Skip to contentSkip to conversation

IB/core: use RCU for uverbs id lookup

PATCH In review Mike Marciniszyn 2 participants 3 messages Nov 2, 2015 – Nov 24, 2015 2 messages not in the archive InfiniBand

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 IB/core: use RCU for uverbs id lookup

The current implementation gets a spin_lock, and at any scale with qib and hfi1 post send, the lock contention grows exponentially with the number of QPs.

idr_find() is RCU compatibile, so read doesn't need the lock.

Change to use rcu_read_lock() and rcu_read_unlock() in __idr_get_uobj().

kfree_rcu() is used to insure a grace period between the idr removal and actual free.

---
 drivers/infiniband/core/uverbs_cmd.c |   12 +++++++-----
 include/rdma/ib_verbs.h              |    1 +
 2 files changed, 8 insertions(+), 5 deletions(-)
drivers/infiniband/core/uverbs_cmd.c +7 -5
index be4cb9f..eb7f74f 100644
@@ -62,9 +62,11 @@ static struct uverbs_lock_class rule_lock_class = { .name = "RULE-uobj" };
  * The ib_uobject locking scheme is as follows:
  *
  * - ib_uverbs_idr_lock protects the uverbs idrs themselves, so it
- *   needs to be held during all idr operations.  When an object is
+ *   needs to be held during all idr write operations.  When an object is
  *   looked up, a reference must be taken on the object's kref before
- *   dropping this lock.
+ *   dropping this lock.  For read operations, the rcu_read_lock()
+ *   and rcu_write_lock() but similarly the kref reference is grabbed
+ *   before the rcu_read_unlock().
  *
  * - Each object also has an rwsem.  This rwsem must be held for
  *   reading while an operation that uses the object is performed.
@@ -96,7 +98,7 @@ static void init_uobj(struct ib_uobject *uobj, u64 user_handle,
 
 static void release_uobj(struct kref *kref)
 {
-	kfree(container_of(kref, struct ib_uobject, ref));
+	kfree_rcu(container_of(kref, struct ib_uobject, ref), rcu);
 }
 
 static void put_uobj(struct ib_uobject *uobj)
@@ -145,7 +147,7 @@ static struct ib_uobject *__idr_get_uobj(struct idr *idr, int id,
 {
 	struct ib_uobject *uobj;
 
-	spin_lock(&ib_uverbs_idr_lock);
+	rcu_read_lock();
 	uobj = idr_find(idr, id);
 	if (uobj) {
 		if (uobj->context == context)
@@ -153,7 +155,7 @@ static struct ib_uobject *__idr_get_uobj(struct idr *idr, int id,
 		else
 			uobj = NULL;
 	}
-	spin_unlock(&ib_uverbs_idr_lock);
+	rcu_read_unlock();
 
 	return uobj;
 }
include/rdma/ib_verbs.h +1 -0
index 7845fae..f46092f 100644
@@ -1246,6 +1246,7 @@ struct ib_uobject {
 	int			id;		/* index into kernel idr */
 	struct kref		ref;
 	struct rw_semaphore	mutex;		/* protects .live */
+	struct rcu_head		rcu;		/* kfree_rcu() overhead */
 	int			live;
 };
 

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

A message that is not in the archive (sent off-list or not yet received): 20151102171325.7601.96477.stgit-K+u1se/DcYrLESAwzcoQNrvm/[email protected]

Reviewed-by
#2 lore raw
Quoted 11 lines · Mike Marciniszyn

The current implementation gets a spin_lock, and at any scale with qib and hfi1 post send, the lock contention grows exponentially with the number of QPs.

idr_find() is RCU compatibile, so read doesn't need the lock.

Change to use rcu_read_lock() and rcu_read_unlock() in __idr_get_uobj().

kfree_rcu() is used to insure a grace period between the idr removal and actual free.

Looks OK to me.

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

A message that is not in the archive (sent off-list or not yet received): 20151102174056.GA13982-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org

#4 lore raw
Quoted 10 lines · Mike Marciniszyn
Quoted 9 lines

The current implementation gets a spin_lock, and at any scale with qib and hfi1 post send, the lock contention grows exponentially with the number of QPs.

idr_find() is RCU compatibile, so read doesn't need the lock.

Change to use rcu_read_lock() and rcu_read_unlock() in __idr_get_uobj().

Quoted 6 lines
Quoted 2 lines

kfree_rcu() is used to insure a grace period between the idr removal and actual free.

Looks OK to me.

Doug,

What is the status of this patch?

I don't see it any of the public repos.

Mike
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html