[Devel] [PATCH RHEL7 COMMIT] ms/percpu-refcount: use RCU-sched insted of normal RCU

[Konstantin Khorenko]

The commit is pushed to "branch-rh7-3.10.0-229.7.2-ovz" and will appear at https://src.openvz.org/scm/ovz/vzkernel.git
after rh7-3.10.0-229.7.2.vz7.6.3
------>
commit 932bf29b63b1e7c74669a8847d7c69cc8b8ba919
Author: Vladimir Davydov <vdavydov at parallels.com>
Date:   Fri Aug 28 14:49:25 2015 +0400

    ms/percpu-refcount: use RCU-sched insted of normal RCU
    
    Patchset description:
    
    Pulling upstream patches converting css refcnt to percpu_ref.
    
    https://jira.sw.ru/browse/PSBM-34174
    
    Kent Overstreet (2):
      percpu: implement generic percpu refcounting
      percpu-refcount: Don't use silly cmpxchg()
    
    Tejun Heo (9):
      percpu-refcount: consistently use plain (non-sched) RCU
      percpu-refcount: cosmetic updates
      percpu-refcount: add __must_check to percpu_ref_init() and don't use
        ACCESS_ONCE() in percpu_ref_kill_rcu()
      percpu-refcount: implement percpu_ref_cancel_init()
      percpu-refcount: implement percpu_tryget() along with
        percpu_ref_kill_and_confirm()
      percpu-refcount: use RCU-sched insted of normal RCU
      cgroup: reorder the operations in cgroup_destroy_locked()
      cgroup: split cgroup destruction into two steps
      cgroup: use percpu refcnt for cgroup_subsys_states
    
    ===
    This patch description:
    
    From: Tejun Heo <tj at kernel.org>
    
    percpu-refcount was incorrectly using preempt_disable/enable() for RCU
    critical sections against call_rcu().  6a24474da8 ("percpu-refcount:
    consistently use plain (non-sched) RCU") fixed it by converting the
    preepmtion operations with rcu_read_[un]lock() citing that there isn't
    any advantage in using sched-RCU over using the usual one; however,
    rcu_read_[un]lock() for the preemptible RCU implementation -
    CONFIG_TREE_PREEMPT_RCU, chosen when CONFIG_PREEMPT - are slightly
    more expensive than preempt_disable/enable().
    
    In a contrived microbench which repeats the followings,
    
     - percpu_ref_get()
     - copy 32 bytes of data into percpu buffer
     - percpu_put_get()
     - copy 32 bytes of data into percpu buffer
    
    rcu_read_[un]lock() used in percpu_ref_get/put() makes it go slower by
    about 15% when compared to using sched-RCU.
    
    As the RCU critical sections are extremely short, using sched-RCU
    shouldn't have any latency implications.  Convert to RCU-sched.
    
    Signed-off-by: Tejun Heo <tj at kernel.org>
    Acked-by: Kent Overstreet <koverstreet at google.com>
    Acked-by: "Paul E. McKenney" <paulmck at linux.vnet.ibm.com>
    Cc: Michal Hocko <mhocko at suse.cz>
    Cc: Rusty Russell <rusty at rustcorp.com.au>
    (cherry picked from commit a4244454df1296e90cc961c1b636b1176ef0d9a0)
    Signed-off-by: Vladimir Davydov <vdavydov at parallels.com>
---
 include/linux/percpu-refcount.h | 12 ++++++------
 lib/percpu-refcount.c           |  2 +-
 2 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/include/linux/percpu-refcount.h b/include/linux/percpu-refcount.h
index dd2a086..95961f0 100644
--- a/include/linux/percpu-refcount.h
+++ b/include/linux/percpu-refcount.h
@@ -105,7 +105,7 @@ static inline void percpu_ref_get(struct percpu_ref *ref)
 {
 	unsigned __percpu *pcpu_count;
 
-	rcu_read_lock();
+	rcu_read_lock_sched();
 
 	pcpu_count = ACCESS_ONCE(ref->pcpu_count);
 
@@ -114,7 +114,7 @@ static inline void percpu_ref_get(struct percpu_ref *ref)
 	else
 		atomic_inc(&ref->count);
 
-	rcu_read_unlock();
+	rcu_read_unlock_sched();
 }
 
 /**
@@ -134,7 +134,7 @@ static inline bool percpu_ref_tryget(struct percpu_ref *ref)
 	unsigned __percpu *pcpu_count;
 	int ret = false;
 
-	rcu_read_lock();
+	rcu_read_lock_sched();
 
 	pcpu_count = ACCESS_ONCE(ref->pcpu_count);
 
@@ -143,7 +143,7 @@ static inline bool percpu_ref_tryget(struct percpu_ref *ref)
 		ret = true;
 	}
 
-	rcu_read_unlock();
+	rcu_read_unlock_sched();
 
 	return ret;
 }
@@ -159,7 +159,7 @@ static inline void percpu_ref_put(struct percpu_ref *ref)
 {
 	unsigned __percpu *pcpu_count;
 
-	rcu_read_lock();
+	rcu_read_lock_sched();
 
 	pcpu_count = ACCESS_ONCE(ref->pcpu_count);
 
@@ -168,7 +168,7 @@ static inline void percpu_ref_put(struct percpu_ref *ref)
 	else if (unlikely(atomic_dec_and_test(&ref->count)))
 		ref->release(ref);
 
-	rcu_read_unlock();
+	rcu_read_unlock_sched();
 }
 
 #endif
diff --git a/lib/percpu-refcount.c b/lib/percpu-refcount.c
index 8bf9e71..7deeb62 100644
--- a/lib/percpu-refcount.c
+++ b/lib/percpu-refcount.c
@@ -154,5 +154,5 @@ void percpu_ref_kill_and_confirm(struct percpu_ref *ref,
 		(((unsigned long) ref->pcpu_count)|PCPU_REF_DEAD);
 	ref->confirm_kill = confirm_kill;
 
-	call_rcu(&ref->rcu, percpu_ref_kill_rcu);
+	call_rcu_sched(&ref->rcu, percpu_ref_kill_rcu);
 }