API reference

Main HPX library reference

template<typename Action>
struct async_result

#include <colocating_distribution_policy.hpp>

Note

This function is part of the invocation policy implemented by this class

Public Types

template<>
using type = hpx::future<typename traits::promise_local_result<typename hpx::traits::extract_action<Action>::remote_result_type>::type>

template<typename Action>
struct async_result

#include <default_distribution_policy.hpp>

Note

This function is part of the invocation policy implemented by this class

Public Types

template<>
using type = hpx::future<typename traits::promise_local_result<typename hpx::traits::extract_action<Action>::remote_result_type>::type>

struct auto_chunk_size

#include <auto_chunk_size.hpp>

Loop iterations are divided into pieces and then assigned to threads. The number of loop iterations combined is determined based on measurements of how long the execution of 1% of the overall number of iterations takes. This executor parameters type makes sure that as many loop iterations are combined as necessary to run for the amount of time specified.

Public Functions

auto_chunk_size()

Construct an auto_chunk_size executor parameters object

Note

Default constructed auto_chunk_size executor parameter types will use 80 microseconds as the minimal time for which any of the scheduled chunks should run.

auto_chunk_size(hpx::util::steady_duration const &rel_time)

Construct an auto_chunk_size executor parameters object

Parameters

• rel_time: [in] The time duration to use as the minimum to decide how many loop iterations should be combined.

class barrier

#include <barrier.hpp>

The barrier is an implementation performing a barrier over a number of participating threads. The different threads don’t have to be on the same locality. This barrier can be invoked in a distributed application.

For a local only barrier

See

hpx::lcos::local::barrier.

Public Functions

barrier(std::string const &base_name)

Creates a barrier, rank is locality id, size is number of localities

A barrier

base_name is created. It expects that hpx::get_num_localities() participate and the local rank is hpx::get_locality_id().

Parameters

• base_name: The name of the barrier

barrier(std::string const &base_name, std::size_t num)

Creates a barrier with a given size, rank is locality id

A barrier

base_name is created. It expects that num participate and the local rank is hpx::get_locality_id().

Parameters

• base_name: The name of the barrier
• num: The number of participating threads

barrier(std::string const &base_name, std::size_t num, std::size_t rank)

Creates a barrier with a given size and rank

A barrier

base_name is created. It expects that num participate and the local rank is rank.

Parameters

• base_name: The name of the barrier
• num: The number of participating threads
• rank: The rank of the calling site for this invocation

barrier(std::string const &base_name, std::vector<std::size_t> const &ranks, std::size_t rank)

Creates a barrier with a vector of ranks

A barrier

base_name is created. It expects that ranks.size() and the local rank is rank (must be contained in ranks).

Parameters

• base_name: The name of the barrier
• ranks: Gives a list of participating ranks (this could be derived from a list of locality ids
• rank: The rank of the calling site for this invocation

void wait()

Wait until each participant entered the barrier. Must be called by all participants

Return

This function returns once all participants have entered the barrier (have called wait).

hpx::future<void> wait(hpx::launch::async_policy)

Wait until each participant entered the barrier. Must be called by all participants

Return

a future that becomes ready once all participants have entered the barrier (have called wait).

Public Static Functions

static void synchronize()

Perform a global synchronization using the default global barrier The barrier is created once at startup and can be reused throughout the lifetime of an HPX application.

Note

This function currently does not support dynamic connection and disconnection of localities.

struct binpacking_distribution_policy

#include <binpacking_distribution_policy.hpp>

This class specifies the parameters for a binpacking distribution policy to use for creating a given number of items on a given set of localities. The binpacking policy will distribute the new objects in a way such that each of the localities will equalize the number of overall objects of this type based on a given criteria (by default this criteria is the overall number of objects of this type).

Public Functions

binpacking_distribution_policy()

Default-construct a new instance of a binpacking_distribution_policy. This policy will represent one locality (the local locality).

binpacking_distribution_policy operator()(std::vector<id_type> const &locs, char const *perf_counter_name = default_binpacking_counter_name) const

Create a new default_distribution policy representing the given set of localities.

Parameters

• locs: [in] The list of localities the new instance should represent
• perf_counter_name: [in] The name of the performance counter which should be used as the distribution criteria (by default the overall number of existing instances of the given component type will be used).

binpacking_distribution_policy operator()(std::vector<id_type> &&locs, char const *perf_counter_name = default_binpacking_counter_name) const

Create a new default_distribution policy representing the given set of localities.

Parameters

• locs: [in] The list of localities the new instance should represent
• perf_counter_name: [in] The name of the performance counter which should be used as the distribution criteria (by default the overall number of existing instances of the given component type will be used).

binpacking_distribution_policy operator()(id_type const &loc, char const *perf_counter_name = default_binpacking_counter_name) const

Create a new default_distribution policy representing the given locality

Parameters

• loc: [in] The locality the new instance should represent
• perf_counter_name: [in] The name of the performance counter which should be used as the distribution criteria (by default the overall number of existing instances of the given component type will be used).

template<typename Component, typename ...Ts>
hpx::future<hpx::id_type> create(Ts&&... vs) const

Create one object on one of the localities associated by this policy instance

Return

A future holding the global address which represents the newly created object

Parameters

• vs: [in] The arguments which will be forwarded to the constructor of the new object.

template<typename Component, typename ...Ts>
hpx::future<std::vector<bulk_locality_result>> bulk_create(std::size_t count, Ts&&... vs) const

Create multiple objects on the localities associated by this policy instance

Return

A future holding the list of global addresses which represent the newly created objects

Parameters

• count: [in] The number of objects to create
• vs: [in] The arguments which will be forwarded to the constructors of the new objects.

std::string const &get_counter_name() const

Returns the name of the performance counter associated with this policy instance.

std::size_t get_num_localities() const

Returns the number of associated localities for this distribution policy

Note

This function is part of the creation policy implemented by this class

class checkpoint

#include <checkpoint.hpp>

Checkpoint Object

Checkpoint is the container object which is produced by save_checkpoint and is consumed by a restore_checkpoint. A checkpoint may be moved into the save_checkpoint object to write the byte stream to the pre-created checkpoint object.

Public Types

using const_iterator = std::vector::const_iterator

Public Functions

checkpoint()

checkpoint(checkpoint const &c)

checkpoint(checkpoint &&c)

~checkpoint()

checkpoint(std::vector<char> const &vec)

checkpoint(std::vector<char> &&vec)

checkpoint &operator=(checkpoint const &c)

checkpoint &operator=(checkpoint &&c)

bool operator==(checkpoint const &c) const

bool operator!=(checkpoint const &c) const

const_iterator begin() const

const_iterator end() const

size_t size() const

Private Functions

template<typename Archive>
void serialize(Archive &arch, const unsigned int version)

Private Members

std::vector<char> data

Friends

friend hpx::util::checkpoint::hpx::serialization::access

std::ostream &operator<<(std::ostream &ost, checkpoint const &ckp)

Operator<< Overload

This overload is the main way to write data from a checkpoint to an object such as a file. Inside the function, the size of the checkpoint will be written to the stream before the checkpoint’s data. The operator>> overload uses this to read the correct number of bytes. Be mindful of this additional write and read when you use different facilities to write out or read in data to a checkpoint!

Parameters

• ost: Output stream to write to.
• ckp: Checkpoint to copy from.

Return

Operator<< returns the ostream object.

std::istream &operator>>(std::istream &ist, checkpoint &ckp)

Operator>> Overload

This overload is the main way to read in data from an object such as a file to a checkpoint. It is important to note that inside the function, the first variable to be read is the size of the checkpoint. This size variable is written to the stream before the checkpoint’s data in the operator<< overload. Be mindful of this additional read and write when you use different facilities to read in or write out data from a checkpoint!

Parameters

• ist: Input stream to write from.
• ckp: Checkpoint to write to.

Return

Operator>> returns the ostream object.

template<typename T, typename ...Ts>
void restore_checkpoint(checkpoint const &c, T &t, Ts&... ts)

Resurrect

Restore_checkpoint takes a checkpoint object as a first argument and the containers which will be filled from the byte stream (in the same order as they were placed in save_checkpoint).

Return

Restore_checkpoint returns void.

Template Parameters

• T: A container to restore.
• Ts: Other containers to restore. Containers must be in the same order that they were inserted into the checkpoint.

Parameters

• c: The checkpoint to restore.
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

struct colocating_distribution_policy

#include <colocating_distribution_policy.hpp>

This class specifies the parameters for a distribution policy to use for creating a given number of items on the locality where a given object is currently placed.

Public Functions

colocating_distribution_policy()

Default-construct a new instance of a colocating_distribution_policy. This policy will represent the local locality.

colocating_distribution_policy operator()(id_type const &id) const

Create a new colocating_distribution_policy representing the locality where the given object os current located

Parameters

• id: [in] The global address of the object with which the new instances should be colocated on

template<typename Client, typename Stub>
colocating_distribution_policy operator()(client_base<Client, Stub> const &client) const

Create a new colocating_distribution_policy representing the locality where the given object os current located

Parameters

• client: [in] The client side representation of the object with which the new instances should be colocated on

template<typename Component, typename ...Ts>
hpx::future<hpx::id_type> create(Ts&&... vs) const

Create one object on the locality of the object this distribution policy instance is associated with

Note

This function is part of the placement policy implemented by this class

Return

A future holding the global address which represents the newly created object

Parameters

• vs: [in] The arguments which will be forwarded to the constructor of the new object.

template<typename Component, typename ...Ts>
hpx::future<std::vector<bulk_locality_result>> bulk_create(std::size_t count, Ts&&... vs) const

Create multiple objects colocated with the object represented by this policy instance

Note

This function is part of the placement policy implemented by this class

Return

A future holding the list of global addresses which represent the newly created objects

Parameters

• count: [in] The number of objects to create
• vs: [in] The arguments which will be forwarded to the constructors of the new objects.

template<typename Action, typename ...Ts>
async_result<Action>::type async(launch policy, Ts&&... vs) const

template<typename Action, typename Callback, typename ...Ts>
async_result<Action>::type async_cb(launch policy, Callback &&cb, Ts&&... vs) const

Note

This function is part of the invocation policy implemented by this class

template<typename Action, typename Continuation, typename ...Ts>
bool apply(Continuation &&c, threads::thread_priority priority, Ts&&... vs) const

Note

This function is part of the invocation policy implemented by this class

template<typename Action, typename ...Ts>
bool apply(threads::thread_priority priority, Ts&&... vs) const

template<typename Action, typename Continuation, typename Callback, typename ...Ts>
bool apply_cb(Continuation &&c, threads::thread_priority priority, Callback &&cb, Ts&&... vs) const

Note

This function is part of the invocation policy implemented by this class

template<typename Action, typename Callback, typename ...Ts>
bool apply_cb(threads::thread_priority priority, Callback &&cb, Ts&&... vs) const

std::size_t get_num_localities() const

Returns the number of associated localities for this distribution policy

Note

This function is part of the creation policy implemented by this class

hpx::id_type get_next_target() const

Returns the locality which is anticipated to be used for the next async operation

class core

#include <partitioner.hpp>

Public Functions

core(std::size_t id = invalid_core_id, numa_domain *domain = nullptr)

std::vector<pu> const &pus() const

std::size_t id() const

Private Functions

std::vector<core> cores_sharing_numa_domain()

Private Members

std::size_t id_

numa_domain *domain_

std::vector<pu> pus_

Private Static Attributes

const std::size_t invalid_core_id = std::size_t(-1)

Friends

friend hpx::resource::core::pu

friend hpx::resource::core::numa_domain

struct default_distribution_policy

#include <default_distribution_policy.hpp>

This class specifies the parameters for a simple distribution policy to use for creating (and evenly distributing) a given number of items on a given set of localities.

Public Functions

default_distribution_policy()

Default-construct a new instance of a default_distribution_policy. This policy will represent one locality (the local locality).

default_distribution_policy operator()(std::vector<id_type> const &locs) const

Create a new default_distribution policy representing the given set of localities.

