> The other problem that was harming all people is that mmget() was
> incrasing the counter unconditionally. But it could increase it even if it
> was just 0. So we could have an mm with count 1 (increased by /proc), but
Before somebody will notice this: my fix is not the right one (probably
due lack to sleep ;). Incidentally it worked _fine_ because the race is so
fast that the slab cache had not the time to reuse the mm_struct (the one
where I hold the mm_lock even if it's just deallocated) for another task.
I am fixing it right this time. The fix is to make atomic the tsk->mm =
&init_mm and the mmput.
A real fix is coming soon...
Andrea Arcangeli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/