Multi-Threading

Base.Threads.@threads — Macro

  1. Threads.@threads

A macro to parallelize a for-loop to run with multiple threads. This spawns nthreads() number of threads, splits the iteration space amongst them, and iterates in parallel. A barrier is placed at the end of the loop which waits for all the threads to finish execution, and the loop returns.

source

Base.Threads.@spawn — Macro

  1. Threads.@spawn expr

Create and run a Task on any available thread. To wait for the task to finish, call wait on the result of this macro, or call fetch to wait and then obtain its return value.

Note

This feature is currently considered experimental.

Julia 1.3

This macro is available as of Julia 1.3.

source

Base.Threads.threadid — Function

  1. Threads.threadid()

Get the ID number of the current thread of execution. The master thread has ID 1.

source

Base.Threads.nthreads — Function

  1. Threads.nthreads()

Get the number of threads available to the Julia process. This is the inclusive upper bound on threadid().

source

Synchronization

Base.Threads.Condition — Type

  1. Threads.Condition([lock])

A thread-safe version of Base.Condition.

Julia 1.2

This functionality requires at least Julia 1.2.

source

Base.Event — Type

  1. Event()

Create a level-triggered event source. Tasks that call wait on an Event are suspended and queued until notify is called on the Event. After notify is called, the Event remains in a signaled state and tasks will no longer block when waiting for it.

Julia 1.1

This functionality requires at least Julia 1.1.

source

See also Synchronization.

Atomic operations

Warning

The API for atomic operations has not yet been finalized and is likely to change.

Base.Threads.Atomic — Type

  1. Threads.Atomic{T}

Holds a reference to an object of type T, ensuring that it is only accessed atomically, i.e. in a thread-safe manner.

Only certain “simple” types can be used atomically, namely the primitive boolean, integer, and float-point types. These are Bool, Int8Int128, UInt8UInt128, and Float16Float64.

New atomic objects can be created from a non-atomic values; if none is specified, the atomic object is initialized with zero.

Atomic objects can be accessed using the [] notation:

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> x[] = 1
  4. 1
  5. julia> x[]
  6. 1

Atomic operations use an atomic_ prefix, such as atomic_add!, atomic_xchg!, etc.

source

Base.Threads.atomic_cas! — Function

  1. Threads.atomic_cas!(x::Atomic{T}, cmp::T, newval::T) where T

Atomically compare-and-set x

Atomically compares the value in x with cmp. If equal, write newval to x. Otherwise, leaves x unmodified. Returns the old value in x. By comparing the returned value to cmp (via ===) one knows whether x was modified and now holds the new value newval.

For further details, see LLVM’s cmpxchg instruction.

This function can be used to implement transactional semantics. Before the transaction, one records the value in x. After the transaction, the new value is stored only if x has not been modified in the mean time.

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> Threads.atomic_cas!(x, 4, 2);
  4. julia> x
  5. Base.Threads.Atomic{Int64}(3)
  6. julia> Threads.atomic_cas!(x, 3, 2);
  7. julia> x
  8. Base.Threads.Atomic{Int64}(2)

source

Base.Threads.atomic_xchg! — Function

  1. Threads.atomic_xchg!(x::Atomic{T}, newval::T) where T

Atomically exchange the value in x

Atomically exchanges the value in x with newval. Returns the old value.

For further details, see LLVM’s atomicrmw xchg instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> Threads.atomic_xchg!(x, 2)
  4. 3
  5. julia> x[]
  6. 2

source

Base.Threads.atomic_add! — Function

  1. Threads.atomic_add!(x::Atomic{T}, val::T) where T <: ArithmeticTypes

Atomically add val to x

Performs x[] += val atomically. Returns the old value. Not defined for Atomic{Bool}.

For further details, see LLVM’s atomicrmw add instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> Threads.atomic_add!(x, 2)
  4. 3
  5. julia> x[]
  6. 5

source

Base.Threads.atomic_sub! — Function

  1. Threads.atomic_sub!(x::Atomic{T}, val::T) where T <: ArithmeticTypes

Atomically subtract val from x

Performs x[] -= val atomically. Returns the old value. Not defined for Atomic{Bool}.

