sched: Use CPUPRI_NR_PRIORITIES instead of MAX_RT_PRIO in cpupri check
authorSteven Rostedt (Red Hat) <rostedt@goodmis.org>
Sun, 13 Apr 2014 13:34:53 +0000 (09:34 -0400)
committerIngo Molnar <mingo@kernel.org>
Wed, 7 May 2014 09:51:33 +0000 (11:51 +0200)
The check at the beginning of cpupri_find() makes sure that the task_pri
variable does not exceed the cp->pri_to_cpu array length. But that length
is CPUPRI_NR_PRIORITIES not MAX_RT_PRIO, where it will miss the last two
priorities in that array.

As task_pri is computed from convert_prio() which should never be bigger
than CPUPRI_NR_PRIORITIES, if the check should cause a panic if it is
hit.

Reported-by: Mike Galbraith <umgwanakikbuti@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/1397015410.5212.13.camel@marge.simpson.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kernel/sched/cpupri.c

index 8b836b376d9129760066326eabf5040f72b2e4f3..3031bac8aa3ea990bc7425e835675c7a5a386ab5 100644 (file)
@@ -70,8 +70,7 @@ int cpupri_find(struct cpupri *cp, struct task_struct *p,
        int idx = 0;
        int task_pri = convert_prio(p->prio);
 
-       if (task_pri >= MAX_RT_PRIO)
-               return 0;
+       BUG_ON(task_pri >= CPUPRI_NR_PRIORITIES);
 
        for (idx = 0; idx < task_pri; idx++) {
                struct cpupri_vec *vec  = &cp->pri_to_cpu[idx];