okay -- it seems that things are working (a little bit).
on http://www.rootshell.be/~nwaibel/docprobe.txt you can find the output of a modified version of docprobe.
you can see the plain mem map of the md2802-d08 there. it is different to the md2800-d08!!!
0x0000: 512byte IPL -- is someone able to rev eng this? 0x01ff
0x0200: 512byte mirror of IPL 0x03ff
0x0400: 512byte mirror of IPL 0x05ff
0x0600: 512byte mirror of IPL 0x07ff
0x0800: 512byte IPL -- this seems to be the first 512byte of page 0 of the 8mbyte flash memory 0x09ff
0x0a00: 512byte -- not sure but could be 2nd part of page 0 of the 8mbyte 0x0bff
0x0c00: 512byte failsave IPL? -- not sure but could be 3rd part of page 0 of the 8mbyte 0x0dff
0x0e00: 512byte -- not sure but could be 4th part of page 0 of the 8mbyte 0x0fff
0x1000: the doc 2000 and doc millennium (2001) control registers are here. but it seems that this is not true for the md2802-d08 chip! everything is set to 0x00 -> docprobe does *NOT* recognize this chip as 0x11ff DoC Millennium!!!
0x1200: hmmm -- control registers? very strange. it is like IPL but (almost) all even bytes are set to 0x00 (not all!) maybe the even bytes are the control regs? 0x13ff
0x1400: no idea -- 512 empty bytes 0x15ff
0x1600: like 0x1200, but first 3 bytes are different! control registers? 0x17ff
0x1800: expected 512byte mirror of IPL here -- but it is a mirror of 0x1200! 0x19ff
0x1a00: 512byte mirror of IPL 0x1bff
0x1c00: expected 512byte mirror of IPL here -- but it is a mirror of 0x1200! 0x1dff
0x1e00: 512byte mirror of IPL 0x1fff
any ideas or more info about this md2802-d08?
niki
BTW: the epia-m does not boot with epia-m + bios savior + dip2plcc + doc (md2802-d08). even not if bios savior is switched to the normal flash containing the bios. even not if there is only epia-m + bios savior + dip2plcc. so i booted with epia-m + bios savior and then (when linux was running) i placed the dip2plcc + doc (md2802-d08) into bios savior and changed the switch.