[UNTESTED] moved library into a single header, added basic SPSC test

This commit is contained in:
2026-01-02 07:14:27 -05:00
parent 0a73580a80
commit ffd5ce5b12
9 changed files with 457 additions and 635 deletions

View File

@@ -1,22 +1,23 @@
LIB_NAME ?= weaver
BUILD_DIR ?= build BUILD_DIR ?= build
SRC_DIR ?= src DEPS_DIR ?= $(BUILD_DIR)/deps
INCLUDE_DIRS := include src/include TEST_DIR ?= tests
INCLUDE_DIR := include
MODE ?= debug MODE ?= debug
RELEASE_OPT ?= 2 RELEASE_OPT ?= 3
CLINE_SIZE ?= 64 CLINE_SIZE ?= 64
CXX ?= clang++ CXX ?= clang++
AR ?= ar AR ?= ar
ARFLAGS := rcs ARFLAGS := rcs
SRCS := $(shell find $(SRC_DIR) -name '*.cc') TESTS := $(shell find $(TEST_DIR) -name '*.cc')
OBJS := $(patsubst $(SRC_DIR)/%.cc,$(BUILD_DIR)/%.o,$(SRCS))
DEPS := $(OBJS:.o=.d)
INCLUDES := $(addprefix -I,$(INCLUDE_DIRS)) TEST_BINS := $(patsubst $(TEST_DIR)/%.cc,$(BUILD_DIR)/tests/%,$(TESTS))
TEST_DEPS := $(patsubst $(TEST_DIR)/%.cc,$(DEPS_DIR)/tests/%.d,$(TESTS))
INCLUDES := $(addprefix -I,$(INCLUDE_DIR))
WARN_FLAGS := \ WARN_FLAGS := \
-Wall \ -Wall \
@@ -26,7 +27,6 @@ WARN_FLAGS := \
COMMON_CXXFLAGS := \ COMMON_CXXFLAGS := \
-std=c++20 \ -std=c++20 \
-MMD -MP \
-flto \ -flto \
--param=destructive-interference-size=$(CLINE_SIZE) \ --param=destructive-interference-size=$(CLINE_SIZE) \
$(WARN_FLAGS) \ $(WARN_FLAGS) \
@@ -50,21 +50,17 @@ else
$(error Unknown MODE '$(MODE)'; use MODE=debug or MODE=release) $(error Unknown MODE '$(MODE)'; use MODE=debug or MODE=release)
endif endif
LIB := $(BUILD_DIR)/lib$(LIB_NAME).a .PHONY: all clean tests
.PHONY: all clean all: tests
tests: $(TEST_BINS)
all: $(LIB) $(BUILD_DIR)/tests/%: $(TEST_DIR)/%.cc
$(LIB): $(OBJS)
@mkdir -p $(dir $@/bin)
$(AR) $(ARFLAGS) $@/bin $^
$(BUILD_DIR)/%.o: $(SRC_DIR)/%.cc
@mkdir -p $(dir $@) @mkdir -p $(dir $@)
$(CXX) $(CXXFLAGS) -c $< -o $@ @mkdir -p $(DEPS_DIR)/tests/$(dir $*)
$(CXX) $(CXXFLAGS) -MMD -MP -MF $(DEPS_DIR)/tests/$*.d $< -o $@
-include $(DEPS)
clean: clean:
rm -rf $(BUILD_DIR) rm -rf $(BUILD_DIR)
-include $(TEST_DEPS)