Parameters

• locs: [in] The list of localities the new instance should represent

default_distribution_policy operator()(std::vector<id_type> &&locs) const

Create a new default_distribution policy representing the given set of localities.

Parameters

• locs: [in] The list of localities the new instance should represent

default_distribution_policy operator()(id_type const &loc) const

Create a new default_distribution policy representing the given locality

Parameters

• loc: [in] The locality the new instance should represent

template<typename Component, typename ...Ts>
hpx::future<hpx::id_type> create(Ts&&... vs) const

Create one object on one of the localities associated by this policy instance

Note

This function is part of the placement policy implemented by this class

Return

A future holding the global address which represents the newly created object

Parameters

• vs: [in] The arguments which will be forwarded to the constructor of the new object.

template<typename Component, typename ...Ts>
hpx::future<std::vector<bulk_locality_result>> bulk_create(std::size_t count, Ts&&... vs) const

Create multiple objects on the localities associated by this policy instance

Note

This function is part of the placement policy implemented by this class

Return

A future holding the list of global addresses which represent the newly created objects

Parameters

• count: [in] The number of objects to create
• vs: [in] The arguments which will be forwarded to the constructors of the new objects.

template<typename Action, typename ...Ts>
async_result<Action>::type async(launch policy, Ts&&... vs) const

template<typename Action, typename Callback, typename ...Ts>
async_result<Action>::type async_cb(launch policy, Callback &&cb, Ts&&... vs) const

Note

This function is part of the invocation policy implemented by this class

template<typename Action, typename Continuation, typename ...Ts>
bool apply(Continuation &&c, threads::thread_priority priority, Ts&&... vs) const

Note

This function is part of the invocation policy implemented by this class

template<typename Action, typename ...Ts>
bool apply(threads::thread_priority priority, Ts&&... vs) const

template<typename Action, typename Continuation, typename Callback, typename ...Ts>
bool apply_cb(Continuation &&c, threads::thread_priority priority, Callback &&cb, Ts&&... vs) const

Note

This function is part of the invocation policy implemented by this class

template<typename Action, typename Callback, typename ...Ts>
bool apply_cb(threads::thread_priority priority, Callback &&cb, Ts&&... vs) const

std::size_t get_num_localities() const

Returns the number of associated localities for this distribution policy

Note

This function is part of the creation policy implemented by this class

hpx::id_type get_next_target() const

Returns the locality which is anticipated to be used for the next async operation

struct dynamic_chunk_size

#include <dynamic_chunk_size.hpp>

Loop iterations are divided into pieces of size chunk_size and then dynamically scheduled among the threads; when a thread finishes one chunk, it is dynamically assigned another If chunk_size is not specified, the default chunk size is 1.

Note

This executor parameters type is equivalent to OpenMP’s DYNAMIC scheduling directive.

Public Functions

dynamic_chunk_size(std::size_t chunk_size = 1)

Construct a dynamic_chunk_size executor parameters object

Parameters

• chunk_size: [in] The optional chunk size to use as the number of loop iterations to schedule together. The default chunk size is 1.

class error_code : public error_code

#include <error_code.hpp>

A hpx::error_code represents an arbitrary error condition.

The class hpx::error_code describes an object used to hold error code values, such as those originating from the operating system or other low-level application program interfaces.

Note

Class hpx::error_code is an adjunct to error reporting by exception

Public Functions

error_code(throwmode mode = plain)

Construct an object of type error_code.

Parameters

• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• nothing:

error_code(error e, throwmode mode = plain)

Construct an object of type error_code.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• nothing:

error_code(error e, char const *func, char const *file, long line, throwmode mode = plain)

Construct an object of type error_code.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• func: The name of the function where the error was raised.
• file: The file name of the code where the error was raised.
• line: The line number of the code line where the error was raised.
• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• nothing:

error_code(error e, char const *msg, throwmode mode = plain)

Construct an object of type error_code.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• msg: The parameter msg holds the error message the new exception should encapsulate.
• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• std::bad_alloc: (if allocation of a copy of the passed string fails).

error_code(error e, char const *msg, char const *func, char const *file, long line, throwmode mode = plain)

Construct an object of type error_code.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• msg: The parameter msg holds the error message the new exception should encapsulate.
• func: The name of the function where the error was raised.
• file: The file name of the code where the error was raised.
• line: The line number of the code line where the error was raised.
• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• std::bad_alloc: (if allocation of a copy of the passed string fails).

error_code(error e, std::string const &msg, throwmode mode = plain)

Construct an object of type error_code.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• msg: The parameter msg holds the error message the new exception should encapsulate.
• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• std::bad_alloc: (if allocation of a copy of the passed string fails).

error_code(error e, std::string const &msg, char const *func, char const *file, long line, throwmode mode = plain)

Construct an object of type error_code.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• msg: The parameter msg holds the error message the new exception should encapsulate.
• func: The name of the function where the error was raised.
• file: The file name of the code where the error was raised.
• line: The line number of the code line where the error was raised.
• mode: The parameter mode specifies whether the constructed hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

Exceptions

• std::bad_alloc: (if allocation of a copy of the passed string fails).

std::string get_message() const

Return a reference to the error message stored in the hpx::error_code.

Exceptions

• nothing:

void clear()

Clear this error_code object. The postconditions of invoking this method are.

• value() == hpx::success and category() == hpx::get_hpx_category()

error_code(error_code const &rhs)

Copy constructor for error_code

Note

This function maintains the error category of the left hand side if the right hand side is a success code.

error_code &operator=(error_code const &rhs)

Assignment operator for error_code

Note

This function maintains the error category of the left hand side if the right hand side is a success code.

Private Functions

error_code(int err, hpx::exception const &e)

error_code(std::exception_ptr const &e)

Private Members

std::exception_ptr exception_

Friends

friend hpx::error_code::exception

error_code make_error_code(std::exception_ptr const &e)

class exception : public system_error

#include <exception.hpp>

A hpx::exception is the main exception type used by HPX to report errors.

The hpx::exception type is the main exception type used by HPX to report errors. Any exceptions thrown by functions in the HPX library are either of this type or of a type derived from it. This implies that it is always safe to use this type only in catch statements guarding HPX library calls.

Subclassed by hpx::exception_list, hpx::parallel::v2::task_canceled_exception

Public Functions

exception(error e = success)

Construct a hpx::exception from a hpx::error.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.

exception(boost::system::system_error const &e)

Construct a hpx::exception from a boost::system_error.

exception(boost::system::error_code const &e)

Construct a hpx::exception from a boost::system::error_code (this is new for Boost V1.69). This constructor is required to compensate for the changes introduced as a resolution to LWG3162 (https://cplusplus.github.io/LWG/issue3162).

exception(error e, char const *msg, throwmode mode = plain)

Construct a hpx::exception from a hpx::error and an error message.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• msg: The parameter msg holds the error message the new exception should encapsulate.
• mode: The parameter mode specifies whether the returned hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

exception(error e, std::string const &msg, throwmode mode = plain)

Construct a hpx::exception from a hpx::error and an error message.

Parameters

• e: The parameter e holds the hpx::error code the new exception should encapsulate.
• msg: The parameter msg holds the error message the new exception should encapsulate.
• mode: The parameter mode specifies whether the returned hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

~exception()

Destruct a hpx::exception

Exceptions

• nothing:

error get_error() const

The function get_error() returns the hpx::error code stored in the referenced instance of a hpx::exception. It returns the hpx::error code this exception instance was constructed from.

Exceptions

• nothing:

error_code get_error_code(throwmode mode = plain) const

The function get_error_code() returns a hpx::error_code which represents the same error condition as this hpx::exception instance.

Parameters

• mode: The parameter mode specifies whether the returned hpx::error_code belongs to the error category hpx_category (if mode is plain, this is the default) or to the category hpx_category_rethrow (if mode is rethrow).

class exception_list : public hpx::exception

#include <exception_list.hpp>

The class exception_list is a container of exception_ptr objects parallel algorithms may use to communicate uncaught exceptions encountered during parallel execution to the caller of the algorithm

The type exception_list::const_iterator fulfills the requirements of a forward iterator.

Public Types

typedef exception_list_type::const_iterator iterator

bidirectional iterator

Public Functions

std::size_t size() const

The number of exception_ptr objects contained within the exception_list.

Note

Complexity: Constant time.

exception_list_type::const_iterator begin() const

An iterator referring to the first exception_ptr object contained within the exception_list.

exception_list_type::const_iterator end() const

An iterator which is the past-the-end value for the exception_list.

Private Types

typedef hpx::lcos::local::spinlock mutex_type

typedef std::list<std::exception_ptr> exception_list_type

Private Members

exception_list_type exceptions_

mutex_type mtx_

struct guided_chunk_size

#include <guided_chunk_size.hpp>

Iterations are dynamically assigned to threads in blocks as threads request them until no blocks remain to be assigned. Similar to dynamic_chunk_size except that the block size decreases each time a number of loop iterations is given to a thread. The size of the initial block is proportional to number_of_iterations / number_of_cores. Subsequent blocks are proportional to number_of_iterations_remaining / number_of_cores. The optional chunk size parameter defines the minimum block size. The default chunk size is 1.

Note

This executor parameters type is equivalent to OpenMP’s GUIDED scheduling directive.

Public Functions

guided_chunk_size(std::size_t min_chunk_size = 1)

Construct a guided_chunk_size executor parameters object

Parameters

• min_chunk_size: [in] The optional minimal chunk size to use as the minimal number of loop iterations to schedule together. The default minimal chunk size is 1.

struct invoke

#include <invoke.hpp>

Public Functions

template<typename F, typename... Ts>HPX_HOST_DEVICE util::invoke_result<F, Ts...>::type hpx::util::functional::invoke::operator()(F && f, Ts &&... vs) const

template<typename R>
struct invoke_r

#include <invoke.hpp>

Public Functions

template<typename F, typename... Ts>HPX_HOST_DEVICE R hpx::util::functional::invoke_r::operator()(F && f, Ts &&... vs) const

template<typename T>
struct is_async_execution_policy : public execution::detail::is_async_execution_policy<hpx::util::decay<T>::type>

#include <is_execution_policy.hpp>

Extension: Detect whether given execution policy makes algorithms asynchronous

1. The type is_async_execution_policy can be used to detect asynchronous execution policies for the purpose of excluding function signatures from otherwise ambiguous overload resolution participation.
2. If T is the type of a standard or implementation-defined execution policy, is_async_execution_policy<T> shall be publicly derived from integral_constant<bool, true>, otherwise from integral_constant<bool, false>.
3. The behavior of a program that adds specializations for is_async_execution_policy is undefined.

template<typename T>
struct is_execution_policy : public execution::detail::is_execution_policy<hpx::util::decay<T>::type>

#include <is_execution_policy.hpp>

1. The type is_execution_policy can be used to detect execution policies for the purpose of excluding function signatures from otherwise ambiguous overload resolution participation.
2. If T is the type of a standard or implementation-defined execution policy, is_execution_policy<T> shall be publicly derived from integral_constant<bool, true>, otherwise from integral_constant<bool, false>.
3. The behavior of a program that adds specializations for is_execution_policy is undefined.

template<typename T>
struct is_parallel_execution_policy : public execution::detail::is_parallel_execution_policy<hpx::util::decay<T>::type>

#include <is_execution_policy.hpp>

Extension: Detect whether given execution policy enables parallelization

1. The type is_parallel_execution_policy can be used to detect parallel execution policies for the purpose of excluding function signatures from otherwise ambiguous overload resolution participation.
2. If T is the type of a standard or implementation-defined execution policy, is_parallel_execution_policy<T> shall be publicly derived from integral_constant<bool, true>, otherwise from integral_constant<bool, false>.
3. The behavior of a program that adds specializations for is_parallel_execution_policy is undefined.

template<typename T>
struct is_sequenced_execution_policy : public execution::detail::is_sequenced_execution_policy<hpx::util::decay<T>::type>

#include <is_execution_policy.hpp>

Extension: Detect whether given execution policy does not enable parallelization

1. The type is_sequenced_execution_policy can be used to detect non-parallel execution policies for the purpose of excluding function signatures from otherwise ambiguous overload resolution participation.
2. If T is the type of a standard or implementation-defined execution policy, is_sequenced_execution_policy<T> shall be publicly derived from integral_constant<bool, true>, otherwise from integral_constant<bool, false>.
3. The behavior of a program that adds specializations for is_sequenced_execution_policy is undefined.

struct launch : public detail::policy_holder<>

#include <launch_policy.hpp>

Launch policies for hpx::async etc.

