The return value of the get_temp function is not checked when doing a
thermal zone update.  This may lead to a critical shutdown if get_temp
fails and the content of the temp variable is incorrectly set higher than
the critical trip point.
This has been observed on a system with incorrect ACPI implementation
where the corresponding methods were not serialized and therefore
sometimes triggered ACPI errors (AE_ALREADY_EXISTS).  The following
critical shutdowns indicated a temperature of 2097 C, which was obviously
wrong.
The patch adds a return value check that jumps over all trip point
evaluations printing a warning if get_temp fails.  The trip points are
evaluated again on the next polling interval with successful get_temp
execution.
Signed-off-by: Michael Brunner <mibru@gmx.de>
Acked-by: Zhang Rui <rui.zhang@intel.com>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
 
        mutex_lock(&tz->lock);
 
-       tz->ops->get_temp(tz, &temp);
+       if (tz->ops->get_temp(tz, &temp)) {
+               /* get_temp failed - retry it later */
+               printk(KERN_WARNING PREFIX "failed to read out thermal zone "
+                      "%d\n", tz->id);
+               goto leave;
+       }
 
        for (count = 0; count < tz->trips; count++) {
                tz->ops->get_trip_type(tz, count, &trip_type);
                                            THERMAL_TRIPS_NONE);
 
        tz->last_temperature = temp;
+
+      leave:
        if (tz->passive)
                thermal_zone_device_set_polling(tz, tz->passive_delay);
        else if (tz->polling_delay)