365
include/weaver.h Normal file
View File

@@ -0,0 +1,365 @@
#ifndef THWEAVER_FABRIC_H
#define THWEAVER_FABRIC_H
#include <array>
#include <atomic>
#include <cstddef>
#include <cstdint>
#include <new>
namespace THWeaver {
static constexpr size_t CLS = std::hardware_destructive_interference_size;
enum class EndpointQueueSendResult : uint32_t { Ok = 0, ErrFull = 1 };
enum class EndpointQueueRecvResult { Ok, ErrEmpty };
enum class FanInFabricSendResult : uint32_t { Ok = 0, ErrFull = 1 };
enum class FanInFabricRecvBitmapResult { Ok, ErrBitmapEmpty, ErrBitmapFailed };
enum class FanInFabricRecvTryResult { Ok, ErrTryFailed };
enum class FanInFabricRecvAggrResult { Ok, ErrAggrFailed };
template <typename MessageType, std::size_t qsize_expt> class EndpointQueue {
public:
static constexpr std::size_t QSIZE = 1 << qsize_expt;
static constexpr std::size_t QSIZE_MASK = QSIZE - 1;
EndpointQueue() noexcept {};
EndpointQueue(const EndpointQueue<MessageType, qsize_expt> &) = delete;
EndpointQueue &
operator=(const EndpointQueue<MessageType, qsize_expt> &) = delete;
EndpointQueue(EndpointQueue<MessageType, qsize_expt> &&) = delete;
EndpointQueue &
operator=(EndpointQueue<MessageType, qsize_expt> &&) = delete;
void init() noexcept {
head.store(0);
tail.store(0);
}
void flush() noexcept {
head.store(0);
tail.store(0);
}
std::size_t size() const noexcept {
return ((tail.load() + QSIZE_MASK) - head.load()) & QSIZE_MASK;
}
bool is_empty() const noexcept { return head.load() == tail.load(); }
bool is_full() const noexcept {
return head.load() == ((tail.load() + 1) & QSIZE_MASK);
}
std::size_t get_head() const noexcept { return head.load(); }
std::size_t get_tail() const noexcept { return tail.load(); }
EndpointQueueSendResult send(MessageType &&message) noexcept {
std::size_t h = head.load();
std::size_t t = tail.load();
if (h == ((t + 1) & QSIZE_MASK))
return EndpointQueueSendResult::ErrFull;
buffer[t] = std::move(message);
t = (t + 1) & QSIZE_MASK;
tail.store(t);
return EndpointQueueSendResult::Ok;
}
EndpointQueueRecvResult recv(MessageType &buffer) noexcept {
std::size_t h = head.load();
std::size_t t = tail.load();
if (h == t)
return EndpointQueueRecvResult::ErrEmpty;
buffer = std::move(this->buffer[h]);
h = (h + 1) & QSIZE_MASK;
head.store(h);
return EndpointQueueRecvResult::Ok;
}
void recv_unsafe(MessageType &buffer, std::size_t old_head) noexcept {
buffer = std::move(this->buffer[old_head]);
head.store((old_head + 1) & QSIZE_MASK);
}
private:
alignas(CLS) std::array<MessageType, QSIZE> buffer;
alignas(CLS) std::atomic<std::size_t> head;
alignas(CLS) std::atomic<std::size_t> tail;
};
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
class FanInFabric {
public:
using BitmapType = uint64_t;
static constexpr std::size_t QSIZE = 1 << qsize_expt;
static constexpr std::size_t QSIZE_MASK = QSIZE - 1;
static constexpr std::size_t BITMAP_SIZE = sizeof(BitmapType) * 8;
static constexpr BitmapType THREAD_BITMAP_MASK =
IN_THREAD_CNT == BITMAP_SIZE
? ~BitmapType(0)
: (BitmapType(1) << IN_THREAD_CNT) - 1;
static_assert(IN_THREAD_CNT <= BITMAP_SIZE, "Producer limit exceeded");
FanInFabric() noexcept {}
FanInFabric(const FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>
&) = delete;
FanInFabric &
operator=(const FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT> &) =
delete;
FanInFabric(FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT> &&) =
delete;
FanInFabric &
operator=(FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT> &&) =
delete;
void init() noexcept {
for (int i = 0; i < IN_THREAD_CNT; ++i)
in_queues[i].init();
hint.store(0);
dl.init();
rotation = 0;
}
template <std::size_t thread_id> void flush() noexcept {
static_assert(thread_id < IN_THREAD_CNT,
"Thread ID out of bounds");
in_queues[thread_id].flush();
hint.fetch_and(~(1 << thread_id), std::memory_order_relaxed);
}
void flush_all() noexcept {
for (int i = 0; i < IN_THREAD_CNT; ++i)
in_queues[i].flush();
hint.store(0, std::memory_order_relaxed);
dl.init();
rotation = 0;
}
template <std::size_t thread_id>
FanInFabricSendResult send(MessageType &&message) noexcept {
static_assert(thread_id < IN_THREAD_CNT,
"Thread ID out of bounds");
return static_cast<FanInFabricSendResult>(static_cast<uint32_t>(
in_queues[thread_id].send(message)));
}
FanInFabricRecvBitmapResult
recv_bitmap_only(MessageType &buffer) noexcept {
BitmapType bitmap = hint.load(std::memory_order_relaxed);
if (!bitmap)
return FanInFabricRecvBitmapResult::ErrBitmapEmpty;
