Andreas Dilger wrote:
>
> Am I misunderstanding what lock_kernel() does, or would calling this
> function twice from within the same syscall cause the system to lock up
> because you are trying to get the same lock twice?
>
Just check the implementation of lock_kernel():
it only acquires the lock if the current thread doesn't own it already.
lock_kernel() , cli() and read_lock() are recursive, the rest is not.
Btw,
* could someone add a "remove before 2.4" marker to the BUG statement in
unlock_kernel()?
* why it the asm version of lock_kernel()/unlock_kernel() currently
disabled?
--
Manfred
