On Tue, Apr 21, 2015 at 05:13:04PM +0300, Ander Conselvan de Oliveira wrote:
> @@ -11548,26 +11479,36 @@ intel_modeset_update_state(struct drm_device *dev, unsigned prepare_pipes)
> if (!intel_encoder->base.crtc)
> continue;
>
> - intel_crtc = to_intel_crtc(intel_encoder->base.crtc);
> + for_each_crtc_in_state(state, crtc, crtc_state, i)
> + if (crtc == intel_encoder->base.crtc)
> + break;
>
> - if (prepare_pipes & (1 << intel_crtc->pipe))
> + if (crtc != intel_encoder->base.crtc)
> + continue;
> +
> + if (crtc_state->enable && needs_modeset(crtc_state))
> intel_encoder->connectors_active = false;
This makes the compiler gasp, and even if the compiler is wrong, the
code isn't perhaps the most clear - and this pattern gets repeated.
For example,
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 067b1dee1b90..3b049cdbb7e5 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11859,15 +11859,15 @@ intel_modeset_update_state(struct drm_atomic_state *state)
if (!intel_encoder->base.crtc)
continue;
- for_each_crtc_in_state(state, crtc, crtc_state, i)
- if (crtc == intel_encoder->base.crtc)
- break;
+ for_each_crtc_in_state(state, crtc, crtc_state, i) {
+ if (crtc != intel_encoder->base.crtc)
+ continue;
- if (crtc != intel_encoder->base.crtc)
- continue;
+ if (crtc_state->enable && needs_modeset(crtc_state))
+ intel_encoder->connectors_active = false;
- if (crtc_state->enable && needs_modeset(crtc_state))
- intel_encoder->connectors_active = false;
+ break;
+ }
}
-Chris
--
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
