Use synchronize_sched before calling call_rcu in CONFIG_PREEMPT_RCU until we
have call_rcu_sched and rcu_barrier_sched in mainline. It will slow down the
marker operations in CONFIG_PREEMPT_RCU, but it fixes the current race against
the preempt_disable/enable() protected code paths.

Paul, is this ok ? It would be good to get this in for 2.6.25 final.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
CC: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/marker.c |    9 +++++++++
 1 file changed, 9 insertions(+)

Index: linux-2.6-lttng/kernel/marker.c
===================================================================
--- linux-2.6-lttng.orig/kernel/marker.c	2008-03-31 08:22:24.000000000 -0400
+++ linux-2.6-lttng/kernel/marker.c	2008-03-31 08:43:30.000000000 -0400
@@ -671,6 +671,9 @@ int marker_probe_register(const char *na
 	entry->rcu_pending = 1;
 	/* write rcu_pending before calling the RCU callback */
 	smp_wmb();
+#ifdef CONFIG_PREEMPT_RCU
+	synchronize_sched();	/* Until we have the call_rcu_sched() */
+#endif
 	call_rcu(&entry->rcu, free_old_closure);
 end:
 	mutex_unlock(&markers_mutex);
@@ -714,6 +717,9 @@ int marker_probe_unregister(const char *
 	entry->rcu_pending = 1;
 	/* write rcu_pending before calling the RCU callback */
 	smp_wmb();
+#ifdef CONFIG_PREEMPT_RCU
+	synchronize_sched();	/* Until we have the call_rcu_sched() */
+#endif
 	call_rcu(&entry->rcu, free_old_closure);
 	remove_marker(name);	/* Ignore busy error message */
 	ret = 0;
@@ -792,6 +798,9 @@ int marker_probe_unregister_private_data
 	entry->rcu_pending = 1;
 	/* write rcu_pending before calling the RCU callback */
 	smp_wmb();
+#ifdef CONFIG_PREEMPT_RCU
+	synchronize_sched();	/* Until we have the call_rcu_sched() */
+#endif
 	call_rcu(&entry->rcu, free_old_closure);
 	remove_marker(entry->name);	/* Ignore busy error message */
 end:
-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 ...
[ message continues ]

On Mon, 31 Mar 2008 09:16:09 -0400


Paul seems to have nodded off.  I'll merge it.
--

[ message continues ]

Paul seems to have fat-fingered the earlier message.  But he does see the
message adding this to -mm, and the patch looks good to me.

							Thanx, Paul
--

[ message continues ]

Hi Andrew,

Right, this could be appended to the changelog then :

Markers do not mix well with CONFIG_PREEMPT_RCU because it uses
preempt_disable/enable() and not rcu_read_lock/unlock for minimal
intrusiveness. We would need call_sched and sched_barrier primitives.

Currently, the modification (connection and disconnection) of probes
from markers requires changes to the data structure done in RCU-style :
a new data structure is created, the pointer is changed atomically, a
quiescent state is reached and then the old data structure is freed.

The quiescent state is reached once all the currently running
preempt_disable regions are done running. We use the call_rcu mechanism
to execute kfree() after such quiescent state has been reached. However,
the new CONFIG_PREEMPT_RCU version of call_rcu and rcu_barrier does not
guarantee that all preempt_disable code regions have finished, hence the
race.

The "proper" way to do this is to use rcu_read_lock/unlock, but we don't
want to use it to minimize intrusiveness on the traced system. (we do
not want the marker code to call into much of the OS code, because it
would quickly restrict what can and cannot be instrumented, such as the
scheduler).

The temporary fix, until we get call_rcu_sched and rcu_barrier_sched in
mainline, is to use synchronize_sched before each call_rcu calls, so we
wait for the quiescent state in the system call code path. It will slow
down batch marker enable/disable, but will make sure the race is gone.

Thanks,


-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
--

[ message continues ]