rcu/context-tracking: Move RCU-dynticks internal functions to context…

…_tracking

Move the core RCU eqs/dynticks functions to context tracking so that
we can later merge all that code within context tracking.

Acked-by: Paul E. McKenney <[email protected]>
Signed-off-by: Frederic Weisbecker <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Neeraj Upadhyay <[email protected]>
Cc: Uladzislau Rezki <[email protected]>
Cc: Joel Fernandes <[email protected]>
Cc: Boqun Feng <[email protected]>
Cc: Nicolas Saenz Julienne <[email protected]>
Cc: Marcelo Tosatti <[email protected]>
Cc: Xiongfeng Wang <[email protected]>
Cc: Yu Liao <[email protected]>
Cc: Phil Auld <[email protected]>
Cc: Paul Gortmaker<[email protected]>
Cc: Alex Belits <[email protected]>
Signed-off-by: Paul E. McKenney <[email protected]>
Reviewed-by: Nicolas Saenz Julienne <[email protected]>
Tested-by: Nicolas Saenz Julienne <[email protected]>

include/linux/context_tracking.h

@@ -122,6 +122,26 @@ static inline void context_tracking_init(void) { }
 #ifdef CONFIG_CONTEXT_TRACKING_IDLE
 extern void ct_idle_enter(void);
 extern void ct_idle_exit(void);
+
+/*
+ * Is the current CPU in an extended quiescent state?
+ *
+ * No ordering, as we are sampling CPU-local information.
+ */
+static __always_inline bool rcu_dynticks_curr_cpu_in_eqs(void)
+{
+	return !(arch_atomic_read(this_cpu_ptr(&context_tracking.dynticks)) & 0x1);
+}
+
+/*
+ * Increment the current CPU's context_tracking structure's ->dynticks field
+ * with ordering.  Return the new value.
+ */
+static __always_inline unsigned long rcu_dynticks_inc(int incby)
+{
+	return arch_atomic_add_return(incby, this_cpu_ptr(&context_tracking.dynticks));
+}
+
 #else
 static inline void ct_idle_enter(void) { }
 static inline void ct_idle_exit(void) { }

include/linux/rcutree.h

@@ -55,6 +55,9 @@ void rcu_irq_exit_check_preempt(void);
 static inline void rcu_irq_exit_check_preempt(void) { }
 #endif
 
+struct task_struct;
+void rcu_preempt_deferred_qs(struct task_struct *t);
+
 void exit_rcu(void);
 
 void rcu_scheduler_starting(void);

kernel/context_tracking.c

@@ -21,6 +21,7 @@
 #include <linux/hardirq.h>
 #include <linux/export.h>
 #include <linux/kprobes.h>
+#include <trace/events/rcu.h>
 
 
 DEFINE_PER_CPU(struct context_tracking, context_tracking) = {
@@ -33,6 +34,309 @@ DEFINE_PER_CPU(struct context_tracking, context_tracking) = {
 EXPORT_SYMBOL_GPL(context_tracking);
 
