Found by the new Stanford locking checker.

Minimal fix for a deadlock in sysvfs: get_branch() can take
read_lock(&pointers_lock), but one caller already has a write_lock.

Perhaps some of the "oh we raced, drop everything and try again" logic in
there can go away now, but this is enugh to fix the obvious deadlock.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>