[PATCH tip/core/rcu 46/55] rcu: Remove rcu_needs_cpu_flush() to avoid false quiescent states

From: Paul E. McKenney
Date: Wed Sep 07 2011 - 02:25:53 EST


The purpose of rcu_needs_cpu_flush() was to iterate on pushing the
current grace period in order to help the current CPU enter dyntick-idle
mode. However, this can result in failures if the CPU starts entering
dyntick-idle mode, but then backs out. In this case, the call to
rcu_pending() from rcu_needs_cpu_flush() might end up announcing a
non-existing quiescent state.

This commit therefore removes rcu_needs_cpu_flush() in favor of letting
the dyntick-idle machinery at the end of the softirq handler push the
loop along via its call to rcu_pending().

Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
---
kernel/rcutree.c | 3 ---
kernel/rcutree.h | 1 -
kernel/rcutree_plugin.h | 25 -------------------------
3 files changed, 0 insertions(+), 29 deletions(-)

diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index 743a658..ab37c19 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -1528,9 +1528,6 @@ static void rcu_process_callbacks(struct softirq_action *unused)
&__get_cpu_var(rcu_sched_data));
__rcu_process_callbacks(&rcu_bh_state, &__get_cpu_var(rcu_bh_data));
rcu_preempt_process_callbacks();
-
- /* If we are last CPU on way to dyntick-idle mode, accelerate it. */
- rcu_needs_cpu_flush();
trace_rcu_utilization("End RCU core");
}

diff --git a/kernel/rcutree.h b/kernel/rcutree.h
index f509f72..849ce9e 100644
--- a/kernel/rcutree.h
+++ b/kernel/rcutree.h
@@ -458,7 +458,6 @@ static int rcu_preempt_needs_cpu(int cpu);
static void __cpuinit rcu_preempt_init_percpu_data(int cpu);
static void rcu_preempt_send_cbs_to_online(void);
static void __init __rcu_init_preempt(void);
-static void rcu_needs_cpu_flush(void);
static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags);
static void rcu_preempt_boost_start_gp(struct rcu_node *rnp);
static void invoke_rcu_callbacks_kthread(void);
diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h
index f751a2d..e4fb1ff 100644
--- a/kernel/rcutree_plugin.h
+++ b/kernel/rcutree_plugin.h
@@ -1949,15 +1949,6 @@ int rcu_needs_cpu(int cpu)
return rcu_needs_cpu_quick_check(cpu);
}

-/*
- * Check to see if we need to continue a callback-flush operations to
- * allow the last CPU to enter dyntick-idle mode. But fast dyntick-idle
- * entry is not configured, so we never do need to.
- */
-static void rcu_needs_cpu_flush(void)
-{
-}
-
#else /* #if !defined(CONFIG_RCU_FAST_NO_HZ) */

#define RCU_NEEDS_CPU_FLUSHES 5
@@ -2033,20 +2024,4 @@ int rcu_needs_cpu(int cpu)
return c;
}

-/*
- * Check to see if we need to continue a callback-flush operations to
- * allow the last CPU to enter dyntick-idle mode.
- */
-static void rcu_needs_cpu_flush(void)
-{
- int cpu = smp_processor_id();
- unsigned long flags;
-
- if (per_cpu(rcu_dyntick_drain, cpu) <= 0)
- return;
- local_irq_save(flags);
- (void)rcu_needs_cpu(cpu);
- local_irq_restore(flags);
-}
-
#endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */
--
1.7.3.2

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/