[Arm-netbook] EOMA-68 passthrough implementation

lkcl luke luke.leighton at gmail.com
Wed Jul 11 19:47:56 BST 2012


On Wed, Jul 11, 2012 at 6:47 PM, Henrik Nordström
<henrik at henriknordstrom.net> wrote:

>> > The HDMI receiver chip that converts HDMI to LCD TTL signalling.
>
>> It sounds like you're saying the passthrough has to handle it.
>
> Depends on how easy you want it's usage to be. You could require that
> something external controls the passthrough card to configure it proper.
> It's mainly the LCD controller that needs configuration.

 yeah.  moving this more towards an "LCD Monitor" style discussion,
which is where the passthrough card originated from, some of these
HDMI ICs which are used inside LCD Monitors have those kinds of
configuration capabilities, and i gather some even have a basic CPU
on-board.  i mean - they have to.  how else will they do the on-screen
menus for brightness etc.?

 normally those ICs you'd re-program them or add an external ROM with
the LCD configuration options, job done.

 unfortunately for those ICs the use-case here is rather different:
the LCD's configuration could change.  in *some* cases that data could
be read over the LCD's I2C port (the EDID data).  however, that
pre-supposes that the LCD manufacturer a) bothered to add an EDID ROM
on the LCD panel b) bothered to make it available over the I2C port in
the standard fashion c) actually put the correct data on it.

 so this is why there has to be an I2C EEPROM as part of the EOMA-68
spec.  we haven't *actually* got round to specifying what's on it yet
though :)  this was discussed some time back, and it's likely to be up
to 17 sets of device-tree data.  1 for the LCD and up to 16 for the
GPIO pins.  heck, possibly even more.

 anyway: all of this is moot as far as a "simple" passthrough card is concerned.

 l.



More information about the arm-netbook mailing list