sched: convert nohz_cpu_mask to cpumask_var_t.

Impact: (future) size reduction for large NR_CPUS.

Dynamically allocating cpumasks (when CONFIG_CPUMASK_OFFSTACK) saves
space for small nr_cpu_ids but big CONFIG_NR_CPUS.  cpumask_var_t
is just a struct cpumask for !CONFIG_CPUMASK_OFFSTACK.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/include/linux/sched.h b/include/linux/sched.h
index c5be6c6..1e33e2c 100644 (file)
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -249,7 +249,7 @@ extern void init_idle_bootup_task(struct task_struct *idle);
 extern int runqueue_is_locked(void);
 extern void task_rq_unlock_wait(struct task_struct *p);
 
-extern cpumask_t nohz_cpu_mask;
+extern cpumask_var_t nohz_cpu_mask;
 #if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ)
 extern int select_nohz_load_balancer(int cpu);
 #else

diff --git a/kernel/rcuclassic.c b/kernel/rcuclassic.c
index e503a00..c03ca3e 100644 (file)
--- a/kernel/rcuclassic.c
+++ b/kernel/rcuclassic.c
@@ -393,7 +393,7 @@ static void rcu_start_batch(struct rcu_ctrlblk *rcp)
                 * unnecessarily.
                 */
                smp_mb();
-               cpus_andnot(rcp->cpumask, cpu_online_map, nohz_cpu_mask);
+               cpumask_andnot(&rcp->cpumask, cpu_online_mask, nohz_cpu_mask);
 
                rcp->signaled = 0;
        }

diff --git a/kernel/sched.c b/kernel/sched.c
index 6b9606a..2723d7a 100644 (file)
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -5870,9 +5870,9 @@ void __cpuinit init_idle(struct task_struct *idle, int cpu)
  * indicates which cpus entered this state. This is used
  * in the rcu update to wait only for active cpus. For system
  * which do not switch off the HZ timer nohz_cpu_mask should
- * always be CPU_MASK_NONE.
+ * always be CPU_BITS_NONE.
  */
-cpumask_t nohz_cpu_mask = CPU_MASK_NONE;
+cpumask_var_t nohz_cpu_mask;
 
 /*
  * Increase the granularity value when there are more CPUs,
@@ -8274,6 +8274,9 @@ void __init sched_init(void)
         */
        current->sched_class = &fair_sched_class;
 
+       /* Allocate the nohz_cpu_mask if CONFIG_CPUMASK_OFFSTACK */
+       alloc_bootmem_cpumask_var(&nohz_cpu_mask);
+
        scheduler_running = 1;
 }
 

diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 342fc9c..70f872c 100644 (file)
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -144,7 +144,7 @@ void tick_nohz_update_jiffies(void)
        if (!ts->tick_stopped)
                return;
 
-       cpu_clear(cpu, nohz_cpu_mask);
+       cpumask_clear_cpu(cpu, nohz_cpu_mask);
        now = ktime_get();
        ts->idle_waketime = now;
 
@@ -283,7 +283,7 @@ void tick_nohz_stop_sched_tick(int inidle)
        if ((long)delta_jiffies >= 1) {
 
                if (delta_jiffies > 1)
-                       cpu_set(cpu, nohz_cpu_mask);
+                       cpumask_set_cpu(cpu, nohz_cpu_mask);
                /*
                 * nohz_stop_sched_tick can be called several times before
                 * the nohz_restart_sched_tick is called. This happens when
@@ -296,7 +296,7 @@ void tick_nohz_stop_sched_tick(int inidle)
                                /*
                                 * sched tick not stopped!
                                 */
-                               cpu_clear(cpu, nohz_cpu_mask);
+                               cpumask_clear_cpu(cpu, nohz_cpu_mask);
                                goto out;
                        }
 
@@ -354,7 +354,7 @@ void tick_nohz_stop_sched_tick(int inidle)
                 * softirq.
                 */
                tick_do_update_jiffies64(ktime_get());
-               cpu_clear(cpu, nohz_cpu_mask);
+               cpumask_clear_cpu(cpu, nohz_cpu_mask);
        }
        raise_softirq_irqoff(TIMER_SOFTIRQ);
 out:
@@ -432,7 +432,7 @@ void tick_nohz_restart_sched_tick(void)
        select_nohz_load_balancer(0);
        now = ktime_get();
        tick_do_update_jiffies64(now);
-       cpu_clear(cpu, nohz_cpu_mask);
+       cpumask_clear_cpu(cpu, nohz_cpu_mask);
 
        /*
         * We stopped the tick in idle. Update process times would miss the