2 * Copyright 2017 Facebook, Inc.
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
8 * http://www.apache.org/licenses/LICENSE-2.0
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
23 #include <folly/Portability.h>
24 #include <folly/detail/Futex.h>
26 #if defined(__clang__)
27 #define NO_SANITIZE_ADDRESS __attribute__((no_sanitize_address))
29 #define NO_SANITIZE_ADDRESS
35 * Tiny exclusive lock that packs four lock slots into a single
36 * byte. Each slot is an independent real, sleeping lock. The default
37 * lock and unlock functions operate on slot zero, which modifies only
38 * the low two bits of the host byte.
40 * You should zero-initialize the bits of a MicroLock that you intend
43 * If you're not space-constrained, prefer std::mutex, which will
44 * likely be faster, since it has more than two bits of information to
47 * You are free to put a MicroLock in a union with some other object.
48 * If, for example, you want to use the bottom two bits of a pointer
49 * as a lock, you can put a MicroLock in a union with the pointer and
50 * limit yourself to MicroLock slot zero, which will use the two
51 * least-significant bits in the bottom byte.
53 * (Note that such a union is safe only because MicroLock is based on
54 * a character type, and even under a strict interpretation of C++'s
55 * aliasing rules, character types may alias anything.)
57 * MicroLock uses a dirty trick: it actually operates on the full
58 * 32-bit, four-byte-aligned bit of memory into which it is embedded.
59 * It never modifies bits outside the ones it's defined to modify, but
60 * it _accesses_ all the bits in the 32-bit memory location for
61 * purposes of futex management.
63 * The MaxSpins template parameter controls the number of times we
64 * spin trying to acquire the lock. MaxYields controls the number of
65 * times we call sched_yield; once we've tried to acquire the lock
66 * MaxSpins + MaxYields times, we sleep on the lock futex.
67 * By adjusting these parameters, you can make MicroLock behave as
68 * much or as little like a conventional spinlock as you'd like.
73 * With the default template options, the timings for uncontended
74 * acquire-then-release come out as follows on Intel(R) Xeon(R) CPU
75 * E5-2660 0 @ 2.20GHz, in @mode/opt, as of the master tree at Tue, 01
78 * ========================================================================
79 * folly/test/SmallLocksBenchmark.cpp relative time/iter iters/s
80 * ========================================================================
81 * MicroSpinLockUncontendedBenchmark 13.46ns 74.28M
82 * PicoSpinLockUncontendedBenchmark 14.99ns 66.71M
83 * MicroLockUncontendedBenchmark 27.06ns 36.96M
84 * StdMutexUncontendedBenchmark 25.18ns 39.72M
85 * VirtualFunctionCall 1.72ns 579.78M
86 * ========================================================================
88 * (The virtual dispatch benchmark is provided for scale.)
90 * While the uncontended case for MicroLock is competitive with the
91 * glibc 2.2.0 implementation of std::mutex, std::mutex is likely to be
92 * faster in the contended case, because we need to wake up all waiters
95 * Make sure to benchmark your particular workload.
101 #if defined(__SANITIZE_ADDRESS__) && !defined(__clang__) && \
102 (defined(__GNUC__) || defined(__GNUG__))
107 inline detail::Futex<>* word() const; // Well, halfword on 64-bit systems
108 inline uint32_t baseShift(unsigned slot) const;
109 inline uint32_t heldBit(unsigned slot) const;
110 inline uint32_t waitBit(unsigned slot) const;
111 static void lockSlowPath(uint32_t oldWord,
112 detail::Futex<>* wordPtr,
113 uint32_t slotHeldBit,
118 inline void unlock(unsigned slot) NO_SANITIZE_ADDRESS;
119 inline void unlock() { unlock(0); }
120 // Initializes all the slots.
