On Fri, Mar 10, 2017 at 06:33:28PM +0000, Matthew Wilcox wrote:

From cd5401d81633d5e48e39d67d4e65156e6759537e Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <redacted>
Date: Fri, 10 Mar 2017 13:22:53 -0500
Subject: [PATCH] thermal: Fix potential deadlock in cpu_cooling

I expanded the scope of cooling_list_lock a little too far; it was
not just covering cpufreq_dev_count, it was also covering the calls
to cpufreq_register_notifier() and cpufreq_unregister_notifier().
Since cooling_list_lock is also used within cpufreq_thermal_notifier(),
lockdep reports a potential deadlock.  I don't think that's actually
possible, but it's easy enough to make it impossible by testing the
condition under cooling_list_lock and dropping the lock before calling
cpufreq_register_notifier().

As a bonus, I noticed that cpufreq_dev_count is only used for the purpose
of knowing whether this is the first or last cooling device registered,
and we know that anyway because we know whether the list transitioned
between empty and not-empty.  So we can delete that variable too.

Fixes: ae606089621ef0349402cfcbeca33a82abbd0fd0
Reported-by: Russell King <linux@armlinux.org.uk>
Signed-off-by: Matthew Wilcox <redacted>

Thanks Matthew, appears to solve the problem.

Tested-by: Russell King <redacted>

-- 
RMK's Patch system: http://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.