 #ifdef CONFIG_CONTEXT_TRACKING_IDLE
+#define TPS(x)  tracepoint_string(x)
+
+/* Record the current task on dyntick-idle entry. */
+static __always_inline void rcu_dynticks_task_enter(void)
+{
+#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL)
+	WRITE_ONCE(current->rcu_tasks_idle_cpu, smp_processor_id());
+#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */
+}
+
+/* Record no current task on dyntick-idle exit. */
+static __always_inline void rcu_dynticks_task_exit(void)
+{
+#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL)
+	WRITE_ONCE(current->rcu_tasks_idle_cpu, -1);
+#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */
+}
+
+/* Turn on heavyweight RCU tasks trace readers on idle/user entry. */
+static __always_inline void rcu_dynticks_task_trace_enter(void)
+{
+#ifdef CONFIG_TASKS_TRACE_RCU
+	if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB))
+		current->trc_reader_special.b.need_mb = true;
+#endif /* #ifdef CONFIG_TASKS_TRACE_RCU */
+}
+
+/* Turn off heavyweight RCU tasks trace readers on idle/user exit. */
+static __always_inline void rcu_dynticks_task_trace_exit(void)
+{
+#ifdef CONFIG_TASKS_TRACE_RCU
+	if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB))
+		current->trc_reader_special.b.need_mb = false;
+#endif /* #ifdef CONFIG_TASKS_TRACE_RCU */
+}
+
+/*
+ * Record entry into an extended quiescent state.  This is only to be
+ * called when not already in an extended quiescent state, that is,
+ * RCU is watching prior to the call to this function and is no longer
+ * watching upon return.
+ */
+static noinstr void rcu_dynticks_eqs_enter(void)
+{
+	int seq;
+
+	/*
+	 * CPUs seeing atomic_add_return() must see prior RCU read-side
+	 * critical sections, and we also must force ordering with the
+	 * next idle sojourn.
+	 */
+	rcu_dynticks_task_trace_enter();  // Before ->dynticks update!
+	seq = rcu_dynticks_inc(1);
+	// RCU is no longer watching.  Better be in extended quiescent state!
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && (seq & 0x1));
+}
+
+/*
+ * Record exit from an extended quiescent state.  This is only to be
+ * called from an extended quiescent state, that is, RCU is not watching
+ * prior to the call to this function and is watching upon return.
+ */
+static noinstr void rcu_dynticks_eqs_exit(void)
+{
+	int seq;
+
+	/*
+	 * CPUs seeing atomic_add_return() must see prior idle sojourns,
+	 * and we also must force ordering with the next RCU read-side
+	 * critical section.
+	 */
+	seq = rcu_dynticks_inc(1);
+	// RCU is now watching.  Better not be in an extended quiescent state!
+	rcu_dynticks_task_trace_exit();  // After ->dynticks update!
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !(seq & 0x1));
+}
+
+/*
+ * Enter an RCU extended quiescent state, which can be either the
+ * idle loop or adaptive-tickless usermode execution.
+ *
+ * We crowbar the ->dynticks_nmi_nesting field to zero to allow for
+ * the possibility of usermode upcalls having messed up our count
+ * of interrupt nesting level during the prior busy period.
+ */
+static void noinstr rcu_eqs_enter(bool user)
+{
+	struct context_tracking *ct = this_cpu_ptr(&context_tracking);
+
+	WARN_ON_ONCE(ct_dynticks_nmi_nesting() != DYNTICK_IRQ_NONIDLE);
+	WRITE_ONCE(ct->dynticks_nmi_nesting, 0);
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) &&
+		     ct_dynticks_nesting() == 0);
+	if (ct_dynticks_nesting() != 1) {
+		// RCU will still be watching, so just do accounting and leave.
+		ct->dynticks_nesting--;
+		return;
+	}
+
+	instrumentation_begin();
+	lockdep_assert_irqs_disabled();
+	trace_rcu_dyntick(TPS("Start"), ct_dynticks_nesting(), 0, ct_dynticks());
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current));
+	rcu_preempt_deferred_qs(current);
+
+	// instrumentation for the noinstr rcu_dynticks_eqs_enter()
+	instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks));
+
+	instrumentation_end();
+	WRITE_ONCE(ct->dynticks_nesting, 0); /* Avoid irq-access tearing. */
+	// RCU is watching here ...
+	rcu_dynticks_eqs_enter();
+	// ... but is no longer watching here.
+	rcu_dynticks_task_enter();
+}
+
+/*
+ * Exit an RCU extended quiescent state, which can be either the
+ * idle loop or adaptive-tickless usermode execution.
+ *
+ * We crowbar the ->dynticks_nmi_nesting field to DYNTICK_IRQ_NONIDLE to
+ * allow for the possibility of usermode upcalls messing up our count of
+ * interrupt nesting level during the busy period that is just now starting.
+ */
+static void noinstr rcu_eqs_exit(bool user)
+{
+	struct context_tracking *ct = this_cpu_ptr(&context_tracking);
+	long oldval;
+
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled());
+	oldval = ct_dynticks_nesting();
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0);
+	if (oldval) {
+		// RCU was already watching, so just do accounting and leave.
+		ct->dynticks_nesting++;
+		return;
+	}
+	rcu_dynticks_task_exit();
+	// RCU is not watching here ...
+	rcu_dynticks_eqs_exit();
+	// ... but is watching here.
+	instrumentation_begin();
+
+	// instrumentation for the noinstr rcu_dynticks_eqs_exit()
+	instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks));
+
+	trace_rcu_dyntick(TPS("End"), ct_dynticks_nesting(), 1, ct_dynticks());
