diff options
-rw-r--r-- | Documentation/memory-barriers.txt | 4 | ||||
-rw-r--r-- | arch/blackfin/include/asm/cmpxchg.h | 1 | ||||
-rw-r--r-- | arch/c6x/include/asm/cmpxchg.h | 2 | ||||
-rw-r--r-- | arch/frv/include/asm/cmpxchg.h | 2 | ||||
-rw-r--r-- | arch/ia64/include/asm/barrier.h | 2 | ||||
-rw-r--r-- | arch/powerpc/include/asm/barrier.h | 2 | ||||
-rw-r--r-- | arch/s390/include/asm/barrier.h | 2 | ||||
-rw-r--r-- | arch/tile/include/asm/cmpxchg.h | 2 | ||||
-rw-r--r-- | arch/x86/Kconfig | 8 | ||||
-rw-r--r-- | arch/x86/include/asm/qspinlock_paravirt.h | 59 | ||||
-rw-r--r-- | include/asm-generic/barrier.h | 2 | ||||
-rw-r--r-- | include/asm-generic/qspinlock.h | 9 | ||||
-rw-r--r-- | include/linux/compiler.h | 17 | ||||
-rw-r--r-- | kernel/futex.c | 83 | ||||
-rw-r--r-- | kernel/locking/qspinlock.c | 82 | ||||
-rw-r--r-- | kernel/locking/qspinlock_paravirt.h | 252 | ||||
-rw-r--r-- | kernel/locking/qspinlock_stat.h | 300 | ||||
-rw-r--r-- | kernel/sched/core.c | 99 | ||||
-rw-r--r-- | kernel/sched/sched.h | 2 | ||||
-rw-r--r-- | lib/atomic64_test.c | 120 |
20 files changed, 904 insertions, 146 deletions
diff --git a/Documentation/memory-barriers.txt b/Documentation/memory-barriers.txt index 85304ebd187c..a61be39c7b51 100644 --- a/Documentation/memory-barriers.txt +++ b/Documentation/memory-barriers.txt @@ -1673,8 +1673,8 @@ There are some more advanced barrier functions: (*) smp_store_mb(var, value) This assigns the value to the variable and then inserts a full memory - barrier after it, depending on the function. It isn't guaranteed to - insert anything more than a compiler barrier in a UP compilation. + barrier after it. It isn't guaranteed to insert anything more than a + compiler barrier in a UP compilation. (*) smp_mb__before_atomic(); diff --git a/arch/blackfin/include/asm/cmpxchg.h b/arch/blackfin/include/asm/cmpxchg.h index c05868cc61c1..253928854299 100644 --- a/arch/blackfin/include/asm/cmpxchg.h +++ b/arch/blackfin/include/asm/cmpxchg.h @@ -128,6 +128,5 @@ static inline unsigned long __xchg(unsigned long x, volatile void *ptr, #endif /* !CONFIG_SMP */ #define xchg(ptr, x) ((__typeof__(*(ptr)))__xchg((unsigned long)(x), (ptr), sizeof(*(ptr)))) -#define tas(ptr) ((void)xchg((ptr), 1)) #endif /* __ARCH_BLACKFIN_CMPXCHG__ */ diff --git a/arch/c6x/include/asm/cmpxchg.h b/arch/c6x/include/asm/cmpxchg.h index b27c8cefb8c3..93d0a5a047a2 100644 --- a/arch/c6x/include/asm/cmpxchg.h +++ b/arch/c6x/include/asm/cmpxchg.h @@ -47,8 +47,6 @@ static inline unsigned int __xchg(unsigned int x, volatile void *ptr, int size) #define xchg(ptr, x) \ ((__typeof__(*(ptr)))__xchg((unsigned int)(x), (void *) (ptr), \ sizeof(*(ptr)))) -#define tas(ptr) xchg((ptr), 1) - #include <asm-generic/cmpxchg-local.h> diff --git a/arch/frv/include/asm/cmpxchg.h b/arch/frv/include/asm/cmpxchg.h index 5b04dd0aecab..a899765102ea 100644 --- a/arch/frv/include/asm/cmpxchg.h +++ b/arch/frv/include/asm/cmpxchg.h @@ -69,8 +69,6 @@ extern uint32_t __xchg_32(uint32_t i, volatile void *v); #endif -#define tas(ptr) (xchg((ptr), 1)) - /*****************************************************************************/ /* * compare and conditionally exchange value with memory diff --git a/arch/ia64/include/asm/barrier.h b/arch/ia64/include/asm/barrier.h index df896a1c41d3..209c4b817c95 100644 --- a/arch/ia64/include/asm/barrier.h +++ b/arch/ia64/include/asm/barrier.h @@ -77,7 +77,7 @@ do { \ ___p1; \ }) -#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); mb(); } while (0) +#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); smp_mb(); } while (0) /* * The group barrier in front of the rsm & ssm are necessary to ensure diff --git