printk: Fix rq->lock vs logbuf_lock unlock lock inversion

commit dbda92d1 upstream.

commit 07354eb1 ("locking printk: Annotate logbuf_lock as raw")
reintroduced a lock inversion problem which was fixed in commit
0b5e1c52 ("printk: Release console_sem after logbuf_lock"). This
happened probably when fixing up patch rejects.

Restore the ordering and unlock logbuf_lock before releasing
console_sem.
Signed-off-by: ybu <ybu@qti.qualcomm.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/E807E903FE6CBE4D95E420FBFCC273B827413C@nasanexd01h.na.qualcomm.comSigned-off-by: Thomas Gleixner <tglx@linutronix.de>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Cc: Qiang Huang <h.huangqiang@huawei.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Jianguo Wu <wujianguo@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

kernel/printk.c

@@ -822,9 +822,9 @@ static int console_trylock_for_printk(unsigned int cpu)
 		}
 	}
 	printk_cpu = UINT_MAX;
+	raw_spin_unlock(&logbuf_lock);
 	if (wake)
 		up(&console_sem);
-	raw_spin_unlock(&logbuf_lock);
 	return retval;
 }
 static const char recursion_bug_msg [] =