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current));
+	WRITE_ONCE(ct->dynticks_nesting, 1);
+	WARN_ON_ONCE(ct_dynticks_nmi_nesting());
+	WRITE_ONCE(ct->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE);
+	instrumentation_end();
+}
+
+/**
+ * rcu_nmi_exit - inform RCU of exit from NMI context
+ *
+ * If we are returning from the outermost NMI handler that interrupted an
+ * RCU-idle period, update ct->dynticks and ct->dynticks_nmi_nesting
+ * to let the RCU grace-period handling know that the CPU is back to
+ * being RCU-idle.
+ *
+ * If you add or remove a call to rcu_nmi_exit(), be sure to test
+ * with CONFIG_RCU_EQS_DEBUG=y.
+ */
+void noinstr rcu_nmi_exit(void)
+{
+	struct context_tracking *ct = this_cpu_ptr(&context_tracking);
+
+	instrumentation_begin();
+	/*
+	 * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks.
+	 * (We are exiting an NMI handler, so RCU better be paying attention
+	 * to us!)
+	 */
+	WARN_ON_ONCE(ct_dynticks_nmi_nesting() <= 0);
+	WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs());
+
+	/*
+	 * If the nesting level is not 1, the CPU wasn't RCU-idle, so
+	 * leave it in non-RCU-idle state.
+	 */
+	if (ct_dynticks_nmi_nesting() != 1) {
+		trace_rcu_dyntick(TPS("--="), ct_dynticks_nmi_nesting(), ct_dynticks_nmi_nesting() - 2,
+				  ct_dynticks());
+		WRITE_ONCE(ct->dynticks_nmi_nesting, /* No store tearing. */
+			   ct_dynticks_nmi_nesting() - 2);
+		instrumentation_end();
+		return;
+	}
+
+	/* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */
+	trace_rcu_dyntick(TPS("Startirq"), ct_dynticks_nmi_nesting(), 0, ct_dynticks());
+	WRITE_ONCE(ct->dynticks_nmi_nesting, 0); /* Avoid store tearing. */
+
+	// instrumentation for the noinstr rcu_dynticks_eqs_enter()
+	instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks));
+	instrumentation_end();
+
+	// RCU is watching here ...
+	rcu_dynticks_eqs_enter();
+	// ... but is no longer watching here.
+
+	if (!in_nmi())
+		rcu_dynticks_task_enter();
+}
+
+/**
+ * rcu_nmi_enter - inform RCU of entry to NMI context
+ *
+ * If the CPU was idle from RCU's viewpoint, update ct->dynticks and
+ * ct->dynticks_nmi_nesting to let the RCU grace-period handling know
+ * that the CPU is active.  This implementation permits nested NMIs, as
+ * long as the nesting level does not overflow an int.  (You will probably
+ * run out of stack space first.)
+ *
+ * If you add or remove a call to rcu_nmi_enter(), be sure to test
+ * with CONFIG_RCU_EQS_DEBUG=y.
+ */
+void noinstr rcu_nmi_enter(void)
+{
+	long incby = 2;
+	struct context_tracking *ct = this_cpu_ptr(&context_tracking);
+
+	/* Complain about underflow. */
+	WARN_ON_ONCE(ct_dynticks_nmi_nesting() < 0);
+
+	/*
+	 * If idle from RCU viewpoint, atomically increment ->dynticks
+	 * to mark non-idle and increment ->dynticks_nmi_nesting by one.
+	 * Otherwise, increment ->dynticks_nmi_nesting by two.  This means
+	 * if ->dynticks_nmi_nesting is equal to one, we are guaranteed
+	 * to be in the outermost NMI handler that interrupted an RCU-idle
+	 * period (observation due to Andy Lutomirski).
+	 */
+	if (rcu_dynticks_curr_cpu_in_eqs()) {
+
+		if (!in_nmi())
+			rcu_dynticks_task_exit();
+
+		// RCU is not watching here ...
+		rcu_dynticks_eqs_exit();
+		// ... but is watching here.
+
+		instrumentation_begin();
+		// instrumentation for the noinstr rcu_dynticks_curr_cpu_in_eqs()
+		instrument_atomic_read(&ct->dynticks, sizeof(ct->dynticks));
+		// instrumentation for the noinstr rcu_dynticks_eqs_exit()
+		instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks));
+
+		incby = 1;
+	} else if (!in_nmi()) {
+		instrumentation_begin();
+		rcu_irq_enter_check_tick();
+	} else  {
+		instrumentation_begin();
+	}
+
+	trace_rcu_dyntick(incby == 1 ? TPS("Endirq") : TPS("++="),
+			  ct_dynticks_nmi_nesting(),
+			  ct_dynticks_nmi_nesting() + incby, ct_dynticks());
+	instrumentation_end();
+	WRITE_ONCE(ct->dynticks_nmi_nesting, /* Prevent store tearing. */
+		   ct_dynticks_nmi_nesting() + incby);
+	barrier();
+}
+
+/**
+ * rcu_idle_enter - inform RCU that current CPU is entering idle
+ *
+ * Enter idle mode, in other words, -leave- the mode in which RCU
+ * read-side critical sections can occur.  (Though RCU read-side
+ * critical sections can occur in irq handlers in idle, a possibility
+ * handled by irq_enter() and irq_exit().)
+ *
+ * If you add or remove a call to rcu_idle_enter(), be sure to test with
+ * CONFIG_RCU_EQS_DEBUG=y.
+ */
+void noinstr rcu_idle_enter(void)
+{
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled());
+	rcu_eqs_enter(false);
+}
+
+/**
+ * rcu_idle_exit - inform RCU that current CPU is leaving idle
+ *
+ * Exit idle mode, in other words, -enter- the mode in which RCU
+ * read-side critical sections can occur.
+ *
+ * If you add or remove a call to rcu_idle_exit(), be sure to test with
+ * CONFIG_RCU_EQS_DEBUG=y.
+ */
+void noinstr rcu_idle_exit(void)
+{
+	unsigned long flags;
+
+	raw_local_irq_save(flags);
+	rcu_eqs_exit(false);
+	raw_local_irq_restore(flags);
+}
+EXPORT_SYMBOL_GPL(rcu_idle_exit);
+
 noinstr void ct_idle_enter(void)
 {
 	rcu_idle_enter();
@@ -139,6 +443,38 @@ noinstr void ct_nmi_exit(void)
 }
 #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */
 