bitmap = rotate(bitmap);
int hinted_id =
(std::countr_zero(bitmap) + rotation) % IN_THREAD_CNT;
rotation = (hinted_id + 1) % IN_THREAD_CNT;
std::size_t head = in_queues[hinted_id].get_head();
std::size_t tail = in_queues[hinted_id].get_tail();
std::size_t qsize = ((tail + QSIZE_MASK) - head) & QSIZE_MASK;
if (qsize <= 1)
hint.fetch_and(~(1 << hinted_id),
std::memory_order_relaxed);
if (qsize) {
in_queues[hinted_id].recv_unsafe(buffer, head);
dl.move_to_last(hinted_id);
return FanInFabricRecvBitmapResult::Ok;
}
return FanInFabricRecvBitmapResult::ErrBitmapFailed;
}
FanInFabricRecvTryResult recv_try(MessageType &buffer) noexcept {
BitmapType bitmap = hint.load(std::memory_order_relaxed);
if (bitmap) {
bitmap = rotate(bitmap);
int hinted_id = (std::countr_zero(bitmap) + rotation) %
IN_THREAD_CNT;
rotation = (hinted_id + 1) % IN_THREAD_CNT;
std::size_t head = in_queues[hinted_id].get_head();
std::size_t tail = in_queues[hinted_id].get_tail();
std::size_t qsize =
((tail + QSIZE_MASK) - head) & QSIZE_MASK;
if (qsize <= 1)
hint.fetch_and(~(1 << hinted_id),
std::memory_order_relaxed);
if (qsize) {
in_queues[hinted_id].recv_unsafe(buffer, head);
dl.move_to_last(hinted_id);
return FanInFabricRecvTryResult::Ok;
}
}
std::size_t head = in_queues[dl.first].get_head();
std::size_t tail = in_queues[dl.first].get_tail();
if (head == tail)
return FanInFabricRecvTryResult::ErrTryFailed;
in_queues[dl.first].recv_unsafe(buffer, head);
dl.move_first_to_last();
return FanInFabricRecvTryResult::Ok;
}
FanInFabricRecvAggrResult recv_aggr(MessageType &buffer) noexcept {
BitmapType bitmap = hint.load(std::memory_order_relaxed);
if (bitmap) {
bitmap = rotate(bitmap);
int hinted_id = (std::countr_zero(bitmap) + rotation) %
IN_THREAD_CNT;
rotation = (hinted_id + 1) % IN_THREAD_CNT;
std::size_t head = in_queues[hinted_id].get_head();
std::size_t tail = in_queues[hinted_id].get_tail();
std::size_t qsize =
((tail + QSIZE_MASK) - head) & QSIZE_MASK;
if (qsize <= 1)
hint.fetch_and(~(1 << hinted_id),
std::memory_order_relaxed);
if (qsize) {
in_queues[hinted_id].recv_unsafe(buffer, head);
dl.move_to_last(hinted_id);
return FanInFabricRecvAggrResult::Ok;
}
}
uint8_t last = dl.last;
do {
uint8_t first = dl.first;
std::size_t head = in_queues[dl.first].get_head();
std::size_t tail = in_queues[dl.first].get_tail();
dl.move_first_to_last();
if (head != tail) {
in_queues[first].recv_unsafe(buffer, head);
return FanInFabricRecvAggrResult::Ok;
}
} while (dl.last != last);
return FanInFabricRecvAggrResult::ErrAggrFailed;
}
void full_scan(BitmapType &bitmap) noexcept {
bitmap = 0;
uint8_t curr = dl.first;
if (in_queues[curr].size())
bitmap |= 1 << curr;
do {
curr = dl.next[curr];
if (in_queues[curr].size())
bitmap |= 1 << curr;
} while (curr != dl.last);
hint.fetch_or(bitmap, std::memory_order_release);
}
template <std::size_t thread_id> std::size_t size() const noexcept {
static_assert(thread_id < IN_THREAD_CNT,
"Thread ID out of bounds");
return in_queues[thread_id].size();
}
template <std::size_t thread_id> bool is_empty() const noexcept {
static_assert(thread_id < IN_THREAD_CNT,
"Thread ID out of bounds");
return in_queues[thread_id].is_empty();
}
template <std::size_t thread_id> bool is_full() const noexcept {
static_assert(thread_id < IN_THREAD_CNT,
"Thread ID out of bounds");
return in_queues[thread_id].is_full();
}
private:
std::array<EndpointQueue<MessageType, qsize_expt>, IN_THREAD_CNT>
in_queues;
alignas(CLS) std::atomic<BitmapType> hint;
struct alignas(CLS) DiscoveryList {
std::array<uint8_t, IN_THREAD_CNT> next;
std::array<uint8_t, IN_THREAD_CNT> prev;
uint8_t first;
uint8_t last;
void init() noexcept {
for (int i = 0; i < IN_THREAD_CNT; ++i) {
next[i] = (i + 1) % IN_THREAD_CNT;
prev[i] =
(i + IN_THREAD_CNT - 1) % IN_THREAD_CNT;
}
first = 0;
last = IN_THREAD_CNT - 1;
}
void move_first_to_last() noexcept {
prev[first] = last;
next[last] = first;
last = next[last];
first = next[last];
}
void move_to_last(uint8_t index) noexcept {
if (index == last)
return;
else if (index == first) {
move_first_to_last();
} else {
prev[next[index]] = prev[index];
next[prev[index]] = next[index];
next[last] = index;
prev[index] = last;
last = index;
}
}
} dl;
std::size_t rotation;
BitmapType rotate(BitmapType bitmap) const noexcept {
bitmap &= THREAD_BITMAP_MASK;
if (rotation == 0)
return bitmap;
return ((bitmap >> rotation) |
(bitmap << (IN_THREAD_CNT - rotation))) &
THREAD_BITMAP_MASK;
}
};
} // namespace THWeaver
#endif

