> > I'm 98% sure the problems are due to not banking the BIOS correctly for that > > motherboard/chipset. > > It's not a banking problem. > The behavior looks identical to the issue with Savage Quest, which was definitely an > incomplete BIOS. > However, in this case, it's that the BIOS is moving a block of memory out of the way > for a while, and then not putting it back before attempting to call it. I suppose > that one of the ports is not returning the expected value. > > > That is not my area of expertise, however > > > > ETA: Carl, whose area this definitely is, says 0f 3c is a MediaGX-specific opcode. > > The other two are 0f 78 and 0f 79.
Quoting Carl:
"0f 78 is a mediagx smm opcode. It's described on page 92 of this real mediagx doc at http://www.da-cha.jp/files/gxmdb_v20.pdf FWIW, it's called SVDC and saves the full segment register descriptor cache to memory.
0f 79 is RSDC, the reverse operation. 7a,7b,7c,7d do the same with the LDTR and TR respectively."
Edited by R. Belmont (08/17/13 05:02 AM)
|