[Intel-gfx] H55 chipset and KMS with HDMI/VGA

[milk]

On Thu, Oct 7, 2010 at 1:09 AM, Chris Wilson <chris at chris-wilson.co.uk> wrote:
> Seems like you've been bitten by two separate bugs. The first is that we
> always enable audio over HDMI even through a dongle and to a device that
> is not audio capable. That has the effect of making the output green/pink
> or even preventing sync'ing entirely. The second is that for some reason
> we have not been able to get VGA 100% reliable on Ironlake, which is
> itself very peculiar.
>
> To fix the first:
> diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi
> index 9fb9501..084a987 100644
> --- a/drivers/gpu/drm/i915/intel_hdmi.c
> +++ b/drivers/gpu/drm/i915/intel_hdmi.c
> @@ -73,7 +73,7 @@ static void intel_hdmi_mode_set(struct drm_encoder *encoder,
>                sdvox |= SDVO_HSYNC_ACTIVE_HIGH;
>
>        if (intel_hdmi->has_hdmi_sink) {
> -               sdvox |= SDVO_AUDIO_ENABLE;
> +               //sdvox |= SDVO_AUDIO_ENABLE;
>                if (HAS_PCH_CPT(dev))
>                        sdvox |= HDMI_MODE_SELECT;
>        }
>
> and we hope to have a complete patch shortly. (Zhenyu did you get a chance
> to update the patches?)

I rebuilt the kernel using the arch linux x86_64 config and the patch,
but the color issue remains for the DVI monitor using an HDMI->DVI
adapter.

-milki

-- 
Hope is a dimension of the spirit. It is not outside us, but within
us. When you lose it, you must seek it again within yourself and in
people around you -- not in objects or even events.    -
-Vaclav Havel