Public Functions

launch()

Default constructor. This creates a launch policy representing all possible launch modes

Public Static Attributes

const detail::fork_policy fork

Predefined launch policy representing asynchronous execution.The new thread is executed in a preferred way

const detail::sync_policy sync

Predefined launch policy representing synchronous execution.

const detail::deferred_policy deferred

Predefined launch policy representing deferred execution.

const detail::apply_policy apply

Predefined launch policy representing fire and forget execution.

const detail::select_policy_generator select

Predefined launch policy representing delayed policy selection.

class numa_domain

#include <partitioner.hpp>

Public Functions

numa_domain(std::size_t id = invalid_numa_domain_id)

std::vector<core> const &cores() const

std::size_t id() const

Private Members

std::size_t id_

std::vector<core> cores_

Private Static Attributes

const std::size_t invalid_numa_domain_id = std::size_t(-1)

Friends

friend hpx::resource::numa_domain::pu

friend hpx::resource::numa_domain::core

struct parallel_execution_tag

#include <execution_fwd.hpp>

Function invocations executed by a group of parallel execution agents execute in unordered fashion. Any such invocations executing in the same thread are indeterminately sequenced with respect to each other.

Note

parallel_execution_tag is weaker than sequenced_execution_tag.

struct parallel_policy

#include <execution_policy.hpp>

The class parallel_policy is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and indicate that a parallel algorithm’s execution may be parallelized.

Subclassed by hpx::parallel::execution::parallel_policy_shim< Executor, Parameters >

Public Types

typedef parallel_executor executor_type

The type of the executor associated with this execution policy.

typedef execution::extract_executor_parameters<executor_type>::type executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef parallel_execution_tag execution_category

The category of the execution agents created by this execution policy.

Public Functions

parallel_task_policy operator()(task_policy_tag) const

Create a new parallel_policy referencing a chunk size.

Return

The new parallel_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor>
rebind_executor<parallel_policy, Executor, executor_parameters_type>::type on(Executor &&exec) const

Create a new parallel_policy referencing an executor and a chunk size.

Return

The new parallel_policy

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with

template<typename ...Parameters, typename ParametersType = typename executor_parameters_join<Parameters...>::type>
rebind_executor<parallel_policy, executor_type, ParametersType>::type with(Parameters&&... params) const

Create a new parallel_policy from the given execution parameters

Note

Requires: is_executor_parameters<Parameters>::value is true

Return

The new parallel_policy

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

executor_type &executor()

Return the associated executor object.

executor_type const &executor() const

Return the associated executor object.

executor_parameters_type &parameters()

Return the associated executor parameters object.

executor_parameters_type const &parameters() const

Return the associated executor parameters object.

Private Functions

template<typename Archive>
void serialize(Archive &ar, const unsigned int version)

Private Members

executor_type exec_

executor_parameters_type params_

Friends

friend hpx::parallel::execution::parallel_policy::hpx::serialization::access

template<typename Policy>
struct parallel_policy_executor

#include <parallel_executor.hpp>

A parallel_executor creates groups of parallel execution agents which execute in threads implicitly created by the executor. This executor prefers continuing with the creating thread first before executing newly created threads.

This executor conforms to the concepts of a TwoWayExecutor, and a BulkTwoWayExecutor

Public Types

typedef parallel_execution_tag execution_category

Associate the parallel_execution_tag executor tag type as a default with this executor.

typedef static_chunk_size executor_parameters_type

Associate the static_chunk_size executor parameters type as a default with this executor.

Public Functions

parallel_policy_executor(Policy l = detail::get_default_policy<Policy>::call(), std::size_t spread = 4, std::size_t tasks = std::size_t(-1))

Create a new parallel executor.

template<typename Executor, typename Parameters>
struct parallel_policy_shim : public hpx::parallel::execution::parallel_policy

#include <execution_policy.hpp>

The class parallel_policy_shim is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and indicate that a parallel algorithm’s execution may be parallelized.

Public Types

typedef Executor executor_type

The type of the executor associated with this execution policy.

typedef Parameters executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef hpx::traits::executor_execution_category<executor_type>::type execution_category

The category of the execution agents created by this execution policy.

Public Functions

parallel_task_policy_shim<Executor, Parameters> operator()(task_policy_tag tag) const

Create a new parallel_policy referencing a chunk size.

Return

The new parallel_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor_>
rebind_executor<parallel_policy_shim, Executor_, executor_parameters_type>::type on(Executor_ &&exec) const

Create a new parallel_policy from the given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new parallel_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters_, typename ParametersType = typename executor_parameters_join<Parameters_...>::type>
rebind_executor<parallel_policy_shim, executor_type, ParametersType>::type with(Parameters_&&... params) const

Create a new parallel_policy_shim from the given execution parameters

Note

Requires: is_executor_parameters<Parameters>::value is true

Return

The new parallel_policy_shim

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

Executor &executor()

Return the associated executor object.

Executor const &executor() const

Return the associated executor object.

Parameters &parameters()

Return the associated executor parameters object.

Parameters const &parameters() const

Return the associated executor parameters object.

struct parallel_task_policy

#include <execution_policy.hpp>

Extension: The class parallel_task_policy is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and indicate that a parallel algorithm’s execution may be parallelized.

The algorithm returns a future representing the result of the corresponding algorithm when invoked with the parallel_policy.

Subclassed by hpx::parallel::execution::parallel_task_policy_shim< Executor, Parameters >

Public Types

typedef parallel_executor executor_type

The type of the executor associated with this execution policy.

typedef execution::extract_executor_parameters<executor_type>::type executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef parallel_execution_tag execution_category

The category of the execution agents created by this execution policy.

Public Functions

parallel_task_policy operator()(task_policy_tag) const

Create a new parallel_task_policy from itself

Return

The new parallel_task_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor>
rebind_executor<parallel_task_policy, Executor, executor_parameters_type>::type on(Executor &&exec) const

Create a new parallel_task_policy from given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new parallel_task_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters, typename ParametersType = typename executor_parameters_join<Parameters...>::type>
rebind_executor<parallel_task_policy, executor_type, ParametersType>::type with(Parameters&&... params) const

Create a new parallel_policy_shim from the given execution parameters

Note

Requires: all parameters are executor_parameters, different parameter types can’t be duplicated

Return

The new parallel_policy_shim

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

executor_type &executor()

Return the associated executor object.

executor_type const &executor() const

Return the associated executor object.

executor_parameters_type &parameters()

Return the associated executor parameters object.

executor_parameters_type const &parameters() const

Return the associated executor parameters object.

Private Functions

template<typename Archive>
void serialize(Archive &ar, const unsigned int version)

Private Members

executor_type exec_

executor_parameters_type params_

Friends

friend hpx::parallel::execution::parallel_task_policy::hpx::serialization::access

template<typename Executor, typename Parameters>
struct parallel_task_policy_shim : public hpx::parallel::execution::parallel_task_policy

#include <execution_policy.hpp>

Extension: The class parallel_task_policy_shim is an execution policy type used as a unique type to disambiguate parallel algorithm overloading based on combining a underlying parallel_task_policy and an executor and indicate that a parallel algorithm’s execution may be parallelized.

Public Types

typedef Executor executor_type

The type of the executor associated with this execution policy.

typedef Parameters executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef hpx::traits::executor_execution_category<executor_type>::type execution_category

The category of the execution agents created by this execution policy.

Public Functions

parallel_task_policy_shim operator()(task_policy_tag tag) const

Create a new parallel_task_policy_shim from itself

Return

The new sequenced_task_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor_>
rebind_executor<parallel_task_policy_shim, Executor_, executor_parameters_type>::type on(Executor_ &&exec) const

Create a new parallel_task_policy from the given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new parallel_task_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters_, typename ParametersType = typename executor_parameters_join<Parameters_...>::type>
rebind_executor<parallel_task_policy_shim, executor_type, ParametersType>::type with(Parameters_&&... params) const

Create a new parallel_policy_shim from the given execution parameters

Note

Requires: all parameters are executor_parameters, different parameter types can’t be duplicated

Return

The new parallel_policy_shim

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

Executor &executor()

Return the associated executor object.

Executor const &executor() const

Return the associated executor object.

Parameters &parameters()

Return the associated executor parameters object.

Parameters const &parameters() const

Return the associated executor parameters object.

struct parallel_unsequenced_policy

#include <execution_policy.hpp>

The class parallel_unsequenced_policy is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and indicate that a parallel algorithm’s execution may be vectorized.

Public Types

typedef parallel_executor executor_type

The type of the executor associated with this execution policy.

typedef execution::extract_executor_parameters<executor_type>::type executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef parallel_execution_tag execution_category

The category of the execution agents created by this execution policy.

Public Functions

parallel_unsequenced_policy operator()(task_policy_tag) const

Create a new parallel_unsequenced_policy from itself

Return

The new parallel_unsequenced_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

executor_type &executor()

Return the associated executor object.

executor_type const &executor() const

Return the associated executor object.

executor_parameters_type &parameters()

Return the associated executor parameters object.

executor_parameters_type const &parameters() const

Return the associated executor parameters object.

Private Functions

template<typename Archive>
void serialize(Archive &ar, const unsigned int version)

Private Members

executor_type exec_

executor_parameters_type params_

Friends

friend hpx::parallel::execution::parallel_unsequenced_policy::hpx::serialization::access

class partitioner

#include <partitioner.hpp>

Public Functions

partitioner(util::function_nonser<int(boost::program_options::variables_map &vm)> const &f, boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> ini_config, resource::partitioner_mode rpmode = resource::mode_default, runtime_mode mode = runtime_mode_default, )

partitioner(util::function_nonser<int(int, char **)> const &f, int argc, char **argv, resource::partitioner_mode rpmode = resource::mode_default, hpx::runtime_mode mode = hpx::runtime_mode_default, )

partitioner(util::function_nonser<int(int, char **)> const &f, int argc, char **argv, std::vector<std::string> const &cfg, resource::partitioner_mode rpmode = resource::mode_default, hpx::runtime_mode mode = hpx::runtime_mode_default, )

partitioner(int argc, char **argv, resource::partitioner_mode rpmode = resource::mode_default, runtime_mode mode = runtime_mode_default)

partitioner(int argc, char **argv, std::vector<std::string> ini_config, resource::partitioner_mode rpmode = resource::mode_default, runtime_mode mode = runtime_mode_default)

partitioner(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, resource::partitioner_mode rpmode = resource::mode_default, runtime_mode mode = runtime_mode_default)

partitioner(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> ini_config, resource::partitioner_mode rpmode = resource::mode_default, runtime_mode mode = runtime_mode_default)

partitioner(std::nullptr_t f, int argc, char **argv, resource::partitioner_mode rpmode = resource::mode_default, hpx::runtime_mode mode = hpx::runtime_mode_default)

partitioner(std::nullptr_t f, int argc, char **argv, std::vector<std::string> const &cfg, resource::partitioner_mode rpmode = resource::mode_default, hpx::runtime_mode mode = hpx::runtime_mode_default)

partitioner(std::nullptr_t f, boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> ini_config, resource::partitioner_mode rpmode = resource::mode_default, runtime_mode mode = runtime_mode_default)

void create_thread_pool(std::string const &name, scheduling_policy sched = scheduling_policy::unspecified, hpx::threads::policies::scheduler_mode = hpx::threads::policies::scheduler_mode::default_mode)

void create_thread_pool(std::string const &name, scheduler_function scheduler_creation)

void set_default_pool_name(std::string const &name)

const std::string &get_default_pool_name() const

void add_resource(hpx::resource::pu const &p, std::string const &pool_name, std::size_t num_threads = 1)

void add_resource(hpx::resource::pu const &p, std::string const &pool_name, bool exclusive, std::size_t num_threads = 1)

void add_resource(std::vector<hpx::resource::pu> const &pv, std::string const &pool_name, bool exclusive = true)

void add_resource(hpx::resource::core const &c, std::string const &pool_name, bool exclusive = true)

void add_resource(std::vector<hpx::resource::core> &cv, std::string const &pool_name, bool exclusive = true)

void add_resource(hpx::resource::numa_domain const &nd, std::string const &pool_name, bool exclusive = true)

void add_resource(std::vector<hpx::resource::numa_domain> const &ndv, std::string const &pool_name, bool exclusive = true)

std::vector<numa_domain> const &numa_domains() const

std::size_t get_number_requested_threads()

hpx::threads::topology const &get_topology() const

Private Members

detail::partitioner &partitioner_

struct persistent_auto_chunk_size

#include <persistent_auto_chunk_size.hpp>

Loop iterations are divided into pieces and then assigned to threads. The number of loop iterations combined is determined based on measurements of how long the execution of 1% of the overall number of iterations takes. This executor parameters type makes sure that as many loop iterations are combined as necessary to run for the amount of time specified.

