I'll reply to that more detailed later on (it's early morning over here in Germany and I want to check a few things first).
SLP Activation doesn't check the DMI tables, it searches through a specific memory address range looking for a particular string.
Even if SLP activation doesn't check the DMI tables itself, the manufacturer string for sure derives from the DMI table data.
Otherwise I wouldn't have been successfull in patching the string in the DMI table and keep the pre-activation working on atotally different hardware leaving OEMBIOS files untouched.
In my opinion this is a more than sufficient practical proof.
To verify this (theoretically) we have to ask how and from where the WMI does collect it's infos.
With the function you use, you retrieve the "Manufacturer-" and the "Model-String".
This "Model-String" is non existent in the DMI tables, but instead I discovered it in the ACPI APIC table ("OEM TABLE ID
") - where you (vice versa) won't find a "Manufacturer-String".
Even though the "Model-String" is (yet) of no interest for the pre-activation (but it will become important in concern of Trusted Computing
), it leads me to the conclusion that the WMI data is collected from different sources (DMI table, ACPI APIC table and so on) and compiled to the WMI database.
So my intension is to find out what data sources should be examined and how the chances are to modify them.
Some thoughts on "Trusted Computing":
Edited by mittern8, 23 May 2006 - 10:43 PM.