thermal: power_allocator: req_range multiplication should be a 64 bit type
authorJavi Merino <javi.merino@arm.com>
Wed, 6 Apr 2016 18:30:18 +0000 (19:30 +0100)
committerEduardo Valentin <edubezval@gmail.com>
Wed, 20 Apr 2016 23:22:45 +0000 (16:22 -0700)
req_range is declared as a u64 to cope with overflows in the
multiplication of two u32.  As both req_power and power_range are u32,
we need to make sure the multiplication is done with u64 types.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
drivers/thermal/power_allocator.c

index 1246aa6fcab0caeda03ace6806c4df2d46c1f164..2f1a863a8e15bc834e007fea264d262d180a04cf 100644 (file)
@@ -301,7 +301,7 @@ static void divvy_up_power(u32 *req_power, u32 *max_power, int num_actors,
        capped_extra_power = 0;
        extra_power = 0;
        for (i = 0; i < num_actors; i++) {
-               u64 req_range = req_power[i] * power_range;
+               u64 req_range = (u64)req_power[i] * power_range;
 
                granted_power[i] = DIV_ROUND_CLOSEST_ULL(req_range,
                                                         total_req_power);