > that support very high refresh rates, but report a lower refresh rate > mode as the preferred mode. It's perhaps a safe default from a power > usage standpoint. Does this series make using those modes harder or > impossible? This doesn't really affect the amount of work we'd need to do to expose those modes. That would still involve replacing the single fixed mode with a list/etc. and looking for the best match for the user's requested mode during a modeset. I'd actually like to do that at some point, if for no other reason allowing me to run at a fixed lower refresh rate while on battery power. Could save a few mW. But haven't gotten bored enough to do the actual work. -- Ville Syrjälä Intel
