diff --git a/Documentation/RCU/Design/Data-Structures/Data-Structures.rst b/Documentation/RCU/Design/Data-Structures/Data-Structures.rst index 04e16775c752..1b0aad184dd7 100644 --- a/Documentation/RCU/Design/Data-Structures/Data-Structures.rst +++ b/Documentation/RCU/Design/Data-Structures/Data-Structures.rst @@ -286,6 +286,39 @@ in order to detect the beginnings and ends of grace periods in a distributed fashion. The values flow from ``rcu_state`` to ``rcu_node`` (down the tree from the root to the leaves) to ``rcu_data``. ++-----------------------------------------------------------------------+ +| **Quick Quiz**: | ++-----------------------------------------------------------------------+ +| Given that the root rcu_node structure has a gp_seq field, | +| why does RCU maintain a separate gp_seq in the rcu_state structure? | +| Why not just use the root rcu_node's gp_seq as the official record | +| and update it directly when starting a new grace period? | ++-----------------------------------------------------------------------+ +| **Answer**: | ++-----------------------------------------------------------------------+ +| On single-node RCU trees (where the root node is also a leaf), | +| updating the root node's gp_seq immediately would create unnecessary | +| lock contention. Here's why: | +| | +| If we did rcu_seq_start() directly on the root node's gp_seq: | +| | +| 1. All CPUs would immediately see their node's gp_seq from their rdp's| +| gp_seq, in rcu_pending(). They would all then invoke the RCU-core. | +| 2. Which calls note_gp_changes() and try to acquire the node lock. | +| 3. But rnp->qsmask isn't initialized yet (happens later in | +| rcu_gp_init()) | +| 4. So each CPU would acquire the lock, find it can't determine if it | +| needs to report quiescent state (no qsmask), update rdp->gp_seq, | +| and release the lock. | +| 5. Result: Lots of lock acquisitions with no grace period progress | +| | +| By having a separate rcu_state.gp_seq, we can increment the official | +| grace period counter without immediately affecting what CPUs see in | +| their nodes. The hierarchical propagation in rcu_gp_init() then | +| updates the root node's gp_seq and qsmask together under the same lock| +| acquisition, avoiding this useless contention. | ++-----------------------------------------------------------------------+ + Miscellaneous ''''''''''''' diff --git a/Documentation/RCU/Design/Requirements/Requirements.rst b/Documentation/RCU/Design/Requirements/Requirements.rst index 6125e7068d2c..b0395540296b 100644 --- a/Documentation/RCU/Design/Requirements/Requirements.rst +++ b/Documentation/RCU/Design/Requirements/Requirements.rst @@ -1970,6 +1970,134 @@ corresponding CPU's leaf node lock is held. This avoids race conditions between RCU's hotplug notifier hooks, the grace period initialization code, and the FQS loop, all of which refer to or modify this bookkeeping. +Note that grace period initialization (rcu_gp_init()) must carefully sequence +CPU hotplug scanning with grace period state changes. For example, the +following race could occur in rcu_gp_init() if rcu_seq_start() were to happen +after the CPU hotplug scanning. + +.. code-block:: none + + CPU0 (rcu_gp_init) CPU1 CPU2 + --------------------- ---- ---- + // Hotplug scan first (WRONG ORDER) + rcu_for_each_leaf_node(rnp) { + rnp->qsmaskinit = rnp->qsmaskinitnext; + } + rcutree_report_cpu_starting() + rnp->qsmaskinitnext |= mask; + rcu_read_lock() + r0 = *X; + r1 = *X; + X = NULL; + cookie = get_state_synchronize_rcu(); + // cookie = 8 (future GP) + rcu_seq_start(&rcu_state.gp_seq); + // gp_seq = 5 + + // CPU1 now invisible to this GP! + rcu_for_each_node_breadth_first() { + rnp->qsmask = rnp->qsmaskinit; + // CPU1 not included! + } + + // GP completes without CPU1 + rcu_seq_end(&rcu_state.gp_seq); + // gp_seq = 8 + poll_state_synchronize_rcu(cookie); + // Returns true! + kfree(r1); + r2 = *r0; // USE-AFTER-FREE! + +By incrementing gp_seq first, CPU1's RCU read-side critical section +is guaranteed to not be missed by CPU2. + +**Concurrent Quiescent State Reporting for Offline CPUs** + +RCU must ensure that CPUs going offline report quiescent states to avoid +blocking grace periods. This requires careful synchronization to handle +race conditions + +**Race condition causing Offline CPU to hang GP** + +A race between CPU offlining and new GP initialization (gp_init) may occur +because `rcu_report_qs_rnp()` in `rcutree_report_cpu_dead()` must temporarily +release the `rcu_node` lock to wake the RCU grace-period kthread: + +.. code-block:: none + + CPU1 (going offline) CPU0 (GP kthread) + -------------------- ----------------- + rcutree_report_cpu_dead() + rcu_report_qs_rnp() + // Must release rnp->lock to wake GP kthread + raw_spin_unlock_irqrestore_rcu_node() + // Wakes up and starts new GP + rcu_gp_init() + // First loop: + copies qsmaskinitnext->qsmaskinit + // CPU1 still in qsmaskinitnext! + + // Second loop: + rnp->qsmask = rnp->qsmaskinit + mask = rnp->qsmask & ~rnp->qsmaskinitnext + // mask is 0! CPU1 still in both masks + // Reacquire lock (but too late) + rnp->qsmaskinitnext &= ~mask // Finally clears bit + +Without `ofl_lock`, the new grace period includes the offline CPU and waits +forever for its quiescent state causing a GP hang. + +**A solution with ofl_lock** + +The `ofl_lock` (offline lock) prevents `rcu_gp_init()` from running during +the vulnerable window when `rcu_report_qs_rnp()` has released `rnp->lock`: + +.. code-block:: none + + CPU0 (rcu_gp_init) CPU1 (rcutree_report_cpu_dead) + ------------------ ------------------------------ + rcu_for_each_leaf_node(rnp) { + arch_spin_lock(&ofl_lock) -----> arch_spin_lock(&ofl_lock) [BLOCKED] + + // Safe: CPU1 can't interfere + rnp->qsmaskinit = rnp->qsmaskinitnext + + arch_spin_unlock(&ofl_lock) ---> // Now CPU1 can proceed + } // But snapshot already taken + +**Another race causing GP hangs in rcu_gpu_init(): Reporting QS for Now-offline CPUs** + +After the first loop takes an atomic snapshot of online CPUs, as shown above, +the second loop in `rcu_gp_init()` detects CPUs that went offline between +releasing `ofl_lock` and acquiring the per-node `rnp->lock`. This detection is +crucial because: + +1. The CPU might have gone offline after the snapshot but before the second loop +2. The offline CPU cannot report its own QS if it's already dead +3. Without this detection, the grace period would wait forever for CPUs that + are now offline. + +The second loop performs this detection safely: + +.. code-block:: none + + rcu_for_each_node_breadth_first(rnp) { + raw_spin_lock_irqsave_rcu_node(rnp, flags); + rnp->qsmask = rnp->qsmaskinit; // Apply the snapshot + + // Detect CPUs offline after snapshot + mask = rnp->qsmask & ~rnp->qsmaskinitnext; + + if (mask && rcu_is_leaf_node(rnp)) + rcu_report_qs_rnp(mask, ...) // Report QS for offline CPUs + } + +This approach ensures atomicity: quiescent state reporting for offline CPUs +happens either in `rcu_gp_init()` (second loop) or in `rcutree_report_cpu_dead()`, +never both and never neither. The `rnp->lock` held throughout the sequence +prevents races - `rcutree_report_cpu_dead()` also acquires this lock when +clearing `qsmaskinitnext`, ensuring mutual exclusion. + Scheduler and RCU ~~~~~~~~~~~~~~~~~ diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 0d2ea9a60145..d3f5a1c69dab 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -5508,7 +5508,8 @@ echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=1" - Default is 0. + Default is 1 if num_possible_cpus() <= 16 and it is not explicitly + disabled by the boot parameter passing 0. rcuscale.gp_async= [KNL] Measure performance of asynchronous diff --git a/include/linux/srcu.h b/include/linux/srcu.h index 900b0d5c05f5..f179700fecaf 100644 --- a/include/linux/srcu.h +++ b/include/linux/srcu.h @@ -46,11 +46,11 @@ int init_srcu_struct(struct srcu_struct *ssp); /* Values for SRCU Tree srcu_data ->srcu_reader_flavor, but also used by rcutorture. */ #define SRCU_READ_FLAVOR_NORMAL 0x1 // srcu_read_lock(). #define SRCU_READ_FLAVOR_NMI 0x2 // srcu_read_lock_nmisafe(). -#define SRCU_READ_FLAVOR_LITE 0x4 // srcu_read_lock_lite(). +// 0x4 // SRCU-lite is no longer with us. #define SRCU_READ_FLAVOR_FAST 0x8 // srcu_read_lock_fast(). #define SRCU_READ_FLAVOR_ALL (SRCU_READ_FLAVOR_NORMAL | SRCU_READ_FLAVOR_NMI | \ - SRCU_READ_FLAVOR_LITE | SRCU_READ_FLAVOR_FAST) // All of the above. -#define SRCU_READ_FLAVOR_SLOWGP (SRCU_READ_FLAVOR_LITE | SRCU_READ_FLAVOR_FAST) + SRCU_READ_FLAVOR_FAST) // All of the above. +#define SRCU_READ_FLAVOR_SLOWGP SRCU_READ_FLAVOR_FAST // Flavors requiring synchronize_rcu() // instead of smp_mb(). void __srcu_read_unlock(struct srcu_struct *ssp, int idx) __releases(ssp); @@ -299,33 +299,6 @@ static inline struct srcu_ctr __percpu *srcu_down_read_fast(struct srcu_struct * return __srcu_read_lock_fast(ssp); } -/** - * srcu_read_lock_lite - register a new reader for an SRCU-protected structure. - * @ssp: srcu_struct in which to register the new reader. - * - * Enter an SRCU read-side critical section, but for a light-weight - * smp_mb()-free reader. See srcu_read_lock() for more information. - * - * If srcu_read_lock_lite() is ever used on an srcu_struct structure, - * then none of the other flavors may be used, whether before, during, - * or after. Note that grace-period auto-expediting is disabled for _lite - * srcu_struct structures because auto-expedited grace periods invoke - * synchronize_rcu_expedited(), IPIs and all. - * - * Note that srcu_read_lock_lite() can be invoked only from those contexts - * where RCU is watching, that is, from contexts where