> > WinNT Explorer crashes with an access to 07ffe0000 too. > > Well, then we should debug that and implement whatever it needs. We > are never going to have a 100% compatible user shared data anyway, so > there's no point in implementing things unless we know they are > needed.
You're right. I still like the idea of a fault handler, but I agree with you :-) Uwe, could you tell me which memory location is accessed exactly (and if possible a short assembly context) ? If WinNT Explorer reads the 32-bit word at 0x7ffe0000 (TickCout) my SafeDisc patch should fix the problem. Laurent Pinchart