121 inline void init() { lock_ = 0; }
124 inline detail::Futex<>* MicroLockCore::word() const {
125 uintptr_t lockptr = (uintptr_t)&lock_;
126 lockptr &= ~(sizeof(uint32_t) - 1);
127 return (detail::Futex<>*)lockptr;
130 inline unsigned MicroLockCore::baseShift(unsigned slot) const {
131 assert(slot < CHAR_BIT / 2);
133 unsigned offset_bytes = (unsigned)((uintptr_t)&lock_ - (uintptr_t)word());
136 unsigned)(kIsLittleEndian ? offset_bytes * CHAR_BIT + slot * 2 : CHAR_BIT * (sizeof(uint32_t) - offset_bytes - 1) + slot * 2);
139 inline uint32_t MicroLockCore::heldBit(unsigned slot) const {
140 return 1U << (baseShift(slot) + 0);
143 inline uint32_t MicroLockCore::waitBit(unsigned slot) const {
144 return 1U << (baseShift(slot) + 1);
147 void MicroLockCore::unlock(unsigned slot) {
148 detail::Futex<>* wordPtr = word();
152 oldWord = wordPtr->load(std::memory_order_relaxed);
154 assert(oldWord & heldBit(slot));
155 newWord = oldWord & ~(heldBit(slot) | waitBit(slot));
156 } while (!wordPtr->compare_exchange_weak(
157 oldWord, newWord, std::memory_order_release, std::memory_order_relaxed));
159 if (oldWord & waitBit(slot)) {
160 // We don't track the number of waiters, so wake everyone
161 (void)wordPtr->futexWake(std::numeric_limits<int>::max(), heldBit(slot));
165 template <unsigned MaxSpins = 1000, unsigned MaxYields = 0>
166 class MicroLockBase : public MicroLockCore {
168 inline void lock(unsigned slot) NO_SANITIZE_ADDRESS;
169 inline void lock() { lock(0); }
170 inline bool try_lock(unsigned slot) NO_SANITIZE_ADDRESS;
171 inline bool try_lock() { return try_lock(0); }
174 template <unsigned MaxSpins, unsigned MaxYields>
175 bool MicroLockBase<MaxSpins, MaxYields>::try_lock(unsigned slot) {
177 // N.B. You might think that try_lock is just the fast path of lock,
178 // but you'd be wrong. Keep in mind that other parts of our host
179 // word might be changing while we take the lock! We're not allowed
180 // to fail spuriously if the lock is in fact not held, even if other
181 // people are concurrently modifying other parts of the word.
183 // We need to loop until we either see firm evidence that somebody
184 // else has the lock (by looking at heldBit) or see our CAS succeed.
185 // A failed CAS by itself does not indicate lock-acquire failure.
187 detail::Futex<>* wordPtr = word();
188 uint32_t oldWord = wordPtr->load(std::memory_order_relaxed);
190 if (oldWord & heldBit(slot)) {
193 } while (!wordPtr->compare_exchange_weak(oldWord,
194 oldWord | heldBit(slot),
195 std::memory_order_acquire,
196 std::memory_order_relaxed));
201 template <unsigned MaxSpins, unsigned MaxYields>
202 void MicroLockBase<MaxSpins, MaxYields>::lock(unsigned slot) {
204 static_assert(MaxSpins + MaxYields < (unsigned)-1, "overflow");
206 detail::Futex<>* wordPtr = word();
208 oldWord = wordPtr->load(std::memory_order_relaxed);
209 if ((oldWord & heldBit(slot)) == 0 &&
210 wordPtr->compare_exchange_weak(oldWord,
211 oldWord | heldBit(slot),
212 std::memory_order_acquire,
213 std::memory_order_relaxed)) {
214 // Fast uncontended case: memory_order_acquire above is our barrier
216 // lockSlowPath doesn't have any slot-dependent computation; it
217 // just shifts the input bit. Make sure its shifting produces the
218 // same result a call to waitBit for our slot would.
219 assert(heldBit(slot) << 1 == waitBit(slot));
220 // lockSlowPath emits its own memory barrier
221 lockSlowPath(oldWord, wordPtr, heldBit(slot), MaxSpins, MaxYields);
225 typedef MicroLockBase<> MicroLock;