31 Atomic operations library [atomics]

31.4 Order and consistency [atomics.order]

namespace std { enum class memory_order : unspecified { relaxed, consume, acquire, release, acq_rel, seq_cst }; inline constexpr memory_order memory_order_relaxed = memory_order::relaxed; inline constexpr memory_order memory_order_consume = memory_order::consume; inline constexpr memory_order memory_order_acquire = memory_order::acquire; inline constexpr memory_order memory_order_release = memory_order::release; inline constexpr memory_order memory_order_acq_rel = memory_order::acq_rel; inline constexpr memory_order memory_order_seq_cst = memory_order::seq_cst; }
The enumeration memory_­order specifies the detailed regular (non-atomic) memory synchronization order as defined in [intro.multithread] and may provide for operation ordering.
Its enumerated values and their meanings are as follows:
  • memory_­order​::​relaxed: no operation orders memory.
  • memory_­order​::​release, memory_­order​::​acq_­rel, and memory_­order​::​seq_­cst: a store operation performs a release operation on the affected memory location.
  • memory_­order​::​consume: a load operation performs a consume operation on the affected memory location.
    [Note 1:
    Prefer memory_­order​::​acquire, which provides stronger guarantees than memory_­order​::​consume.
    Implementations have found it infeasible to provide performance better than that of memory_­order​::​acquire.
    Specification revisions are under consideration.
    — end note]
  • memory_­order​::​acquire, memory_­order​::​acq_­rel, and memory_­order​::​seq_­cst: a load operation performs an acquire operation on the affected memory location.
[Note 2:
Atomic operations specifying memory_­order​::​relaxed are relaxed with respect to memory ordering.
Implementations must still guarantee that any given atomic access to a particular atomic object be indivisible with respect to all other atomic accesses to that object.
— end note]
An atomic operation A that performs a release operation on an atomic object M synchronizes with an atomic operation B that performs an acquire operation on M and takes its value from any side effect in the release sequence headed by A.
An atomic operation A on some atomic object M is coherence-ordered before another atomic operation B on M if
  • A is a modification, and B reads the value stored by A, or
  • A precedes B in the modification order of M, or
  • A and B are not the same atomic read-modify-write operation, and there exists an atomic modification X of M such that A reads the value stored by X and X precedes B in the modification order of M, or
  • there exists an atomic modification X of M such that A is coherence-ordered before X and X is coherence-ordered before B.
There is a single total order S on all memory_­order​::​seq_­cst operations, including fences, that satisfies the following constraints.
First, if A and B are memory_­order​::​seq_­cst operations and A strongly happens before B, then A precedes B in S.
Second, for every pair of atomic operations A and B on an object M, where A is coherence-ordered before B, the following four conditions are required to be satisfied by S:
  • if A and B are both memory_­order​::​seq_­cst operations, then A precedes B in S; and
  • if A is a memory_­order​::​seq_­cst operation and B happens before a memory_­order​::​seq_­cst fence Y, then A precedes Y in S; and
  • if a memory_­order​::​seq_­cst fence X happens before A and B is a memory_­order​::​seq_­cst operation, then X precedes B in S; and
  • if a memory_­order​::​seq_­cst fence X happens before A and B happens before a memory_­order​::​seq_­cst fence Y, then X precedes Y in S.
[Note 3:
This definition ensures that S is consistent with the modification order of any atomic object M.
It also ensures that a memory_­order​::​seq_­cst load A of M gets its value either from the last modification of M that precedes A in S or from some non-memory_­order​::​seq_­cst modification of M that does not happen before any modification of M that precedes A in S.
— end note]
[Note 4:
We do not require that S be consistent with “happens before” ([intro.races]).
This allows more efficient implementation of memory_­order​::​acquire and memory_­order​::​release on some machine architectures.
It can produce surprising results when these are mixed with memory_­order​::​seq_­cst accesses.
— end note]
[Note 5:
memory_­order​::​seq_­cst ensures sequential consistency only for a program that is free of data races and uses exclusively memory_­order​::​seq_­cst atomic operations.
Any use of weaker ordering will invalidate this guarantee unless extreme care is used.
In many cases, memory_­order​::​seq_­cst atomic operations are reorderable with respect to other atomic operations performed by the same thread.
— end note]
Implementations should ensure that no “out-of-thin-air” values are computed that circularly depend on their own computation.
[Note 6:
For example, with x and y initially zero, // Thread 1: r1 = y.load(memory_order::relaxed); x.store(r1, memory_order::relaxed);
// Thread 2: r2 = x.load(memory_order::relaxed); y.store(r2, memory_order::relaxed); should not produce r1 == r2 == 42, since the store of 42 to y is only possible if the store to x stores 42, which circularly depends on the store to y storing 42.
Note that without this restriction, such an execution is possible.
— end note]
[Note 7:
The recommendation similarly disallows r1 == r2 == 42 in the following example, with x and y again initially zero: // Thread 1: r1 = x.load(memory_order::relaxed); if (r1 == 42) y.store(42, memory_order::relaxed);
// Thread 2: r2 = y.load(memory_order::relaxed); if (r2 == 42) x.store(42, memory_order::relaxed); — end note]
Atomic read-modify-write operations shall always read the last value (in the modification order) written before the write associated with the read-modify-write operation.
Implementations should make atomic stores visible to atomic loads within a reasonable amount of time.
template<class T> T kill_dependency(T y) noexcept;
Effects: The argument does not carry a dependency to the return value.
Returns: y.