"Stefan Tauner" wrote :
Someone with more knowledge about FWH than me has to look at it. It would certainly help if we would know which flash chip is really inside it...
I'll strip-search that machine as soon as possible ;=)
@ "Andrew Goodbody" :
I'm pretty sure that the detection of FWH devices requires writing to the address space used and you cannot do that as you cannot set the BIOS WE bit in the chipset. So unless you can get around the SMI protection of that bit then there is no way to detect the chip in use. Even if you did detect it, you still could not program it.
I'll check whether the BIOS also has locked access to SMRAM - usually it wasn't done at the time. If the SMRAM is accessible from outside SMM, it would be straightforward to bypass the protection (just replace an RSM instruction as the SMI "handler" ;-) Even otherwise, it might be possible to shunt BIOS initialisation by capturing execution after a hot CPU reset (not just init) to the CPU - although processor and chipset dependent, I don't know if it will be possible with this rig.
Happy new year to all !