OSDev.org

My O.S. doesn't uses virtual memory, but I would like to use the pageing hardware on the x86 to give each application one long spread of memory (only within physical limitaions of course). I understand the mechanics of this (thanks to the intel manuals), but there is one thing I'am getting stumped on.
My O.S. I uses call gates to enter the O.S. would I have to turn off pageing as I don't want my O.S. to be part of the pageing mechanisum.
The O.S. resides in the first 1mb, with the rest being set aside for application use.

Thanks for any help

Jason wrote: My O.S. I uses call gates to enter the O.S. would I have to turn off pageing as I don't want my O.S. to be part of the pageing mechanisum.
The O.S. resides in the first 1mb, with the rest being set aside for application use.

You could "identity map" the first 1 Mb of memory, so that the first 1 Mb of every address space is the same as physical memory. That way your call gates will still work.

Cheers,

Brendan

Thanks for the help