a/arch/powerpc/include/asm/barrier.h b/arch/powerpc/include/asm/barrier.h index 0eca6efc0631..a7af5fb7b914 100644 --- a/arch/powerpc/include/asm/barrier.h +++ b/arch/powerpc/include/asm/barrier.h @@ -34,7 +34,7 @@ #define rmb() __asm__ __volatile__ ("sync" : : : "memory") #define wmb() __asm__ __volatile__ ("sync" : : : "memory") -#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); mb(); } while (0) +#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); smp_mb(); } while (0) #ifdef __SUBARCH_HAS_LWSYNC # define SMPWMB LWSYNC diff --git a/arch/s390/include/asm/barrier.h b/arch/s390/include/asm/barrier.h index d68e11e0df5e..7ffd0b19135c 100644 --- a/arch/s390/include/asm/barrier.h +++ b/arch/s390/include/asm/barrier.h @@ -36,7 +36,7 @@ #define smp_mb__before_atomic() smp_mb() #define smp_mb__after_atomic() smp_mb() -#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); mb(); } while (0) +#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); smp_mb(); } while (0) #define smp_store_release(p, v) \ do { \ diff --git a/arch/tile/include/asm/cmpxchg.h b/arch/tile/include/asm/cmpxchg.h index 0ccda3c425be..25d5899497be 100644 --- a/arch/tile/include/asm/cmpxchg.h +++ b/arch/tile/include/asm/cmpxchg.h @@ -127,8 +127,6 @@ long long _atomic64_cmpxchg(long long *v, long long o, long long n); #endif -#define tas(ptr) xchg((ptr), 1) - #endif /* __ASSEMBLY__ */ #endif /* _ASM_TILE_CMPXCHG_H */ diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index db3622f22b61..965fc4216f76 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -687,6 +687,14 @@ config PARAVIRT_SPINLOCKS If you are unsure how to answer this question, answer Y. +config QUEUED_LOCK_STAT + bool "Paravirt queued spinlock statistics" + depends on PARAVIRT_SPINLOCKS && DEBUG_FS && QUEUED_SPINLOCKS + ---help--- + Enable the collection of statistical data on the slowpath + behavior of paravirtualized queued spinlocks and report + them on debugfs. + source "arch/x86/xen/Kconfig" config KVM_GUEST diff --git a/arch/x86/include/asm/qspinlock_paravirt.h b/arch/x86/include/asm/qspinlock_paravirt.h index b002e711ba88..9f92c180ed2f 100644 --- a/arch/x86/include/asm/qspinlock_paravirt.h +++ b/arch/x86/include/asm/qspinlock_paravirt.h @@ -1,6 +1,65 @@ #ifndef __ASM_QSPINLOCK_PARAVIRT_H #define __ASM_QSPINLOCK_PARAVIRT_H +/* + * For x86-64, PV_CALLEE_SAVE_REGS_THUNK() saves and restores 8 64-bit + * registers. For i386, however, only 1 32-bit register needs to be saved + * and restored. So an optimized version of __pv_queued_spin_unlock() is + * hand-coded for 64-bit, but it isn't worthwhile to do it for 32-bit. + */ +#ifdef CONFIG_64BIT + +PV_CALLEE_SAVE_REGS_THUNK(__pv_queued_spin_unlock_slowpath); +#define __pv_queued_spin_unlock __pv_queued_spin_unlock +#define PV_UNLOCK "__raw_callee_save___pv_queued_spin_unlock" +#define PV_UNLOCK_SLOWPATH "__raw_callee_save___pv_queued_spin_unlock_slowpath" + +/* + * Optimized assembly version of __raw_callee_save___pv_queued_spin_unlock + * which combines the registers saving trunk and the body of the following + * C code: + * + * void __pv_queued_spin_unlock(struct qspinlock *lock) + * { + * struct __qspinlock *l = (void *)lock; + * u8 lockval = cmpxchg(&l->locked, _Q_LOCKED_VAL, 0); + * + * if (likely(lockval == _Q_LOCKED_VAL)) + * return; + * pv_queued_spin_unlock_slowpath(lock, lockval); + * } + * + * For x86-64, + * rdi = lock (first argument) + * rsi = lockval (second argument) + * rdx = internal variable (set to 0) + */ +asm (".pushsection .text;" + ".globl " PV_UNLOCK ";" + ".align 4,0x90;" + PV_UNLOCK ": " + "push %rdx;" + "mov $0x1,%eax;" + "xor %edx,%edx;" + "lock cmpxchg %dl,(%rdi);" + "cmp $0x1,%al;" + "jne .slowpath;" + "pop %rdx;" + "ret;" + ".slowpath: " + "push %rsi;" + "movzbl %al,%esi;" + "call " PV_UNLOCK_SLOWPATH ";" + "pop %rsi;" + "pop %rdx;" + "ret;" + ".size " PV_UNLOCK ", .