>Eg if we only allow a process 512MB of VA space, we can map each new
>pmap (sorry, a BSDism although I don't work on that any more either) at
>say 4KB + (512 * (i%6)) or something like that. Even with 1GB per proc
>you'd have enough room for 3...
Yes, I've thought about this in the past. In fact by putting each task in a
different domain you could avoid even having to flush the TLB. I think it
would always be a bit of a specialist interest and it's probably not suitable
for "mainstream" kernels but it's certainly an idea worth pursuing if someone
has the time and inclination. Compiling everything as PIC would obviously
cost a small amount but it's not the end of the world.
p.
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