it would be legal - * to invoke rcu_read_lock(). Otherwise, lockdep will complain. - */ -static inline int srcu_read_lock_lite(struct srcu_struct *ssp) __acquires(ssp) -{ - int retval; - - srcu_check_read_flavor_force(ssp, SRCU_READ_FLAVOR_LITE); - retval = __srcu_read_lock_lite(ssp); - rcu_try_lock_acquire(&ssp->dep_map); - return retval; -} - /** * srcu_read_lock_nmisafe - register a new reader for an SRCU-protected structure. * @ssp: srcu_struct in which to register the new reader. @@ -434,22 +407,6 @@ static inline void srcu_up_read_fast(struct srcu_struct *ssp, struct srcu_ctr __ __srcu_read_unlock_fast(ssp, scp); } -/** - * srcu_read_unlock_lite - unregister a old reader from an SRCU-protected structure. - * @ssp: srcu_struct in which to unregister the old reader. - * @idx: return value from corresponding srcu_read_lock_lite(). - * - * Exit a light-weight SRCU read-side critical section. - */ -static inline void srcu_read_unlock_lite(struct srcu_struct *ssp, int idx) - __releases(ssp) -{ - WARN_ON_ONCE(idx & ~0x1); - srcu_check_read_flavor(ssp, SRCU_READ_FLAVOR_LITE); - srcu_lock_release(&ssp->dep_map); - __srcu_read_unlock_lite(ssp, idx); -} - /** * srcu_read_unlock_nmisafe - unregister a old reader from an SRCU-protected structure. * @ssp: srcu_struct in which to unregister the old reader. @@ -524,4 +481,9 @@ DEFINE_LOCK_GUARD_1(srcu, struct srcu_struct, srcu_read_unlock(_T->lock, _T->idx), int idx) +DEFINE_LOCK_GUARD_1(srcu_fast, struct srcu_struct, + _T->scp = srcu_read_lock_fast(_T->lock), + srcu_read_unlock_fast(_T->lock, _T->scp), + struct srcu_ctr __percpu *scp) + #endif diff --git a/include/linux/srcutiny.h b/include/linux/srcutiny.h index 380260317d98..51ce25f07930 100644 --- a/include/linux/srcutiny.h +++ b/include/linux/srcutiny.h @@ -93,9 +93,6 @@ static inline void __srcu_read_unlock_fast(struct srcu_struct *ssp, struct srcu_ __srcu_read_unlock(ssp, __srcu_ptr_to_ctr(ssp, scp)); } -#define __srcu_read_lock_lite __srcu_read_lock -#define __srcu_read_unlock_lite __srcu_read_unlock - static inline void synchronize_srcu_expedited(struct srcu_struct *ssp) { synchronize_srcu(ssp); diff --git a/include/linux/srcutree.h b/include/linux/srcutree.h index 8bed7e6cc4c1..bf44d8d1e69e 100644 --- a/include/linux/srcutree.h +++ b/include/linux/srcutree.h @@ -278,44 +278,6 @@ static inline void __srcu_read_unlock_fast(struct srcu_struct *ssp, struct srcu_ RCU_LOCKDEP_WARN(!rcu_is_watching(), "RCU must be watching srcu_read_unlock_fast()."); } -/* - * Counts the new reader in the appropriate per-CPU element of the - * srcu_struct. Returns an index that must be passed to the matching - * srcu_read_unlock_lite(). - * - * Note that this_cpu_inc() is an RCU read-side critical section either - * because it disables interrupts, because it is a single instruction, - * or because it is a read-modify-write atomic operation, depending on - * the whims of the architecture. - */ -static inline int __srcu_read_lock_lite(struct srcu_struct *ssp) -{ - struct srcu_ctr __percpu *scp = READ_ONCE(ssp->srcu_ctrp); - - RCU_LOCKDEP_WARN(!rcu_is_watching(), "RCU must be watching srcu_read_lock_lite()."); - this_cpu_inc(scp->srcu_locks.counter); /* Y */ - barrier(); /* Avoid leaking the critical section. */ - return __srcu_ptr_to_ctr(ssp, scp); -} - -/* - * Removes the count for the old reader from the appropriate - * per-CPU element of the srcu_struct. Note that this may well be a - * different CPU than that which was incremented by the corresponding - * srcu_read_lock_lite(), but it must be within the same task. - * - * Note that this_cpu_inc() is an RCU read-side critical section either - * because it disables interrupts, because it is a single instruction, - * or because it is a read-modify-write atomic operation, depending on - * the whims of the architecture. - */ -static inline void __srcu_read_unlock_lite(struct srcu_struct *ssp, int idx) -{ - barrier(); /* Avoid leaking the critical section. */ - this_cpu_inc(__srcu_ctr_to_ptr(ssp, idx)->srcu_unlocks.counter); /* Z */ - RCU_LOCKDEP_WARN(!rcu_is_watching(), "RCU must be watching srcu_read_unlock_lite()."); -} - void __srcu_check_read_flavor(struct srcu_struct *ssp, int read_flavor); // Record reader usage even for CONFIG_PROVE_RCU=n kernels. This is diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c index 70ec0f21abc3..7a893d51d02b 100644 --- a/kernel/rcu/rcutorture.c +++ b/kernel/rcu/rcutorture.c @@ -55,22 +55,24 @@ MODULE_DESCRIPTION("Read-Copy Update module-based torture test facility"); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Paul E. McKenney and Josh Triplett "); -/* Bits for ->extendables field, extendables param, and related definitions. */ -#define RCUTORTURE_RDR_SHIFT_1 8 /* Put SRCU index in upper bits. */ -#define RCUTORTURE_RDR_MASK_1 (0xff << RCUTORTURE_RDR_SHIFT_1) -#define RCUTORTURE_RDR_SHIFT_2 16 /* Put SRCU index in upper bits. */ -#define RCUTORTURE_RDR_MASK_2 (0xff << RCUTORTURE_RDR_SHIFT_2) -#define RCUTORTURE_RDR_BH 0x01 /* Extend readers by disabling bh. */ -#define RCUTORTURE_RDR_IRQ 0x02 /* ... disabling interrupts. */ -#define RCUTORTURE_RDR_PREEMPT 0x04 /* ... disabling preemption. */ -#define RCUTORTURE_RDR_RBH 0x08 /* ... rcu_read_lock_bh(). */ -#define RCUTORTURE_RDR_SCHED 0x10 /* ... rcu_read_lock_sched(). */ -#define RCUTORTURE_RDR_RCU_1 0x20 /* ... entering another RCU reader. */ -#define RCUTORTURE_RDR_RCU_2 0x40 /* ... entering another RCU reader. */ -#define RCUTORTURE_RDR_NBITS 7 /* Number of bits defined above. */ -#define RCUTORTURE_MAX_EXTEND \ +// Bits for ->extendables field, extendables param, and related definitions. +#define RCUTORTURE_RDR_SHIFT_1 8 // Put SRCU index in upper bits. +#define RCUTORTURE_RDR_MASK_1 (0xff << RCUTORTURE_RDR_SHIFT_1) +#define RCUTORTURE_RDR_SHIFT_2 16 // Put SRCU index in upper bits. +#define RCUTORTURE_RDR_MASK_2 (0xff << RCUTORTURE_RDR_SHIFT_2) +#define RCUTORTURE_RDR_BH 0x01 // Extend readers by disabling bh. +#define RCUTORTURE_RDR_IRQ 0x02 // ... disabling interrupts. +#define RCUTORTURE_RDR_PREEMPT 0x04 // ... disabling preemption. +#define RCUTORTURE_RDR_RBH 0x08 // ... rcu_read_lock_bh(). +#define RCUTORTURE_RDR_SCHED 0x10 // ... rcu_read_lock_sched(). +#define RCUTORTURE_RDR_RCU_1 0x20 // ... entering another RCU reader. +#define RCUTORTURE_RDR_RCU_2 0x40 // ... entering another RCU reader. +#define RCUTORTURE_RDR_UPDOWN 0x80 // ... up-read from task, down-read from timer. + // Note: Manual start, automatic end. +#define RCUTORTURE_RDR_NBITS 8 // Number of bits defined above. +#define RCUTORTURE_MAX_EXTEND \ (RCUTORTURE_RDR_BH | RCUTORTURE_RDR_IRQ | RCUTORTURE_RDR_PREEMPT | \ - RCUTORTURE_RDR_RBH | RCUTORTURE_RDR_SCHED) + RCUTORTURE_RDR_RBH | RCUTORTURE_RDR_SCHED) // Intentionally omit RCUTORTURE_RDR_UPDOWN. #define RCUTORTURE_RDR_ALLBITS \ (RCUTORTURE_MAX_EXTEND | RCUTORTURE_RDR_RCU_1 | RCUTORTURE_RDR_RCU_2 | \ RCUTORTURE_RDR_MASK_1 | RCUTORTURE_RDR_MASK_2) @@ -110,6 +112,7 @@ torture_param(bool, gp_sync, false, "Use synchronous GP wait primitives"); torture_param(int, irqreader, 1, "Allow RCU readers from irq handlers"); torture_param(int, leakpointer, 0, "Leak pointer dereferences from readers"); torture_param(int, n_barrier_cbs, 0, "# of callbacks/kthreads for barrier testing"); +torture_param(int, n_up_down, 32, "# of concurrent up/down hrtimer-based RCU readers"); torture_param(int, nfakewriters, 4, "Number of RCU fake writer threads"); torture_param(int, nreaders, -1, "Number of RCU reader threads"); torture_param(int, object_debug, 0, "Enable debug-object double call_rcu() testing"); @@ -156,6 +159,7 @@ static int nrealfakewriters; static struct task_struct *writer_task; static struct task_struct **fakewriter_tasks; static struct task_struct **reader_tasks; +static struct task_struct *updown_task; static struct task_struct **nocb_tasks; static struct task_struct *stats_task; static struct task_struct *fqs_task; @@ -378,6 +382,8 @@ struct rcu_torture_ops { void (*readunlock)(int idx); int (*readlock_held)(void); // lockdep. int (*readlock_nesting)(void); // actual nesting, if available, -1 if not. + int (*down_read)(void); + void (*up_read)(int idx); unsigned long (*get_gp_seq)(void); unsigned long (*gp_diff)(unsigned long new, unsigned long old); void (*deferred_free)(struct rcu_torture *p); @@ -427,6 +433,7 @@ struct rcu_torture_ops { int no_pi_lock; int debug_objects; int start_poll_irqsoff; + int have_up_down; const char *name; }; @@ -464,7 +471,7 @@ rcu_read_delay(struct torture_random_state *rrsp, struct rt_read_seg *rtrsp) !(torture_random(rrsp) % (nrealreaders * 2000 * longdelay_ms))) { started = cur_ops->get_gp_seq(); ts = rcu_trace_clock_local(); - if (preempt_count() & (SOFTIRQ_MASK | HARDIRQ_MASK)) + if ((preempt_count() & HARDIRQ_MASK) || softirq_count()) longdelay_ms = 5; /* Avoid triggering BH limits. */ mdelay(longdelay_ms); rtrsp->rt_delay_ms = longdelay_ms; @@ -711,11 +718,6 @@ static int srcu_torture_read_lock(void) WARN_ON_ONCE(idx & ~0x1); ret += idx << 1; } - if (reader_flavor & SRCU_READ_FLAVOR_LITE) { - idx = srcu_read_lock_lite(srcu_ctlp); - WARN_ON_ONCE(idx & ~0x1); - ret += idx << 2; - } if (reader_flavor & SRCU_READ_FLAVOR_FAST) { scp = srcu_read_lock_fast(srcu_ctlp); idx = __srcu_ptr_to_ctr(srcu_ctlp, scp); @@ -749,8 +751,6 @@ static void srcu_torture_read_unlock(int idx) WARN_ON_ONCE((reader_flavor && (idx & ~reader_flavor)) || (!reader_flavor && (idx & ~0x1))); if (reader_flavor & SRCU_READ_FLAVOR_FAST) srcu_read_unlock_fast(srcu_ctlp, __srcu_ctr_to_ptr(srcu_ctlp, (idx & 0x8) >> 3)); - if (reader_flavor & SRCU_READ_FLAVOR_LITE) - srcu_read_unlock_lite(srcu_ctlp, (idx & 0x4) >> 2); if (reader_flavor & SRCU_READ_FLAVOR_NMI) srcu_read_unlock_nmisafe(srcu_ctlp, (idx & 0x2) >> 1); if ((reader_flavor & SRCU_READ_FLAVOR_NORMAL) || !