clk: Really fix deadlock with mmap_sem

Commit 6314b67 (clk: Don't hold prepare_lock across debugfs
creation, 2014-09-04) forgot to update one place where we hold
the prepare_lock while creating debugfs directories. This means
we still have the chance of a deadlock that the commit was trying
to fix. Actually fix it by moving the debugfs creation outside
the prepare_lock.

Cc: <stable@vger.kernel.org> # 3.18
Reported-by: Russell King <rmk+kernel@arm.linux.org.uk>
Fixes: 6314b67 "clk: Don't hold prepare_lock across debugfs creation"
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Michael Turquette <mturquette@linaro.org>
[mturquette@linaro.org: removed lockdep_assert]

drivers/clk/clk.c

@@ -240,7 +240,6 @@ static const struct file_operations clk_dump_fops = {
 	.release	= single_release,
 };
 
-/* caller must hold prepare_lock */
 static int clk_debug_create_one(struct clk *clk, struct dentry *pdentry)
 {
 	struct dentry *d;
@@ -1944,7 +1943,6 @@ int __clk_init(struct device *dev, struct clk *clk)
 	else
 		clk->rate = 0;
 
-	clk_debug_register(clk);
 	/*
 	 * walk the list of orphan clocks and reparent any that are children of
 	 * this clock
@@ -1979,6 +1977,9 @@ int __clk_init(struct device *dev, struct clk *clk)
 out:
 	clk_prepare_unlock();
 
+	if (!ret)
+		clk_debug_register(clk);
+
 	return ret;
 }