-" PV_UNLOCK ";" + ".popsection"); + +#else /* CONFIG_64BIT */ + +extern void __pv_queued_spin_unlock(struct qspinlock *lock); PV_CALLEE_SAVE_REGS_THUNK(__pv_queued_spin_unlock); +#endif /* CONFIG_64BIT */ #endif diff --git a/include/asm-generic/barrier.h b/include/asm-generic/barrier.h index b42afada1280..0f45f93ef692 100644 --- a/include/asm-generic/barrier.h +++ b/include/asm-generic/barrier.h @@ -93,7 +93,7 @@ #endif /* CONFIG_SMP */ #ifndef smp_store_mb -#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); mb(); } while (0) +#define smp_store_mb(var, value) do { WRITE_ONCE(var, value); smp_mb(); } while (0) #endif #ifndef smp_mb__before_atomic diff --git a/include/asm-generic/qspinlock.h b/include/asm-generic/qspinlock.h index e2aadbc7151f..39e1cb201b8e 100644 --- a/include/asm-generic/qspinlock.h +++ b/include/asm-generic/qspinlock.h @@ -12,8 +12,9 @@ * GNU General Public License for more details. * * (C) Copyright 2013-2015 Hewlett-Packard Development Company, L.P. + * (C) Copyright 2015 Hewlett-Packard Enterprise Development LP * - * Authors: Waiman Long <waiman.long@hp.com> + * Authors: Waiman Long <waiman.long@hpe.com> */ #ifndef __ASM_GENERIC_QSPINLOCK_H #define __ASM_GENERIC_QSPINLOCK_H @@ -62,7 +63,7 @@ static __always_inline int queued_spin_is_contended(struct qspinlock *lock) static __always_inline int queued_spin_trylock(struct qspinlock *lock) { if (!atomic_read(&lock->val) && - (atomic_cmpxchg(&lock->val, 0, _Q_LOCKED_VAL) == 0)) + (atomic_cmpxchg_acquire(&lock->val, 0, _Q_LOCKED_VAL) == 0)) return 1; return 0; } @@ -77,7 +78,7 @@ static __always_inline void queued_spin_lock(struct qspinlock *lock) { u32 val; - val = atomic_cmpxchg(&lock->val, 0, _Q_LOCKED_VAL); + val = atomic_cmpxchg_acquire(&lock->val, 0, _Q_LOCKED_VAL); if (likely(val == 0)) return; queued_spin_lock_slowpath(lock, val); @@ -93,7 +94,7 @@ static __always_inline void queued_spin_unlock(struct qspinlock *lock) /* * smp_mb__before_atomic() in order to guarantee release semantics */ - smp_mb__before_atomic_dec(); + smp_mb__before_atomic(); atomic_sub(_Q_LOCKED_VAL, &lock->val); } #endif diff --git a/include/linux/compiler.h b/include/linux/compiler.h index 4dac1036594f..00b042c49ccd 100644 --- a/include/linux/compiler.h +++ b/include/linux/compiler.h @@ -299,6 +299,23 @@ static __always_inline void __write_once_size(volatile void *p, void *res, int s __u.__val; \ }) +/** + * smp_cond_acquire() - Spin wait for cond with ACQUIRE ordering + * @cond: boolean expression to wait for + * + * Equivalent to using smp_load_acquire() on the condition variable but employs + * the control dependency of the wait to reduce the barrier on many platforms. + * + * The control dependency provides a LOAD->STORE order, the additional RMB + * provides LOAD->LOAD order, together they provide LOAD->{LOAD,STORE} order, + * aka. ACQUIRE. + */ +#define smp_cond_acquire(cond) do { \ + while (!(cond)) \ + cpu_relax(); \ + smp_rmb(); /* ctrl + rmb := acquire */ \ +} while (0) + #endif /* __KERNEL__ */ #endif /* __ASSEMBLY__ */ diff --git a/kernel/futex.c b/kernel/futex.c index 684d7549825a..8a310e240cda 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -725,9 +725,12 @@ static struct futex_pi_state * alloc_pi_state(void) } /* + * Drops a reference to the pi_state object and frees or caches it + * when the last reference is gone. + * * Must be called with the hb lock held. */ -static void free_pi_state(struct futex_pi_state *pi_state) +static void put_pi_state(struct futex_pi_state *pi_state) { if (!pi_state) return; @@ -1706,31 +1709,35 @@ retry_private: * exist yet, look it up one more time to ensure we have a * reference to it. If the lock was taken, ret contains the * vpid of the top waiter task. + * If the lock was not taken, we have pi_state and an initial + * refcount on it. In case of an error we have nothing. */ if (ret > 0) { WARN_ON(pi_state); drop_count++; task_count++; /* - * If we acquired the lock, then the user - * space value of uaddr2 should be vpid. It - * cannot be changed by the top waiter as it - * is blocked on hb2 lock if it tries to do - * so. If something fiddled with it behind our - * back the pi state lookup might unearth - * it. So we rather use the known value than - * rereading and handing potential crap to - * lookup_pi_state. + * If we acquired the lock, then the user space value + * of uaddr2 should be vpid. It cannot be changed by + * the top waiter as it is blocked on hb2 lock if it + * tries to do so. If something fiddled with it behind + * our back the pi state lookup might unearth it. So + * we rather use the known value than rereading and + * handing potential crap to lookup_pi_state. + * + * If that call succeeds then we have pi_state and an + * initial refcount on it. */ ret = lookup_pi_state(ret, hb2, &key2, &pi_state); } switch (ret) { case 0: + /* We hold a reference on the pi state. */ break; + + /* If the above failed, then pi_state is NULL */ case -EFAULT: - free_pi_state(pi_state); - pi_state = NULL; double_unlock_hb(hb1, hb2); hb_waiters_dec(hb2); put_futex_key(&key2); @@ -1746,8 +1753,6 @@ retry_private: * exit to complete. * - The user space value changed. */ - free_pi_state(pi_state); - pi_state = NULL; double_unlock_hb(hb1, hb2); hb_waiters_dec(hb2); put_futex_key(&key2); @@ -1801,30 +1806,58 @@ retry_private: * of requeue_pi if we couldn't acquire the lock atomically. */ if (requeue_pi) { - /* Prepare the waiter to take the rt_mutex. */ + /* + * Prepare the waiter to take the rt_mutex. Take a + * refcount on the pi_state and store the pointer in + * the futex_q object of the waiter. + */ atomic_inc(&pi_state->refcount); this->pi_state = pi_state; ret = rt_mutex_start_proxy_lock(&pi_state->pi_mutex, this->rt_waiter, this->task); if (ret == 1) { - /* We got the lock. */ + /* + * We got the lock. We do neither drop the + * refcount on pi_state nor clear + * this->pi_state because the waiter needs the + * pi_state for cleaning up the user space + * value. It will drop the refcount after + * doing so. + */ requeue_pi_wake_futex(this, &key2, hb2); drop_count++; continue; } else if (ret) { - /* -EDEADLK */ + /* + * rt_mutex_start_proxy_lock() detected a + * potential deadlock when we tried to queue + * that waiter. Drop the pi_state reference + * which we took above and remove the pointer + * to the state from the waiters futex_q + * object. + */ this->pi_state = NULL; - free_pi_state(pi_state); - goto out_unlock; + put_pi_state(pi_state); + /* + * We stop queueing more waiters and let user + * space deal with the mess. + */ + break; } } requeue_futex(this, hb1, hb2, &key2); drop_count++; } + /* + * We took an extra initial reference to the pi_state either + * in futex_proxy_trylock_atomic() or in lookup_pi_state(). We + * need to drop it here again. + */ + put_pi_state(pi_state); + out_unlock: - free_pi_state(pi_state); double_unlock_hb(hb1, hb2); wake_up_q(&wake_q); hb_waiters_dec(hb2); @@ -1973,7 +2006,7 @@ static void unqueue_me_pi(struct futex_q *q) __unqueue_futex(q); BUG_ON(!q->pi_state); - free_pi_state(q->pi_state); + put_pi_state(q->pi_state); q->pi_state = NULL; spin_unlock(q->lock_ptr); @@ -2755,6 +2788,11 