[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20200131151540.041600199@infradead.org>
Date: Fri, 31 Jan 2020 16:07:06 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: peterz@...radead.org, mingo@...nel.org, will@...nel.org
Cc: oleg@...hat.com, tglx@...utronix.de, linux-kernel@...r.kernel.org,
bigeasy@...utronix.de, juri.lelli@...hat.com, williams@...hat.com,
bristot@...hat.com, longman@...hat.com, dave@...olabs.net,
jack@...e.com
Subject: [PATCH -v2 3/7] locking/percpu-rwsem: Move __this_cpu_inc() into the slowpath
As preparation to rework __percpu_down_read() move the
__this_cpu_inc() into it.
Signed-off-by: Peter Zijlstra (Intel) <peterz@...radead.org>
Tested-by: Juri Lelli <juri.lelli@...hat.com>
---
include/linux/percpu-rwsem.h | 10 ++++++----
kernel/locking/percpu-rwsem.c | 2 ++
2 files changed, 8 insertions(+), 4 deletions(-)
--- a/include/linux/percpu-rwsem.h
+++ b/include/linux/percpu-rwsem.h
@@ -59,8 +59,9 @@ static inline void percpu_down_read(stru
* and that once the synchronize_rcu() is done, the writer will see
* anything we did within this RCU-sched read-size critical section.
*/
- __this_cpu_inc(*sem->read_count);
- if (unlikely(!rcu_sync_is_idle(&sem->rss)))
+ if (likely(rcu_sync_is_idle(&sem->rss)))
+ __this_cpu_inc(*sem->read_count);
+ else
__percpu_down_read(sem, false); /* Unconditional memory barrier */
/*
* The preempt_enable() prevents the compiler from
@@ -77,8 +78,9 @@ static inline bool percpu_down_read_tryl
/*
* Same as in percpu_down_read().
*/
- __this_cpu_inc(*sem->read_count);
- if (unlikely(!rcu_sync_is_idle(&sem->rss)))
+ if (likely(rcu_sync_is_idle(&sem->rss)))
+ __this_cpu_inc(*sem->read_count);
+ else
ret = __percpu_down_read(sem, true); /* Unconditional memory barrier */
preempt_enable();
/*
--- a/kernel/locking/percpu-rwsem.c
+++ b/kernel/locking/percpu-rwsem.c
@@ -47,6 +47,8 @@ EXPORT_SYMBOL_GPL(percpu_free_rwsem);
bool __percpu_down_read(struct percpu_rw_semaphore *sem, bool try)
{
+ __this_cpu_inc(*sem->read_count);
+
/*
* Due to having preemption disabled the decrement happens on
* the same CPU as the increment, avoiding the
Powered by blists - more mailing lists