On Sun, Jun 9, 2013 at 7:23 PM, Chris Wilson <chris at chris-wilson.co.uk> wrote:
> Reviewing the dotclock computation from the refclk and dividers is a
> very useful step in manually checking the register configuration.
>
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> ---
> drivers/gpu/drm/i915/intel_display.c | 23 +++++++++++++++++++++++
> 1 file changed, 23 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index e2e2173..1333a56 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -4844,6 +4844,19 @@ static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
> POSTING_READ(PIPECONF(intel_crtc->pipe));
> }
>
> +static u64 compute_dotclock(int refclk, const intel_clock_t *clock)
> +{
> + u64 dotclock;
> +
> + dotclock = refclk;
> + dotclock *= 5 * (clock->m1 + 2) + clock->m2 + 2;
> + do_div(dotclock, 2+clock->n);
> + do_div(dotclock, clock->p1);
> + do_div(dotclock, clock->p2);
> +
> + return dotclock;
> +}
Why is intel_clock (and pineview_clock fwiw) not good enough? We
should probably shovel them into a vfunc somewhere, but I kinda want
to hold off with that until we've wrestled the haswell wrpll clocks a
bit more into submission. Then we could integrate this into the hw
state readout and pipe config dump code (which would kill a fastboot
special case as a nice side effect).
Cheers, Daniel
> +
> static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
> int x, int y,
> struct drm_framebuffer *fb)
> @@ -4888,6 +4901,10 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
> DRM_ERROR("Couldn't find PLL settings for mode!\n");
> return -EINVAL;
> }
> + DRM_DEBUG_KMS("refclk=%d, clock=(.m1=%d, .m2=%d, n=%d, .p1=%d, .p2=%d), computed dotclock = %lld\n",
> + refclk, clock.m1, clock.m2, clock.n, clock.p1, clock.p2,
> + compute_dotclock(refclk, &clock));
> +
>
> /* Ensure that the cursor is valid for the new mode before changing... */
> intel_crtc_update_cursor(crtc, true);
> @@ -5736,6 +5753,12 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
> DRM_ERROR("Couldn't find PLL settings for mode!\n");
> return -EINVAL;
> }
> +
> + DRM_DEBUG_KMS("refclk=%d, clock=(.m1=%d, .m2=%d, n=%d, .p1=%d, .p2=%d), computed dotclock = %lld\n",
> + ironlake_get_refclk(crtc),
> + clock.m1, clock.m2, clock.n, clock.p1, clock.p2,
> + compute_dotclock(ironlake_get_refclk(crtc), &clock));
> +
> /* Compat-code for transition, will disappear. */
> if (!intel_crtc->config.clock_set) {
> intel_crtc->config.dpll.n = clock.n;
> --
> 1.7.10.4
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