Public Functions

persistent_auto_chunk_size()

Construct an persistent_auto_chunk_size executor parameters object

Note

Default constructed persistent_auto_chunk_size executor parameter types will use 0 microseconds as the execution time for each chunk and 80 microseconds as the minimal time for which any of the scheduled chunks should run.

persistent_auto_chunk_size(hpx::util::steady_duration const &time_cs)

Construct an persistent_auto_chunk_size executor parameters object

Parameters

• time_cs: The execution time for each chunk.

persistent_auto_chunk_size(hpx::util::steady_duration const &time_cs, hpx::util::steady_duration const &rel_time)

Construct an persistent_auto_chunk_size executor parameters object

Parameters

• rel_time: [in] The time duration to use as the minimum to decide how many loop iterations should be combined.
• time_cs: The execution time for each chunk.

class pu

#include <partitioner.hpp>

Public Functions

pu(std::size_t id = invalid_pu_id, core *core = nullptr, std::size_t thread_occupancy = 0)

std::size_t id() const

Private Functions

std::vector<pu> pus_sharing_core()

std::vector<pu> pus_sharing_numa_domain()

Private Members

std::size_t id_

core *core_

std::size_t thread_occupancy_

std::size_t thread_occupancy_count_

Private Static Attributes

const std::size_t invalid_pu_id = std::size_t(-1)

Friends

friend hpx::resource::pu::core

friend hpx::resource::pu::numa_domain

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

typedef parallel_task_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

typedef sequenced_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

template<>
typedef sequenced_task_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

template<>
typedef parallel_task_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

template<>
typedef parallel_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

typedef parallel_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

typedef sequenced_task_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

template<typename Executor_, typename Parameters_>
struct rebind

#include <execution_policy.hpp>

Rebind the type of executor used by this execution policy. The execution category of Executor shall not be weaker than that of this execution policy

Public Types

template<>
typedef sequenced_policy_shim<Executor_, Parameters_> type

The type of the rebound execution policy.

struct sequenced_execution_tag

#include <execution_fwd.hpp>

Function invocations executed by a group of sequential execution agents execute in sequential order.

struct sequenced_executor

#include <sequenced_executor.hpp>

A sequential_executor creates groups of sequential execution agents which execute in the calling thread. The sequential order is given by the lexicographical order of indices in the index space.

struct sequenced_policy

#include <execution_policy.hpp>

The class sequenced_policy is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and require that a parallel algorithm’s execution may not be parallelized.

Subclassed by hpx::parallel::execution::sequenced_policy_shim< Executor, Parameters >

Public Types

typedef sequenced_executor executor_type

The type of the executor associated with this execution policy.

typedef execution::extract_executor_parameters<executor_type>::type executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef sequenced_execution_tag execution_category

The category of the execution agents created by this execution policy.

Public Functions

sequenced_task_policy operator()(task_policy_tag) const

Create a new sequenced_task_policy.

Return

The new sequenced_task_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor>
rebind_executor<sequenced_policy, Executor, executor_parameters_type>::type on(Executor &&exec) const

Create a new sequenced_policy from the given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new sequenced_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters, typename ParametersType = typename executor_parameters_join<Parameters...>::type>
rebind_executor<sequenced_policy, executor_type, ParametersType>::type with(Parameters&&... params) const

Create a new sequenced_policy from the given execution parameters

Note

Requires: all parameters are executor_parameters, different parameter types can’t be duplicated

Return

The new sequenced_policy

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

executor_type &executor()

Return the associated executor object. Return the associated executor object.

executor_type const &executor() const

Return the associated executor object.

executor_parameters_type &parameters()

Return the associated executor parameters object.

executor_parameters_type const &parameters() const

Return the associated executor parameters object.

Private Functions

template<typename Archive>
void serialize(Archive &ar, const unsigned int version)

Private Members

executor_type exec_

executor_parameters_type params_

Friends

friend hpx::parallel::execution::sequenced_policy::hpx::serialization::access

template<typename Executor, typename Parameters>
struct sequenced_policy_shim : public hpx::parallel::execution::sequenced_policy

#include <execution_policy.hpp>

The class sequenced_policy is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and require that a parallel algorithm’s execution may not be parallelized.

Public Types

typedef Executor executor_type

The type of the executor associated with this execution policy.

typedef Parameters executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef hpx::traits::executor_execution_category<executor_type>::type execution_category

The category of the execution agents created by this execution policy.

Public Functions

sequenced_task_policy_shim<Executor, Parameters> operator()(task_policy_tag tag) const

Create a new sequenced_task_policy.

Return

The new sequenced_task_policy_shim

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor_>
rebind_executor<sequenced_policy_shim, Executor_, executor_parameters_type>::type on(Executor_ &&exec) const

Create a new sequenced_policy from the given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new sequenced_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters_, typename ParametersType = typename executor_parameters_join<Parameters_...>::type>
rebind_executor<sequenced_policy_shim, executor_type, ParametersType>::type with(Parameters_&&... params) const

Create a new sequenced_policy_shim from the given execution parameters

Note

Requires: all parameters are executor_parameters, different parameter types can’t be duplicated

Return

The new sequenced_policy_shim

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

Executor &executor()

Return the associated executor object.

Executor const &executor() const

Return the associated executor object.

Parameters &parameters()

Return the associated executor parameters object.

Parameters const &parameters() const

Return the associated executor parameters object.

struct sequenced_task_policy

#include <execution_policy.hpp>

Extension: The class sequenced_task_policy is an execution policy type used as a unique type to disambiguate parallel algorithm overloading and indicate that a parallel algorithm’s execution may not be parallelized (has to run sequentially).

The algorithm returns a future representing the result of the corresponding algorithm when invoked with the sequenced_policy.

Subclassed by hpx::parallel::execution::sequenced_task_policy_shim< Executor, Parameters >

Public Types

typedef sequenced_executor executor_type

The type of the executor associated with this execution policy.

typedef execution::extract_executor_parameters<executor_type>::type executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef sequenced_execution_tag execution_category

The category of the execution agents created by this execution policy.

Public Functions

sequenced_task_policy operator()(task_policy_tag) const

Create a new sequenced_task_policy from itself

Return

The new sequenced_task_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor>
rebind_executor<sequenced_task_policy, Executor, executor_parameters_type>::type on(Executor &&exec) const

Create a new sequenced_task_policy from the given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new sequenced_task_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters, typename ParametersType = typename executor_parameters_join<Parameters...>::type>
rebind_executor<sequenced_task_policy, executor_type, ParametersType>::type with(Parameters&&... params) const

Create a new sequenced_task_policy from the given execution parameters

Note

Requires: all parameters are executor_parameters, different parameter types can’t be duplicated

Return

The new sequenced_task_policy

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

executor_type &executor()

Return the associated executor object.

executor_type const &executor() const

Return the associated executor object.

executor_parameters_type &parameters()

Return the associated executor parameters object.

executor_parameters_type const &parameters() const

Return the associated executor parameters object.

Private Functions

template<typename Archive>
void serialize(Archive &ar, const unsigned int version)

Private Members

executor_type exec_

executor_parameters_type params_

Friends

friend hpx::parallel::execution::sequenced_task_policy::hpx::serialization::access

template<typename Executor, typename Parameters>
struct sequenced_task_policy_shim : public hpx::parallel::execution::sequenced_task_policy

#include <execution_policy.hpp>

Extension: The class sequenced_task_policy_shim is an execution policy type used as a unique type to disambiguate parallel algorithm overloading based on combining a underlying sequenced_task_policy and an executor and indicate that a parallel algorithm’s execution may not be parallelized (has to run sequentially).

The algorithm returns a future representing the result of the corresponding algorithm when invoked with the sequenced_policy.

Public Types

typedef Executor executor_type

The type of the executor associated with this execution policy.

typedef Parameters executor_parameters_type

The type of the associated executor parameters object which is associated with this execution policy

typedef hpx::traits::executor_execution_category<executor_type>::type execution_category

The category of the execution agents created by this execution policy.

Public Functions

sequenced_task_policy_shim const &operator()(task_policy_tag tag) const

Create a new sequenced_task_policy from itself

Return

The new sequenced_task_policy

Parameters

• tag: [in] Specify that the corresponding asynchronous execution policy should be used

template<typename Executor_>
rebind_executor<sequenced_task_policy_shim, Executor_, executor_parameters_type>::type on(Executor_ &&exec) const

Create a new sequenced_task_policy from the given executor

Note

Requires: is_executor<Executor>::value is true

Return

The new sequenced_task_policy

Template Parameters

• Executor: The type of the executor to associate with this execution policy.

Parameters

• exec: [in] The executor to use for the execution of the parallel algorithm the returned execution policy is used with.

template<typename ...Parameters_, typename ParametersType = typename executor_parameters_join<Parameters_...>::type>
rebind_executor<sequenced_task_policy_shim, executor_type, ParametersType>::type with(Parameters_&&... params) const

Create a new sequenced_task_policy_shim from the given execution parameters

Note

Requires: all parameters are executor_parameters, different parameter types can’t be duplicated

Return

The new sequenced_task_policy_shim

Template Parameters

• Parameters: The type of the executor parameters to associate with this execution policy.

Parameters

• params: [in] The executor parameters to use for the execution of the parallel algorithm the returned execution policy is used with.

Executor &executor()

Return the associated executor object.

Executor const &executor() const

Return the associated executor object.

Parameters &parameters()

Return the associated executor parameters object.

Parameters const &parameters() const

Return the associated executor parameters object.

struct static_chunk_size

#include <static_chunk_size.hpp>

Loop iterations are divided into pieces of size chunk_size and then assigned to threads. If chunk_size is not specified, the iterations are evenly (if possible) divided contiguously among the threads.

Note

This executor parameters type is equivalent to OpenMP’s STATIC scheduling directive.

Public Functions

static_chunk_size()

Construct a static_chunk_size executor parameters object

Note

By default the number of loop iterations is determined from the number of available cores and the overall number of loop iterations to schedule.

static_chunk_size(std::size_t chunk_size)

Construct a static_chunk_size executor parameters object

Parameters

• chunk_size: [in] The optional chunk size to use as the number of loop iterations to run on a single thread.

template<typename ExPolicy = parallel::execution::parallel_policy>
class task_block

#include <task_block.hpp>

The class task_block defines an interface for forking and joining parallel tasks. The define_task_block and define_task_block_restore_thread function templates create an object of type task_block and pass a reference to that object to a user-provided callable object.

An object of class task_block cannot be constructed, destroyed, copied, or moved except by the implementation of the task region library. Taking the address of a task_block object via operator& or addressof is ill formed. The result of obtaining its address by any other means is unspecified.