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, if (q.pi_state && (q.pi_state->owner != current)) { spin_lock(q.lock_ptr); ret = fixup_pi_state_owner(uaddr2, &q, current); + /* + * Drop the reference to the pi state which + * the requeue_pi() code acquired for us. + */ + put_pi_state(q.pi_state); spin_unlock(q.lock_ptr); } } else { @@ -3046,7 +3084,8 @@ long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout, if (op & FUTEX_CLOCK_REALTIME) { flags |= FLAGS_CLOCKRT; - if (cmd != FUTEX_WAIT_BITSET && cmd != FUTEX_WAIT_REQUEUE_PI) + if (cmd != FUTEX_WAIT && cmd != FUTEX_WAIT_BITSET && \ + cmd != FUTEX_WAIT_REQUEUE_PI) return -ENOSYS; } diff --git a/kernel/locking/qspinlock.c b/kernel/locking/qspinlock.c index 87e9ce6a63c5..393d1874b9e0 100644 --- a/kernel/locking/qspinlock.c +++ b/kernel/locking/qspinlock.c @@ -14,8 +14,9 @@ * (C) Copyright 2013-2015 Hewlett-Packard Development Company, L.P. * (C) Copyright 2013-2014 Red Hat, Inc. * (C) Copyright 2015 Intel Corp. + * (C) Copyright 2015 Hewlett-Packard Enterprise Development LP * - * Authors: Waiman Long <waiman.long@hp.com> + * Authors: Waiman Long <waiman.long@hpe.com> * Peter Zijlstra <peterz@infradead.org> */ @@ -176,7 +177,12 @@ static __always_inline u32 xchg_tail(struct qspinlock *lock, u32 tail) { struct __qspinlock *l = (void *)lock; - return (u32)xchg(&l->tail, tail >> _Q_TAIL_OFFSET) << _Q_TAIL_OFFSET; + /* + * Use release semantics to make sure that the MCS node is properly + * initialized before changing the tail code. + */ + return (u32)xchg_release(&l->tail, + tail >> _Q_TAIL_OFFSET) << _Q_TAIL_OFFSET; } #else /* _Q_PENDING_BITS == 8 */ @@ -208,7 +214,11 @@ static __always_inline u32 xchg_tail(struct qspinlock *lock, u32 tail) for (;;) { new = (val & _Q_LOCKED_PENDING_MASK) | tail; - old = atomic_cmpxchg(&lock->val, val, new); + /* + * Use release semantics to make sure that the MCS node is + * properly initialized before changing the tail code. + */ + old = atomic_cmpxchg_release(&lock->val, val, new); if (old == val) break; @@ -238,18 +248,20 @@ static __always_inline void set_locked(struct qspinlock *lock) */ static __always_inline void __pv_init_node(struct mcs_spinlock *node) { } -static __always_inline void __pv_wait_node(struct mcs_spinlock *node) { } +static __always_inline void __pv_wait_node(struct mcs_spinlock *node, + struct mcs_spinlock *prev) { } static __always_inline void __pv_kick_node(struct qspinlock *lock, struct mcs_spinlock *node) { } -static __always_inline void __pv_wait_head(struct qspinlock *lock, - struct mcs_spinlock *node) { } +static __always_inline u32 __pv_wait_head_or_lock(struct qspinlock *lock, + struct mcs_spinlock *node) + { return 0; } #define pv_enabled() false #define pv_init_node __pv_init_node #define pv_wait_node __pv_wait_node #define pv_kick_node __pv_kick_node -#define pv_wait_head __pv_wait_head +#define pv_wait_head_or_lock __pv_wait_head_or_lock #ifdef CONFIG_PARAVIRT_SPINLOCKS #define queued_spin_lock_slowpath native_queued_spin_lock_slowpath @@ -319,7 +331,11 @@ void queued_spin_lock_slowpath(struct qspinlock *lock, u32 val) if (val == new) new |= _Q_PENDING_VAL; - old = atomic_cmpxchg(&lock->val, val, new); + /* + * Acquire semantic is required here as the function may + * return immediately if the lock was free. + */ + old = atomic_cmpxchg_acquire(&lock->val, val, new); if (old == val) break; @@ -382,6 +398,7 @@ queue: * p,*,* -> n,*,* */ old = xchg_tail(lock, tail); + next = NULL; /* * if there was a previous node; link it and wait until reaching the @@ -391,8 +408,18 @@ queue: prev = decode_tail(old); WRITE_ONCE(prev->next, node); - pv_wait_node(node); + pv_wait_node(node, prev); arch_mcs_spin_lock_contended(&node->locked); + + /* + * While