From mboxrd@z Thu Jan  1 00:00:00 1970
From: davem@davemloft.net (David Miller)
Date: Mon, 21 Sep 2015 12:01:59 -0700 (PDT)
Subject: [PATCH 1/7] phy: fix of_mdio_find_bus() device refcount leak
In-Reply-To: <E1ZcsNb-0001Bf-U5@rmk-PC.arm.linux.org.uk>
References: <20150918094625.GB21084@n2100.arm.linux.org.uk>
 <E1ZcsNb-0001Bf-U5@rmk-PC.arm.linux.org.uk>
Message-ID: <20150921.120159.326170798236851436.davem@davemloft.net>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

From: Russell King <rmk+kernel@arm.linux.org.uk>
Date: Fri, 18 Sep 2015 10:54:55 +0100

> Update the comment, and arrange for the only user of this function
> to drop this refcount when disposing of a reference to it.

mdio_mux is not the only user of of_mdio_find_bus(), DSA uses it as
well.

So if anything this commit message is inaccurate.

I also wonder about this refcounting scheme.

If you are going to drop the inner device reference, then we take the
mdio bus returned from of_mdio_find_bus() what holds onto it and keeps
it from disappearing on us?

Don't we have to hold onto some reference count of some kind here?