View File

@@ -1,88 +0,0 @@
#ifndef THWEAVER_ATOMIC_H
#define THWEAVER_ATOMIC_H
#include <cstdint>
#include <concepts>
#include <limits>
#include <type_traits>
template <typename T>
concept Additive = std::is_integral_v<T> && !std::is_same_v<T, bool>;
template <typename A, typename T>
concept AtomicAdditive =
requires(A a, T v) {
{
a.load()
}
-> std::same_as<T>;
{
a.store(v)
};
{
a.fetch_add(v)
}
-> std::same_as<T>;
{
a.fetch_sub(v)
}
-> std::same_as<T>;
};
template <typename T>
concept Bitwise =
(std::is_integral_v<T> || std::is_enum_v<T>) &&
!std::is_same_v<T, bool>;
template <typename A, typename T>
concept AtomicBitwise =
requires(A a, T v) {
{
a.load()
}
-> std::same_as<T>;
{
a.store(v)
};
{
a.fetch_or(v)
}
-> std::same_as<T>;
{
a.fetch_and(v)
}
-> std::same_as<T>;
{
a.fetch_xor(v)
}
-> std::same_as<T>;
};
// Uses the full bitmap
template<typename Atomic, typename T, uint32_t NumBits, typename Index = uint32_t>
requires Bitwise<T> && AtomicBitwise<Atomic, T> &&
Additive<Index> && std::is_unsigned_v<Index>
class AtomicBitmap {
public:
static_assert(NumBits <= std::numeric_limits<T>::digits);
AtomicBitmap() noexcept
: bits(T {}), rotation(0) {}
void reset(const T &bits, const Index rotation) noexcept;
bool get_bit(const Index n) const noexcept;
void set_bit(const Index n) noexcept;
void clear_bit(const Index n) noexcept;
void rotate(const Index n) noexcept;
Index get_rotation() const noexcept;
Index get_low_bit() const noexcept;
Index rotate_to_low_bit(const Index offset = 0) noexcept;
private:
Atomic bits;
Index rotation;
};
#endif

View File

@@ -1,35 +0,0 @@
#ifndef THWEAVER_BITWISE_H
#define THWEAVER_BITWISE_H
#include <concepts>
template <typename T>
concept Bitwise =
requires(T a, T b) {
{
a & b
}
-> std::same_as<T>;
{
a | b
}
-> std::same_as<T>;
{
a ^ b
}
-> std::same_as<T>;
{
~a
}
-> std::same_as<T>;
{
a << 1
}
-> std::same_as<T>;
{
a >> 1
}
-> std::same_as<T>;
};
#endif

