On 1/5/14, Miles Fidelman <mfidel...@meetinghouse.net> wrote: > If you have a memory hole, chances are there wouldn't be a BIOS option > to help. It has to do with the way that memory-mapped i/o is handled > with some families of chips and their associated motherboards. > > For example, I have two older servers - fully loaded with 4Gig of RAM > (as I said, older servers) - but no way, no how does the system see more > than 3G. The other 1G is taken up by memory mapped i/o space. It's a > hardware design issue, not a BIOS issue. (For reference: P4 640 > processor, Supermicro P8SCT motherboard). > > In addition to the reference I sent earlier, this sort of describes the > issue: http://en.wikipedia.org/wiki/3_GB_barrier > and this: http://www.dansdata.com/askdan00015.htm > If you do some googling, you'll also find an Intel design note that's > the definitive description of the issue (I can't seem to find it right now) > > Some BIOSs support "memory hole remapping," but others don't - and that > assumes the underlying chipset and motherboard will support it. A lot > don't. > > Of course this might not be the problem you're seeing. What CPU, > chipset, motherboard, and BIOS are you running? >
Thanks! CPU is P4/2.9G, motherboard is 848P-M7 it's quite like 848P-M Deluxe: http://www.ecs.com.cn/ECSWebSite/Product/Product_Detail.aspx?CategoryID=1&DetailID=402&DetailName=Feature&MenuID=24&LanID=0 actually I might try to update BIOS from site above model numbers are the same, it's not recommended to update but it seems I have no other choice if motherboard become unusable, my trouble also ends I can't fully understand technical details you describe the point I want to repeat is the machine can run Windows XP and in memtest it's OK after configuration Linux fail here, (or does it?) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/cab-gxza3viblvtsrknchrffblixkcevohpyoknk3podokyf...@mail.gmail.com