1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
|
template <typename T>
struct MPSCNode {
std::atomic<MPSCNode<T> *> next;
T value;
};
//
// Multiple Producer Single Consumer Lockless Queue
// URL: https://www.1024cores.net
//
template <typename T>
struct MPSCQueue {
MPSCNode<T> sentinel;
std::atomic<MPSCNode<T> *> head;
std::atomic<MPSCNode<T> *> tail;
std::atomic<isize> count;
};
template <typename T> gb_internal void mpsc_init (MPSCQueue<T> *q, gbAllocator const &allocator);
template <typename T> gb_internal void mpsc_destroy(MPSCQueue<T> *q);
template <typename T> gb_internal isize mpsc_enqueue(MPSCQueue<T> *q, T const &value);
template <typename T> gb_internal bool mpsc_dequeue(MPSCQueue<T> *q, T *value_);
template <typename T>
gb_internal void mpsc_init(MPSCQueue<T> *q, gbAllocator const &allocator) {
q->sentinel.next.store(nullptr, std::memory_order_relaxed);
q->head.store(&q->sentinel, std::memory_order_relaxed);
q->tail.store(&q->sentinel, std::memory_order_relaxed);
q->count.store(0, std::memory_order_relaxed);
}
template <typename T>
gb_internal void mpsc_destroy(MPSCQueue<T> *q) {
GB_ASSERT(q->count.load() == 0);
}
template <typename T>
gb_internal MPSCNode<T> *mpsc_alloc_node(MPSCQueue<T> *q, T const &value) {
// auto new_node = gb_alloc_item(heap_allocator(), MPSCNode<T>);
auto new_node = gb_alloc_item(permanent_allocator(), MPSCNode<T>);
new_node->value = value;
return new_node;
}
template <typename T>
gb_internal void mpsc_free_node(MPSCQueue<T> *q, MPSCNode<T> *node) {
// TODO(bill): determine a good way to handle the freed nodes rather than letting them leak
}
template <typename T>
gb_internal isize mpsc_enqueue(MPSCQueue<T> *q, MPSCNode<T> *node) {
node->next.store(nullptr, std::memory_order_relaxed);
auto prev = q->head.exchange(node, std::memory_order_acq_rel);
prev->next.store(node, std::memory_order_release);
isize count = 1 + q->count.fetch_add(1, std::memory_order_relaxed);
return count;
}
template <typename T>
gb_internal isize mpsc_enqueue(MPSCQueue<T> *q, T const &value) {
auto node = mpsc_alloc_node(q, value);
return mpsc_enqueue(q, node);
}
template <typename T>
gb_internal bool mpsc_dequeue(MPSCQueue<T> *q, T *value_) {
auto tail = q->tail.load(std::memory_order_relaxed);
auto next = tail->next.load(std::memory_order_relaxed);
if (next) {
q->tail.store(next, std::memory_order_relaxed);
if (value_) *value_ = next->value;
q->count.fetch_sub(1, std::memory_order_relaxed);
mpsc_free_node(q, tail);
return true;
}
GB_ASSERT(q->count.load(std::memory_order_acquire) == 0);
return false;
}
////////////////////////////
#define MPMC_CACHE_LINE_SIZE 64
typedef std::atomic<i32> MPMCQueueAtomicIdx;
// Multiple Producer Multiple Consumer Queue
template <typename T>
struct MPMCQueue {
static size_t const PAD0_OFFSET = (sizeof(T *) + sizeof(MPMCQueueAtomicIdx *) + sizeof(gbAllocator) + sizeof(BlockingMutex) + sizeof(i32) + sizeof(i32));
T * nodes;
MPMCQueueAtomicIdx *indices;
BlockingMutex mutex;
MPMCQueueAtomicIdx count;
i32 mask; // capacity-1, because capacity must be a power of 2
char pad0[(MPMC_CACHE_LINE_SIZE*2 - PAD0_OFFSET) % MPMC_CACHE_LINE_SIZE];
MPMCQueueAtomicIdx head_idx;
char pad1[MPMC_CACHE_LINE_SIZE - sizeof(i32)];
MPMCQueueAtomicIdx tail_idx;
};
gb_internal gbAllocator mpmc_allocator(void) {