View File

@@ -1,114 +0,0 @@
#ifndef THWEAVER_FABRIC_H
#define THWEAVER_FABRIC_H
#include <array>
#include <atomic>
#include <cstddef>
#include <cstdint>
#include <new>
#include <weaver/results.h>
namespace THWeaver {
template <typename MessageType, std::size_t qsize_expt> class EndpointQueue {
public:
static constexpr size_t CLS =
std::hardware_destructive_interference_size;
static constexpr std::size_t QSIZE = 1 << qsize_expt;
static constexpr std::size_t QSIZE_MASK = QSIZE - 1;
EndpointQueue(const EndpointQueue<MessageType, qsize_expt> &) = delete;
EndpointQueue &
operator=(const EndpointQueue<MessageType, qsize_expt> &) = delete;
EndpointQueue(EndpointQueue<MessageType, qsize_expt> &&) = delete;
EndpointQueue &
operator=(EndpointQueue<MessageType, qsize_expt> &&) = delete;
void init() noexcept;
void flush() noexcept;
std::size_t size() const noexcept;
bool is_empty() const noexcept;
bool is_full() const noexcept;
std::size_t get_head() const noexcept;
std::size_t get_tail() const noexcept;
EndpointQueueSendResult send(MessageType &&message) noexcept;
EndpointQueueRecvResult recv(MessageType &buffer) noexcept;
void recv_unsafe(MessageType &buffer, std::size_t old_head) noexcept;
private:
alignas(CLS) std::array<MessageType, QSIZE> buffer;
alignas(CLS) std::atomic<std::size_t> head;
alignas(CLS) std::atomic<std::size_t> tail;
};
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
class FanInFabric {
public:
using BitmapType = uint64_t;
static constexpr size_t CLS =
std::hardware_destructive_interference_size;
static constexpr std::size_t QSIZE = 1 << qsize_expt;
static constexpr std::size_t QSIZE_MASK = QSIZE - 1;
static constexpr std::size_t BITMAP_SIZE = sizeof(BitmapType) * 8;
static constexpr BitmapType THREAD_BITMAP_MASK =
IN_THREAD_CNT == BITMAP_SIZE
? ~BitmapType(0)
: (BitmapType(1) << IN_THREAD_CNT) - 1;
static_assert(IN_THREAD_CNT <= BITMAP_SIZE, "Producer limit exceeded");
FanInFabric(const FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>
&) = delete;
FanInFabric &
operator=(const FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT> &) =
delete;
FanInFabric(FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT> &&) =
delete;
FanInFabric &
operator=(FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT> &&) =
delete;
void init() noexcept;
template <std::size_t thread_id> void flush() noexcept;
void flush_all() noexcept;
template <std::size_t thread_id>
FanInFabricSendResult send(MessageType &&message) noexcept;
FanInFabricRecvBitmapResult
recv_bitmap_only(MessageType &buffer) noexcept;
FanInFabricRecvTryResult recv_try(MessageType &buffer) noexcept;
FanInFabricRecvAggrResult recv_aggr(MessageType &buffer) noexcept;
void full_scan(BitmapType &bitmap) noexcept;
template <std::size_t thread_id> std::size_t size() const noexcept;
template <std::size_t thread_id> bool is_empty() const noexcept;
template <std::size_t thread_id> bool is_full() const noexcept;
private:
std::array<EndpointQueue<MessageType, qsize_expt>, IN_THREAD_CNT>
in_queues;
alignas(CLS) std::atomic<BitmapType> hint;
struct alignas(CLS) DiscoveryList {
std::array<uint8_t, IN_THREAD_CNT> next;
std::array<uint8_t, IN_THREAD_CNT> prev;
uint8_t first;
uint8_t last;
void init() noexcept;
void move_first_to_last() noexcept;
void move_to_last(uint8_t index) noexcept;
} dl;
std::size_t rotation;
BitmapType rotate(BitmapType bitmap) const noexcept;
};
} // namespace THWeaver
#endif