waiting for the MCS lock, the next pointer may have + * been set by another lock waiter. We optimistically load + * the next pointer & prefetch the cacheline for writing + * to reduce latency in the upcoming MCS unlock operation. + */ + next = READ_ONCE(node->next); + if (next) + prefetchw(next); } /* @@ -406,11 +433,22 @@ queue: * sequentiality; this is because the set_locked() function below * does not imply a full barrier. * + * The PV pv_wait_head_or_lock function, if active, will acquire + * the lock and return a non-zero value. So we have to skip the + * smp_load_acquire() call. As the next PV queue head hasn't been + * designated yet, there is no way for the locked value to become + * _Q_SLOW_VAL. So both the set_locked() and the + * atomic_cmpxchg_relaxed() calls will be safe. + * + * If PV isn't active, 0 will be returned instead. + * */ - pv_wait_head(lock, node); - while ((val = smp_load_acquire(&lock->val.counter)) & _Q_LOCKED_PENDING_MASK) - cpu_relax(); + if ((val = pv_wait_head_or_lock(lock, node))) + goto locked; + smp_cond_acquire(!((val = atomic_read(&lock->val)) & _Q_LOCKED_PENDING_MASK)); + +locked: /* * claim the lock: * @@ -422,11 +460,17 @@ queue: * to grab the lock. */ for (;;) { - if (val != tail) { + /* In the PV case we might already have _Q_LOCKED_VAL set */ + if ((val & _Q_TAIL_MASK) != tail) { set_locked(lock); break; } - old = atomic_cmpxchg(&lock->val, val, _Q_LOCKED_VAL); + /* + * The smp_load_acquire() call above has provided the necessary + * acquire semantics required for locking. At most two + * iterations of this loop may be ran. + */ + old = atomic_cmpxchg_relaxed(&lock->val, val, _Q_LOCKED_VAL); if (old == val) goto release; /* No contention */ @@ -434,10 +478,12 @@ queue: } /* - * contended path; wait for next, release. + * contended path; wait for next if not observed yet, release. */ - while (!(next = READ_ONCE(node->next))) - cpu_relax(); + if (!next) { + while (!(next = READ_ONCE(node->next))) + cpu_relax(); + } arch_mcs_spin_unlock_contended(&next->locked); pv_kick_node(lock, next); @@ -462,7 +508,7 @@ EXPORT_SYMBOL(queued_spin_lock_slowpath); #undef pv_init_node #undef pv_wait_node #undef pv_kick_node -#undef pv_wait_head +#undef pv_wait_head_or_lock #undef queued_spin_lock_slowpath #define queued_spin_lock_slowpath __pv_queued_spin_lock_slowpath diff --git a/kernel/locking/qspinlock_paravirt.h b/kernel/locking/qspinlock_paravirt.h index f0450ff4829b..87bb235c3448 100644 --- a/kernel/locking/qspinlock_paravirt.h +++ b/kernel/locking/qspinlock_paravirt.h @@ -23,6 +23,20 @@ #define _Q_SLOW_VAL (3U << _Q_LOCKED_OFFSET) /* + * Queue Node Adaptive Spinning + * + * A queue node vCPU will stop spinning if the vCPU in the previous node is + * not running. The one lock stealing attempt allowed at slowpath entry + * mitigates the slight slowdown for non-overcommitted guest with this + * aggressive wait-early mechanism. + * + * The status of the previous node will be checked at fixed interval + * controlled by PV_PREV_CHECK_MASK. This is to ensure that we won't + * pound on the cacheline of the previous node too heavily. + */ +#define PV_PREV_CHECK_MASK 0xff + +/* * Queue node uses: vcpu_running & vcpu_halted. * Queue head uses: vcpu_running & vcpu_hashed. */ @@ -41,6 +55,94 @@ struct pv_node { }; /* + * By replacing the regular queued_spin_trylock() with the function below, + * it will be called once when a lock waiter enter the PV slowpath before + * being queued. By allowing one lock stealing attempt here when the pending + * bit is off, it helps to reduce the performance impact of lock waiter + * preemption without the drawback of lock starvation. + */ +#define queued_spin_trylock(l) pv_queued_spin_steal_lock(l) +static inline bool pv_queued_spin_steal_lock(struct qspinlock *lock) +{ + struct __qspinlock *l = (void *)lock; + + return !