(reader_flavor & SRCU_READ_FLAVOR_ALL)) @@ -762,6 +762,50 @@ static int torture_srcu_read_lock_held(void) return srcu_read_lock_held(srcu_ctlp); } +static bool srcu_torture_have_up_down(void) +{ + int rf = reader_flavor; + + if (!rf) + rf = SRCU_READ_FLAVOR_NORMAL; + return !!(cur_ops->have_up_down & rf); +} + +static int srcu_torture_down_read(void) +{ + int idx; + struct srcu_ctr __percpu *scp; + + WARN_ON_ONCE(reader_flavor & ~SRCU_READ_FLAVOR_ALL); + WARN_ON_ONCE(reader_flavor & (reader_flavor - 1)); + + if ((reader_flavor & SRCU_READ_FLAVOR_NORMAL) || !(reader_flavor & SRCU_READ_FLAVOR_ALL)) { + idx = srcu_down_read(srcu_ctlp); + WARN_ON_ONCE(idx & ~0x1); + return idx; + } + if (reader_flavor & SRCU_READ_FLAVOR_FAST) { + scp = srcu_down_read_fast(srcu_ctlp); + idx = __srcu_ptr_to_ctr(srcu_ctlp, scp); + WARN_ON_ONCE(idx & ~0x1); + return idx << 3; + } + WARN_ON_ONCE(1); + return 0; +} + +static void srcu_torture_up_read(int idx) +{ + WARN_ON_ONCE((reader_flavor && (idx & ~reader_flavor)) || (!reader_flavor && (idx & ~0x1))); + if (reader_flavor & SRCU_READ_FLAVOR_FAST) + srcu_up_read_fast(srcu_ctlp, __srcu_ctr_to_ptr(srcu_ctlp, (idx & 0x8) >> 3)); + else if ((reader_flavor & SRCU_READ_FLAVOR_NORMAL) || + !(reader_flavor & SRCU_READ_FLAVOR_ALL)) + srcu_up_read(srcu_ctlp, idx & 0x1); + else + WARN_ON_ONCE(1); +} + static unsigned long srcu_torture_completed(void) { return srcu_batches_completed(srcu_ctlp); @@ -819,6 +863,8 @@ static struct rcu_torture_ops srcu_ops = { .readlock = srcu_torture_read_lock, .read_delay = srcu_read_delay, .readunlock = srcu_torture_read_unlock, + .down_read = srcu_torture_down_read, + .up_read = srcu_torture_up_read, .readlock_held = torture_srcu_read_lock_held, .get_gp_seq = srcu_torture_completed, .gp_diff = rcu_seq_diff, @@ -839,6 +885,8 @@ static struct rcu_torture_ops srcu_ops = { .irq_capable = 1, .no_pi_lock = IS_ENABLED(CONFIG_TINY_SRCU), .debug_objects = 1, + .have_up_down = IS_ENABLED(CONFIG_TINY_SRCU) + ? 0 : SRCU_READ_FLAVOR_NORMAL | SRCU_READ_FLAVOR_FAST, .name = "srcu" }; @@ -864,6 +912,8 @@ static struct rcu_torture_ops srcud_ops = { .read_delay = srcu_read_delay, .readunlock = srcu_torture_read_unlock, .readlock_held = torture_srcu_read_lock_held, + .down_read = srcu_torture_down_read, + .up_read = srcu_torture_up_read, .get_gp_seq = srcu_torture_completed, .gp_diff = rcu_seq_diff, .deferred_free = srcu_torture_deferred_free, @@ -883,6 +933,8 @@ static struct rcu_torture_ops srcud_ops = { .irq_capable = 1, .no_pi_lock = IS_ENABLED(CONFIG_TINY_SRCU), .debug_objects = 1, + .have_up_down = IS_ENABLED(CONFIG_TINY_SRCU) + ? 0 : SRCU_READ_FLAVOR_NORMAL | SRCU_READ_FLAVOR_FAST, .name = "srcud" }; @@ -910,7 +962,8 @@ static struct rcu_torture_ops busted_srcud_ops = { /* * Definitions for trivial CONFIG_PREEMPT=n-only torture testing. - * This implementation does not necessarily work well with CPU hotplug. + * This implementation does not work well with CPU hotplug nor + * with rcutorture's shuffling. */ static void synchronize_rcu_trivial(void) @@ -923,6 +976,16 @@ static void synchronize_rcu_trivial(void) } } +static void rcu_sync_torture_init_trivial(void) +{ + rcu_sync_torture_init(); + // if (onoff_interval || shuffle_interval) { + if (WARN_ONCE(onoff_interval || shuffle_interval, "%s: Non-zero onoff_interval (%d) or shuffle_interval (%d) breaks trivial RCU, resetting to zero", __func__, onoff_interval, shuffle_interval)) { + onoff_interval = 0; + shuffle_interval = 0; + } +} + static int rcu_torture_read_lock_trivial(void) { preempt_disable(); @@ -936,7 +999,7 @@ static void rcu_torture_read_unlock_trivial(int idx) static struct rcu_torture_ops trivial_ops = { .ttype = RCU_TRIVIAL_FLAVOR, - .init = rcu_sync_torture_init, + .init = rcu_sync_torture_init_trivial, .readlock = rcu_torture_read_lock_trivial, .read_delay = rcu_read_delay, /* just reuse rcu's version. */ .readunlock = rcu_torture_read_unlock_trivial, @@ -1722,6 +1785,7 @@ rcu_torture_writer(void *arg) cur_ops->gp_kthread_dbg(); WARN(1, "%s: rtort_pipe_count: %d\n", __func__, rcu_tortures[i].rtort_pipe_count); rcu_ftrace_dump(DUMP_ALL); + break; } if (stutter_waited) sched_set_normal(current, oldnice); @@ -1915,14 +1979,14 @@ static void rcu_torture_reader_do_mbchk(long myid, struct rcu_torture *rtp, // Verify the specified RCUTORTURE_RDR* state. #define ROEC_ARGS "%s %s: Current %#x To add %#x To remove %#x preempt_count() %#x\n", __func__, s, curstate, new, old, preempt_count() -static void rcutorture_one_extend_check(char *s, int curstate, int new, int old, bool insoftirq) +static void rcutorture_one_extend_check(char *s, int curstate, int new, int old) { int mask; - if (!IS_ENABLED(CONFIG_RCU_TORTURE_TEST_CHK_RDR_STATE)) + if (!IS_ENABLED(CONFIG_RCU_TORTURE_TEST_CHK_RDR_STATE) || in_nmi()) return; - WARN_ONCE(!(curstate & RCUTORTURE_RDR_IRQ) && irqs_disabled(), ROEC_ARGS); + WARN_ONCE(!(curstate & RCUTORTURE_RDR_IRQ) && irqs_disabled() && !in_hardirq(), ROEC_ARGS); WARN_ONCE((curstate & RCUTORTURE_RDR_IRQ) && !irqs_disabled(), ROEC_ARGS); // If CONFIG_PREEMPT_COUNT=n, further checks are unreliable. @@ -1930,21 +1994,21 @@ static void rcutorture_one_extend_check(char *s, int curstate, int new, int old, return; WARN_ONCE((curstate & (RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH)) && - !(preempt_count() & SOFTIRQ_MASK), ROEC_ARGS); + !softirq_count(), ROEC_ARGS); WARN_ONCE((curstate & (RCUTORTURE_RDR_PREEMPT | RCUTORTURE_RDR_SCHED)) && !(preempt_count() & PREEMPT_MASK), ROEC_ARGS); WARN_ONCE(cur_ops->readlock_nesting && (curstate & (RCUTORTURE_RDR_RCU_1 | RCUTORTURE_RDR_RCU_2)) && cur_ops->readlock_nesting() == 0, ROEC_ARGS); - // Timer handlers have all sorts of stuff disabled, so ignore + // Interrupt handlers have all sorts of stuff disabled, so ignore // unintended disabling. - if (insoftirq) + if (in_serving_softirq() || in_hardirq()) return; WARN_ONCE(cur_ops->extendables && !(curstate & (RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH)) && - (preempt_count() & SOFTIRQ_MASK), ROEC_ARGS); + softirq_count(), ROEC_ARGS); /* * non-preemptible RCU in a preemptible kernel uses preempt_disable() @@ -1965,6 +2029,9 @@ static void rcutorture_one_extend_check(char *s, int curstate, int new, int old, if (!IS_ENABLED(CONFIG_PREEMPT_RCU)) mask |= RCUTORTURE_RDR_PREEMPT | RCUTORTURE_RDR_SCHED; + if (IS_ENABLED(CONFIG_PREEMPT_RT) && softirq_count()) + mask |= RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH; + WARN_ONCE(cur_ops->readlock_nesting && !(curstate & mask) && cur_ops->readlock_nesting() > 0, ROEC_ARGS); } @@ -1978,8 +2045,7 @@ static void rcutorture_one_extend_check(char *s, int curstate, int new, int old, * beginning or end of the critical section and if there was actually a * change, do a ->read_delay(). */ -static void rcutorture_one_extend(int *readstate, int newstate, bool insoftirq, - struct torture_random_state *trsp, +static void rcutorture_one_extend(int *readstate, int newstate, struct torture_random_state *trsp, struct rt_read_seg *rtrsp) { bool first; @@ -1993,8 +2059,8 @@ static void rcutorture_one_extend(int *readstate, int newstate, bool insoftirq, first = idxold1 == 0; WARN_ON_ONCE(idxold2 < 0); - WARN_ON_ONCE(idxold2 & ~RCUTORTURE_RDR_ALLBITS); - rcutorture_one_extend_check("before change", idxold1, statesnew, statesold, insoftirq); + WARN_ON_ONCE(idxold2 & ~(RCUTORTURE_RDR_ALLBITS | RCUTORTURE_RDR_UPDOWN)); + rcutorture_one_extend_check("before change", idxold1, statesnew, statesold); rtrsp->rt_readstate = newstate; /* First, put new protection in place to avoid critical-section gap. */ @@ -2014,8 +2080,7 @@ static void rcutorture_one_extend(int *readstate, int newstate, bool insoftirq, idxnew2 = (cur_ops->readlock() << RCUTORTURE_RDR_SHIFT_2) & RCUTORTURE_RDR_MASK_2; // Complain unless both the old and the new protection is in place. - rcutorture_one_extend_check("during change", - idxold1 | statesnew, statesnew, statesold, insoftirq); + rcutorture_one_extend_check("during change", idxold1 | statesnew, statesnew, statesold); // Sample CPU under both sets of protections to reduce confusion. if (IS_ENABLED(CONFIG_RCU_TORTURE_TEST_LOG_CPU)) { @@ -2069,6 +2134,11 @@ static void rcutorture_one_extend(int *readstate, int newstate, bool insoftirq, if (lockit) raw_spin_unlock_irqrestore(¤t->pi_lock, flags); } + if (statesold & RCUTORTURE_RDR_UPDOWN) { + cur_ops->up_read((idxold1 & RCUTORTURE_RDR_MASK_1) >> RCUTORTURE_RDR_SHIFT_1); + WARN_ON_ONCE(idxnew1 != -1); + idxold1 = 0; + } /* Delay if neither beginning nor end and there was a change. */ if ((statesnew || statesold) && *readstate && newstate) @@ -2085,7 +2155,7 @@ static void rcutorture_one_extend(int *readstate, int newstate, bool insoftirq, WARN_ON_ONCE(*readstate < 0); if (WARN_ON_ONCE(*readstate & ~RCUTORTURE_RDR_ALLBITS)) pr_info("Unexpected readstate value of %#x\n", *readstate); - rcutorture_one_extend_check("after change", *readstate, statesnew, statesold, insoftirq); + rcutorture_one_extend_check("after change", *readstate, statesnew, statesold); } /* Return the biggest extendables mask given current RCU and boot parameters. */ @@ -2152,8 +2222,7 @@ rcutorture_extend_mask(int oldmask, struct torture_random_state *trsp) * critical section. */ static struct rt_read_seg * -rcutorture_loop_extend(int *readstate, bool insoftirq, struct torture_random_state *trsp, - struct rt_read_seg *rtrsp) +rcutorture_loop_extend(int *readstate, struct torture_random_state *trsp, struct rt_read_seg *rtrsp) { int i; int j; @@ -2167,7 +2236,8 @@ rcutorture_loop_extend(int *readstate, bool insoftirq, struct torture_random_sta i = ((i | (i >> 3)) & RCUTORTURE_RDR_MAX_LOOPS) + 1; for (j = 0; j < i; j++) { mask = rcutorture_extend_mask(*readstate, trsp); - rcutorture_one_extend(readstate, mask, insoftirq, trsp, &rtrsp[j]); + WARN_ON_ONCE(mask & RCUTORTURE_RDR_UPDOWN); + rcutorture_one_extend(readstate, mask, trsp, &rtrsp[j]); } return &rtrsp[j]; } @@ -2209,10 +2279,11 @@ static bool rcu_torture_one_read_start(struct rcu_torture_one_read_state *rtorsp rtorsp->started = cur_ops->get_gp_seq(); rtorsp->ts = rcu_trace_clock_local(); rtorsp->p = rcu_dereference_check(rcu_torture_current, - !cur_ops->readlock_held || cur_ops->readlock_held()); + !cur_ops->readlock_held || cur_ops->readlock_held() || + (rtorsp->readstate & RCUTORTURE_RDR_UPDOWN)); if (rtorsp->p == NULL) { /* Wait for rcu_torture_writer to get underway */ - rcutorture_one_extend(&rtorsp->readstate, 0, myid < 0, trsp, rtorsp->rtrsp); + rcutorture_one_extend(&rtorsp->readstate, 0, trsp, rtorsp->rtrsp); return false; } if (rtorsp->p->rtort_mbtest == 0) @@ -2226,7 +2297,7 @@ static bool rcu_torture_one_read_start(struct rcu_torture_one_read_state *rtorsp * critical sections and check for errors. */ static void rcu_torture_one_read_end(struct rcu_torture_one_read_state *rtorsp, - struct torture_random_state *trsp, long myid) + struct torture_random_state *trsp) { int i; unsigned long completed; @@ -2273,7 +2344,7 @@ static void rcu_torture_one_read_end(struct rcu_torture_one_read_state *rtorsp, } if (cur_ops->reader_blocked) preempted = cur_ops->reader_blocked(); - rcutorture_one_extend(&rtorsp->readstate, 0, myid < 0, trsp, rtorsp->rtrsp); + rcutorture_one_extend(&rtorsp->readstate, 0, trsp, rtorsp->rtrsp); WARN_ON_ONCE(rtorsp->readstate); // This next splat is expected behavior if leakpointer, especially // for CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels. @@ -2302,13 +2373,14 @@ static bool rcu_torture_one_read(struct torture_random_state *trsp, long myid) WARN_ON_ONCE(!rcu_is_watching()); init_rcu_torture_one_read_state(&rtors, trsp); newstate = rcutorture_extend_mask(rtors.readstate, trsp); - rcutorture_one_extend(&rtors.readstate, newstate, myid < 0, trsp, rtors.rtrsp++); + WARN_ON_ONCE(newstate & RCUTORTURE_RDR_UPDOWN); + rcutorture_one_extend(&rtors.readstate, newstate, trsp, rtors.rtrsp++); if (!rcu_torture_one_read_start(&rtors, trsp, myid)) { - rcutorture_one_extend(&rtors.readstate, 0, myid < 0, trsp, rtors.rtrsp); + rcutorture_one_extend(&rtors.readstate, 0, trsp, rtors.rtrsp); return false; } - rtors.rtrsp = rcutorture_loop_extend(&rtors.readstate, myid < 0, trsp, rtors.rtrsp); - rcu_torture_one_read_end(&rtors, trsp, myid); + rtors.rtrsp = rcutorture_loop_extend(&rtors.readstate, trsp, rtors.rtrsp); + rcu_torture_one_read_end(&rtors, trsp); return true; } @@ -2378,6 +2450,152 @@ rcu_torture_reader(void *arg) return 0; } +struct rcu_torture_one_read_state_updown { + struct hrtimer rtorsu_hrt; + bool rtorsu_inuse; + ktime_t rtorsu_kt; + int rtorsu_cpu; + unsigned long rtorsu_j; + unsigned long rtorsu_ndowns; + unsigned long rtorsu_nups; + unsigned long rtorsu_nmigrates; + struct torture_random_state rtorsu_trs; + struct rcu_torture_one_read_state rtorsu_rtors; +}; + +static struct rcu_torture_one_read_state_updown *updownreaders; +static DEFINE_TORTURE_RANDOM(rcu_torture_updown_rand); +static int rcu_torture_updown(void *arg); + +static enum hrtimer_restart rcu_torture_updown_hrt(struct hrtimer *hrtp) +{ + int cpu = raw_smp_processor_id(); + struct rcu_torture_one_read_state_updown *rtorsup; + + rtorsup = container_of(hrtp, struct rcu_torture_one_read_state_updown, rtorsu_hrt); + rcu_torture_one_read_end(&rtorsup->rtorsu_rtors, &rtorsup->rtorsu_trs); + WARN_ONCE(rtorsup->rtorsu_nups >= rtorsup->rtorsu_ndowns, "%s: Up without matching down #%zu.\n", __func__, rtorsup - updownreaders); + WRITE_ONCE(rtorsup->rtorsu_nups, rtorsup->rtorsu_nups + 1); + WRITE_ONCE(rtorsup->rtorsu_nmigrates, + rtorsup->rtorsu_nmigrates + (cpu != rtorsup->rtorsu_cpu)); + smp_store_release(&rtorsup->rtorsu_inuse, false); + return HRTIMER_NORESTART; +} + +static int rcu_torture_updown_init(void) +{ + int i; + struct torture_random_state *rand = &rcu_torture_updown_rand; + int ret; + + if (n_up_down < 0) + return 0; + if (!srcu_torture_have_up_down()) { + VERBOSE_TOROUT_STRING("rcu_torture_updown_init: Disabling up/down reader tests due to lack of primitives"); + return 0; + } + updownreaders = kcalloc(n_up_down, sizeof(*updownreaders), GFP_KERNEL); + if (!updownreaders) { + VERBOSE_TOROUT_STRING("rcu_torture_updown_init: Out of memory, disabling up/down reader tests"); + return -ENOMEM; + } + for (i = 0; i < n_up_down; i++) { + init_rcu_torture_one_read_state(&updownreaders[i].rtorsu_rtors, rand); + hrtimer_setup(&updownreaders[i].rtorsu_hrt, rcu_torture_updown_hrt, CLOCK_MONOTONIC, + HRTIMER_MODE_REL | HRTIMER_MODE_HARD); + torture_random_init(&updownreaders[i].rtorsu_trs); + init_rcu_torture_one_read_state(&updownreaders[i].rtorsu_rtors, + &updownreaders[i].rtorsu_trs); + } + ret = torture_create_kthread(rcu_torture_updown, rand, updown_task); + if (ret) { + kfree(updownreaders); + updownreaders = NULL; + } + return ret; +} + +static void rcu_torture_updown_cleanup(void) +{ + struct rcu_torture_one_read_state_updown *rtorsup; + + for (rtorsup = updownreaders; rtorsup < &updownreaders[n_up_down]; rtorsup++) { + if (!smp_load_acquire(&rtorsup->rtorsu_inuse)) + continue; + if (hrtimer_cancel(&rtorsup->rtorsu_hrt) || WARN_ON_ONCE(rtorsup->rtorsu_inuse)) { + rcu_torture_one_read_end(&rtorsup->rtorsu_rtors, &rtorsup->rtorsu_trs); + WARN_ONCE(rtorsup->rtorsu_nups >= rtorsup->rtorsu_ndowns, "%s: Up without matching down #%zu.\n", __func__, rtorsup - updownreaders); + WRITE_ONCE(rtorsup->rtorsu_nups, rtorsup->rtorsu_nups + 1); + smp_store_release(&rtorsup->rtorsu_inuse, false); + } + + } + kfree(updownreaders); + updownreaders = NULL; +} + +// Do one reader for rcu_torture_updown(). +static void rcu_torture_updown_one(struct rcu_torture_one_read_state_updown *rtorsup) +{ + int idx; + int rawidx; + ktime_t t; + + init_rcu_torture_one_read_state(&rtorsup->rtorsu_rtors, &rtorsup->rtorsu_trs); + rawidx = cur_ops->down_read(); + WRITE_ONCE(rtorsup->rtorsu_ndowns, rtorsup->rtorsu_ndowns + 1); + idx = (rawidx << RCUTORTURE_RDR_SHIFT_1) & RCUTORTURE_RDR_MASK_1; + rtorsup->rtorsu_rtors.readstate = idx | RCUTORTURE_RDR_UPDOWN; + rtorsup->rtorsu_rtors.rtrsp++; + rtorsup->rtorsu_cpu = raw_smp_processor_id(); + if (!rcu_torture_one_read_start(&rtorsup->rtorsu_rtors, &rtorsup->rtorsu_trs, -1)) { + WARN_ONCE(rtorsup->rtorsu_nups >= rtorsup->rtorsu_ndowns, "%s: Up without matching down #%zu.\n", __func__, rtorsup - updownreaders); + WRITE_ONCE(rtorsup->rtorsu_nups, rtorsup->rtorsu_nups + 1); + schedule_timeout_idle(HZ); + return; + } + smp_store_release(&rtorsup->rtorsu_inuse, true); + t = torture_random(&rtorsup->rtorsu_trs) & 0xfffff; // One per million. + if (t < 10 * 1000) + t = 200 * 1000 * 1000; + hrtimer_start(&rtorsup->rtorsu_hrt, t, HRTIMER_MODE_REL | HRTIMER_MODE_HARD); + smp_mb(); // Sample jiffies after posting hrtimer. + rtorsup->rtorsu_j = jiffies; // Not used by hrtimer handler. + rtorsup->rtorsu_kt = t; +} + +/* + * RCU torture up/down reader kthread, starting RCU readers in kthread + * context and ending them in hrtimer handlers. Otherwise similar to + * rcu_torture_reader(). + */ +static int +rcu_torture_updown(void *arg) +{ + unsigned long j; + struct rcu_torture_one_read_state_updown *rtorsup; + + VERBOSE_TOROUT_STRING("rcu_torture_updown task started"); + do { + for (rtorsup = updownreaders; rtorsup < &updownreaders[n_up_down]; rtorsup++) { + if (torture_must_stop()) + break; + j = smp_load_acquire(&jiffies); // Time before ->rtorsu_inuse. + if (smp_load_acquire(&rtorsup->rtorsu_inuse)) { + WARN_ONCE(time_after(j, rtorsup->rtorsu_j + 1 + HZ * 10), + "hrtimer queued at jiffies %lu for %lld ns took %lu jiffies\n", rtorsup->rtorsu_j, rtorsup->rtorsu_kt, j - rtorsup->rtorsu_j); + continue; + } + rcu_torture_updown_one(rtorsup); + } + torture_hrtimeout_ms(1, 1000, &rcu_torture_updown_rand); + stutter_wait("rcu_torture_updown"); + } while (!torture_must_stop()); + rcu_torture_updown_cleanup(); + torture_kthread_stopping("rcu_torture_updown"); + return 0; +} + /* * Randomly Toggle CPUs' callback-offload state. This uses hrtimers to * increase race probabilities and fuzzes the interval between toggling. @@ -2441,6 +2659,10 @@ rcu_torture_stats_print(void) long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 }; long batchsummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 }; long n_gpwraps = 0; + unsigned long ndowns = 0; + unsigned long nunexpired = 0; + unsigned long nmigrates = 0; + unsigned long nups = 0; struct rcu_torture *rtcp; static unsigned long rtcv_snap = ULONG_MAX; static bool splatted; @@ -2454,10 +2676,18 @@ rcu_torture_stats_print(void) if (cur_ops->get_gpwrap_count) n_gpwraps += cur_ops->get_gpwrap_count(cpu); } + if (updownreaders) { + for (i = 0; i < n_up_down; i++) { + ndowns += READ_ONCE(updownreaders[i].rtorsu_ndowns); + nups += READ_ONCE(updownreaders[i].rtorsu_nups); + nunexpired += READ_ONCE(updownreaders[i].rtorsu_inuse); + nmigrates += READ_ONCE(updownreaders[i].rtorsu_nmigrates); + } + } for (i = RCU_TORTURE_PIPE_LEN; i >= 0; i--) { if (pipesummary[i] != 0) break; - } + } // The value of variable "i" is used later, so don't clobber it! pr_alert("%s%s ", torture_type, TORTURE_FLAG); rtcp = rcu_access_pointer(rcu_torture_current); @@ -2478,6 +2708,8 @@ rcu_torture_stats_print(void) n_rcu_torture_boost_failure, n_rcu_torture_boosts, atomic_long_read(&n_rcu_torture_timers)); + if (updownreaders) + pr_cont("ndowns: %lu nups: %lu nhrt: %lu nmigrates: %lu ", ndowns, nups, nunexpired, nmigrates); torture_onoff_stats(); pr_cont("barrier: %ld/%ld:%ld ", data_race(n_barrier_successes), @@ -2632,7 +2864,7 @@ rcu_torture_print_module_parms(struct rcu_torture_ops *cur_ops, const char *tag) "reader_flavor=%x " "nocbs_nthreads=%d nocbs_toggle=%d " "test_nmis=%d " - "preempt_duration=%d preempt_interval=%d\n", + "preempt_duration=%d preempt_interval=%d n_up_down=%d\n", torture_type, tag, nrealreaders, nrealfakewriters, stat_interval, verbose, test_no_idle_hz, shuffle_interval, stutter, irqreader, fqs_duration, fqs_holdoff, fqs_stutter, @@ -2646,7 +2878,7 @@ rcu_torture_print_module_parms(struct rcu_torture_ops *cur_ops, const char *tag) reader_flavor, nocbs_nthreads, nocbs_toggle, test_nmis, - preempt_duration, preempt_interval); + preempt_duration, preempt_interval, n_up_down); } static int rcutorture_booster_cleanup(unsigned int cpu) @@ -3749,6 +3981,10 @@ rcu_torture_cleanup(void) nocb_tasks = NULL; } + if (updown_task) { + torture_stop_kthread(rcu_torture_updown, updown_task); + updown_task = NULL; + } if (reader_tasks) { for (i = 0; i < nrealreaders; i++) torture_stop_kthread(rcu_torture_reader, @@ -4245,11 +4481,6 @@ rcu_torture_init(void) /* Start up the kthreads. */ rcu_torture_write_types(); - firsterr = torture_create_kthread(rcu_torture_writer, NULL, - writer_task); - if (torture_init_error(firsterr)) - goto unwind; - if (nrealfakewriters > 0) { fakewriter_tasks = kcalloc(nrealfakewriters, sizeof(fakewriter_tasks[0]), @@ -4282,6 +4513,15 @@ rcu_torture_init(void) if (torture_init_error(firsterr)) goto unwind; } + + firsterr = torture_create_kthread(rcu_torture_writer, NULL, + writer_task); + if (torture_init_error(firsterr)) + goto unwind; + + firsterr = rcu_torture_updown_init(); + if (torture_init_error(firsterr)) + goto unwind; nrealnocbers = nocbs_nthreads; if (WARN_ON(nrealnocbers < 0)) nrealnocbers = 1; diff --git a/kernel/rcu/refscale.c b/kernel/rcu/refscale.c index f11a7c2af778..df646e0694a8 100644 --- a/kernel/rcu/refscale.c +++ b/kernel/rcu/refscale.c @@ -85,7 +85,7 @@ torture_param(int, holdoff, IS_BUILTIN(CONFIG_RCU_REF_SCALE_TEST) ? 10 : 0, // Number of typesafe_lookup structures, that is, the degree of concurrency. torture_param(long, lookup_instances, 0, "Number of typesafe_lookup structures."); // Number of loops per experiment, all readers execute operations concurrently. -torture_param(long, loops, 10000, "Number of loops per experiment."); +torture_param(int, loops, 10000, "Number of loops per experiment."); // Number of readers, with -1 defaulting to about 75% of the CPUs. torture_param(int, nreaders, -1, "Number of readers, -1 for 75% of CPUs."); // Number of runs. @@ -246,36 +246,6 @@ static const struct ref_scale_ops srcu_fast_ops = { .name = "srcu-fast" }; -static void srcu_lite_ref_scale_read_section(const int nloops) -{ - int i; - int idx; - - for (i = nloops; i >= 0; i--) { - idx = srcu_read_lock_lite(srcu_ctlp); - srcu_read_unlock_lite(srcu_ctlp, idx); - } -} - -static void srcu_lite_ref_scale_delay_section(const int nloops, const int udl, const int ndl) -{ - int i; - int idx; - - for (i = nloops; i >= 0; i--) { - idx = srcu_read_lock_lite(srcu_ctlp); - un_delay(udl, ndl); - srcu_read_unlock_lite(srcu_ctlp, idx); - } -} - -static const struct ref_scale_ops srcu_lite_ops = { - .init = rcu_sync_scale_init, - .readsection = srcu_lite_ref_scale_read_section, - .delaysection = srcu_lite_ref_scale_delay_section, - .name = "srcu-lite" -}; - #ifdef CONFIG_TASKS_RCU // Definitions for RCU Tasks ref scale testing: Empty read markers. @@ -1140,7 +1110,7 @@ static void ref_scale_print_module_parms(const struct ref_scale_ops *cur_ops, const char *tag) { pr_alert("%s" SCALE_FLAG - "--- %s: verbose=%d verbose_batched=%d shutdown=%d holdoff=%d lookup_instances=%ld loops=%ld nreaders=%d nruns=%d readdelay=%d\n", scale_type, tag, + "--- %s: verbose=%d verbose_batched=%d shutdown=%d holdoff=%d lookup_instances=%ld loops=%d nreaders=%d nruns=%d readdelay=%d\n", scale_type, tag, verbose, verbose_batched, shutdown, holdoff, lookup_instances, loops, nreaders, nruns, readdelay); } @@ -1193,7 +1163,7 @@ ref_scale_init(void) long i; int firsterr = 0; static const struct ref_scale_ops *scale_ops[] = { - &rcu_ops, &srcu_ops, &srcu_fast_ops, &srcu_lite_ops, RCU_TRACE_OPS RCU_TASKS_OPS + &rcu_ops, &srcu_ops, &srcu_fast_ops, RCU_TRACE_OPS RCU_TASKS_OPS &refcnt_ops, &rwlock_ops, &rwsem_ops, &lock_ops, &lock_irq_ops, &acqrel_ops, &sched_clock_ops, &clock_ops, &jiffies_ops, &typesafe_ref_ops, &typesafe_lock_ops, &typesafe_seqlock_ops, @@ -1238,12 +1208,16 @@ ref_scale_init(void) // Reader tasks (default to ~75% of online CPUs). if (nreaders < 0) nreaders = (num_online_cpus() >> 1) + (num_online_cpus() >> 2); - if (WARN_ONCE(loops <= 0, "%s: loops = %ld, adjusted to 1\n", __func__, loops)) + if (WARN_ONCE(loops <= 0, "%s: loops = %d, adjusted to 1\n", __func__, loops)) loops = 1; if (WARN_ONCE(nreaders <= 0, "%s: nreaders = %d, adjusted to 1\n", __func__, nreaders)) nreaders = 1; if (WARN_ONCE(nruns <= 0, "%s: nruns = %d, adjusted to 1\n", __func__, nruns)) nruns = 1; + if (WARN_ONCE(loops > INT_MAX / nreaders, + "%s: nreaders * loops will overflow, adjusted loops to %d", + __func__, INT_MAX / nreaders)) + loops = INT_MAX / nreaders; reader_tasks = kcalloc(nreaders, sizeof(reader_tasks[0]), GFP_KERNEL); if (!reader_tasks) { diff --git a/kernel/rcu/srcutree.c b/kernel/rcu/srcutree.c index 48047260697e..c5e8ebc493d5 100644 --- a/kernel/rcu/srcutree.c +++ b/kernel/rcu/srcutree.c @@ -502,6 +502,8 @@ static bool srcu_readers_active_idx_check(struct srcu_struct *ssp, int idx) */ if (!did_gp) smp_mb(); /* A */ + else if (srcu_gp_is_expedited(ssp)) + synchronize_rcu_expedited(); /* X */ else synchronize_rcu(); /* X */ diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 14d4499c6fc3..174ee243b349 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -160,7 +160,6 @@ static void rcu_report_qs_rnp(unsigned long mask, struct rcu_node *rnp, unsigned long gps, unsigned long flags); static void invoke_rcu_core(void); static void rcu_report_exp_rdp(struct rcu_data *rdp); -static void sync_sched_exp_online_cleanup(int cpu); static void check_cb_ovld_locked(struct rcu_data *rdp, struct rcu_node *rnp); static bool rcu_rdp_is_offloaded(struct rcu_data *rdp); static bool rcu_rdp_cpu_online(struct rcu_data *rdp); @@ -377,7 +376,7 @@ EXPORT_SYMBOL_GPL(rcu_momentary_eqs); */ static int rcu_is_cpu_rrupt_from_idle(void) { - long nesting; + long nmi_nesting = ct_nmi_nesting(); /* * Usually called from the tick; but also used from smp_function_call() @@ -389,21 +388,28 @@ static int rcu_is_cpu_rrupt_from_idle(void) /* Check for counter underflows */ RCU_LOCKDEP_WARN(ct_nesting() < 0, "RCU nesting counter underflow!"); - RCU_LOCKDEP_WARN(ct_nmi_nesting() <= 0, - "RCU nmi_nesting counter underflow/zero!"); - /* Are we at first interrupt nesting level? */ - nesting = ct_nmi_nesting(); - if (nesting > 1) + /* Non-idle interrupt or nested idle interrupt */ + if (nmi_nesting > 1) return false; /* - * If we're not in an interrupt, we must be in the idle task! + * Non nested idle interrupt (interrupting section where RCU + * wasn't watching). */ - WARN_ON_ONCE(!nesting && !is_idle_task(current)); + if (nmi_nesting == 1) + return true; - /* Does CPU appear to be idle from an RCU standpoint? */ - return ct_nesting() == 0; + /* Not in an interrupt */ + if (!nmi_nesting) { + RCU_LOCKDEP_WARN(!in_task() || !is_idle_task(current), + "RCU nmi_nesting counter not in idle task!"); + return !rcu_is_watching_curr_cpu(); + } + + RCU_LOCKDEP_WARN(1, "RCU nmi_nesting counter underflow/zero!"); + + return false; } #define DEFAULT_RCU_BLIMIT (IS_ENABLED(CONFIG_RCU_STRICT_GRACE_PERIOD) ? 1000 : 10) @@ -1625,8 +1631,10 @@ static void rcu_sr_put_wait_head(struct llist_node *node) atomic_set_release(&sr_wn->inuse, 0); } -/* Disabled by default. */ -static int rcu_normal_wake_from_gp; +/* Enable rcu_normal_wake_from_gp automatically on small systems. */ +#define WAKE_FROM_GP_CPU_THRESHOLD 16 + +static int rcu_normal_wake_from_gp = -1; module_param(rcu_normal_wake_from_gp, int, 0644); static struct workqueue_struct *sync_wq; @@ -1829,6 +1837,18 @@ static noinline_for_stack bool rcu_gp_init(void) start_new_poll = rcu_sr_normal_gp_init(); /* Record GP times before starting GP, hence rcu_seq_start(). */ old_gp_seq = rcu_state.gp_seq; + /* + * Critical ordering: rcu_seq_start() must happen BEFORE the CPU hotplug + * scan below. Otherwise we risk a race where a newly onlining CPU could + * be missed by the current grace period, potentially leading to + * use-after-free errors. For a detailed explanation of this race, see + * Documentation/RCU/Design/Requirements/Requirements.rst in the + * "Hotplug CPU" section. + * + * Also note that the root rnp's gp_seq is kept separate from, and lags, + * the rcu_state's gp_seq, for a reason. See the Quick-Quiz on + * Single-node systems for more details (in Data-Structures.rst). + */ rcu_seq_start(&rcu_state.gp_seq); /* Ensure that rcu_seq_done_exact() guardband doesn't give false positives. */ WARN_ON_ONCE(IS_ENABLED(CONFIG_PROVE_RCU) && @@ -1865,6 +1885,10 @@ static noinline_for_stack bool rcu_gp_init(void) /* Exclude CPU hotplug operations. */ rcu_for_each_leaf_node(rnp) { local_irq_disable(); + /* + * Serialize with CPU offline. See Requirements.rst > Hotplug CPU > + * Concurrent Quiescent State Reporting for Offline CPUs. + */ arch_spin_lock(&rcu_state.ofl_lock); raw_spin_lock_rcu_node(rnp); if (rnp->qsmaskinit == rnp->qsmaskinitnext && @@ -1939,7 +1963,12 @@ static noinline_for_stack bool rcu_gp_init(void) trace_rcu_grace_period_init(rcu_state.name, rnp->gp_seq, rnp->level, rnp->grplo, rnp->grphi, rnp->qsmask); - /* Quiescent states for tasks on any now-offline CPUs. */ + /* + * Quiescent states for tasks on any now-offline CPUs. Since we + * released the ofl and rnp lock before this loop, CPUs might + * have gone offline and we have to report QS on their behalf. + * See Requirements.rst > Hotplug CPU > Concurrent QS Reporting. + */ mask = rnp->qsmask & ~rnp->qsmaskinitnext; rnp->rcu_gp_init_mask = mask; if ((mask || rnp->wait_blkd_tasks) && rcu_is_leaf_node(rnp)) @@ -3243,7 +3272,7 @@ static void synchronize_rcu_normal(void) trace_rcu_sr_normal(rcu_state.name, &rs.head, TPS("request")); - if (!READ_ONCE(rcu_normal_wake_from_gp)) { + if (READ_ONCE(rcu_normal_wake_from_gp) < 1) { wait_rcu_gp(call_rcu_hurry); goto trace_complete_out; } @@ -4268,7 +4297,6 @@ int rcutree_online_cpu(unsigned int cpu) raw_spin_unlock_irqrestore_rcu_node(rnp, flags); if (rcu_scheduler_active == RCU_SCHEDULER_INACTIVE) return 0; /* Too early in boot for scheduler work. */ - sync_sched_exp_online_cleanup(cpu); // Stop-machine done, so allow nohz_full to disable tick. tick_dep_clear(TICK_DEP_BIT_RCU); @@ -4358,6 +4386,12 @@ void rcutree_report_cpu_dead(void) * may introduce a new READ-side while it is actually off the QS masks. */ lockdep_assert_irqs_disabled(); + /* + * CPUHP_AP_SMPCFD_DYING was the last call for rcu_exp_handler() execution. + * The requested QS must have been reported on the last context switch + * from stop machine to idle. + */ + WARN_ON_ONCE(rdp->cpu_no_qs.b.exp); // Do any dangling deferred wakeups. do_nocb_deferred_wakeup(rdp); @@ -4365,6 +4399,13 @@ void rcutree_report_cpu_dead(void) /* Remove outgoing CPU from mask in the leaf rcu_node structure. */ mask = rdp->grpmask; + + /* + * Hold the ofl_lock and rnp lock to avoid races between CPU going + * offline and doing a QS report (as below), versus rcu_gp_init(). + * See Requirements.rst > Hotplug CPU > Concurrent QS Reporting section + * for more details. + */ arch_spin_lock(&rcu_state.ofl_lock); raw_spin_lock_irqsave_rcu_node(rnp, flags); /* Enforce GP memory-order guarantee. */ rdp->rcu_ofl_gp_seq = READ_ONCE(rcu_state.gp_seq); @@ -4375,6 +4416,7 @@ void rcutree_report_cpu_dead(void) rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags); raw_spin_lock_irqsave_rcu_node(rnp, flags); } + /* Clear from ->qsmaskinitnext to mark offline. */ WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext & ~mask); raw_spin_unlock_irqrestore_rcu_node(rnp, flags); arch_spin_unlock(&rcu_state.ofl_lock); @@ -4847,6 +4889,12 @@ void __init rcu_init(void) sync_wq = alloc_workqueue("sync_wq", WQ_MEM_RECLAIM, 0); WARN_ON(!sync_wq); + /* Respect if explicitly disabled via a boot parameter. */ + if (rcu_normal_wake_from_gp < 0) { + if (num_possible_cpus() <= WAKE_FROM_GP_CPU_THRESHOLD) + rcu_normal_wake_from_gp = 1; + } + /* Fill in default value for rcutree.qovld boot parameter. */ /* -After- the rcu_node ->lock fields are initialized! */ if (qovld < 0) diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index 3830c19cf2f6..de6ca13a7b5f 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -174,6 +174,17 @@ struct rcu_snap_record { unsigned long jiffies; /* Track jiffies value */ }; +/* + * An IRQ work (deferred_qs_iw) is used by RCU to get the scheduler's attention. + * to report quiescent states at the soonest possible time. + * The request can be in one of the following states: + * - DEFER_QS_IDLE: An IRQ work is yet to be scheduled. + * - DEFER_QS_PENDING: An IRQ work was scheduled but either not yet run, or it + * ran and we still haven't reported a quiescent state. + */ +#define DEFER_QS_IDLE 0 +#define DEFER_QS_PENDING 1 + /* Per-CPU data for read-copy update. */ struct rcu_data { /* 1) quiescent-state and grace-period handling : */ @@ -192,7 +203,7 @@ struct rcu_data { /* during and after the last grace */ /* period it is aware of. */ struct irq_work defer_qs_iw; /* Obtain later scheduler attention. */ - bool defer_qs_iw_pending; /* Scheduler attention pending? */ + int defer_qs_iw_pending; /* Scheduler attention pending? */ struct work_struct strict_work; /* Schedule readers for strict GPs. */ /* 2) batch handling */ diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h index c36c7d5575ca..6058a734090c 100644 --- a/kernel/rcu/tree_exp.h +++ b/kernel/rcu/tree_exp.h @@ -141,6 +141,13 @@ static void __maybe_unused sync_exp_reset_tree(void) raw_spin_lock_irqsave_rcu_node(rnp, flags); WARN_ON_ONCE(rnp->expmask); WRITE_ONCE(rnp->expmask, rnp->expmaskinit); + /* + * Need to wait for any blocked tasks as well. Note that + * additional blocking tasks will also block the expedited GP + * until such time as the ->expmask bits are cleared. + */ + if (rcu_is_leaf_node(rnp) && rcu_preempt_has_tasks(rnp)) + WRITE_ONCE(rnp->exp_tasks, rnp->blkd_tasks.next); raw_spin_unlock_irqrestore_rcu_node(rnp, flags); } } @@ -393,13 +400,6 @@ static void __sync_rcu_exp_select_node_cpus(struct rcu_exp_work *rewp) } mask_ofl_ipi = rnp->expmask & ~mask_ofl_test; - /* - * Need to wait for any blocked tasks as well. Note that - * additional blocking tasks will also block the expedited GP - * until such time as the ->expmask bits are cleared. - */ - if (rcu_preempt_has_tasks(rnp)) - WRITE_ONCE(rnp->exp_tasks, rnp->blkd_tasks.next); raw_spin_unlock_irqrestore_rcu_node(rnp, flags); /* IPI the remaining CPUs for expedited quiescent state. */ @@ -751,12 +751,8 @@ static void rcu_exp_handler(void *unused) struct task_struct *t = current; /* - * First, is there no need for a quiescent state from this CPU, - * or is this CPU already looking for a quiescent state for the - * current grace period? If either is the case, just leave. - * However, this should not happen due to the preemptible - * sync_sched_exp_online_cleanup() implementation being a no-op, - * so warn if this does happen. + * WARN if the CPU is unexpectedly already looking for a + * QS or has already reported one. */ ASSERT_EXCLUSIVE_WRITER_SCOPED(rdp->cpu_no_qs.b.exp); if (WARN_ON_ONCE(!(READ_ONCE(rnp->expmask) & rdp->grpmask) || @@ -803,11 +799,6 @@ static void rcu_exp_handler(void *unused) WARN_ON_ONCE(1); } -/* PREEMPTION=y, so no PREEMPTION=n expedited grace period to clean up after. */ -static void sync_sched_exp_online_cleanup(int cpu) -{ -} - /* * Scan the current list of tasks blocked within RCU read-side critical * sections, printing out the tid of each that is blocking the current @@ -885,38 +876,6 @@ static void rcu_exp_handler(void *unused) rcu_exp_need_qs(); } -/* Send IPI for expedited cleanup if needed at end of CPU-hotplug operation. */ -static void sync_sched_exp_online_cleanup(int cpu) -{ - unsigned long flags; - int my_cpu; - struct rcu_data *rdp; - int ret; - struct rcu_node *rnp; - - rdp = per_cpu_ptr(&rcu_data, cpu); - rnp = rdp->mynode; - my_cpu = get_cpu(); - /* Quiescent state either not needed or already requested, leave. */ - if (!(READ_ONCE(rnp->expmask) & rdp->grpmask) || - READ_ONCE(rdp->cpu_no_qs.b.exp)) { - put_cpu(); - return; - } - /* Quiescent state needed on current CPU, so set it up locally. */ - if (my_cpu == cpu) { - local_irq_save(flags); - rcu_exp_need_qs(); - local_irq_restore(flags); - put_cpu(); - return; - } - /* Quiescent state needed on some other CPU, send IPI. */ - ret = smp_call_function_single(cpu, rcu_exp_handler, NULL, 0); - put_cpu(); - WARN_ON_ONCE(ret); -} - /* * Because preemptible RCU does not exist, we never have to check for * tasks blocked within RCU read-side critical sections that are diff --git a/kernel/rcu/tree_nocb.h b/kernel/rcu/tree_nocb.h index b473ff056f49..e6cd56603cad 100644 --- a/kernel/rcu/tree_nocb.h +++ b/kernel/rcu/tree_nocb.h @@ -276,7 +276,7 @@ static void wake_nocb_gp_defer(struct rcu_data *rdp, int waketype, * callback storms, no need to wake up too early. */ if (waketype == RCU_NOCB_WAKE_LAZY && - rdp->nocb_defer_wakeup == RCU_NOCB_WAKE_NOT) { + rdp_gp->nocb_defer_wakeup == RCU_NOCB_WAKE_NOT) { mod_timer(&rdp_gp->nocb_timer, jiffies + rcu_get_jiffies_lazy_flush()); WRITE_ONCE(rdp_gp->nocb_defer_wakeup, waketype); } else if (waketype == RCU_NOCB_WAKE_BYPASS) { @@ -1146,7 +1146,6 @@ static bool rcu_nocb_rdp_offload_wait_cond(struct rcu_data *rdp) static int rcu_nocb_rdp_offload(struct rcu_data *rdp) { int wake_gp; - struct rcu_data *rdp_gp = rdp->nocb_gp_rdp; WARN_ON_ONCE(cpu_online(rdp->cpu)); /* @@ -1156,7 +1155,7 @@ static int rcu_nocb_rdp_offload(struct rcu_data *rdp) if (!rdp->nocb_gp_rdp) return -EINVAL; - if (WARN_ON_ONCE(!rdp_gp->nocb_gp_kthread)) + if (WARN_ON_ONCE(!rdp->nocb_gp_kthread)) return -EINVAL; pr_info("Offloading %d\n", rdp->cpu); @@ -1166,7 +1165,7 @@ static int rcu_nocb_rdp_offload(struct rcu_data *rdp) wake_gp = rcu_nocb_queue_toggle_rdp(rdp); if (wake_gp) - wake_up_process(rdp_gp->nocb_gp_kthread); + wake_up_process(rdp->nocb_gp_kthread); swait_event_exclusive(rdp->nocb_state_wq, rcu_nocb_rdp_offload_wait_cond(rdp)); @@ -1564,6 +1563,9 @@ static void show_rcu_nocb_state(struct rcu_data *rdp) if (rdp->nocb_gp_rdp == rdp) show_rcu_nocb_gp_state(rdp); + if (!rcu_segcblist_is_offloaded(&rdp->cblist)) + return; + nocb_next_rdp = list_next_or_null_rcu(&rdp->nocb_gp_rdp->nocb_head_rdp, &rdp->nocb_entry_rdp, typeof(*rdp), diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h index 0b0f56f6abc8..fc14adf15cbb 100644 --- a/kernel/rcu/tree_plugin.h +++ b/kernel/rcu/tree_plugin.h @@ -486,13 +486,16 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags) struct rcu_node *rnp; union rcu_special special; + rdp = this_cpu_ptr(&rcu_data); + if (rdp->defer_qs_iw_pending == DEFER_QS_PENDING) + rdp->defer_qs_iw_pending = DEFER_QS_IDLE; + /* * If RCU core is waiting for this CPU to exit its critical section, * report the fact that it has exited. Because irqs are disabled, * t->rcu_read_unlock_special cannot change. */ special = t->rcu_read_unlock_special; - rdp = this_cpu_ptr(&rcu_data); if (!special.s && !rdp->cpu_no_qs.b.exp) { local_irq_restore(flags); return; @@ -534,7 +537,6 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags) WARN_ON_ONCE(rnp->completedqs == rnp->gp_seq && (!empty_norm || rnp->qsmask)); empty_exp = sync_rcu_exp_done(rnp); - smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */ np = rcu_next_node_entry(t, rnp); list_del_init(&t->rcu_node_entry); t->rcu_blocked_node = NULL; @@ -624,10 +626,98 @@ notrace void rcu_preempt_deferred_qs(struct task_struct *t) */ static void rcu_preempt_deferred_qs_handler(struct irq_work *iwp) { + unsigned long flags; struct rcu_data *rdp; rdp = container_of(iwp, struct rcu_data, defer_qs_iw); - rdp->defer_qs_iw_pending = false; + local_irq_save(flags); + + /* + * If the IRQ work handler happens to run in the middle of RCU read-side + * critical section, it could be ineffective in getting the scheduler's + * attention to report a deferred quiescent state (the whole point of the + * IRQ work). For this reason, requeue the IRQ work. + * + * Basically, we want to avoid following situation: + * 1. rcu_read_unlock() queues IRQ work (state -> DEFER_QS_PENDING) + * 2. CPU enters new rcu_read_lock() + * 3. IRQ work runs but cannot report QS due to rcu_preempt_depth() > 0 + * 4. rcu_read_unlock() does not re-queue work (state still PENDING) + * 5. Deferred QS reporting does not happen. + */ + if (rcu_preempt_depth() > 0) + WRITE_ONCE(rdp->defer_qs_iw_pending, DEFER_QS_IDLE); + + local_irq_restore(flags); +} + +/* + * Check if expedited grace period processing during unlock is needed. + * + * This function determines whether expedited handling is required based on: + * 1. Task blocking an expedited grace period (based on a heuristic, could be + * false-positive, see below.) + * 2. CPU participating in an expedited grace period + * 3. Strict grace period mode requiring expedited handling + * 4. RCU priority deboosting needs when interrupts were disabled + * + * @t: The task being checked + * @rdp: The per-CPU RCU data + * @rnp: The RCU node for this CPU + * @irqs_were_disabled: Whether interrupts were disabled before rcu_read_unlock() + * + * Returns true if expedited processing of the rcu_read_unlock() is needed. + */ +static bool rcu_unlock_needs_exp_handling(struct task_struct *t, + struct rcu_data *rdp, + struct rcu_node *rnp, + bool irqs_were_disabled) +{ + /* + * Check if this task is blocking an expedited grace period. If the + * task was preempted within an RCU read-side critical section and is + * on the expedited grace period blockers list (exp_tasks), we need + * expedited handling to unblock the expedited GP. This is not an exact + * check because 't' might not be on the exp_tasks list at all - its + * just a fast heuristic that can be false-positive sometimes. + */ + if (t->rcu_blocked_node && READ_ONCE(t->rcu_blocked_node->exp_tasks)) + return true; + + /* + * Check if this CPU is participating in an expedited grace period. + * The expmask bitmap tracks which CPUs need to check in for the + * current expedited GP. If our CPU's bit is set, we need expedited + * handling to help complete the expedited GP. + */ + if (rdp->grpmask & READ_ONCE(rnp->expmask)) + return true; + + /* + * In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, all grace periods + * are treated as short for testing purposes even if that means + * disturbing the system more. Check if either: + * - This CPU has not yet reported a quiescent state, or + * - This task was preempted within an RCU critical section + * In either case, require expedited handling for strict GP mode. + */ + if (IS_ENABLED(CONFIG_RCU_STRICT_GRACE_PERIOD) && + ((rdp->grpmask & READ_ONCE(rnp->qsmask)) || t->rcu_blocked_node)) + return true; + + /* + * RCU priority boosting case: If a task is subject to RCU priority + * boosting and exits an RCU read-side critical section with interrupts + * disabled, we need expedited handling to ensure timely deboosting. + * Without this, a low-priority task could incorrectly run at high + * real-time priority for an extended period degrading real-time + * responsiveness. This applies to all CONFIG_RCU_BOOST=y kernels, + * not just to PREEMPT_RT. + */ + if (IS_ENABLED(CONFIG_RCU_BOOST) && irqs_were_disabled && t->rcu_blocked_node) + return true; + + return false; } /* @@ -649,18 +739,14 @@ static void rcu_read_unlock_special(struct task_struct *t) local_irq_save(flags); irqs_were_disabled = irqs_disabled_flags(flags); if (preempt_bh_were_disabled || irqs_were_disabled) { - bool expboost; // Expedited GP in flight or possible boosting. + bool needs_exp; // Expedited handling needed. struct rcu_data *rdp = this_cpu_ptr(&rcu_data); struct rcu_node *rnp = rdp->mynode; - expboost = (t->rcu_blocked_node && READ_ONCE(t->rcu_blocked_node->exp_tasks)) || - (rdp->grpmask & READ_ONCE(rnp->expmask)) || - (IS_ENABLED(CONFIG_RCU_STRICT_GRACE_PERIOD) && - ((rdp->grpmask & READ_ONCE(rnp->qsmask)) || t->rcu_blocked_node)) || - (IS_ENABLED(CONFIG_RCU_BOOST) && irqs_were_disabled && - t->rcu_blocked_node); + needs_exp = rcu_unlock_needs_exp_handling(t, rdp, rnp, irqs_were_disabled); + // Need to defer quiescent state until everything is enabled. - if (use_softirq && (in_hardirq() || (expboost && !irqs_were_disabled))) { + if (use_softirq && (in_hardirq() || (needs_exp && !irqs_were_disabled))) { // Using softirq, safe to awaken, and either the // wakeup is free or there is either an expedited // GP in flight or a potential need to deboost. @@ -673,17 +759,13 @@ static void rcu_read_unlock_special(struct task_struct *t) set_tsk_need_resched(current); set_preempt_need_resched(); if (IS_ENABLED(CONFIG_IRQ_WORK) && irqs_were_disabled && - expboost && !rdp->defer_qs_iw_pending && cpu_online(rdp->cpu)) { + needs_exp && rdp->defer_qs_iw_pending != DEFER_QS_PENDING && + cpu_online(rdp->cpu)) { // Get scheduler to re-evaluate and call hooks. // If !IRQ_WORK, FQS scan will eventually IPI. - if (IS_ENABLED(CONFIG_RCU_STRICT_GRACE_PERIOD) && - IS_ENABLED(CONFIG_PREEMPT_RT)) - rdp->defer_qs_iw = IRQ_WORK_INIT_HARD( - rcu_preempt_deferred_qs_handler); - else - init_irq_work(&rdp->defer_qs_iw, - rcu_preempt_deferred_qs_handler); - rdp->defer_qs_iw_pending = true; + rdp->defer_qs_iw = + IRQ_WORK_INIT_HARD(rcu_preempt_deferred_qs_handler); + rdp->defer_qs_iw_pending = DEFER_QS_PENDING; irq_work_queue_on(&rdp->defer_qs_iw, rdp->cpu); } } diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 69482c2f0771..2a5a5329322d 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -982,8 +982,7 @@ void show_rcu_gp_kthreads(void) for_each_possible_cpu(cpu) { rdp = per_cpu_ptr(&rcu_data, cpu); cbs += data_race(READ_ONCE(rdp->n_cbs_invoked)); - if (rcu_segcblist_is_offloaded(&rdp->cblist)) - show_rcu_nocb_state(rdp); + show_rcu_nocb_state(rdp); } pr_info("RCU callbacks invoked since boot: %lu\n", cbs); show_rcu_tasks_gp_kthreads(); diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index f597734d83cc..22a6de59b77b 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -857,8 +857,6 @@ our %deprecated_apis = ( "kunmap" => "kunmap_local", "kmap_atomic" => "kmap_local_page", "kunmap_atomic" => "kunmap_local", - "srcu_read_lock_lite" => "srcu_read_lock_fast", - "srcu_read_unlock_lite" => "srcu_read_unlock_fast", ); #Create a search pattern for all these strings to speed up a loop below diff --git a/tools/testing/selftests/rcutorture/bin/kvm-build.sh b/tools/testing/selftests/rcutorture/bin/kvm-build.sh index 11f8d232b0ee..3edfd064ef81 100755 --- a/tools/testing/selftests/rcutorture/bin/kvm-build.sh +++ b/tools/testing/selftests/rcutorture/bin/kvm-build.sh @@ -44,7 +44,7 @@ fi ncpus="`getconf _NPROCESSORS_ONLN`" make -j$((2 * ncpus)) $TORTURE_KMAKE_ARG > $resdir/Make.out 2>&1 retval=$? -if test $retval -ne 0 || grep "rcu[^/]*": < $resdir/Make.out | grep -E -q "Stop|Error|error:|warning:" || grep -E -q "Stop|Error|error:" < $resdir/Make.out +if test $retval -ne 0 || grep "rcu[^/]*": < $resdir/Make.out | grep -E -q "Stop|ERROR|Error|error:|warning:" || grep -E -q "Stop|ERROR|Error|error:" < $resdir/Make.out then echo Kernel build error grep -E "Stop|Error|error:|warning:" < $resdir/Make.out diff --git a/tools/testing/selftests/rcutorture/bin/kvm.sh b/tools/testing/selftests/rcutorture/bin/kvm.sh index 42e5e8597a1a..617cba339d28 100755 --- a/tools/testing/selftests/rcutorture/bin/kvm.sh +++ b/tools/testing/selftests/rcutorture/bin/kvm.sh @@ -199,7 +199,7 @@ do fi ;; --kconfig|--kconfigs) - checkarg --kconfig "(Kconfig options)" $# "$2" '^\(#CHECK#\)\?CONFIG_[A-Z0-9_]\+=\([ynm]\|[0-9]\+\|"[^"]*"\)\( \(#CHECK#\)\?CONFIG_[A-Z0-9_]\+=\([ynm]\|[0-9]\+\|"[^"]*"\)\)*$' '^error$' + checkarg --kconfig "(Kconfig options)" $# "$2" '^\(#CHECK#\)\?CONFIG_[A-Z0-9_]\+=\([ynm]\|[0-9]\+\|"[^"]*"\)\( \+\(#CHECK#\)\?CONFIG_[A-Z0-9_]\+=\([ynm]\|[0-9]\+\|"[^"]*"\)\)* *$' '^error$' TORTURE_KCONFIG_ARG="`echo "$TORTURE_KCONFIG_ARG $2" | sed -e 's/^ *//' -e 's/ *$//'`" shift ;; @@ -442,18 +442,7 @@ echo $scriptname $args touch $resdir/$ds/log echo $scriptname $args >> $resdir/$ds/log echo ${TORTURE_SUITE} > $resdir/$ds/torture_suite -echo Build directory: `pwd` > $resdir/$ds/testid.txt -if test -d .git -then - echo Current commit: `git rev-parse HEAD` >> $resdir/$ds/testid.txt - echo >> $resdir/$ds/testid.txt - echo ' ---' Output of "'"git status"'": >> $resdir/$ds/testid.txt - git status >> $resdir/$ds/testid.txt - echo >> $resdir/$ds/testid.txt - echo >> $resdir/$ds/testid.txt - echo ' ---' Output of "'"git diff HEAD"'": >> $resdir/$ds/testid.txt - git diff HEAD >> $resdir/$ds/testid.txt -fi +mktestid.sh $resdir/$ds ___EOF___ kvm-assign-cpus.sh /sys/devices/system/node > $T/cpuarray.awk kvm-get-cpus-script.sh $T/cpuarray.awk $T/dumpbatches.awk diff --git a/tools/testing/selftests/rcutorture/bin/mktestid.sh b/tools/testing/selftests/rcutorture/bin/mktestid.sh new file mode 100755 index 000000000000..16f9907a4dae --- /dev/null +++ b/tools/testing/selftests/rcutorture/bin/mktestid.sh @@ -0,0 +1,29 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0+ +# +# Create a testid.txt file in the specified directory. +# +# Usage: mktestid.sh dirpath +# +# Copyright (C) Meta Platforms, Inc. 2025 +# +# Author: Paul E. McKenney + +resdir="$1" +if test -z "${resdir}" || ! test -d "${resdir}" || ! test -w "${resdir}" +then + echo Path '"'${resdir}'"' not writeable directory, no ${resdir}/testid.txt. + exit 1 +fi +echo Build directory: `pwd` > ${resdir}/testid.txt +if test -d .git +then + echo Current commit: `git rev-parse HEAD` >> ${resdir}/testid.txt + echo >> ${resdir}/testid.txt + echo ' ---' Output of "'"git status"'": >> ${resdir}/testid.txt + git status >> ${resdir}/testid.txt + echo >> ${resdir}/testid.txt + echo >> ${resdir}/testid.txt + echo ' ---' Output of "'"git diff HEAD"'": >> ${resdir}/testid.txt + git diff HEAD >> ${resdir}/testid.txt +fi diff --git a/tools/testing/selftests/rcutorture/bin/torture.sh b/tools/testing/selftests/rcutorture/bin/torture.sh index e03fdaca89b3..611bc03a8dc7 100755 --- a/tools/testing/selftests/rcutorture/bin/torture.sh +++ b/tools/testing/selftests/rcutorture/bin/torture.sh @@ -30,6 +30,15 @@ then VERBOSE_BATCH_CPUS=0 fi +# Machine architecture? ("uname -p" is said to be less portable.)1 +thisarch="`uname -m`" +if test "${thisarch}" = aarch64 +then + ifnotaarch64=no +else + ifnotaarch64=yes +fi + # Configurations/scenarios. configs_rcutorture= configs_locktorture= @@ -55,9 +64,9 @@ do_normal=yes explicit_normal=no do_kasan=yes do_kcsan=no -do_clocksourcewd=yes +do_clocksourcewd="${ifnotaarch64}" do_rt=yes -do_rcutasksflavors=yes +do_rcutasksflavors="${ifnotaarch64}" # FIXME: Back to "yes" when SMP=n auto-avoided do_srcu_lockdep=yes do_rcu_rust=no @@ -124,7 +133,7 @@ do ;; --do-all|--doall) do_allmodconfig=yes - do_rcutasksflavor=yes + do_rcutasksflavors="${ifnotaarch64}" # FIXME: Back to "yes" when SMP=n auto-avoided do_rcutorture=yes do_locktorture=yes do_scftorture=yes @@ -136,7 +145,7 @@ do explicit_normal=no do_kasan=yes do_kcsan=yes - do_clocksourcewd=yes + do_clocksourcewd="${ifnotaarch64}" do_srcu_lockdep=yes ;; --do-allmodconfig|--do-no-allmodconfig|--no-allmodconfig) @@ -274,7 +283,7 @@ then configs_rcutorture=CFLIST fi duration_rcutorture=$((duration_base*duration_rcutorture_frac/10)) -if test "$duration_rcutorture" -eq 0 +if test "$duration_rcutorture" -eq 0 && test "$do_locktorture" = "yes" then echo " --- Zero time for rcutorture, disabling" | tee -a $T/log do_rcutorture=no @@ -286,7 +295,7 @@ then configs_locktorture=CFLIST fi duration_locktorture=$((duration_base*duration_locktorture_frac/10)) -if test "$duration_locktorture" -eq 0 +if test "$duration_locktorture" -eq 0 && test "$do_locktorture" = "yes" then echo " --- Zero time for locktorture, disabling" | tee -a $T/log do_locktorture=no @@ -298,12 +307,19 @@ then configs_scftorture=CFLIST fi duration_scftorture=$((duration_base*duration_scftorture_frac/10)) -if test "$duration_scftorture" -eq 0 +if test "$duration_scftorture" -eq 0 && test "$do_scftorture" = "yes" then echo " --- Zero time for scftorture, disabling" | tee -a $T/log do_scftorture=no fi +# CONFIG_EXPERT=y is currently required for arm64 KCSAN runs. +kcsan_expert= +if test "${thisarch}" = aarch64 +then + kcsan_expert="CONFIG_EXPERT=y" +fi + touch $T/failures touch $T/successes @@ -362,13 +378,19 @@ function torture_set { then curflavor=$flavor torture_one "$@" - mv $T/last-resdir $T/last-resdir-nodebug || : + if test -e $T/last-resdir + then + mv $T/last-resdir $T/last-resdir-nodebug || : + fi fi if test "$do_kasan" = "yes" then curflavor=${flavor}-kasan torture_one "$@" --kasan - mv $T/last-resdir $T/last-resdir-kasan || : + if test -e $T/last-resdir + then + mv $T/last-resdir $T/last-resdir-kasan || : + fi fi if test "$do_kcsan" = "yes" then @@ -378,8 +400,16 @@ function torture_set { kcsan_kmake_tag="--kmake-args" cur_kcsan_kmake_args="$kcsan_kmake_args" fi - torture_one "$@" --kconfig "CONFIG_DEBUG_LOCK_ALLOC=y CONFIG_PROVE_LOCKING=y" $kcsan_kmake_tag $cur_kcsan_kmake_args --kcsan - mv $T/last-resdir $T/last-resdir-kcsan || : + chk_rdr_state= + if test "${flavor}" = rcutorture + then + chk_rdr_state="CONFIG_RCU_TORTURE_TEST_CHK_RDR_STATE=y" + fi + torture_one "$@" --kconfig "CONFIG_DEBUG_LOCK_ALLOC=y CONFIG_PROVE_LOCKING=y ${kcsan_expert} ${chk_rdr_state}" $kcsan_kmake_tag $cur_kcsan_kmake_args --kcsan + if test -e $T/last-resdir + then + mv $T/last-resdir $T/last-resdir-kcsan || : + fi fi } @@ -389,6 +419,7 @@ then echo " --- allmodconfig:" Start `date` | tee -a $T/log amcdir="tools/testing/selftests/rcutorture/res/$ds/allmodconfig" mkdir -p "$amcdir" + mktestid.sh "$amcdir" echo " --- make clean" | tee $amcdir/log > "$amcdir/Make.out" 2>&1 make -j$MAKE_ALLOTED_CPUS clean >> "$amcdir/Make.out" 2>&1 retcode=$? @@ -407,6 +438,10 @@ then make -j$MAKE_ALLOTED_CPUS >> "$amcdir/Make.out" 2>&1 retcode="$?" echo $retcode > "$amcdir/Make.exitcode" + if grep -E -q "Stop|ERROR|Error|error:|warning:" < "$amcdir/Make.out" + then + retcode=99 + fi buildphase='"make"' fi if test "$retcode" -eq 0 @@ -495,6 +530,7 @@ then echo " --- do-rcu-rust:" Start `date` | tee -a $T/log rrdir="tools/testing/selftests/rcutorture/res/$ds/results-rcu-rust" mkdir -p "$rrdir" + mktestid.sh "$rrdir" echo " --- make LLVM=1 rustavailable " | tee -a $rrdir/log > $rrdir/rustavailable.out make LLVM=1 rustavailable > $T/rustavailable.out 2>&1 retcode=$? @@ -681,7 +717,14 @@ nfailures=0 echo FAILURES: | tee -a $T/log if test -s "$T/failures" then - awk < "$T/failures" -v sq="'" '{ print "echo " sq $0 sq; print "sed -e " sq "1,/^ --- .* Test summary:$/d" sq " " $2 "/log | grep Summary: | sed -e " sq "s/^[^S]*/ /" sq; }' | sh | tee -a $T/log | tee "$T/failuresum" + awk < "$T/failures" -v sq="'" ' + { + print "echo " sq $0 sq; + if ($2 != "") + print "sed -e " sq "1,/^ --- .* Test summary:$/d" sq " " $2 "/log | grep Summary: | sed -e " sq "s/^[^S]*/ /" sq; + else + print "echo " sq " " sq "Run failed to produce results directory."; + }' | sh | tee -a $T/log | tee "$T/failuresum" nfailures="`wc -l "$T/failures" | awk '{ print $1 }'`" grep "^ Summary: " "$T/failuresum" | grep -v '^ Summary: Bugs: [0-9]* (all bugs kcsan)$' > "$T/nonkcsan" @@ -691,15 +734,18 @@ then fi ret=2 fi -if test "$do_kcsan" = "yes" +if test "$do_kcsan" = "yes" && test -e tools/testing/selftests/rcutorture/res/$ds then TORTURE_KCONFIG_KCSAN_ARG=1 tools/testing/selftests/rcutorture/bin/kcsan-collapse.sh tools/testing/selftests/rcutorture/res/$ds > tools/testing/selftests/rcutorture/res/$ds/kcsan.sum fi echo Started at $startdate, ended at `date`, duration `get_starttime_duration $starttime`. | tee -a $T/log echo Summary: Successes: $nsuccesses Failures: $nfailures. | tee -a $T/log -tdir="`cat $T/successes $T/failures | head -1 | awk '{ print $NF }' | sed -e 's,/[^/]\+/*$,,'`" -find "$tdir" -name 'ConfigFragment.diags' -print > $T/configerrors -find "$tdir" -name 'Make.out.diags' -print > $T/builderrors +tdir="`cat $T/successes $T/failures | awk 'NF > 1 { print $NF }' | head -1 | sed -e 's,/[^/]\+/*$,,'`" +if test -n "$tdir" +then + find "$tdir" -name 'ConfigFragment.diags' -print > $T/configerrors + find "$tdir" -name 'Make.out.diags' -print > $T/builderrors +fi if test -s "$T/configerrors" then echo " Scenarios with .config errors: `wc -l "$T/configerrors" | awk '{ print $1 }'`" diff --git a/tools/testing/selftests/rcutorture/configs/rcu/BUSTED b/tools/testing/selftests/rcutorture/configs/rcu/BUSTED index 48d8a245c7fa..7d75f4b94943 100644 --- a/tools/testing/selftests/rcutorture/configs/rcu/BUSTED +++ b/tools/testing/selftests/rcutorture/configs/rcu/BUSTED @@ -5,3 +5,6 @@ CONFIG_HOTPLUG_CPU=y CONFIG_PREEMPT_NONE=n CONFIG_PREEMPT_VOLUNTARY=n CONFIG_PREEMPT=y +CONFIG_RCU_TORTURE_TEST_CHK_RDR_STATE=y +CONFIG_RCU_TORTURE_TEST_LOG_CPU=y +CONFIG_RCU_TORTURE_TEST_LOG_GP=y diff --git a/tools/testing/selftests/rcutorture/configs/rcu/CFLIST b/tools/testing/selftests/rcutorture/configs/rcu/CFLIST index 45f572570a8c..98b6175e5aa0 100644 --- a/tools/testing/selftests/rcutorture/configs/rcu/CFLIST +++ b/tools/testing/selftests/rcutorture/configs/rcu/CFLIST @@ -5,7 +5,6 @@ TREE04 TREE05 TREE07 TREE09 -SRCU-L SRCU-N SRCU-P SRCU-T diff --git a/tools/testing/selftests/rcutorture/configs/rcu/SRCU-L b/tools/testing/selftests/rcutorture/configs/rcu/SRCU-L deleted file mode 100644 index 3b4fa8dbef8a..000000000000 --- a/tools/testing/selftests/rcutorture/configs/rcu/SRCU-L +++ /dev/null @@ -1,10 +0,0 @@ -CONFIG_RCU_TRACE=n -CONFIG_SMP=y -CONFIG_NR_CPUS=6 -CONFIG_HOTPLUG_CPU=y -CONFIG_PREEMPT_NONE=y -CONFIG_PREEMPT_VOLUNTARY=n -CONFIG_PREEMPT=n -#CHECK#CONFIG_RCU_EXPERT=n -CONFIG_KPROBES=n -CONFIG_FTRACE=n diff --git a/tools/testing/selftests/rcutorture/configs/rcu/SRCU-L.boot b/tools/testing/selftests/rcutorture/configs/rcu/SRCU-L.boot deleted file mode 100644 index 0207b3138c5b..000000000000 --- a/tools/testing/selftests/rcutorture/configs/rcu/SRCU-L.boot +++ /dev/null @@ -1,3 +0,0 @@ -rcutorture.torture_type=srcu -rcutorture.reader_flavor=0x4 -rcutorture.fwd_progress=3