View File

@@ -1,21 +0,0 @@
#ifndef THWEAVER_RESULTS_H
#define THWEAVER_RESULTS_H
#include <cstdint>
namespace THWeaver {
enum class EndpointQueueSendResult : uint32_t { Ok = 0, ErrFull = 1 };
enum class EndpointQueueRecvResult { Ok, ErrEmpty };
enum class FanInFabricSendResult : uint32_t { Ok = 0, ErrFull = 1 };
enum class FanInFabricRecvBitmapResult { Ok, ErrBitmapEmpty, ErrBitmapFailed };
enum class FanInFabricRecvTryResult { Ok, ErrTryFailed };
enum class FanInFabricRecvAggrResult { Ok, ErrAggrFailed };
} // namespace THWeaver
#endif

View File

@@ -1,267 +0,0 @@
#include <atomic>
#include <bit>
#include <cstddef>
#include <cstdint>
#include <weaver/fabric.h>
#include <weaver/results.h>
namespace THWeaver {
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
void FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::init() noexcept {
for (int i = 0; i < IN_THREAD_CNT; ++i)
in_queues[i].init();
hint.store(0);
dl.init();
rotation = 0;
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
template <std::size_t thread_id>
void FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::flush() noexcept {
static_assert(thread_id < IN_THREAD_CNT, "Thread ID out of bounds");
in_queues[thread_id].flush();
hint.fetch_and(~(1 << thread_id), std::memory_order_relaxed);
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
void FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::flush_all() noexcept {
for (int i = 0; i < IN_THREAD_CNT; ++i)
in_queues[i].flush();
hint.store(0, std::memory_order_relaxed);
dl.init();
rotation = 0;
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
template <std::size_t thread_id>
FanInFabricSendResult FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::send(
MessageType &&message) noexcept {
static_assert(thread_id < IN_THREAD_CNT, "Thread ID out of bounds");
return static_cast<FanInFabricSendResult>(
static_cast<uint32_t>(in_queues[thread_id].send(message)));
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
template <std::size_t thread_id>
std::size_t
FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::size() const noexcept {
static_assert(thread_id < IN_THREAD_CNT, "Thread ID out of bounds");
return in_queues[thread_id].size();
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
template <std::size_t thread_id>
bool FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::is_empty()
const noexcept {
static_assert(thread_id < IN_THREAD_CNT, "Thread ID out of bounds");
return in_queues[thread_id].is_empty();
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
template <std::size_t thread_id>
bool FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::is_full()
const noexcept {
static_assert(thread_id < IN_THREAD_CNT, "Thread ID out of bounds");
return in_queues[thread_id].is_full();
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
FanInFabricRecvBitmapResult
FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::recv_bitmap_only(
MessageType &buffer) noexcept {
BitmapType bitmap = hint.load(std::memory_order_relaxed);
if (!bitmap)
return FanInFabricRecvBitmapResult::ErrBitmapEmpty;
bitmap = rotate(bitmap);
int hinted_id = (std::countr_zero(bitmap) + rotation) % IN_THREAD_CNT;
rotation = (hinted_id + 1) % IN_THREAD_CNT;
std::size_t head = in_queues[hinted_id].get_head();
std::size_t tail = in_queues[hinted_id].get_tail();
std::size_t qsize = ((tail + QSIZE_MASK) - head) & QSIZE_MASK;
if (qsize <= 1)
hint.fetch_add(~(1 << hinted_id), std::memory_order_relaxed);
if (qsize) {
in_queues[hinted_id].recv_unsafe(buffer, head);
dl.move_to_last(hinted_id);
return FanInFabricRecvBitmapResult::Ok;
}
return FanInFabricRecvBitmapResult::ErrBitmapFailed;
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
FanInFabricRecvTryResult
FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::recv_try(