A task_block

is active if it was created by the nearest enclosing task block, where “task block” refers to an invocation of define_task_block or define_task_block_restore_thread and “nearest

enclosing” means the most recent invocation that has not yet completed. Code designated for execution in another thread by means other than the facilities in this section (e.g., using thread or async) are not enclosed in the task region and a

task_block passed to (or captured by) such code is not active within that code. Performing any operation on a task_block that is not active results in undefined behavior.

The task_block that is active before a specific call to the run member function is not active within the asynchronous function that invoked run. (The invoked function should not, therefore, capture the task_block from the surrounding block.)

Example:
    define_task_block([&](auto& tr) {
        tr.run([&] {
            tr.run([] { f(); });                // Error: tr is not active
            define_task_block([&](auto& tr) {   // Nested task block
                tr.run(f);                      // OK: inner tr is active
                /// ...
            });
        });
        /// ...
    });

Template Parameters

• ExPolicy: The execution policy an instance of a task_block was created with. This defaults to parallel_policy.

Public Types

typedef ExPolicy execution_policy

Refers to the type of the execution policy used to create the task_block.

Public Functions

execution_policy const &get_execution_policy() const

Return the execution policy instance used to create this task_block

template<typename F, typename ...Ts>
void run(F &&f, Ts&&... ts)

Causes the expression f() to be invoked asynchronously. The invocation of f is permitted to run on an unspecified thread in an unordered fashion relative to the sequence of operations following the call to run(f) (the continuation), or indeterminately sequenced within the same thread as the continuation.

The call to run synchronizes with the invocation of f. The completion of f() synchronizes with the next invocation of wait on the same task_block or completion of the nearest enclosing task block (i.e., the define_task_block or define_task_block_restore_thread that created this task block).

Requires: F shall be MoveConstructible. The expression, (void)f(), shall be well-formed.

Precondition: this shall be the active task_block.

Postconditions: A call to run may return on a different thread than that on which it was called.

Note

The call to run is sequenced before the continuation as if run returns on the same thread. The invocation of the user-supplied callable object f may be immediate or may be delayed until compute resources are available. run might or might not return before invocation of f completes.

Exceptions

• This: function may throw task_canceled_exception, as described in Exception Handling.

template<typename Executor, typename F, typename ...Ts>
void run(Executor &exec, F &&f, Ts&&... ts)

Causes the expression f() to be invoked asynchronously using the given executor. The invocation of f is permitted to run on an unspecified thread associated with the given executor and in an unordered fashion relative to the sequence of operations following the call to run(exec, f) (the continuation), or indeterminately sequenced within the same thread as the continuation.

The call to run synchronizes with the invocation of f. The completion of f() synchronizes with the next invocation of wait on the same task_block or completion of the nearest enclosing task block (i.e., the define_task_block or define_task_block_restore_thread that created this task block).

Requires: Executor shall be a type modeling the Executor concept. F shall be MoveConstructible. The expression, (void)f(), shall be well-formed.

Precondition: this shall be the active task_block.

Postconditions: A call to run may return on a different thread than that on which it was called.

Note

The call to run is sequenced before the continuation as if run returns on the same thread. The invocation of the user-supplied callable object f may be immediate or may be delayed until compute resources are available. run might or might not return before invocation of f completes.

Exceptions

• This: function may throw task_canceled_exception, as described in Exception Handling.

void wait()

Blocks until the tasks spawned using this task_block have finished.

Precondition: this shall be the active task_block.

Postcondition: All tasks spawned by the nearest enclosing task region have finished. A call to wait may return on a different thread than that on which it was called.

Example:
    define_task_block([&](auto& tr) {
        tr.run([&]{ process(a, w, x); }); // Process a[w] through a[x]
        if (y < x) tr.wait();   // Wait if overlap between [w, x) and [y, z)
        process(a, y, z);       // Process a[y] through a[z]
    });

Note

The call to wait is sequenced before the continuation as if wait returns on the same thread.

Exceptions

• This: function may throw task_canceled_exception, as described in Exception Handling.

ExPolicy &policy()

Returns a reference to the execution policy used to construct this object.

Precondition: this shall be the active task_block.

ExPolicy const &policy() const

Returns a reference to the execution policy used to construct this object.

Precondition: this shall be the active task_block.

Private Members

mutex_type mtx_

std::vector<hpx::future<void>> tasks_

parallel::exception_list errors_

threads::thread_id_type id_

ExPolicy policy_

class task_canceled_exception : public hpx::exception

#include <task_block.hpp>

The class task_canceled_exception defines the type of objects thrown by task_block::run or task_block::wait if they detect that an exception is pending within the current parallel region.

Public Functions

task_canceled_exception()

struct thread_interrupted : public exception

#include <exception.hpp>

A hpx::thread_interrupted is the exception type used by HPX to interrupt a running HPX thread.

The hpx::thread_interrupted type is the exception type used by HPX to interrupt a running thread.

A running thread can be interrupted by invoking the interrupt() member function of the corresponding hpx::thread object. When the interrupted thread next executes one of the specified interruption points (or if it is currently blocked whilst executing one) with interruption enabled, then a hpx::thread_interrupted exception will be thrown in the interrupted thread. If not caught, this will cause the execution of the interrupted thread to terminate. As with any other exception, the stack will be unwound, and destructors for objects of automatic storage duration will be executed.

If a thread wishes to avoid being interrupted, it can create an instance of hpx::this_thread::disable_interruption. Objects of this class disable interruption for the thread that created them on construction, and restore the interruption state to whatever it was before on destruction.

void f()
{
    // interruption enabled here
    {
        hpx::this_thread::disable_interruption di;
        // interruption disabled
        {
            hpx::this_thread::disable_interruption di2;
            // interruption still disabled
        } // di2 destroyed, interruption state restored
        // interruption still disabled
    } // di destroyed, interruption state restored
    // interruption now enabled
}

The effects of an instance of hpx::this_thread::disable_interruption can be temporarily reversed by constructing an instance of hpx::this_thread::restore_interruption, passing in the hpx::this_thread::disable_interruption object in question. This will restore the interruption state to what it was when the hpx::this_thread::disable_interruption object was constructed, and then disable interruption again when the hpx::this_thread::restore_interruption object is destroyed.

void g()
{
    // interruption enabled here
    {
        hpx::this_thread::disable_interruption di;
        // interruption disabled
        {
            hpx::this_thread::restore_interruption ri(di);
            // interruption now enabled
        } // ri destroyed, interruption disable again
    } // di destroyed, interruption state restored
    // interruption now enabled
}

At any point, the interruption state for the current thread can be queried by calling hpx::this_thread::interruption_enabled().

class thread_pool_base : public manage_executor

#include <thread_pool_base.hpp>

The base class used to manage a pool of OS threads.

Public Functions

virtual hpx::future<void> resume() = 0

Resumes the thread pool. When the all OS threads on the thread pool have been resumed the returned future will be ready.

Note

Can only be called from an HPX thread. Use resume_cb or resume_direct to suspend the pool from outside HPX.

Return

A future<void> which is ready when the thread pool has been resumed.

Exceptions

• hpx::exception: if called from outside the HPX runtime.

virtual void resume_cb(std::function<void(void)> callback, error_code &ec = throws, ) = 0

Resumes the thread pool. Takes a callback as a parameter which will be called when all OS threads on the thread pool have been resumed.

Parameters

• callback: [in] called when the thread pool has been resumed.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

virtual void resume_direct(error_code &ec = throws) = 0

Resumes the thread pool. Blocks until all OS threads on the thread pool have been resumed.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

virtual hpx::future<void> suspend() = 0

Suspends the thread pool. When the all OS threads on the thread pool have been suspended the returned future will be ready.

Note

Can only be called from an HPX thread. Use suspend_cb or suspend_direct to suspend the pool from outside HPX. A thread pool cannot be suspended from an HPX thread running on the pool itself.

Return

A future<void> which is ready when the thread pool has been suspended.

Exceptions

• hpx::exception: if called from outside the HPX runtime.

virtual void suspend_cb(std::function<void(void)> callback, error_code &ec = throws, ) = 0

Suspends the thread pool. Takes a callback as a parameter which will be called when all OS threads on the thread pool have been suspended.

Note

A thread pool cannot be suspended from an HPX thread running on the pool itself.

Parameters

• callback: [in] called when the thread pool has been suspended.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

Exceptions

• hpx::exception: if called from an HPX thread which is running on the pool itself.

virtual void suspend_direct(error_code &ec = throws) = 0

Suspends the thread pool. Blocks until all OS threads on the thread pool have been suspended.

Note

A thread pool cannot be suspended from an HPX thread running on the pool itself.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

Exceptions

• hpx::exception: if called from an HPX thread which is running on the pool itself.

virtual hpx::future<void> suspend_processing_unit(std::size_t virt_core) = 0

Suspends the given processing unit. When the processing unit has been suspended the returned future will be ready.

Note

Can only be called from an HPX thread. Use suspend_processing_unit_cb or to suspend the processing unit from outside HPX. Requires that the pool has threads::policies::enable_elasticity set.

Return

A future<void> which is ready when the given processing unit has been suspended.

Parameters

• virt_core: [in] The processing unit on the the pool to be suspended. The processing units are indexed starting from 0.

Exceptions

• hpx::exception: if called from outside the HPX runtime.

virtual void suspend_processing_unit_cb(std::function<void(void)> callback, std::size_t virt_core, error_code &ec = throws, ) = 0

Suspends the given processing unit. Takes a callback as a parameter which will be called when the processing unit has been suspended.

Note

Requires that the pool has threads::policies::enable_elasticity set.

Parameters

• callback: [in] Callback which is called when the processing unit has been suspended.
• virt_core: [in] The processing unit to suspend.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

virtual hpx::future<void> resume_processing_unit(std::size_t virt_core) = 0

Resumes the given processing unit. When the processing unit has been resumed the returned future will be ready.

Note

Can only be called from an HPX thread. Use resume_processing_unit_cb or to resume the processing unit from outside HPX. Requires that the pool has threads::policies::enable_elasticity set.

Return

A future<void> which is ready when the given processing unit has been resumed.

Parameters

• virt_core: [in] The processing unit on the the pool to be resumed. The processing units are indexed starting from 0.

virtual void resume_processing_unit_cb(std::function<void(void)> callback, std::size_t virt_core, error_code &ec = throws, ) = 0

Resumes the given processing unit. Takes a callback as a parameter which will be called when the processing unit has been resumed.

Note

Requires that the pool has threads::policies::enable_elasticity set.

Parameters

• callback: [in] Callback which is called when the processing unit has been suspended.
• virt_core: [in] The processing unit to resume.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

struct thread_schedule_hint

#include <thread_enums.hpp>

Public Functions

thread_schedule_hint()

thread_schedule_hint(std::int16_t thread_hint)

thread_schedule_hint(thread_schedule_hint_mode mode, std::int16_t hint)

Public Members

thread_schedule_hint_mode mode

std::int16_t hint

struct unsequenced_execution_tag

#include <execution_fwd.hpp>

Function invocations executed by a group of vector execution agents are permitted to execute in unordered fashion when executed in different threads, and un-sequenced with respect to one another when executed in the same thread.

Note

unsequenced_execution_tag is weaker than parallel_execution_tag.

struct unwrap

#include <unwrap.hpp>

A helper function object for functionally invoking hpx::util::unwrap. For more information please refer to its documentation.

struct unwrap_all

#include <unwrap.hpp>

A helper function object for functionally invoking hpx::util::unwrap_all. For more information please refer to its documentation.

template<std::size_t Depth>
struct unwrap_n

#include <unwrap.hpp>

A helper function object for functionally invoking hpx::util::unwrap_n. For more information please refer to its documentation.

template<typename Sequence>
struct when_any_result

#include <when_any.hpp>

Result type for when_any, contains a sequence of futures and an index pointing to a ready future.

Public Members

std::size_t index

The index of a future which has become ready.

Sequence futures

The sequence of futures as passed to hpx::when_any.

template<typename Sequence>
struct when_some_result

#include <when_some.hpp>

Result type for when_some, contains a sequence of futures and indices pointing to ready futures.

