Re: [PATCH 2/6] drm/i915/dp: Correctly report eDP in the core connector type

From: Jesse Barnes <jbarnes@virtuousgeek.org>
To: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 2/6] drm/i915/dp: Correctly report eDP in the core connector type
Date: Sun, 3 Oct 2010 08:04:58 -0700	[thread overview]
Message-ID: <20101003080458.0e6d2679@jbarnes-x201> (raw)
In-Reply-To: <20100902153726.GB22015@zhen-devel.sh.intel.com>

On Thu, 2 Sep 2010 23:37:26 +0800
Zhenyu Wang <zhenyuw@linux.intel.com> wrote:

> On 2010.07.16 14:46:28 -0400, Adam Jackson wrote:
> > +	if (output_reg == DP_A || IS_PCH_eDP(dp_priv)) {
> > +		type = DRM_MODE_CONNECTOR_eDP;
> > +		intel_encoder->type = INTEL_OUTPUT_EDP;
> > +	} else {
> > +		type = DRM_MODE_CONNECTOR_DisplayPort;
> > +		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
> > +	}
> > +
> 
> Adam, I believe this one causes regression on PCH eDP port as
> for bug 27220. It changes behavior in mode_set path to program
> FDI link properly. 
> 
> I think above part should be reverted for .36 kernel. What's your
> idea? 
> 
> Thanks Yakui to remind me about this recent change.

Did PCH eDP ever work?  I'll try going back to a kernel before this
change, but current code definitely seems broken (we get never try to
get a real clock value for PCH eDP, and I don't think we program the
transcoder regs at all).

Even with those fixed though (among other things), I haven't been able
to get my Vaio with switchable graphics working.  Which platforms did
you have working, and what changeset is most likely to work?

Thanks,
-- 
Jesse Barnes, Intel Open Source Technology Center