#endif
+#if defined(CONFIG_SMP) && defined(CONFIG_PREEMPT_RT)
+
+/*
+ * Migrate-Disable and why it is (strongly) undesired.
+ *
+ * The premise of the Real-Time schedulers we have on Linux
+ * (SCHED_FIFO/SCHED_DEADLINE) is that M CPUs can/will run M tasks
+ * concurrently, provided there are sufficient runnable tasks, also known as
+ * work-conserving. For instance SCHED_DEADLINE tries to schedule the M
+ * earliest deadline threads, and SCHED_FIFO the M highest priority threads.
+ *
+ * The correctness of various scheduling models depends on this, but is it
+ * broken by migrate_disable() that doesn't imply preempt_disable(). Where
+ * preempt_disable() implies an immediate priority ceiling, preemptible
+ * migrate_disable() allows nesting.
+ *
+ * The worst case is that all tasks preempt one another in a migrate_disable()
+ * region and stack on a single CPU. This then reduces the available bandwidth
+ * to a single CPU. And since Real-Time schedulability theory considers the
+ * Worst-Case only, all Real-Time analysis shall revert to single-CPU
+ * (instantly solving the SMP analysis problem).
+ *
+ *
+ * The reason we have it anyway.
+ *
+ * PREEMPT_RT breaks a number of assumptions traditionally held. By forcing a
+ * number of primitives into becoming preemptible, they would also allow
+ * migration. This turns out to break a bunch of per-cpu usage. To this end,
+ * all these primitives employ migirate_disable() to restore this implicit
+ * assumption.
+ *
+ * This is a 'temporary' work-around at best. The correct solution is getting
+ * rid of the above assumptions and reworking the code to employ explicit
+ * per-cpu locking or short preempt-disable regions.
+ *
+ * The end goal must be to get rid of migrate_disable(), alternatively we need
+ * a schedulability theory that does not depend on abritrary migration.
+ *
+ *
+ * Notes on the implementation.
+ *
+ * The implementation is particularly tricky since existing code patterns
+ * dictate neither migrate_disable() nor migrate_enable() is allowed to block.
+ * This means that it cannot use cpus_read_lock() to serialize against hotplug,
+ * nor can it easily migrate itself into a pending affinity mask change on
+ * migrate_enable().
+ *
+ *
+ * Note: even non-work-conserving schedulers like semi-partitioned depends on
+ * migration, so migrate_disable() is not only a problem for
+ * work-conserving schedulers.
+ *
+ */
+extern void migrate_disable(void);
+extern void migrate_enable(void);
+
+#elif defined(CONFIG_PREEMPT_RT)
+
+static inline void migrate_disable(void) { }
+static inline void migrate_enable(void) { }
+
+#else /* !CONFIG_PREEMPT_RT */
+
/**
* migrate_disable - Prevent migration of the current task
*
preempt_enable();
}
+#endif /* CONFIG_SMP && CONFIG_PREEMPT_RT */
+
#endif /* __LINUX_PREEMPT_H */
#ifdef CONFIG_SMP
+#ifdef CONFIG_PREEMPT_RT
+
+static void
+__do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask, u32 flags);
+
+static int __set_cpus_allowed_ptr(struct task_struct *p,
+ const struct cpumask *new_mask,
+ u32 flags);
+
+static void migrate_disable_switch(struct rq *rq, struct task_struct *p)
+{
+ if (likely(!p->migration_disabled))
+ return;
+
+ if (p->cpus_ptr != &p->cpus_mask)
+ return;
+
+ /*
+ * Violates locking rules! see comment in __do_set_cpus_allowed().
+ */
+ __do_set_cpus_allowed(p, cpumask_of(rq->cpu), SCA_MIGRATE_DISABLE);
+}
+
+void migrate_disable(void)
+{
+ if (current->migration_disabled++)
+ return;
+
+ barrier();
+}
+EXPORT_SYMBOL_GPL(migrate_disable);
+
+void migrate_enable(void)
+{
+ struct task_struct *p = current;
+
+ if (--p->migration_disabled)
+ return;
+
+ barrier();
+
+ if (p->cpus_ptr == &p->cpus_mask)
+ return;
+
+ __set_cpus_allowed_ptr(p, &p->cpus_mask, SCA_MIGRATE_ENABLE);
+}
+EXPORT_SYMBOL_GPL(migrate_enable);
+
+static inline bool is_migration_disabled(struct task_struct *p)
+{
+ return p->migration_disabled;
+}
+
+#endif
+
/*
* Per-CPU kthreads are allowed to run on !active && online CPUs, see
* __set_cpus_allowed_ptr() and select_fallback_rq().
if (!cpumask_test_cpu(cpu, p->cpus_ptr))
return false;
- if (is_per_cpu_kthread(p))
+ if (is_per_cpu_kthread(p) || is_migration_disabled(p))
return cpu_online(cpu);
return cpu_active(cpu);
*/
void set_cpus_allowed_common(struct task_struct *p, const struct cpumask *new_mask, u32 flags)
{
+ if (flags & (SCA_MIGRATE_ENABLE | SCA_MIGRATE_DISABLE)) {
+ p->cpus_ptr = new_mask;
+ return;
+ }
+
cpumask_copy(&p->cpus_mask, new_mask);
p->nr_cpus_allowed = cpumask_weight(new_mask);
}
struct rq *rq = task_rq(p);
bool queued, running;
- lockdep_assert_held(&p->pi_lock);
+ /*
+ * This here violates the locking rules for affinity, since we're only
+ * supposed to change these variables while holding both rq->lock and
+ * p->pi_lock.
+ *
+ * HOWEVER, it magically works, because ttwu() is the only code that
+ * accesses these variables under p->pi_lock and only does so after
+ * smp_cond_load_acquire(&p->on_cpu, !VAL), and we're in __schedule()
+ * before finish_task().
+ *
+ * XXX do further audits, this smells like something putrid.
+ */
+ if (flags & SCA_MIGRATE_DISABLE)
+ SCHED_WARN_ON(!p->on_cpu);
+ else
+ lockdep_assert_held(&p->pi_lock);
queued = task_on_rq_queued(p);
running = task_current(rq, p);
rq = task_rq_lock(p, &rf);
update_rq_clock(rq);
- if (p->flags & PF_KTHREAD) {
+ if (p->flags & PF_KTHREAD || is_migration_disabled(p)) {
/*
- * Kernel threads are allowed on online && !active CPUs
+ * Kernel threads are allowed on online && !active CPUs.
+ *
+ * Specifically, migration_disabled() tasks must not fail the
+ * cpumask_any_and_distribute() pick below, esp. so on
+ * SCA_MIGRATE_ENABLE, otherwise we'll not call
+ * set_cpus_allowed_common() and actually reset p->cpus_ptr.
*/
cpu_valid_mask = cpu_online_mask;
}
goto out;
}
- if (cpumask_equal(&p->cpus_mask, new_mask))
+ if (!(flags & SCA_MIGRATE_ENABLE) && cpumask_equal(&p->cpus_mask, new_mask))
goto out;
/*
* Clearly, migrating tasks to offline CPUs is a fairly daft thing.
*/
WARN_ON_ONCE(!cpu_online(new_cpu));
+
+ WARN_ON_ONCE(is_migration_disabled(p));
#endif
trace_sched_migrate_task(p, new_cpu);
}
fallthrough;
case possible:
+ /*
+ * XXX When called from select_task_rq() we only
+ * hold p->pi_lock and again violate locking order.
+ *
+ * More yuck to audit.
+ */
do_set_cpus_allowed(p, cpu_possible_mask);
state = fail;
break;
{
lockdep_assert_held(&p->pi_lock);
- if (p->nr_cpus_allowed > 1)
+ if (p->nr_cpus_allowed > 1 && !is_migration_disabled(p))
cpu = p->sched_class->select_task_rq(p, cpu, sd_flags, wake_flags);
else
cpu = cpumask_any(p->cpus_ptr);
#endif /* CONFIG_SMP */
+#if !defined(CONFIG_SMP) || !defined(CONFIG_PREEMPT_RT)
+
+static inline void migrate_disable_switch(struct rq *rq, struct task_struct *p) { }
+
+static inline bool is_migration_disabled(struct task_struct *p)
+{
+ return false;
+}
+
+#endif
+
static void
ttwu_stat(struct task_struct *p, int cpu, int wake_flags)
{
*/
++*switch_count;
+ migrate_disable_switch(rq, prev);
psi_sched_switch(prev, next, !task_on_rq_queued(prev));
trace_sched_switch(preempt, prev, next);
projects / platform / kernel / linux-rpi.git / commitdiff