Public Members

std::vector<std::size_t> indices

List of indices of futures which became ready.

Sequence futures

The sequence of futures as passed to hpx::when_some.

namespace applier

The namespace applier contains all definitions needed for the class hpx::applier::applier and its related functionality. This namespace is part of the HPX core module.

namespace hpx

Unnamed Group

error_code make_error_code(error e, throwmode mode = plain)

Returns a new error_code constructed from the given parameters.

error_code make_error_code(error e, char const *func, char const *file, long line, throwmode mode = plain)

error_code make_error_code(error e, char const *msg, throwmode mode = plain)

Returns error_code(e, msg, mode).

error_code make_error_code(error e, char const *msg, char const *func, char const *file, long line, throwmode mode = plain)

error_code make_error_code(error e, std::string const &msg, throwmode mode = plain)

Returns error_code(e, msg, mode).

error_code make_error_code(error e, std::string const &msg, char const *func, char const *file, long line, throwmode mode = plain)

error_code make_error_code(std::exception_ptr const &e)

Typedefs

typedef util::function_nonser<void(boost::system::error_code const&, parcelset::parcel const&)> parcel_write_handler_type

The type of a function which can be registered as a parcel write handler using the function hpx::set_parcel_write_handler.

Note

A parcel write handler is a function which is called by the parcel layer whenever a parcel has been sent by the underlying networking library and if no explicit parcel handler function was specified for the parcel.

typedef util::unique_function_nonser<void()> shutdown_function_type

The type of a function which is registered to be executed as a shutdown or pre-shutdown function.

typedef util::unique_function_nonser<void()> startup_function_type

The type of a function which is registered to be executed as a startup or pre-startup function.

Enums

enum error

Possible error conditions.

This enumeration lists all possible error conditions which can be reported from any of the API functions.

Values:

success = 0

The operation was successful.

no_success = 1

The operation did failed, but not in an unexpected manner.

not_implemented = 2

The operation is not implemented.

out_of_memory = 3

The operation caused an out of memory condition.

bad_action_code = 4

bad_component_type = 5

The specified component type is not known or otherwise invalid.

network_error = 6

A generic network error occurred.

version_too_new = 7

The version of the network representation for this object is too new.

version_too_old = 8

The version of the network representation for this object is too old.

version_unknown = 9

The version of the network representation for this object is unknown.

unknown_component_address = 10

duplicate_component_address = 11

The given global id has already been registered.

invalid_status = 12

The operation was executed in an invalid status.

bad_parameter = 13

One of the supplied parameters is invalid.

internal_server_error = 14

service_unavailable = 15

bad_request = 16

repeated_request = 17

lock_error = 18

duplicate_console = 19

There is more than one console locality.

no_registered_console = 20

There is no registered console locality available.

startup_timed_out = 21

uninitialized_value = 22

bad_response_type = 23

deadlock = 24

assertion_failure = 25

null_thread_id = 26

Attempt to invoke a API function from a non-HPX thread.

invalid_data = 27

yield_aborted = 28

The yield operation was aborted.

dynamic_link_failure = 29

commandline_option_error = 30

One of the options given on the command line is erroneous.

serialization_error = 31

There was an error during serialization of this object.

unhandled_exception = 32

An unhandled exception has been caught.

kernel_error = 33

The OS kernel reported an error.

broken_task = 34

The task associated with this future object is not available anymore.

task_moved = 35

The task associated with this future object has been moved.

task_already_started = 36

The task associated with this future object has already been started.

future_already_retrieved = 37

The future object has already been retrieved.

promise_already_satisfied = 38

The value for this future object has already been set.

future_does_not_support_cancellation = 39

The future object does not support cancellation.

future_can_not_be_cancelled = 40

The future can’t be canceled at this time.

no_state = 41

The future object has no valid shared state.

broken_promise = 42

The promise has been deleted.

thread_resource_error = 43

future_cancelled = 44

thread_cancelled = 45

thread_not_interruptable = 46

duplicate_component_id = 47

The component type has already been registered.

unknown_error = 48

An unknown error occurred.

bad_plugin_type = 49

The specified plugin type is not known or otherwise invalid.

filesystem_error = 50

The specified file does not exist or other filesystem related error.

bad_function_call = 51

equivalent of std::bad_function_call

task_canceled_exception = 52

parallel::v2::task_canceled_exception

task_block_not_active = 53

task_region is not active

out_of_range = 54

Equivalent to std::out_of_range.

length_error = 55

Equivalent to std::length_error.

migration_needs_retry = 56

migration failed because of global race, retry

enum throwmode

Encode error category for new error_code.

Values:

plain = 0

rethrow = 1

enum runtime_mode

A HPX runtime can be executed in two different modes: console mode and worker mode.

Values:

runtime_mode_invalid = -1

runtime_mode_console = 0

The runtime is the console locality.

runtime_mode_worker = 1

The runtime is a worker locality.

runtime_mode_connect = 2

The runtime is a worker locality connecting late

runtime_mode_default = 3

The runtime mode will be determined based on the command line arguments

runtime_mode_last

Functions

int init(util::function_nonser<int(boost::program_options::variables_map &vm)> const &f, boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> const &cfg, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is the main entry point for any HPX application. This function (or one of its overloads below) should be called from the users main() function. It will set up the HPX runtime environment and schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied f.

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(int (*f)(boost::program_options::variables_map &vm), boost::program_options::options_description const &desc_cmdline, int argc, char **argv, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is the main entry point for any HPX application. This function (or one of its overloads below) should be called from the users main() function. It will set up the HPX runtime environment and schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied f.

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> const &cfg, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(int argc, char **argv, std::vector<std::string> const &cfg, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, hpx::runtime_mode mode)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

If the parameter mode is runtime_mode_default, the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> const &cfg, hpx::runtime_mode mode)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

If the parameter mode is runtime_mode_default, the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(std::string const &app_name, int argc = 0, char **argv = nullptr, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• app_name: [in] The name of the application.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(int argc = 0, char **argv = nullptr, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. If not command line arguments are passed, console mode is assumed.

Note

If no command line arguments are passed the HPX runtime system will not support any of the default command line options as described in the section ‘HPX Command Line Options’.

Parameters

• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(std::vector<std::string> const &cfg, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings).

Return

The function returns the value, which has been returned from hpx_main (or 0 when executed in worker mode).

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. If not command line arguments are passed, console mode is assumed.

Note

If no command line arguments are passed the HPX runtime system will not support any of the default command line options as described in the section ‘HPX Command Line Options’.

Parameters

• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(int (*f)(boost::program_options::variables_map &vm), std::string const &app_name, int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied function f.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• app_name: [in] The name of the application.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(int (*f)(boost::program_options::variables_map &vm), int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied function f.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(util::function_nonser<int(int, char **)> const &f, std::string const &app_name, int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied function f.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• app_name: [in] The name of the application.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(util::function_nonser<int(int, char **)> const &f, int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied function f.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int init(util::function_nonser<int(int, char **)> const &f, int argc, char **argv, std::vector<std::string> const &cfg, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main entry point for launching the HPX runtime system.

This is a simplified main entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will schedule the function given by f as a HPX thread. This overload will not call hpx_main.

Return

The function returns the value, which has been returned from the user supplied function f.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(util::function_nonser<int(boost::program_options::variables_map &vm)> const &f, boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> const &cfg, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is the main, non-blocking entry point for any HPX application. This function (or one of its overloads below) should be called from the users main() function. It will set up the HPX runtime environment and schedule the function given by f as a HPX thread. It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(int (*f)(boost::program_options::variables_map &vm), boost::program_options::options_description const &desc_cmdline, int argc, char **argv, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is the main, non-blocking entry point for any HPX application. This function (or one of its overloads below) should be called from the users main() function. It will set up the HPX runtime environment and schedule the function given by f as a HPX thread. It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> const &cfg, startup_function_type startup = startup_function_type(), shutdown_function_type shutdown = shutdown_function_type(), hpx::runtime_mode mode = hpx::runtime_mode_default)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is not given (defaulted), the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• startup: [in] A function to be executed inside a HPX thread before f is called. If this parameter is not given no function will be executed.
• shutdown: [in] A function to be executed inside an HPX thread while hpx::finalize is executed. If this parameter is not given no function will be executed.
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(int argc, char **argv, std::vector<std::string> const &cfg, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is runtime_mode_default, the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, hpx::runtime_mode mode)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is runtime_mode_default, the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(boost::program_options::options_description const &desc_cmdline, int argc, char **argv, std::vector<std::string> const &cfg, hpx::runtime_mode mode)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

In console mode it will execute the user supplied function hpx_main, in worker mode it will execute an empty hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

If the parameter mode is runtime_mode_default, the created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. Otherwise it will be executed as specified by the parametermode.

Parameters

• desc_cmdline: [in] This parameter may hold the description of additional command line arguments understood by the application. These options will be prepended to the default command line options understood by hpx::init (see description below).
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(std::string const &app_name, int argc = 0, char **argv = nullptr, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• app_name: [in] The name of the application.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(int argc = 0, char **argv = nullptr, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. If not command line arguments are passed, console mode is assumed.

Note

If no command line arguments are passed the HPX runtime system will not support any of the default command line options as described in the section ‘HPX Command Line Options’.

Parameters

• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(std::vector<std::string> const &cfg, hpx::runtime_mode mode = hpx::runtime_mode_default)

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv. If not command line arguments are passed, console mode is assumed.

Note

If no command line arguments are passed the HPX runtime system will not support any of the default command line options as described in the section ‘HPX Command Line Options’.

Parameters

• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(int (*f)(boost::program_options::variables_map &vm), std::string const &app_name, int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will schedule the function given by f as a HPX thread. It will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application.
• app_name: [in] The name of the application.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(util::function_nonser<int(int, char **)> const &f, std::string const &app_name, int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will schedule the function given by f as a HPX thread. It will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• app_name: [in] The name of the application.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(int (*f)(boost::program_options::variables_map &vm), int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will schedule the function given by f as a HPX thread. It will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(util::function_nonser<int(int, char **)> const &f, int argc, char **argv, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will schedule the function given by f as a HPX thread. It will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

bool start(util::function_nonser<int(int, char **)> const &f, int argc, char **argv, std::vector<std::string> const &cfg, hpx::runtime_mode mode = hpx::runtime_mode_default, )

Main non-blocking entry point for launching the HPX runtime system.

This is a simplified main, non-blocking entry point, which can be used to set up the runtime for an HPX application (the runtime system will be set up in console mode or worker mode depending on the command line settings). It will return immediately after that. Use hpx::wait and hpx::stop to synchronize with the runtime system’s execution. This overload will schedule the function given by f as a HPX thread. It will not call hpx_main.

Return

The function returns true if command line processing succeeded and the runtime system was started successfully. It will return false otherwise.

Note

The created runtime system instance will be executed in console or worker mode depending on the command line arguments passed in argc/argv.

Parameters

• f: [in] The function to be scheduled as an HPX thread. Usually this function represents the main entry point of any HPX application. If f is nullptr the HPX runtime environment will be started without invoking f.
• argc: [in] The number of command line arguments passed in argv. This is usually the unchanged value as passed by the operating system (to main()).
• argv: [in] The command line arguments for this application, usually that is the value as passed by the operating system (to main()).
• cfg: A list of configuration settings which will be added to the system configuration before the runtime instance is run. Each of the entries in this list must have the format of a fully defined key/value pair from an ini-file (for instance ‘hpx.component.enabled=1’)
• mode: [in] The mode the created runtime environment should be initialized in. There has to be exactly one locality in each HPX application which is executed in console mode (hpx::runtime_mode_console), all other localities have to be run in worker mode (hpx::runtime_mode_worker). Normally this is set up automatically, but sometimes it is necessary to explicitly specify the mode.

int finalize(double shutdown_timeout, double localwait = -1.0, error_code &ec = throws)

Main function to gracefully terminate the HPX runtime system.

