Subject: rcu: Make ksoftirqd do RCU quiescent states From: "Paul E. McKenney" Date: Wed, 5 Oct 2011 11:45:18 -0700 Implementing RCU-bh in terms of RCU-preempt makes the system vulnerable to network-based denial-of-service attacks. This patch therefore makes __do_softirq() invoke rcu_bh_qs(), but only when __do_softirq() is running in ksoftirqd context. A wrapper layer in interposed so that other calls to __do_softirq() avoid invoking rcu_bh_qs(). The underlying function __do_softirq_common() does the actual work. The reason that rcu_bh_qs() is bad in these non-ksoftirqd contexts is that there might be a local_bh_enable() inside an RCU-preempt read-side critical section. This local_bh_enable() can invoke __do_softirq() directly, so if __do_softirq() were to invoke rcu_bh_qs() (which just calls rcu_preempt_qs() in the PREEMPT_RT_FULL case), there would be an illegal RCU-preempt quiescent state in the middle of an RCU-preempt read-side critical section. Therefore, quiescent states can only happen in cases where __do_softirq() is invoked directly from ksoftirqd. Signed-off-by: Paul E. McKenney Link: http://lkml.kernel.org/r/20111005184518.GA21601@linux.vnet.ibm.com Signed-off-by: Thomas Gleixner --- include/linux/rcupdate.h | 4 ---- kernel/rcu/tree.c | 9 ++++++++- kernel/rcu/tree_plugin.h | 8 +++++++- 3 files changed, 15 insertions(+), 6 deletions(-) --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -343,11 +343,7 @@ static inline int rcu_preempt_depth(void /* Internal to kernel */ void rcu_init(void); void rcu_sched_qs(void); -#ifdef CONFIG_PREEMPT_RT_FULL -static inline void rcu_bh_qs(void) { } -#else void rcu_bh_qs(void); -#endif void rcu_check_callbacks(int user); void rcu_report_dead(unsigned int cpu); --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -259,7 +259,14 @@ void rcu_sched_qs(void) this_cpu_ptr(&rcu_sched_data), true); } -#ifndef CONFIG_PREEMPT_RT_FULL +#ifdef CONFIG_PREEMPT_RT_FULL +static void rcu_preempt_qs(void); + +void rcu_bh_qs(void) +{ + rcu_preempt_qs(); +} +#else void rcu_bh_qs(void) { if (__this_cpu_read(rcu_bh_data.cpu_no_qs.s)) { --- a/kernel/rcu/tree_plugin.h +++ b/kernel/rcu/tree_plugin.h @@ -28,6 +28,7 @@ #include #include #include +#include #include "../time/tick-internal.h" #ifdef CONFIG_RCU_BOOST @@ -1244,7 +1245,7 @@ static void rcu_prepare_kthreads(int cpu #endif /* #else #ifdef CONFIG_RCU_BOOST */ -#if !defined(CONFIG_RCU_FAST_NO_HZ) +#if !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) /* * Check to see if any future RCU-related work will need to be done @@ -1261,7 +1262,9 @@ int rcu_needs_cpu(u64 basemono, u64 *nex return IS_ENABLED(CONFIG_RCU_NOCB_CPU_ALL) ? 0 : rcu_cpu_has_callbacks(NULL); } +#endif /* !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) */ +#if !defined(CONFIG_RCU_FAST_NO_HZ) /* * Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up * after it. @@ -1357,6 +1360,8 @@ static bool __maybe_unused rcu_try_advan return cbs_ready; } +#ifndef CONFIG_PREEMPT_RT_FULL + /* * Allow the CPU to enter dyntick-idle mode unless it has callbacks ready * to invoke. If the CPU has callbacks, try to advance them. Tell the @@ -1402,6 +1407,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nex *nextevt = basemono + dj * TICK_NSEC; return 0; } +#endif /* #ifndef CONFIG_PREEMPT_RT_FULL */ /* * Prepare a CPU for idle from an RCU perspective. The first major task