return heap_allocator();
}
gb_internal void mpmc_internal_init_indices(MPMCQueueAtomicIdx *indices, i32 offset, i32 size) {
GB_ASSERT(offset % 8 == 0);
GB_ASSERT(size % 8 == 0);
// NOTE(bill): pretend it's not atomic for performance
auto *raw_data = cast(i32 *)indices;
for (i32 i = offset; i < size; i += 8) {
raw_data[i+0] = i+0;
raw_data[i+1] = i+1;
raw_data[i+2] = i+2;
raw_data[i+3] = i+3;
raw_data[i+4] = i+4;
raw_data[i+5] = i+5;
raw_data[i+6] = i+6;
raw_data[i+7] = i+7;
}
}
template <typename T>
gb_internal void mpmc_init(MPMCQueue<T> *q, isize size_i) {
if (size_i < 8) {
size_i = 8;
}
GB_ASSERT(size_i < I32_MAX);
i32 size = cast(i32)size_i;
size = next_pow2(size);
GB_ASSERT(gb_is_power_of_two(size));
q->mask = size-1;
gbAllocator a = mpmc_allocator();
q->nodes = gb_alloc_array(a, T, size);
q->indices = gb_alloc_array(a, MPMCQueueAtomicIdx, size);
mpmc_internal_init_indices(q->indices, 0, q->mask+1);
}
template <typename T>
gb_internal void mpmc_destroy(MPMCQueue<T> *q) {
gbAllocator a = mpmc_allocator();
gb_free(a, q->nodes);
gb_free(a, q->indices);
}
template <typename T>
gb_internal bool mpmc_internal_grow(MPMCQueue<T> *q) {
gbAllocator a = mpmc_allocator();
mutex_lock(&q->mutex);
i32 old_size = q->mask+1;
i32 new_size = old_size*2;
resize_array_raw(&q->nodes, a, old_size, new_size);
if (q->nodes == nullptr) {
GB_PANIC("Unable to resize enqueue: %td -> %td", old_size, new_size);
mutex_unlock(&q->mutex);
return false;
}
resize_array_raw(&q->indices, a, old_size, new_size);
if (q->indices == nullptr) {
GB_PANIC("Unable to resize enqueue: %td -> %td", old_size, new_size);
mutex_unlock(&q->mutex);
return false;
}
mpmc_internal_init_indices(q->indices, old_size, new_size);
q->mask = new_size-1;
mutex_unlock(&q->mutex);
return true;
}
template <typename T>
gb_internal i32 mpmc_enqueue(MPMCQueue<T> *q, T const &data) {
GB_ASSERT(q->mask != 0);
i32 head_idx = q->head_idx.load(std::memory_order_relaxed);
for (;;) {
i32 index = head_idx & q->mask;
auto node = &q->nodes[index];
auto node_idx_ptr = &q->indices[index];
i32 node_idx = node_idx_ptr->load(std::memory_order_acquire);
i32 diff = node_idx - head_idx;
if (diff == 0) {
i32 next_head_idx = head_idx+1;
if (q->head_idx.compare_exchange_weak(head_idx, next_head_idx)) {
*node = data;
node_idx_ptr->store(next_head_idx, std::memory_order_release);
return q->count.fetch_add(1, std::memory_order_release);
}
} else if (diff < 0) {
if (!mpmc_internal_grow(q)) {
return -1;
}
} else {
head_idx = q->head_idx.load(std::memory_order_relaxed);
}
}
}
template <typename T>
gb_internal bool mpmc_dequeue(MPMCQueue<T> *q, T *data_) {
if (q->mask == 0) {
return false;
}
i32 tail_idx = q->tail_idx.load(std::memory_order_relaxed);
for (;;) {
auto node_ptr = &q->nodes[tail_idx & q->mask];
auto node_idx_ptr = &q->indices[tail_idx & q->mask];
i32 node_idx = node_idx_ptr->load(std::memory_order_acquire);
i32 diff = node_idx - (tail_idx+1);
if (diff == 0) {
i32 next_tail_idx = tail_idx+1;
if (q->tail_idx.compare_exchange_weak(tail_idx, next_tail_idx)) {
if (data_) *data_ = *node_ptr;
node_idx_ptr->store(tail_idx + q->mask + 1, std::memory_order_release);
q->count.fetch_sub(1, std::memory_order_release);
return true;
}
} else if (diff < 0) {
return false;
} else {
tail_idx = q->tail_idx.load(std::memory_order_relaxed);
}
}
}
|