The function hpx::finalize is the main way to (gracefully) exit any HPX application. It should be called from one locality only (usually the console) and it will notify all connected localities to finish execution. Only after all other localities have exited this function will return, allowing to exit the console locality as well.

During the execution of this function the runtime system will invoke all registered shutdown functions (see hpx::init) on all localities.

The default value (

-1.0) will try to find a globally set timeout value (can be set as the configuration parameter hpx.shutdown_timeout), and if that is not set or -1.0 as well, it will disable any timeout, each connected locality will wait for all existing HPX-threads to terminate.

Parameters

• shutdown_timeout: This parameter allows to specify a timeout (in microseconds), specifying how long any of the connected localities should wait for pending tasks to be executed. After this timeout, all suspended HPX-threads will be aborted. Note, that this function will not abort any running HPX-threads. In any case the shutdown will not proceed as long as there is at least one pending/running HPX-thread.

The default value (

-1.0) will try to find a globally set wait time value (can be set as the configuration parameter “hpx.finalize_wait_time”), and if this is not set or -1.0 as well, it will disable any addition local wait time before proceeding.

Parameters

• localwait: This parameter allows to specify a local wait time (in microseconds) before the connected localities will be notified and the overall shutdown process starts.

This function will block and wait for all connected localities to exit before returning to the caller. It should be the last HPX-function called by any application.

Return

This function will always return zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

Using this function is an alternative to hpx::disconnect, these functions do not need to be called both.

int finalize(error_code &ec = throws)

Main function to gracefully terminate the HPX runtime system.

The function hpx::finalize is the main way to (gracefully) exit any HPX application. It should be called from one locality only (usually the console) and it will notify all connected localities to finish execution. Only after all other localities have exited this function will return, allowing to exit the console locality as well.

During the execution of this function the runtime system will invoke all registered shutdown functions (see hpx::init) on all localities.

This function will block and wait for all connected localities to exit before returning to the caller. It should be the last HPX-function called by any application.

Return

This function will always return zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

Using this function is an alternative to hpx::disconnect, these functions do not need to be called both.

HPX_NORETURN void hpx::terminate()

Terminate any application non-gracefully.

The function hpx::terminate is the non-graceful way to exit any application immediately. It can be called from any locality and will terminate all localities currently used by the application.

Note

This function will cause HPX to call std::terminate() on all localities associated with this application. If the function is called not from an HPX thread it will fail and return an error using the argument ec.

int disconnect(double shutdown_timeout, double localwait = -1.0, error_code &ec = throws)

Disconnect this locality from the application.

The function hpx::disconnect can be used to disconnect a locality from a running HPX application.

During the execution of this function the runtime system will invoke all registered shutdown functions (see hpx::init) on this locality.

The default value (

-1.0) will try to find a globally set timeout value (can be set as the configuration parameter “hpx.shutdown_timeout”), and if that is not set or -1.0 as well, it will disable any timeout, each connected locality will wait for all existing HPX-threads to terminate.

Parameters

• shutdown_timeout: This parameter allows to specify a timeout (in microseconds), specifying how long this locality should wait for pending tasks to be executed. After this timeout, all suspended HPX-threads will be aborted. Note, that this function will not abort any running HPX-threads. In any case the shutdown will not proceed as long as there is at least one pending/running HPX-thread.

The default value (

-1.0) will try to find a globally set wait time value (can be set as the configuration parameter hpx.finalize_wait_time), and if this is not set or -1.0 as well, it will disable any addition local wait time before proceeding.

Parameters

• localwait: This parameter allows to specify a local wait time (in microseconds) before the connected localities will be notified and the overall shutdown process starts.

This function will block and wait for this locality to finish executing before returning to the caller. It should be the last HPX-function called by any locality being disconnected.

Return

This function will always return zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

int disconnect(error_code &ec = throws)

Disconnect this locality from the application.

The function hpx::disconnect can be used to disconnect a locality from a running HPX application.

During the execution of this function the runtime system will invoke all registered shutdown functions (see hpx::init) on this locality.

This function will block and wait for this locality to finish executing before returning to the caller. It should be the last HPX-function called by any locality being disconnected.

Return

This function will always return zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

int stop(error_code &ec = throws)

Stop the runtime system.

This function will block and wait for this locality to finish executing before returning to the caller. It should be the last HPX-function called on every locality. This function should be used only if the runtime system was started using

hpx::start.

Return

The function returns the value, which has been returned from the user supplied main HPX function (usually hpx_main).

int suspend(error_code &ec = throws)

Suspend the runtime system.

The function hpx::suspend is used to suspend the HPX runtime system. It can only be used when running HPX on a single locality. It will block waiting for all thread pools to be empty. This function only be called when the runtime is running, or already suspended in which case this function will do nothing.

Return

This function will always return zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

int resume(error_code &ec = throws)

Resume the HPX runtime system.

The function hpx::resume is used to resume the HPX runtime system. It can only be used when running HPX on a single locality. It will block waiting for all thread pools to be resumed. This function only be called when the runtime suspended, or already running in which case this function will do nothing.

Return

This function will always return zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

boost::system::error_category const &get_hpx_category()

Returns generic HPX error category used for new errors.

boost::system::error_category const &get_hpx_rethrow_category()

Returns generic HPX error category used for errors re-thrown after the exception has been de-serialized.

error_code make_success_code(throwmode mode = plain)

Returns error_code(hpx::success, “success”, mode).

std::string diagnostic_information(exception_info const &xi)

Extract the diagnostic information embedded in the given exception and return a string holding a formatted message.

The function hpx::diagnostic_information can be used to extract all diagnostic information stored in the given exception instance as a formatted string. This simplifies debug output as it composes the diagnostics into one, easy to use function call. This includes the name of the source file and line number, the sequence number of the OS-thread and the HPX-thread id, the locality id and the stack backtrace of the point where the original exception was thrown.

Return

The formatted string holding all of the available diagnostic information stored in the given exception instance.

See

hpx::get_error_locality_id(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for all diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if any of the required allocation operations fail)

std::string get_error_what(exception_info const &xi)

Return the error message of the thrown exception.

The function hpx::get_error_what can be used to extract the diagnostic information element representing the error message as stored in the given exception instance.

Return

The error message stored in the exception If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error() hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::uint32_t get_error_locality_id(hpx::exception_info const &xi)

Return the locality id where the exception was thrown.

The function hpx::get_error_locality_id can be used to extract the diagnostic information element representing the locality id as stored in the given exception instance.

Return

The locality id of the locality where the exception was thrown. If the exception instance does not hold this information, the function will return hpx::naming::invalid_locality_id.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• nothing:

error get_error(hpx::exception const &e)

Return the locality id where the exception was thrown.

The function hpx::get_error can be used to extract the diagnostic information element representing the error value code as stored in the given exception instance.

Return

The error value code of the locality where the exception was thrown. If the exception instance does not hold this information, the function will return hpx::naming::invalid_locality_id.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• e: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception, hpx::error_code, or std::exception_ptr.

Exceptions

• nothing:

error get_error(hpx::error_code const &e)

Return the locality id where the exception was thrown.

The function hpx::get_error can be used to extract the diagnostic information element representing the error value code as stored in the given exception instance.

Return

The error value code of the locality where the exception was thrown. If the exception instance does not hold this information, the function will return hpx::naming::invalid_locality_id.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• e: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception, hpx::error_code, or std::exception_ptr.

Exceptions

• nothing:

std::string get_error_host_name(hpx::exception_info const &xi)

Return the hostname of the locality where the exception was thrown.

The function hpx::get_error_host_name can be used to extract the diagnostic information element representing the host name as stored in the given exception instance.

Return

The hostname of the locality where the exception was thrown. If the exception instance does not hold this information, the function will return and empty string.

See

hpx::diagnostic_information() hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error() hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::int64_t get_error_process_id(hpx::exception_info const &xi)

Return the (operating system) process id of the locality where the exception was thrown.

The function hpx::get_error_process_id can be used to extract the diagnostic information element representing the process id as stored in the given exception instance.

Return

The process id of the OS-process which threw the exception If the exception instance does not hold this information, the function will return 0.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• nothing:

std::string get_error_env(hpx::exception_info const &xi)

Return the environment of the OS-process at the point the exception was thrown.

The function hpx::get_error_env can be used to extract the diagnostic information element representing the environment of the OS-process collected at the point the exception was thrown.

Return

The environment from the point the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::string get_error_function_name(hpx::exception_info const &xi)

Return the function name from which the exception was thrown.

The function hpx::get_error_function_name can be used to extract the diagnostic information element representing the name of the function as stored in the given exception instance.

Return

The name of the function from which the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id() hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::string get_error_backtrace(hpx::exception_info const &xi)

Return the stack backtrace from the point the exception was thrown.

The function hpx::get_error_backtrace can be used to extract the diagnostic information element representing the stack backtrace collected at the point the exception was thrown.

Return

The stack back trace from the point the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::string get_error_file_name(hpx::exception_info const &xi)

Return the (source code) file name of the function from which the exception was thrown.

The function hpx::get_error_file_name can be used to extract the diagnostic information element representing the name of the source file as stored in the given exception instance.

Return

The name of the source file of the function from which the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

long get_error_line_number(hpx::exception_info const &xi)

Return the line number in the (source code) file of the function from which the exception was thrown.

The function hpx::get_error_line_number can be used to extract the diagnostic information element representing the line number as stored in the given exception instance.

Return

The line number of the place where the exception was thrown. If the exception instance does not hold this information, the function will return -1.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name() hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• nothing:

std::size_t get_error_os_thread(hpx::exception_info const &xi)

Return the sequence number of the OS-thread used to execute HPX-threads from which the exception was thrown.

The function hpx::get_error_os_thread can be used to extract the diagnostic information element representing the sequence number of the OS-thread as stored in the given exception instance.

Return

The sequence number of the OS-thread used to execute the HPX-thread from which the exception was thrown. If the exception instance does not hold this information, the function will return std::size(-1).

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_thread_id(), hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• nothing:

std::size_t get_error_thread_id(hpx::exception_info const &xi)

Return the unique thread id of the HPX-thread from which the exception was thrown.

The function hpx::get_error_thread_id can be used to extract the diagnostic information element representing the HPX-thread id as stored in the given exception instance.

Return

The unique thread id of the HPX-thread from which the exception was thrown. If the exception instance does not hold this information, the function will return std::size_t(0).

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread() hpx::get_error_thread_description(), hpx::get_error(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error_what(), hpx::get_error_config(), hpx::get_error_state()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• nothing:

std::string get_error_thread_description(hpx::exception_info const &xi)

Return any additionally available thread description of the HPX-thread from which the exception was thrown.

The function hpx::get_error_thread_description can be used to extract the diagnostic information element representing the additional thread description as stored in the given exception instance.

Return

Any additionally available thread description of the HPX-thread from which the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error(), hpx::get_error_state(), hpx::get_error_what(), hpx::get_error_config()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::string get_error_config(hpx::exception_info const &xi)

Return the HPX configuration information point from which the exception was thrown.

The function hpx::get_error_config can be used to extract the HPX configuration information element representing the full HPX configuration information as stored in the given exception instance.

Return

Any additionally available HPX configuration information the point from which the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error(), hpx::get_error_state() hpx::get_error_what(), hpx::get_error_thread_description()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

std::string get_error_state(hpx::exception_info const &xi)

Return the HPX runtime state information at which the exception was thrown.

The function hpx::get_error_state can be used to extract the HPX runtime state information element representing the state the runtime system is currently in as stored in the given exception instance.

Return

The point runtime state at the point at which the exception was thrown. If the exception instance does not hold this information, the function will return an empty string.

See

hpx::diagnostic_information(), hpx::get_error_host_name(), hpx::get_error_process_id(), hpx::get_error_function_name(), hpx::get_error_file_name(), hpx::get_error_line_number(), hpx::get_error_os_thread(), hpx::get_error_thread_id(), hpx::get_error_backtrace(), hpx::get_error_env(), hpx::get_error(), hpx::get_error_what(), hpx::get_error_thread_description()

Parameters

• xi: The parameter e will be inspected for the requested diagnostic information elements which have been stored at the point where the exception was thrown. This parameter can be one of the following types: hpx::exception_info, hpx::error_code, std::exception, or std::exception_ptr.