For further details, see LLVM’s atomicrmw sub instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> Threads.atomic_sub!(x, 2)
  4. 3
  5. julia> x[]
  6. 1

source

Base.Threads.atomic_and! — Function

  1. Threads.atomic_and!(x::Atomic{T}, val::T) where T

Atomically bitwise-and x with val

Performs x[] &= val atomically. Returns the old value.

For further details, see LLVM’s atomicrmw and instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> Threads.atomic_and!(x, 2)
  4. 3
  5. julia> x[]
  6. 2

source

Base.Threads.atomic_nand! — Function

  1. Threads.atomic_nand!(x::Atomic{T}, val::T) where T

Atomically bitwise-nand (not-and) x with val

Performs x[] = ~(x[] & val) atomically. Returns the old value.

For further details, see LLVM’s atomicrmw nand instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(3)
  2. Base.Threads.Atomic{Int64}(3)
  3. julia> Threads.atomic_nand!(x, 2)
  4. 3
  5. julia> x[]
  6. -3

source

Base.Threads.atomic_or! — Function

  1. Threads.atomic_or!(x::Atomic{T}, val::T) where T

Atomically bitwise-or x with val

Performs x[] |= val atomically. Returns the old value.

For further details, see LLVM’s atomicrmw or instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(5)
  2. Base.Threads.Atomic{Int64}(5)
  3. julia> Threads.atomic_or!(x, 7)
  4. 5
  5. julia> x[]
  6. 7

source

Base.Threads.atomic_xor! — Function

  1. Threads.atomic_xor!(x::Atomic{T}, val::T) where T

Atomically bitwise-xor (exclusive-or) x with val

Performs x[] $= val atomically. Returns the old value.

For further details, see LLVM’s atomicrmw xor instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(5)
  2. Base.Threads.Atomic{Int64}(5)
  3. julia> Threads.atomic_xor!(x, 7)
  4. 5
  5. julia> x[]
  6. 2

source

Base.Threads.atomic_max! — Function

  1. Threads.atomic_max!(x::Atomic{T}, val::T) where T

Atomically store the maximum of x and val in x

Performs x[] = max(x[], val) atomically. Returns the old value.

For further details, see LLVM’s atomicrmw max instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(5)
  2. Base.Threads.Atomic{Int64}(5)
  3. julia> Threads.atomic_max!(x, 7)
  4. 5
  5. julia> x[]
  6. 7

source

Base.Threads.atomic_min! — Function

  1. Threads.atomic_min!(x::Atomic{T}, val::T) where T

Atomically store the minimum of x and val in x

Performs x[] = min(x[], val) atomically. Returns the old value.

For further details, see LLVM’s atomicrmw min instruction.

Examples

  1. julia> x = Threads.Atomic{Int}(7)
  2. Base.Threads.Atomic{Int64}(7)
  3. julia> Threads.atomic_min!(x, 5)
  4. 7
  5. julia> x[]
  6. 5

source

Base.Threads.atomic_fence — Function

  1. Threads.atomic_fence()

Insert a sequential-consistency memory fence

Inserts a memory fence with sequentially-consistent ordering semantics. There are algorithms where this is needed, i.e. where an acquire/release ordering is insufficient.

This is likely a very expensive operation. Given that all other atomic operations in Julia already have acquire/release semantics, explicit fences should not be necessary in most cases.

For further details, see LLVM’s fence instruction.

source

ccall using a threadpool (Experimental)

Base.@threadcall — Macro

  1. @threadcall((cfunc, clib), rettype, (argtypes...), argvals...)

The @threadcall macro is called in the same way as ccall but does the work in a different thread. This is useful when you want to call a blocking C function without causing the main julia thread to become blocked. Concurrency is limited by size of the libuv thread pool, which defaults to 4 threads but can be increased by setting the UV_THREADPOOL_SIZE environment variable and restarting the julia process.

Note that the called function should never call back into Julia.

source

Low-level synchronization primitives

These building blocks are used to create the regular synchronization objects.

Base.Threads.SpinLock — Type

  1. SpinLock()

Create a non-reentrant lock. Recursive use will result in a deadlock. Each lock must be matched with an unlock.

Test-and-test-and-set spin locks are quickest up to about 30ish contending threads. If you have more contention than that, perhaps a lock is the wrong way to synchronize.

source