Re: 1 GIG limit...
> On Sun, Aug 01, 1999 at 04:35:28PM +0200, Marcus Brinkmann wrote:
> > Nobody asked until now. Now that someone asked, Roland or Thomas will say
> > something about it, I am sure. However, this is even more useful if there is
> > a volunteer to write and test the changes, too. Information is dead if it
> > isn't made use of.
>
> AFAICT, the main problem is that libdiskfs mmap()s the whole partition
> in memory.
My guess, without looking at the actual code, is that it's because of the
limits of the data typed used.
If that were the case (which sounds much too good to be true), the
solution could be to just address the partition via a vector of pointers
to 1 gig maps. this is feasible since partitions -usually- don't change
size while they are mounted.
although i suppose it could confuse operations across the boundaries a
bit. i think i'll go look at the code before i say anything dumber :)
--Michael Bacarella
Reply to: