diff options
Diffstat (limited to 'include/linux/preempt.h')
| -rw-r--r-- | include/linux/preempt.h | 243 |
1 files changed, 219 insertions, 24 deletions
diff --git a/include/linux/preempt.h b/include/linux/preempt.h index dd92b1a93919..d964f965c8ff 100644 --- a/include/linux/preempt.h +++ b/include/linux/preempt.h @@ -8,7 +8,8 @@ */ #include <linux/linkage.h> -#include <linux/list.h> +#include <linux/cleanup.h> +#include <linux/types.h> /* * We put the hardirq and softirq counter into the preemption @@ -26,13 +27,13 @@ * PREEMPT_MASK: 0x000000ff * SOFTIRQ_MASK: 0x0000ff00 * HARDIRQ_MASK: 0x000f0000 - * NMI_MASK: 0x00100000 + * NMI_MASK: 0x00f00000 * PREEMPT_NEED_RESCHED: 0x80000000 */ #define PREEMPT_BITS 8 #define SOFTIRQ_BITS 8 #define HARDIRQ_BITS 4 -#define NMI_BITS 1 +#define NMI_BITS 4 #define PREEMPT_SHIFT 0 #define SOFTIRQ_SHIFT (PREEMPT_SHIFT + PREEMPT_BITS) @@ -77,31 +78,67 @@ /* preempt_count() and related functions, depends on PREEMPT_NEED_RESCHED */ #include <asm/preempt.h> +/** + * interrupt_context_level - return interrupt context level + * + * Returns the current interrupt context level. + * 0 - normal context + * 1 - softirq context + * 2 - hardirq context + * 3 - NMI context + */ +static __always_inline unsigned char interrupt_context_level(void) +{ + unsigned long pc = preempt_count(); + unsigned char level = 0; + + level += !!(pc & (NMI_MASK)); + level += !!(pc & (NMI_MASK | HARDIRQ_MASK)); + level += !!(pc & (NMI_MASK | HARDIRQ_MASK | SOFTIRQ_OFFSET)); + + return level; +} + +/* + * These macro definitions avoid redundant invocations of preempt_count() + * because such invocations would result in redundant loads given that + * preempt_count() is commonly implemented with READ_ONCE(). + */ + +#define nmi_count() (preempt_count() & NMI_MASK) #define hardirq_count() (preempt_count() & HARDIRQ_MASK) -#define softirq_count() (preempt_count() & SOFTIRQ_MASK) -#define irq_count() (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_MASK \ - | NMI_MASK)) +#ifdef CONFIG_PREEMPT_RT +# define softirq_count() (current->softirq_disable_cnt & SOFTIRQ_MASK) +# define irq_count() ((preempt_count() & (NMI_MASK | HARDIRQ_MASK)) | softirq_count()) +#else +# define softirq_count() (preempt_count() & SOFTIRQ_MASK) +# define irq_count() (preempt_count() & (NMI_MASK | HARDIRQ_MASK | SOFTIRQ_MASK)) +#endif /* - * Are we doing bottom half or hardware interrupt processing? + * Macros to retrieve the current execution context: * - * in_irq() - We're in (hard) IRQ context + * in_nmi() - We're in NMI context + * in_hardirq() - We're in hard IRQ context + * in_serving_softirq() - We're in softirq context + * in_task() - We're in task context + */ +#define in_nmi() (nmi_count()) +#define in_hardirq() (hardirq_count()) +#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) +#ifdef CONFIG_PREEMPT_RT +# define in_task() (!((preempt_count() & (NMI_MASK | HARDIRQ_MASK)) | in_serving_softirq())) +#else +# define in_task() (!(preempt_count() & (NMI_MASK | HARDIRQ_MASK | SOFTIRQ_OFFSET))) +#endif + +/* + * The following macros are deprecated and should not be used in new code: * in_softirq() - We have BH disabled, or are processing softirqs * in_interrupt() - We're in NMI,IRQ,SoftIRQ context or have BH disabled - * in_serving_softirq() - We're in softirq context - * in_nmi() - We're in NMI context - * in_task() - We're in task context - * - * Note: due to the BH disabled confusion: in_softirq(),in_interrupt() really - * should not be used in new code. */ -#define in_irq() (hardirq_count()) #define in_softirq() (softirq_count()) #define in_interrupt() (irq_count()) -#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) -#define in_nmi() (preempt_count() & NMI_MASK) -#define in_task() (!(preempt_count() & \ - (NMI_MASK | HARDIRQ_MASK | SOFTIRQ_OFFSET))) /* * The preempt_count offset after preempt_disable(); @@ -115,7 +152,12 @@ /* * The preempt_count offset after spin_lock() */ -#define PREEMPT_LOCK_OFFSET PREEMPT_DISABLE_OFFSET +#if !defined(CONFIG_PREEMPT_RT) +#define PREEMPT_LOCK_OFFSET PREEMPT_DISABLE_OFFSET +#else +/* Locks on RT do not disable preemption */ +#define PREEMPT_LOCK_OFFSET 0 +#endif /* * The preempt_count offset needed for things like: @@ -182,7 +224,7 @@ do { \ #define preemptible() (preempt_count() == 0 && !irqs_disabled()) -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION #define preempt_enable() \ do { \ barrier(); \ @@ -203,7 +245,7 @@ do { \ __preempt_schedule(); \ } while (0) -#else /* !CONFIG_PREEMPT */ +#else /* !CONFIG_PREEMPTION */ #define preempt_enable() \ do { \ barrier(); \ @@ -217,7 +259,7 @@ do { \ } while (0) #define preempt_check_resched() do { } while (0) -#endif /* CONFIG_PREEMPT */ +#endif /* CONFIG_PREEMPTION */ #define preempt_disable_notrace() \ do { \ @@ -275,6 +317,7 @@ do { \ #ifdef CONFIG_PREEMPT_NOTIFIERS struct preempt_notifier; +struct task_struct; /** * preempt_ops - notifiers called when a task is preempted and rescheduled @@ -316,10 +359,162 @@ void preempt_notifier_unregister(struct preempt_notifier *notifier); static inline void preempt_notifier_init(struct preempt_notifier *notifier, struct preempt_ops *ops) { - INIT_HLIST_NODE(¬ifier->link); + /* INIT_HLIST_NODE() open coded, to avoid dependency on list.h */ + notifier->link.next = NULL; + notifier->link.pprev = NULL; notifier->ops = ops; } #endif +/* + * Migrate-Disable and why it is undesired. + * + * When a preempted task becomes eligible to run under the ideal model (IOW it + * becomes one of the M highest priority tasks), it might still have to wait + * for the preemptee's migrate_disable() section to complete. Thereby suffering + * a reduction in bandwidth in the exact duration of the migrate_disable() + * section. + * + * Per this argument, the change from preempt_disable() to migrate_disable() + * gets us: + * + * - a higher priority tasks gains reduced wake-up latency; with preempt_disable() + * it would have had to wait for the lower priority task. + * + * - a lower priority tasks; which under preempt_disable() could've instantly + * migrated away when another CPU becomes available, is now constrained + * by the ability to push the higher priority task away, which might itself be + * in a migrate_disable() section, reducing its available bandwidth. + * + * IOW it trades latency / moves the interference term, but it stays in the + * system, and as long as it remains unbounded, the system is not fully + * deterministic. + * + * + * 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 migrate_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 arbitrary 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. + * + */ + +/** + * preempt_disable_nested - Disable preemption inside a normally preempt disabled section + * + * Use for code which requires preemption protection inside a critical + * section which has preemption disabled implicitly on non-PREEMPT_RT + * enabled kernels, by e.g.: + * - holding a spinlock/rwlock + * - soft interrupt context + * - regular interrupt handlers + * + * On PREEMPT_RT enabled kernels spinlock/rwlock held sections, soft + * interrupt context and regular interrupt handlers are preemptible and + * only prevent migration. preempt_disable_nested() ensures that preemption + * is disabled for cases which require CPU local serialization even on + * PREEMPT_RT. For non-PREEMPT_RT kernels this is a NOP. + * + * The use cases are code sequences which are not serialized by a + * particular lock instance, e.g.: + * - seqcount write side critical sections where the seqcount is not + * associated to a particular lock and therefore the automatic + * protection mechanism does not work. This prevents a live lock + * against a preempting high priority reader. + * - RMW per CPU variable updates like vmstat. + */ +/* Macro to avoid header recursion hell vs. lockdep */ +#define preempt_disable_nested() \ +do { \ + if (IS_ENABLED(CONFIG_PREEMPT_RT)) \ + preempt_disable(); \ + else \ + lockdep_assert_preemption_disabled(); \ +} while (0) + +/** + * preempt_enable_nested - Undo the effect of preempt_disable_nested() + */ +static __always_inline void preempt_enable_nested(void) +{ + if (IS_ENABLED(CONFIG_PREEMPT_RT)) + preempt_enable(); +} + +DEFINE_LOCK_GUARD_0(preempt, preempt_disable(), preempt_enable()) +DEFINE_LOCK_GUARD_0(preempt_notrace, preempt_disable_notrace(), preempt_enable_notrace()) + +#ifdef CONFIG_PREEMPT_DYNAMIC + +extern bool preempt_model_none(void); +extern bool preempt_model_voluntary(void); +extern bool preempt_model_full(void); +extern bool preempt_model_lazy(void); + +#else + +static inline bool preempt_model_none(void) +{ + return IS_ENABLED(CONFIG_PREEMPT_NONE); +} +static inline bool preempt_model_voluntary(void) +{ + return IS_ENABLED(CONFIG_PREEMPT_VOLUNTARY); +} +static inline bool preempt_model_full(void) +{ + return IS_ENABLED(CONFIG_PREEMPT); +} + +static inline bool preempt_model_lazy(void) +{ + return IS_ENABLED(CONFIG_PREEMPT_LAZY); +} + +#endif + +static inline bool preempt_model_rt(void) +{ + return IS_ENABLED(CONFIG_PREEMPT_RT); +} + +extern const char *preempt_model_str(void); + +/* + * Does the preemption model allow non-cooperative preemption? + * + * For !CONFIG_PREEMPT_DYNAMIC kernels this is an exact match with + * CONFIG_PREEMPTION; for CONFIG_PREEMPT_DYNAMIC this doesn't work as the + * kernel is *built* with CONFIG_PREEMPTION=y but may run with e.g. the + * PREEMPT_NONE model. + */ +static inline bool preempt_model_preemptible(void) +{ + return preempt_model_full() || preempt_model_lazy() || preempt_model_rt(); +} + #endif /* __LINUX_PREEMPT_H */ |