Philip Blundell writes:
> The root cause is an assumption in copy_page_range that pagetable structures
> are naturally aligned. Because 26-bit machines have fake pagetables that
> aren't used by the hardware, this isn't necessarily true. The logic that
> works out how large an area is by masking address bits was getting confused.
There's nothing wrong with this assumption. Keep looking at the memory allocation
in the A5k and you'll find some more bits to clean up which will reduce kernel
memory usage dramatically (do we need 64k for the kernel stack and task_struct?)
You may want to try using the small_page stuff to allocate an 8k block from a 32k
page, for example.
_____
|_____| ------------------------------------------------- ---+---+-
| | Russell King [EMAIL PROTECTED] --- ---
| | | | http://www.arm.linux.org.uk/~rmk/armlinux.html / / |
| +-+-+ --- -+-
/ | THE developer of ARM Linux |+| /|\
/ | | | --- |
+-+-+ ------------------------------------------------- /\\\ |
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
