sched: Simple helper functions for find_busiest_group()
authorGautham R Shenoy <ego@in.ibm.com>
Wed, 25 Mar 2009 09:13:35 +0000 (14:43 +0530)
committerIngo Molnar <mingo@elte.hu>
Wed, 25 Mar 2009 09:30:44 +0000 (10:30 +0100)
Impact: cleanup

Currently the load idx calculation code is in find_busiest_group().
Move that to a static inline helper function.

Similary, to find the first cpu of a sched_group we use
cpumask_first(sched_group_cpus(group))

Use a helper to that. It improves readability in some cases.

Signed-off-by: Gautham R Shenoy <ego@in.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: "Balbir Singh" <balbir@in.ibm.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: "Dhaval Giani" <dhaval@linux.vnet.ibm.com>
Cc: Bharata B Rao <bharata@linux.vnet.ibm.com>
Cc: "Vaidyanathan Srinivasan" <svaidy@linux.vnet.ibm.com>
LKML-Reference: <20090325091335.13992.55424.stgit@sofia.in.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
kernel/sched.c

index 7b389c74f8ff904432e0e251ebe211994566fa22..6aec1e7a72a30366a9132941637594efe2f3b876 100644 (file)
@@ -3189,6 +3189,43 @@ static int move_one_task(struct rq *this_rq, int this_cpu, struct rq *busiest,
 
        return 0;
 }
+/********** Helpers for find_busiest_group ************************/
+
+/**
+ * group_first_cpu - Returns the first cpu in the cpumask of a sched_group.
+ * @group: The group whose first cpu is to be returned.
+ */
+static inline unsigned int group_first_cpu(struct sched_group *group)
+{
+       return cpumask_first(sched_group_cpus(group));
+}
+
+/**
+ * get_sd_load_idx - Obtain the load index for a given sched domain.
+ * @sd: The sched_domain whose load_idx is to be obtained.
+ * @idle: The Idle status of the CPU for whose sd load_icx is obtained.
+ */
+static inline int get_sd_load_idx(struct sched_domain *sd,
+                                       enum cpu_idle_type idle)
+{
+       int load_idx;
+
+       switch (idle) {
+       case CPU_NOT_IDLE:
+               load_idx = sd->busy_idx;
+               break;
+
+       case CPU_NEWLY_IDLE:
+               load_idx = sd->newidle_idx;
+               break;
+       default:
+               load_idx = sd->idle_idx;
+               break;
+       }
+
+       return load_idx;
+}
+/******* find_busiest_group() helpers end here *********************/
 
 /*
  * find_busiest_group finds and returns the busiest CPU group within the
@@ -3217,12 +3254,7 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
        busiest_load_per_task = busiest_nr_running = 0;
        this_load_per_task = this_nr_running = 0;
 
-       if (idle == CPU_NOT_IDLE)
-               load_idx = sd->busy_idx;
-       else if (idle == CPU_NEWLY_IDLE)
-               load_idx = sd->newidle_idx;
-       else
-               load_idx = sd->idle_idx;
+       load_idx = get_sd_load_idx(sd, idle);
 
        do {
                unsigned long load, group_capacity, max_cpu_load, min_cpu_load;
@@ -3238,7 +3270,7 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
                                               sched_group_cpus(group));
 
                if (local_group)
-                       balance_cpu = cpumask_first(sched_group_cpus(group));
+                       balance_cpu = group_first_cpu(group);
 
                /* Tally up the load of all CPUs in the group */
                sum_weighted_load = sum_nr_running = avg_load = 0;
@@ -3359,8 +3391,7 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
                 */
                if ((sum_nr_running < min_nr_running) ||
                    (sum_nr_running == min_nr_running &&
-                    cpumask_first(sched_group_cpus(group)) >
-                    cpumask_first(sched_group_cpus(group_min)))) {
+                    group_first_cpu(group) > group_first_cpu(group_min))) {
                        group_min = group;
                        min_nr_running = sum_nr_running;
                        min_load_per_task = sum_weighted_load /
@@ -3375,8 +3406,8 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
                if (sum_nr_running <= group_capacity - 1) {
                        if (sum_nr_running > leader_nr_running ||
                            (sum_nr_running == leader_nr_running &&
-                            cpumask_first(sched_group_cpus(group)) <
-                            cpumask_first(sched_group_cpus(group_leader)))) {
+                            group_first_cpu(group) <
+                            group_first_cpu(group_leader))) {
                                group_leader = group;
                                leader_nr_running = sum_nr_running;
                        }
@@ -3504,7 +3535,7 @@ out_balanced:
                *imbalance = min_load_per_task;
                if (sched_mc_power_savings >= POWERSAVINGS_BALANCE_WAKEUP) {
                        cpu_rq(this_cpu)->rd->sched_mc_preferred_wakeup_cpu =
-                               cpumask_first(sched_group_cpus(group_leader));
+                               group_first_cpu(group_leader);
                }
                return group_min;
        }