Re: [PATCH v4 2/6] apple-gmux: Add switch_ddc support

From: Lukas Wunner <lukas@wunner.de>
To: Alex Deucher <alexdeucher@gmail.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>,
	DRI Development <dri-devel@lists.freedesktop.org>,
	Darren Hart <dvhart@infradead.org>
Subject: Re: [PATCH v4 2/6] apple-gmux: Add switch_ddc support
Date: Thu, 29 Oct 2015 15:58:10 +0100	[thread overview]
Message-ID: <20151029145810.GA20495@wunner.de> (raw)
In-Reply-To: <CADnq5_P6OAkHCUzksdBeSzxZxrgD6XS+HeAvDXr=01oDDhM1TA@mail.gmail.com>

Hi Alex,

On Mon, Oct 12, 2015 at 05:10:20PM -0400, Alex Deucher wrote:
> On Mon, Oct 12, 2015 at 5:07 PM, Alex Deucher <alexdeucher@gmail.com> wrote:
> > On Fri, Aug 14, 2015 at 12:18 PM, Lukas Wunner <lukas@wunner.de> wrote:
> >> Originally by Seth Forshee <seth.forshee@canonical.com>, 2012-10-04:
> >>     The gmux allows muxing the DDC independently from the display, so
> >>     support this functionality. This will allow reading the EDID for the
> >>     inactive GPU, fixing issues with machines that either don't have a
> >>     VBT or have invalid mode data in the VBT.
> >>
> >> Modified by Lukas Wunner <lukas@wunner.de>, 2015-10-07:
> >>     Advertise ->switch_ddc handler callback only on the pre-retina
> >>     Macbook Pro. The retina uses eDP and its mux is incapable of
> >>     switching the AUX channel separately from the main link.
> >>     It's an NXP CBTL06142 (alternate parts: TI HD3SS212,
> >>     Pericom PPI3VDP12412).
> >>
> >>     Sources:
> >>     http://www.electronicproducts.com/-whatsinside_text-145.aspx
> >>     http://slideshare.net/jjwu6266/apple-2012-wwdc-apple-macbook-pro-with-retina-display
> >>     http://www.techrepublic.com/blog/cracking-open/teardown-shows-retina-macbook-pro-is-nearly-impossible-to-upgrade-difficult-to-work-on/
> >>
> >>     Datasheets:
> >>     http://www.nxp.com/documents/data_sheet/CBTL06141.pdf
> >>     http://www.ti.com/lit/ds/symlink/hd3ss212.pdf
> >>     https://www.pericom.com/assets/Datasheets/PI3VDP12412.pdf
> >>
> >> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=88861
> >> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=61115
> >> Tested-by: Lukas Wunner <lukas@wunner.de>
> >>     [MBP  9,1 2012  intel IVB + nvidia GK107  pre-retina  15"]
> >>
> >> Cc: Seth Forshee <seth.forshee@canonical.com>
> >> Signed-off-by: Lukas Wunner <lukas@wunner.de>
> >> ---
> >>  drivers/platform/x86/apple-gmux.c | 21 +++++++++++++++++++--
> >>  1 file changed, 19 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/drivers/platform/x86/apple-gmux.c b/drivers/platform/x86/apple-gmux.c
> >> index 0dec3f5..78997b7 100644
> >> --- a/drivers/platform/x86/apple-gmux.c
> >> +++ b/drivers/platform/x86/apple-gmux.c
> >> @@ -276,11 +276,9 @@ static const struct backlight_ops gmux_bl_ops = {
> >>  static int gmux_switchto(enum vga_switcheroo_client_id id)
> >>  {
> >>         if (id == VGA_SWITCHEROO_IGD) {
> >> -               gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_DDC, 1);
> >>                 gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_DISPLAY, 2);
> >>                 gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_EXTERNAL, 2);
> >>         } else {
> >> -               gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_DDC, 2);
> >>                 gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_DISPLAY, 3);
> >>                 gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_EXTERNAL, 3);
> >>         }
> >
> > Won't this hunk break manual switching until the later patches land?
> > Seems like you might want to break this out as a separate patch later
> > in the series.
> 
> Also. I'm not sure how the gmux works, but this might break ddc on
> external displays that are muxed.

I had to do some experimenting and research to clarify this definitively,
hence the belated reply.

Answer is no, GMUX_PORT_SWITCH_DDC only switches DDC of the panel.

The first two generations of the MacBook Pro (MBP 5 2008/09, MBP 6 2010)
are able to switch the external DP port between GPUs, but only in its
entirety. The mux used for this (NXP CBTL06141) is not capable of
switching DDC separately. It's controlled by GMUX_PORT_SWITCH_EXTERNAL.

The following two generations (MBP 8 2011, MBP 9 2012) replaced the
external DP port with a combined DP/Thunderbolt port which can drive
either a single "dumb" DP display or two daisy-chained Thunderbolt
displays using DP-over-Thunderbolt. The ability to switch the external
port in its entirety was given up but the AUX channel is still switchable,
under the control of GMUX_PORT_SWITCH_EXTERNAL.

The port appears disabled to the Intel GPU, but I hacked i915 to
force-enable pipe B and all DP ports and was then able to retrieve
DPCD and EDID from an external display.

The retina MBP gained a second DP/Thunderbolt port and an HDMI port
and is still able to switch AUX to the integrated GPU on both DP ports!
I wonder why Apple went to these lengths. Maybe they found that HPD is
unreliable, so they let the Intel GPU poll AUX and wake up the discrete
GPU if anything is detected?

Best regards,

Lukas

> 
> Alex
> 
> >
> > Alex
> >
> >
> >> @@ -288,6 +286,18 @@ static int gmux_switchto(enum vga_switcheroo_client_id id)
> >>         return 0;
> >>  }
> >>
> >> +static int gmux_switch_ddc(enum vga_switcheroo_client_id id)
> >> +{
> >> +       pr_debug("Switching gmux DDC to %d\n", id);
> >> +
> >> +       if (id == VGA_SWITCHEROO_IGD)
> >> +               gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_DDC, 1);
> >> +       else
> >> +               gmux_write8(apple_gmux_data, GMUX_PORT_SWITCH_DDC, 2);
> >> +
> >> +       return 0;
> >> +}
> >> +
> >>  static int gmux_set_discrete_state(struct apple_gmux_data *gmux_data,
> >>                                    enum vga_switcheroo_state state)
> >>  {
> >> @@ -588,6 +598,13 @@ static int gmux_probe(struct pnp_dev *pnp, const struct pnp_device_id *id)
> >>                 gmux_data->gpe = -1;
> >>         }
> >>
> >> +       /*
> >> +        * The gmux in pre-retina MacBook Pros can switch DDC separately
> >> +        * from the other pins of the outputs. It's distinguishable from
> >> +        * the gmux in retinas by being non-indexed.
> >> +        */
> >> +       if (!gmux_data->indexed)
> >> +               gmux_handler.switch_ddc = gmux_switch_ddc;
> >>         if (vga_switcheroo_register_handler(&gmux_handler)) {
> >>                 ret = -ENODEV;
> >>                 goto err_register_handler;
> >> --
> >> 1.8.5.2 (Apple Git-48)
> >>
> >> _______________________________________________
> >> dri-devel mailing list
> >> dri-devel@lists.freedesktop.org
> >> http://lists.freedesktop.org/mailman/listinfo/dri-devel
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel