Hadn't picked up that the IEGD is now only verified on the Atom. Although it still has support for older chipsets. I thought the EMGD was for the Atom whilst the IEGD would support the 'bigger brothers'. I was obviously wrong.
We're running a system with the i7 processor on WES2009. But the only graphics driver is the full blown desktop version. I'd used IEGD before on 945 chipset so knew it got rid of a lot of the bloat of the desktop version. We're running a graphics intensive application on our system.
How long will it take for the IEGD to support the i7?
The reason we were looking at the IEGD is that the desktop driver turns off the VGA output if there isn't a monitor connected at boot up. I was hoping the IEGD would give us the option of disabling this 'feature'. We have a display on the LVDS output but the VGA output may have it's monitor plugged in at some point after system start-up. The desktop driver requires a <Ctrl><Alt><F1> key sequence to re-activate the VGA monitor output (We're in clone mode) BUT we don't have a keyboard connected to do that.
I had a customer who had the same requirement, and the IEGD does offer the ability to turn off monitor autodetect, which is a nice feature. As far as support moving forward, someone from Intel will have to advise on their plans.
There are currently no plans to support any non-Atom chipsets beyond the G45 with IEGD. I agree that the feature set provided by our embedded drivers is really good (I've been one who has been working a lot of years to get those sort of features developed). However, it is not all bad news for embedded features for later chipsets. The embedded driver for Sandybridge uses the "desktop" graphics core but there is a configuration options that allows certian embedded features to be turned on in the driver through the BMP utility. Those include multiple LVDS, EDID-less operation on typically edid displays. hybrid multi-monitor support, and additional video decode capbilities on XP. Unfortunately, the "Hot Port" feature you like is not currently implemented although there is a "hot plug" capability that will eneable the display when you plug it in but that likely does not do what you would like to do.
If you disagree with the direction we are going, the best thing you can do is to work with your Intel field rep to make sure your product needs are communicated to platform and product planners and help us to get the right capabilites in place for your needs.
I've been doing some digging around. The BMP would be used to provide a new VBIOS which we don't have the option of doing.
However we've been having a go with the Common User Interface Software Development Kit (CUI SDK) Test utility.
This has some useful functions in it. We found the GetDeviceData and SetDeviceData calls under the Display API menu quite useful.
The IGFX_CONNECTOR_STATUS_GUID_1_0 GUID can be used in a call to GetDeviceData to discover how many displays are connected. As the LCD interface always comes up as connected we can poll this function to look for a second diaply being connected. (at this point the VGA monitor still rmains blank)
Then we can use the IGFX_GET_SET_CONFIGURATION_SET GUID in a call to change the configuration and enable the VGA monitor.
We think there is a bug in the test program because no matter what opmode we set we can't setup the secondary monitor. Although we can sucessfully switch on the VGA monitor or the LCD in single display mode we can't enable both. The structure for the secondary display is un-editable even when both displays are already enabled.
For my next task i have to try and get hold of a copy of the CUI SDK. We'll need this as we're funning XP embedded. The API calls are part of Vista and Win 7 but not XP.
Questions?