MessageType &buffer) noexcept {
BitmapType bitmap = hint.load(std::memory_order_relaxed);
if (bitmap) {
bitmap = rotate(bitmap);
int hinted_id =
(std::countr_zero(bitmap) + rotation) % IN_THREAD_CNT;
rotation = (hinted_id + 1) % IN_THREAD_CNT;
std::size_t head = in_queues[hinted_id].get_head();
std::size_t tail = in_queues[hinted_id].get_tail();
std::size_t qsize = ((tail + QSIZE_MASK) - head) & QSIZE_MASK;
if (qsize <= 1)
hint.fetch_add(~(1 << hinted_id),
std::memory_order_relaxed);
if (qsize) {
in_queues[hinted_id].recv_unsafe(buffer, head);
dl.move_to_last(hinted_id);
return FanInFabricRecvTryResult::Ok;
}
}
std::size_t head = in_queues[dl.first].get_head();
std::size_t tail = in_queues[dl.first].get_tail();
if (head == tail)
return FanInFabricRecvTryResult::ErrTryFailed;
in_queues[dl.first].recv_unsafe(buffer, head);
dl.move_first_to_last();
return FanInFabricRecvTryResult::Ok;
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
FanInFabricRecvAggrResult
FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::recv_aggr(
MessageType &buffer) noexcept {
BitmapType bitmap = hint.load(std::memory_order_relaxed);
if (bitmap) {
bitmap = rotate(bitmap);
int hinted_id =
(std::countr_zero(bitmap) + rotation) % IN_THREAD_CNT;
rotation = (hinted_id + 1) % IN_THREAD_CNT;
std::size_t head = in_queues[hinted_id].get_head();
std::size_t tail = in_queues[hinted_id].get_tail();
std::size_t qsize = ((tail + QSIZE_MASK) - head) & QSIZE_MASK;
if (qsize <= 1)
hint.fetch_add(~(1 << hinted_id),
std::memory_order_relaxed);
if (qsize) {
in_queues[hinted_id].recv_unsafe(buffer, head);
dl.move_to_last(hinted_id);
return FanInFabricRecvAggrResult::Ok;
}
}
uint8_t last = dl.last;
do {
std::size_t head = in_queues[dl.first].get_head();
std::size_t tail = in_queues[dl.first].get_tail();
dl.move_first_to_last();
if (head != tail) {
in_queues[dl.first].recv_unsafe(buffer, head);
return FanInFabricRecvAggrResult::Ok;
}
} while (dl.last != last);
return FanInFabricRecvAggrResult::ErrAggrFailed;
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
void FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::full_scan(
BitmapType &bitmap) noexcept {
bitmap = 0;
uint8_t curr = dl.first;
if (in_queues[curr].size())
bitmap |= 1 << curr;
do {
curr = dl.next[curr];
if (in_queues[curr].size())
bitmap |= 1 << curr;
} while (curr != dl.last);
hint.fetch_or(bitmap, std::memory_order_release);
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
void FanInFabric<MessageType, qsize_expt,
IN_THREAD_CNT>::DiscoveryList::init() noexcept {
for (int i = 0; i < IN_THREAD_CNT; ++i) {
next[i] = (i + 1) % IN_THREAD_CNT;
prev[i] = (i + IN_THREAD_CNT - 1) % IN_THREAD_CNT;
}
first = 0;
last = IN_THREAD_CNT - 1;
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
void FanInFabric<MessageType, qsize_expt,
IN_THREAD_CNT>::DiscoveryList::move_first_to_last() noexcept {
prev[first] = last;
next[last] = first;
last = next[last];
first = next[last];
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
void FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::DiscoveryList::
move_to_last(uint8_t index) noexcept {
if (index == last)
return;
else if (index == first) {
move_first_to_last();
} else {
prev[next[index]] = prev[index];
next[prev[index]] = next[index];
next[last] = index;
prev[index] = last;
last = index;
}
}
template <typename MessageType, std::size_t qsize_expt,
std::size_t IN_THREAD_CNT>
FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::BitmapType
FanInFabric<MessageType, qsize_expt, IN_THREAD_CNT>::rotate(
BitmapType bitmap) const noexcept {
bitmap &= THREAD_BITMAP_MASK;
if (rotation == 0)
return bitmap;
return ((bitmap >> rotation) | (bitmap << (IN_THREAD_CNT - rotation))) &
THREAD_BITMAP_MASK;
}
} // namespace THWeaver

