Re: > 1GB physical on x86: Summary so far

Matthias Urlichs (smurf@nova.noris.de)
3 Jul 1997 21:41:44 +0200

"Larry M. Augustin" <lma@varesearch.com> writes:
>
> However, Ingo Molvar suggests
> that 4GB physical should be possible with a minor change to
> PAGE_OFFSET. I don't understand this. Is he suggesting that a
> 2GB/2GB user/kernel address space split as a possible solution, thus
> decreasing the user space virtual memory and providing more address
> space for the kernel to map physical pages?
>
You can't have 4 GB of kernel memory because the kernel relies on being
able to map the virtual address space of the current process into kernel
space too.

The memory map is rather simple: Below PAGE_OFFSET => virtual address space
of user process. Above PAGE_OFFSET => real memory. The kernel also needs
some space above its real memory space for vmalloc(), I/O remapping, and
possibly some others I've conveniently forgotten. ;-)

A 2 GB/2 GB split should be possible. You'd probably have to do some work
on the code used to set up the page tables, though -- I didn't check.

On the other hand, I'd be surprised to see a mainboard that supports even
1 GByte of memory reliably... most max out at 512 these days, and their
performance sucks above the 64-MB threshold if you rely on your vendor's
claim that the tag RAM they installed is actually big enough. :-(

-- 
Don't start something you would be afraid to see finished.

-- 
Matthias Urlichs         \  noris network GmbH  /  Xlink-POP Nürnberg 
Schleiermacherstraße 12   \   Linux+Internet   /   EMail: urlichs@noris.de
90491 Nürnberg (Germany)   \    Consulting+Programming+Networking+etc'ing
   PGP: 1024/4F578875   1B 89 E2 1C 43 EA 80 44  15 D2 29 CF C6 C7 E0 DE
       Click <A HREF="http://info.noris.de/~smurf/finger">here</A>.    42