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 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68 --
| Arjan van de Ven | [Patch v2] Make PCI extended config space (MMCONFIG) a driver opt-in |
| Tilman Schmidt | git guidance |
| Vu Pham | Re: [Scst-devel] Integration of SCST in the mainstream Linux kernel |
| Greg KH | [GIT PATCH] driver core patches against 2.6.24 |
git: | |
| David Miller | Re: Git and GCC |
| Mike | I don't want the .git directory next to my code. |
| Steffen Prohaska | merge vs rebase: Is visualization in gitk the only problem? |
| David Kastrup | What is the idea for bare repositories? |
| Richard Stallman | Real men don't attack straw men |
| GVG GVG | ssh_exchange_identification: Connection closed by remote host |
| Falk Brockerhoff | ftp-proxy and no route to host issue |
| Pieter Verberne | Remove escape characters from file |
| Chuck Lever | Re: [bug?] tg3: Failed to load firmware "tigon/tg3_tso.bin" |
| David Miller | Re: [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
| Stefan Richter | Re: [GIT]: Networking |
| jamal | Re: [LARTC] ifb and ppp |