View File

@@ -1,89 +0,0 @@
#include <atomic>
#include <cstddef>
#include <weaver/fabric.h>
#include <weaver/results.h>
namespace THWeaver {
template <typename MessageType, std::size_t qsize_expt>
void EndpointQueue<MessageType, qsize_expt>::init() noexcept {
head.store(0);
tail.store(0);
}
template <typename MessageType, std::size_t qsize_expt>
void EndpointQueue<MessageType, qsize_expt>::flush() noexcept {
head.store(0);
tail.store(0);
}
template <typename MessageType, std::size_t qsize_expt>
EndpointQueueSendResult
EndpointQueue<MessageType, qsize_expt>::send(MessageType &&message) noexcept {
std::size_t h = head.load(std::memory_order_acquire);
std::size_t t = tail.load(std::memory_order_acquire);
if (h == ((t + 1) & QSIZE_MASK))
return EndpointQueueSendResult::ErrFull;
buffer[t] = std::move(message);
t = (t + 1) & QSIZE_MASK;
tail.store(t);
return EndpointQueueSendResult::Ok;
}
template <typename MessageType, std::size_t qsize_expt>
EndpointQueueRecvResult
EndpointQueue<MessageType, qsize_expt>::recv(MessageType &buffer) noexcept {
std::size_t h = head.load();
std::size_t t = tail.load();
if (h == t)
return EndpointQueueRecvResult::ErrEmpty;
buffer = std::move(this->buffer[h]);
h = (h + 1) & QSIZE_MASK;
head.store(h);
return EndpointQueueRecvResult::Ok;
}
template <typename MessageType, std::size_t qsize_expt>
void EndpointQueue<MessageType, qsize_expt>::recv_unsafe(
MessageType &buffer, std::size_t old_head) noexcept {
buffer = std::move(this->buffer[old_head]);
head.store((old_head + 1) & QSIZE_MASK);
}
template <typename MessageType, std::size_t qsize_expt>
std::size_t EndpointQueue<MessageType, qsize_expt>::size() const noexcept {
return ((tail.load() + QSIZE_MASK) - head.load()) & QSIZE_MASK;
}
template <typename MessageType, std::size_t qsize_expt>
bool EndpointQueue<MessageType, qsize_expt>::is_empty() const noexcept {
return head.load() == tail.load();
}
template <typename MessageType, std::size_t qsize_expt>
bool EndpointQueue<MessageType, qsize_expt>::is_full() const noexcept {
return head.load() != ((tail.load() + 1) & QSIZE_MASK);
}
template <typename MessageType, std::size_t qsize_expt>
std::size_t EndpointQueue<MessageType, qsize_expt>::get_head() const noexcept {
return head.load();
}
template <typename MessageType, std::size_t qsize_expt>
std::size_t EndpointQueue<MessageType, qsize_expt>::get_tail() const noexcept {
return tail.load();
}
} // namespace THWeaver

View File

@@ -0,0 +1,75 @@
#include <chrono>
#include <cstdint>
#include <iostream>
#include <thread>
#include <vector>
#include <weaver.h>
using std::chrono::nanoseconds;
using std::chrono::steady_clock;
int init() {
int n = 0;
std::cin >> n;
return n;
}
int main() {
std::vector<steady_clock::time_point> send_time;
std::vector<steady_clock::time_point> recv_time;
std::vector<uint64_t> received;
int n = init();
if (n <= 0) {
std::cerr << "n must be positive\n";
return 1;
}
received.resize(n);
send_time.resize(n);
recv_time.resize(n);
THWeaver::EndpointQueue<uint64_t, 10> queue;
queue.init();
std::thread producer([&]() {
for (int i = 0; i < n; ++i) {
while (queue.send(static_cast<uint64_t>(i + 1)) !=
THWeaver::EndpointQueueSendResult::Ok) {
}
send_time[i] = steady_clock::now();
}
});
std::thread consumer([&]() {
for (int i = 0; i < n; ++i) {
uint64_t value = 0;
while (queue.recv(value) !=
THWeaver::EndpointQueueRecvResult::Ok) {
}
recv_time[i] = steady_clock::now();
received[i] = value;
}
});
producer.join();
consumer.join();
for (int i = 0; i < n; ++i) {
uint64_t expected = static_cast<uint64_t>(i + 1);
if (received[i] != expected) {
std::cerr << "mismatch at " << i << ": expected "
<< expected << ", got " << received[i]
<< "\n";
return 1;
}
}
for (int i = 0; i < n; ++i) {
nanoseconds duration = std::chrono::duration_cast<nanoseconds>(
recv_time[i] - send_time[i]);
std::cout << (i + 1) << "," << duration.count() << "\n";
}
return 0;
}