> This one we could use. For some reason we hardcode it to
> 1 now, which is correct for our use cases but I don't really> see a reason to > hardcode it here. We are supposed to calculate> it correctly after all, and > chv_crtc_clock_get() also just blindly> reads it out.> > > bestm2_frac = > crtc_state->dpll.m2 & 0x3fffff;> > - bestm1 = crtc_state->dpll.m1;> > This > one is a bit trickier since I don't think the spec even> gives us other > values. But we could assert that it's correct.> > Some something along these > lines I think would be best:> + drm_WARN_ON(&dev_priv->drm, bestm1 != 2);> > vlv_dpio_write(dev_priv, pipe, CHV_PLL_DW1(port),> > DPIO_CHV_M1_DIV_BY_2 |> - 1 << DPIO_CHV_N_DIV_SHIFT);> + > bestn << DPIO_CHV_N_DIV_SHIFT); Thanks for your kind response ! Does that means the variable will be used by the hardware？if so as far as I see it, I don't seem to see the relevant interface.