(atomic_read(&lock->val) & _Q_LOCKED_PENDING_MASK) && + (cmpxchg(&l->locked, 0, _Q_LOCKED_VAL) == 0); +} + +/* + * The pending bit is used by the queue head vCPU to indicate that it + * is actively spinning on the lock and no lock stealing is allowed. + */ +#if _Q_PENDING_BITS == 8 +static __always_inline void set_pending(struct qspinlock *lock) +{ + struct __qspinlock *l = (void *)lock; + + WRITE_ONCE(l->pending, 1); +} + +static __always_inline void clear_pending(struct qspinlock *lock) +{ + struct __qspinlock *l = (void *)lock; + + WRITE_ONCE(l->pending, 0); +} + +/* + * The pending bit check in pv_queued_spin_steal_lock() isn't a memory + * barrier. Therefore, an atomic cmpxchg() is used to acquire the lock + * just to be sure that it will get it. + */ +static __always_inline int trylock_clear_pending(struct qspinlock *lock) +{ + struct __qspinlock *l = (void *)lock; + + return !READ_ONCE(l->locked) && + (cmpxchg(&l->locked_pending, _Q_PENDING_VAL, _Q_LOCKED_VAL) + == _Q_PENDING_VAL); +} +#else /* _Q_PENDING_BITS == 8 */ +static __always_inline void set_pending(struct qspinlock *lock) +{ + atomic_set_mask(_Q_PENDING_VAL, &lock->val); +} + +static __always_inline void clear_pending(struct qspinlock *lock) +{ + atomic_clear_mask(_Q_PENDING_VAL, &lock->val); +} + +static __always_inline int trylock_clear_pending(struct qspinlock *lock) +{ + int val = atomic_read(&lock->val); + + for (;;) { + int old, new; + + if (val & _Q_LOCKED_MASK) + break; + + /* + * Try to clear pending bit & set locked bit + */ + old = val; + new = (val & ~_Q_PENDING_MASK) | _Q_LOCKED_VAL; + val = atomic_cmpxchg(&lock->val, old, new); + + if (val == old) + return 1; + } + return 0; +} +#endif /* _Q_PENDING_BITS == 8 */ + +/* + * Include queued spinlock statistics code + */ +#include "qspinlock_stat.h" + +/* * Lock and MCS node addresses hash table for fast lookup * * Hashing is done on a per-cacheline basis to minimize the need to access @@ -100,10 +202,13 @@ static struct qspinlock **pv_hash(struct qspinlock *lock, struct pv_node *node) { unsigned long offset, hash = hash_ptr(lock, pv_lock_hash_bits); struct pv_hash_entry *he; + int hopcnt = 0; for_each_hash_entry(he, offset, hash) { + hopcnt++; if (!cmpxchg(&he->lock, NULL, lock)) { WRITE_ONCE(he->node, node); + qstat_hop(hopcnt); return &he->lock; } } @@ -144,6 +249,20 @@ static struct pv_node *pv_unhash(struct qspinlock *lock) } /* + * Return true if when it is time to check the previous node which is not + * in a running state. + */ +static inline bool +pv_wait_early(struct pv_node *prev, int loop) +{ + + if ((loop & PV_PREV_CHECK_MASK) != 0) + return false; + + return READ_ONCE(prev->state) != vcpu_running; +} + +/* * Initialize the PV part of the mcs_spinlock node. */ static void pv_init_node(struct mcs_spinlock *node) @@ -161,15 +280,23 @@ static void pv_init_node(struct mcs_spinlock *node) * pv_kick_node() is used to set _Q_SLOW_VAL and fill in hash table on its * behalf. */ -static void pv_wait_node(struct mcs_spinlock *node) +static void pv_wait_node(struct mcs_spinlock *node, struct mcs_spinlock *prev) { struct pv_node *pn = (struct pv_node *)node; + struct pv_node *pp = (struct pv_node *)prev; + int waitcnt = 0; int loop; + bool wait_early; - for (;;) { - for (loop = SPIN_THRESHOLD; loop; loop--) { + /* waitcnt processing will be compiled out if !QUEUED_LOCK_STAT */ + for (;; waitcnt++) { + for (wait_early = false, loop = SPIN_THRESHOLD; loop; loop--) { if (READ_ONCE(node->locked)) return; + if (pv_wait_early(pp, loop)) { + wait_early = true; + break; + } cpu_relax(); } @@ -184,12 +311,17 @@ static void pv_wait_node(struct mcs_spinlock *node) */ smp_store_mb(pn->state, vcpu_halted); - if (!READ_ONCE(node->locked)) + if (!READ_ONCE(node->locked)) { + qstat_inc(qstat_pv_wait_node, true); + qstat_inc(qstat_pv_wait_again, waitcnt); + qstat_inc(qstat_pv_wait_early, wait_early); pv_wait(&pn->state, vcpu_halted); + } /* - * If pv_kick_node() changed us to vcpu_hashed, retain that value - * so that pv_wait_head() knows to not also try to hash this lock. + * If pv_kick_node() changed us to vcpu_hashed, retain that + * value so that pv_wait_head_or_lock() knows to not also try + * to hash this lock. */ cmpxchg(&pn->state, vcpu_halted, vcpu_running); @@ -200,6 +332,7 @@ static void pv_wait_node(struct mcs_spinlock *node) * So it is better to spin for a while in the hope that the * MCS lock will be released soon. */ + qstat_inc(qstat_pv_spurious_wakeup, !READ_ONCE(node->locked)); } /* @@ -212,8 +345,9 @@ static void pv_wait_node(struct mcs_spinlock *node) /* * Called after setting next->locked = 1 when we're the lock owner. * - * Instead of waking the waiters stuck in pv_wait_node() advance their state such - * that they're waiting in pv_wait_head(), this avoids a wake/sleep cycle. + * Instead of waking the waiters stuck in pv_wait_node() advance their state + * such that they're waiting in pv_wait_head_or_lock(), this avoids a + * wake/sleep cycle. */ static void pv_kick_node(struct qspinlock *lock, struct mcs_spinlock *node) { @@ -242,14 +376,19 @@ static void pv_kick_node(struct qspinlock *lock, struct mcs_spinlock *node) } /* - * Wait for l->locked to become clear; halt the vcpu after a short spin. + * Wait for l->locked to become clear and acquire the lock; + * halt the vcpu after a short spin. * __pv_queued_spin_unlock() will wake us. + * + * The current value of the lock will be returned for additional processing. */ -static void pv_wait_head(struct qspinlock *lock, struct mcs_spinlock *node) +static u32 +pv_wait_head_or_lock(struct qspinlock *lock, struct mcs_spinlock *node) { struct pv_node *pn = (struct pv_node *)node; struct __qspinlock *l = (void *)lock; struct qspinlock **lp = NULL; + int waitcnt = 0; int loop; /* @@ -259,12 +398,25 @@ static void pv_wait_head(struct qspinlock *lock, struct mcs_spinlock *node) if (READ_ONCE(pn->state) == vcpu_hashed) lp = (struct qspinlock **)1; - for (;;) { + for (;; waitcnt++) { + /* + * Set correct vCPU state to be used by queue node wait-early + * mechanism. + */ + WRITE_ONCE(pn->state, vcpu_running); + + /* + * Set the pending bit in the active lock spinning loop to + * disable lock stealing before attempting to acquire the lock. + */ + set_pending(lock); for (loop = SPIN_THRESHOLD; loop; loop--) { - if (!READ_ONCE(l->locked)) - return; + if (trylock_clear_pending(lock)) + goto gotlock; cpu_relax(); } + clear_pending(lock); + if (!lp) { /* ONCE */ lp = pv_hash(lock, pn); @@ -280,51 +432,50 @@ static void pv_wait_head(struct qspinlock *lock, struct mcs_spinlock *node) * * Matches the smp_rmb() in __pv_queued_spin_unlock(). */ - if (!cmpxchg(&l->locked, _Q_LOCKED_VAL, _Q_SLOW_VAL)) { + if (xchg(&l->locked, _Q_SLOW_VAL) == 0) { /* - * The lock is free and _Q_SLOW_VAL has never - * been set. Therefore we need to unhash before - * getting the lock. + * The lock was free and now we own the lock. + * Change the lock value back to _Q_LOCKED_VAL + * and unhash the table. */ + WRITE_ONCE(l->locked, _Q_LOCKED_VAL); WRITE_ONCE(*lp, NULL); - return; + goto gotlock; } } + WRITE_ONCE(pn->state, vcpu_halted); + qstat_inc(qstat_pv_wait_head, true); + |