perf: Fix perf_lock_task_context() vs RCU
Peter Zijlstra
maintainerPATCH 1/1 perf: Fix perf_lock_task_context() vs RCU
Subject: perf: Fix perf_lock_task_context() vs RCU
Jiri managed to trigger:
[] ======================================================
[] [ INFO: possible circular locking dependency detected ]
[] 3.10.0+ #228 Tainted: G W
[] -------------------------------------------------------
[] p/6613 is trying to acquire lock:
[] (rcu_node_0){..-...}, at: [<ffffffff810ca797>] rcu_read_unlock_special+0xa7/0x250 []
[] but task is already holding lock:
[] (&ctx->lock){-.-...}, at: [<ffffffff810f2879>] perf_lock_task_context+0xd9/0x2c0 []
[] which lock already depends on the new lock.
[]
[] the existing dependency chain (in reverse order) is:
[]
[] -> #4 (&ctx->lock){-.-...}:
[] -> #3 (&rq->lock){-.-.-.}:
[] -> #2 (&p->pi_lock){-.-.-.}:
[] -> #1 (&rnp->nocb_gp_wq[1]){......}:
[] -> #0 (rcu_node_0){..-...}:
Paul was quick to explain that due to preemptible RCU we cannot call rcu_read_unlock() while holding scheduler (or nested) locks when part of the read side critical section was preemptible.
Therefore solve it by making the entire RCU read side non-preemptible.
Also pull out the retry from under the non-preempt to play nice with RT.
--- kernel/events/core.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
kernel/events/core.c +14 -1
@@ -947,8 +947,18 @@ perf_lock_task_context(struct task_struc {struct perf_event_context *ctx; -rcu_read_lock (); retry: +/* +* One of the few rules of preemptible RCU is that one cannot do +* rcu_read_unlock() while holding a scheduler (or nested) lock when +* part of the read side critical section was preemptible -- see +* rcu_read_unlock_special(). +* +* Since ctx->lock nests under rq->lock we must ensure the entire read +* side critical section is non-preemptible. +*/ +preempt_disable (); +rcu_read_lock (); ctx =rcu_dereference (task->perf_event_ctxp[ctxn]);if (ctx) {/*
@@ -964,6 +974,8 @@ perf_lock_task_context(struct task_strucraw_spin_lock_irqsave (&ctx->lock, *flags);if (ctx !=rcu_dereference (task->perf_event_ctxp[ctxn])) {raw_spin_unlock_irqrestore (&ctx->lock, *flags); +rcu_read_unlock (); +preempt_enable (); goto retry; }
@@ -973,6 +985,7 @@ perf_lock_task_context(struct task_struc } }rcu_read_unlock (); +preempt_enable ();return ctx; }
Quoted 28 lines · Peter Zijlstra
Subject: perf: Fix
perf_lock_task_context()vs RCUJiri managed to trigger:
[] ======================================================
[] [ INFO: possible circular locking dependency detected ]
[] 3.10.0+ #228 Tainted: G W
[] -------------------------------------------------------
[] p/6613 is trying to acquire lock:
[] (rcu_node_0){..-...}, at: [<ffffffff810ca797>]rcu_read_unlock_special+0xa7/0x250[]
[] but task is already holding lock:
[] (&ctx->lock){-.-...}, at: [<ffffffff810f2879>]perf_lock_task_context+0xd9/0x2c0[]
[] which lock already depends on the new lock.
[]
[] the existing dependency chain (in reverse order) is:
[]
[] -> #4 (&ctx->lock){-.-...}:
[] -> #3 (&rq->lock){-.-.-.}:
[] -> #2 (&p->pi_lock){-.-.-.}:
[] -> #1 (&rnp->nocb_gp_wq[1]){......}:
[] -> #0 (rcu_node_0){..-...}:Paul was quick to explain that due to preemptible RCU we cannot call
rcu_read_unlock()while holding scheduler (or nested) locks when part of the read side critical section was preemptible.
Quoted 7 lines
Therefore solve it by making the entire RCU read side non-preemptible.
Also pull out the retry from under the non-preempt to play nice with RT.
Cc Paul E. McKenney <[email protected]>Reported-by Jiri Olsa <[email protected]>Signed-off-by Peter Zijlstra <[email protected]>
Quoted 51 lines
--- kernel/events/core.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)kernel/events/core.c +14
-2@@ -947,8 +947,18 @@ perf_lock_task_context(struct task_struc {struct perf_event_context *ctx; -rcu_read_lock (); retry: +/* +* One of the few rules of preemptible RCU is that one cannot do +* rcu_read_unlock() while holding a scheduler (or nested) lock when +* part of the read side critical section was preemptible -- see +* rcu_read_unlock_special(). +* +* Since ctx->lock nests under rq->lock we must ensure the entire read +* side critical section is non-preemptible. +*/ +preempt_disable (); +rcu_read_lock (); ctx =rcu_dereference (task->perf_event_ctxp[ctxn]);if (ctx) {/* @@ -964,6 +974,8 @@ perf_lock_task_context(struct task_strucraw_spin_lock_irqsave (&ctx->lock, *flags);if (ctx !=rcu_dereference (task->perf_event_ctxp[ctxn])) {raw_spin_unlock_irqrestore (&ctx->lock, *flags); +rcu_read_unlock (); +preempt_enable (); goto retry; }@@ -973,6 +985,7 @@ perf_lock_task_context(struct task_struc } }rcu_read_unlock (); +preempt_enable ();return ctx; } --To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/