Driver Injection

Umm just mount the wim and then copy the inf files to Windows\Inf folder

　

1) Create a folder on your workstation's C: drive (call it PE). Create a subfolder called Mount.

2) Copy the WIM from the server to your C:\PE folder.

3) From a CMD window run the following command: C:\Program Files\Windows AIK\Tools\x86>imagex /mountrw c:\pe\boot.wim 1 c:\pe\mount

4) Next, make sure you have the drivers you want to integrate unpacked and local on your workstation.

5) Run the following command to integrate those drivers:

C:\Program Files\Windows AIK\Tools\PETools>peimg /inf=C:\Driver\driver.inf c:\pe\mount <<= replace Driver\driver.inf with the absolute path to your INF.

6) Make sure the command shows successful then run the following command:

C:\Program Files\Windows AIK\Tools\x86>imagex /unmount /commit c:\pe\boot

7) If that's successful then you can copy back the WIM to the server and restart WDS (you may need to stop WDS before you can copy it back.

peimg /inf=C:\Driver\driver.inf c:\pe\mount

As ( bad ) luck would have it, guide after guide (even direct from Microsoft) have you load the driver using:
•imagex /mountrw yourwimpath\boot.wim 1 yourmountpath\yourmountfolder


The problem is that there are two images in the boot.wim, and you need to select the second one to inject the drivers into:
•imagex /mountrw yourwimpath\boot.wim 2 yourmountpath\yourmountfolder

After some searching, I have found references to this on the Internet, and even on Microsoft’s website. It all depends on which documentation you are following. If you get the right one, you are fine. If not, well, you go slowly insane as you try over and over to inject new drivers, and try it again.


Bottom line: if you are creating a custom Windows PE image, then you do want to mount image 1. If you are using WDS to deploy an OS, it is image 2 that needs the drivers.