+#ifdef CONFIG_NO_HZ_FULL
+/**
+ * rcu_user_enter - inform RCU that we are resuming userspace.
+ *
+ * Enter RCU idle mode right before resuming userspace.  No use of RCU
+ * is permitted between this call and rcu_user_exit(). This way the
+ * CPU doesn't need to maintain the tick for RCU maintenance purposes
+ * when the CPU runs in userspace.
+ *
+ * If you add or remove a call to rcu_user_enter(), be sure to test with
+ * CONFIG_RCU_EQS_DEBUG=y.
+ */
+noinstr void rcu_user_enter(void)
+{
+	rcu_eqs_enter(true);
+}
+
+/**
+ * rcu_user_exit - inform RCU that we are exiting userspace.
+ *
+ * Exit RCU idle mode while entering the kernel because it can
+ * run a RCU read side critical section anytime.
+ *
+ * If you add or remove a call to rcu_user_exit(), be sure to test with
+ * CONFIG_RCU_EQS_DEBUG=y.
+ */
+void noinstr rcu_user_exit(void)
+{
+	rcu_eqs_exit(true);
+}
+#endif /* #ifdef CONFIG_NO_HZ_FULL */
+
 #ifdef CONFIG_CONTEXT_TRACKING_USER
 
 #define CREATE_TRACE_POINTS