Exceptions

• std::bad_alloc: (if one of the required allocations fails)

bool register_thread(runtime *rt, char const *name, error_code &ec = throws)

Register the current kernel thread with HPX, this should be done once for each external OS-thread intended to invoke HPX functionality. Calling this function more than once will silently fail.

void unregister_thread(runtime *rt)

Unregister the thread from HPX, this should be done once in the end before the external thread exists.

naming::gid_type const &get_locality()

The function get_locality returns a reference to the locality prefix.

std::size_t get_runtime_instance_number()

The function get_runtime_instance_number returns a unique number associated with the runtime instance the current thread is running in.

bool register_on_exit(util::function_nonser<void()> const&)

Register a function to be called during system shutdown.

bool is_starting()

Test whether the runtime system is currently being started.

This function returns whether the runtime system is currently being started or not, e.g. whether the current state of the runtime system is hpx::state_startup

Note

This function needs to be executed on a HPX-thread. It will return false otherwise.

bool tolerate_node_faults()

Test if HPX runs in fault-tolerant mode.

This function returns whether the runtime system is running in fault-tolerant mode

bool is_running()

Test whether the runtime system is currently running.

This function returns whether the runtime system is currently running or not, e.g. whether the current state of the runtime system is hpx::state_running

Note

This function needs to be executed on a HPX-thread. It will return false otherwise.

bool is_stopped()

Test whether the runtime system is currently stopped.

This function returns whether the runtime system is currently stopped or not, e.g. whether the current state of the runtime system is hpx::state_stopped

Note

This function needs to be executed on a HPX-thread. It will return false otherwise.

bool is_stopped_or_shutting_down()

Test whether the runtime system is currently being shut down.

This function returns whether the runtime system is currently being shut down or not, e.g. whether the current state of the runtime system is hpx::state_stopped or hpx::state_shutdown

Note

This function needs to be executed on a HPX-thread. It will return false otherwise.

std::size_t get_num_worker_threads()

Return the number of worker OS- threads used to execute HPX threads.

This function returns the number of OS-threads used to execute HPX threads. If the function is called while no HPX runtime system is active, it will return zero.

std::uint64_t get_system_uptime()

Return the system uptime measure on the thread executing this call.

This function returns the system uptime measured in nanoseconds for the thread executing this call. If the function is called while no HPX runtime system is active, it will return zero.

void start_active_counters(error_code &ec = throws)

Start all active performance counters, optionally naming the section of code.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

The active counters are those which have been specified on the command line while executing the application (see command line option –hpx:print-counter)

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void reset_active_counters(error_code &ec = throws)

Resets all active performance counters.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

The active counters are those which have been specified on the command line while executing the application (see command line option –hpx:print-counter)

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void reinit_active_counters(bool reset = true, error_code &ec = throws)

Re-initialize all active performance counters.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

The active counters are those which have been specified on the command line while executing the application (see command line option –hpx:print-counter)

Parameters

• reset: [in] Reset the current values before re-initializing counters (default: true)
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void stop_active_counters(error_code &ec = throws)

Stop all active performance counters.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

The active counters are those which have been specified on the command line while executing the application (see command line option –hpx:print-counter)

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void evaluate_active_counters(bool reset = false, char const *description = nullptr, error_code &ec = throws)

Evaluate and output all active performance counters, optionally naming the point in code marked by this function.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

The output generated by this function is redirected to the destination specified by the corresponding command line options (see –hpx:print-counter-destination).

Note

The active counters are those which have been specified on the command line while executing the application (see command line option –hpx:print-counter)

Parameters

• reset: [in] this is an optional flag allowing to reset the counter value after it has been evaluated.
• description: [in] this is an optional value naming the point in the code marked by the call to this function.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

serialization::binary_filter *create_binary_filter(char const *binary_filter_type, bool compress, serialization::binary_filter *next_filter = nullptr, error_code &ec = throws)

Create an instance of a binary filter plugin.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• binary_filter_type: [in] The type of the binary filter to create
• compress: [in] The created filter should support compression
• next_filter: [in] Use this as the filter to dispatch the invocation into.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::vector<Client> find_all_from_basename(std::string base_name, std::size_t num_ids)

Return all registered ids from all localities from the given base name.

This function locates all ids which were registered with the given base name. It returns a list of futures representing those ids.

Return all registered clients from all localities from the given base name.

Return

A list of futures representing the ids which were registered using the given base name.

Note

The futures will become ready even if the event (for instance, binding the name to an id) has already happened in the past. This is important in order to reliably retrieve ids from a name, even if the name was already registered.

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• num_ids: [in] The number of registered ids to expect.

This function locates all ids which were registered with the given base name. It returns a list of futures representing those ids.

Return

A list of futures representing the ids which were registered using the given base name.

Note

The futures embedded in the returned client objects will become ready even if the event (for instance, binding the name to an id) has already happened in the past. This is important in order to reliably retrieve ids from a name, even if the name was already registered.

Template Parameters

• Client: The client type to return

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• num_ids: [in] The number of registered ids to expect.

std::vector<Client> find_from_basename(std::string base_name, std::vector<std::size_t> const &ids)

Return registered ids from the given base name and sequence numbers.

This function locates the ids which were registered with the given base name and the given sequence numbers. It returns a list of futures representing those ids.

Return registered clients from the given base name and sequence numbers.

Return

A list of futures representing the ids which were registered using the given base name and sequence numbers.

Note

The futures will become ready even if the event (for instance, binding the name to an id) has already happened in the past. This is important in order to reliably retrieve ids from a name, even if the name was already registered.

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• ids: [in] The sequence numbers of the registered ids.

This function locates the ids which were registered with the given base name and the given sequence numbers. It returns a list of futures representing those ids.

Return

A list of futures representing the ids which were registered using the given base name and sequence numbers.

Note

The futures embedded in the returned client objects will become ready even if the event (for instance, binding the name to an id) has already happened in the past. This is important in order to reliably retrieve ids from a name, even if the name was already registered.

Template Parameters

• Client: The client type to return

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• ids: [in] The sequence numbers of the registered ids.

Client find_from_basename(std::string base_name, std::size_t sequence_nr = ~0U)

Return registered id from the given base name and sequence number.

This function locates the id which was registered with the given base name and the given sequence number. It returns a future representing those id.

This function locates the id which was registered with the given base name and the given sequence number. It returns a future representing those id.

Return

A representing the id which was registered using the given base name and sequence numbers.

Note

The future will become ready even if the event (for instance, binding the name to an id) has already happened in the past. This is important in order to reliably retrieve ids from a name, even if the name was already registered.

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• sequence_nr: [in] The sequence number of the registered id.

Return

A representing the id which was registered using the given base name and sequence numbers.

Note

The future embedded in the returned client object will become ready even if the event (for instance, binding the name to an id) has already happened in the past. This is important in order to reliably retrieve ids from a name, even if the name was already registered.

Template Parameters

• Client: The client type to return

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• sequence_nr: [in] The sequence number of the registered id.

hpx::future<bool> register_with_basename(std::string base_name, hpx::id_type id, std::size_t sequence_nr = ~0U)

Register the given id using the given base name.

The function registers the given ids using the provided base name.

Return

A future representing the result of the registration operation itself.

Note

The operation will fail if the given sequence number is not unique.

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• id: [in] The id to register using the given base name.
• sequence_nr: [in, optional] The sequential number to use for the registration of the id. This number has to be unique system wide for each registration using the same base name. The default is the current locality identifier. Also, the sequence numbers have to be consecutive starting from zero.

hpx::future<bool> register_with_basename(std::string base_name, hpx::future<hpx::id_type> f, std::size_t sequence_nr = ~0U)

Register the id wrapped in the given future using the given base name.

The function registers the object the given future refers to using the provided base name.

Return

A future representing the result of the registration operation itself.

Note

The operation will fail if the given sequence number is not unique.

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• f: [in] The future which should be registered using the given base name.
• sequence_nr: [in, optional] The sequential number to use for the registration of the id. This number has to be unique system wide for each registration using the same base name. The default is the current locality identifier. Also, the sequence numbers have to be consecutive starting from zero.

template<typename Client, typename Stub>
hpx::future<bool> register_with_basename(std::string base_name, components::client_base<Client, Stub> &client, std::size_t sequence_nr = ~0U)

Register the id wrapped in the given client using the given base name.

The function registers the object the given client refers to using the provided base name.

Return

A future representing the result of the registration operation itself.

Note

The operation will fail if the given sequence number is not unique.

Template Parameters

• Client: The client type to register

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• client: [in] The client which should be registered using the given base name.
• sequence_nr: [in, optional] The sequential number to use for the registration of the id. This number has to be unique system wide for each registration using the same base name. The default is the current locality identifier. Also, the sequence numbers have to be consecutive starting from zero.

Client unregister_with_basename(std::string base_name, std::size_t sequence_nr = ~0U)

Unregister the given id using the given base name.

The function unregisters the given ids using the provided base name.

Unregister the given base name.

Return

A future representing the result of the un-registration operation itself.

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• sequence_nr: [in, optional] The sequential number to use for the un-registration. This number has to be the same as has been used with register_with_basename before.

The function unregisters the given ids using the provided base name.

Return

A future representing the result of the un-registration operation itself.

Template Parameters

• Client: The client type to return

Parameters

• base_name: [in] The base name for which to retrieve the registered ids.
• sequence_nr: [in, optional] The sequential number to use for the un-registration. This number has to be the same as has been used with register_with_basename before.

naming::id_type find_here(error_code &ec = throws)

Return the global id representing this locality.

The function find_here() can be used to retrieve the global id usable to refer to the current locality.

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global id representing the locality this function has been called on.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return hpx::naming::invalid_id otherwise.

See

hpx::find_all_localities(), hpx::find_locality()

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

naming::id_type find_root_locality(error_code &ec = throws)

Return the global id representing the root locality.

The function find_root_locality() can be used to retrieve the global id usable to refer to the root locality. The root locality is the locality where the main AGAS service is hosted.

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global id representing the root locality for this application.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return hpx::naming::invalid_id otherwise.

See

hpx::find_all_localities(), hpx::find_locality()

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::vector<naming::id_type> find_all_localities(error_code &ec = throws)

Return the list of global ids representing all localities available to this application.

The function find_all_localities() can be used to retrieve the global ids of all localities currently available to this application.

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global ids representing the localities currently available to this application.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return an empty vector otherwise.

See

hpx::find_here(), hpx::find_locality()

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::vector<naming::id_type> find_all_localities(components::component_type type, error_code &ec = throws)

Return the list of global ids representing all localities available to this application which support the given component type.

The function find_all_localities() can be used to retrieve the global ids of all localities currently available to this application which support the creation of instances of the given component type.

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global ids representing the localities currently available to this application which support the creation of instances of the given component type. If no localities supporting the given component type are currently available, this function will return an empty vector.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return an empty vector otherwise.

See

hpx::find_here(), hpx::find_locality()

Parameters

• type: [in] The type of the components for which the function should return the available localities.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::vector<naming::id_type> find_remote_localities(error_code &ec = throws)

Return the list of locality ids of remote localities supporting the given component type. By default this function will return the list of all remote localities (all but the current locality).

The function find_remote_localities() can be used to retrieve the global ids of all remote localities currently available to this application (i.e. all localities except the current one).

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global ids representing the remote localities currently available to this application.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return an empty vector otherwise.

See

hpx::find_here(), hpx::find_locality()

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::vector<naming::id_type> find_remote_localities(components::component_type type, error_code &ec = throws)

Return the list of locality ids of remote localities supporting the given component type. By default this function will return the list of all remote localities (all but the current locality).

The function find_remote_localities() can be used to retrieve the global ids of all remote localities currently available to this application (i.e. all localities except the current one) which support the creation of instances of the given component type.

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global ids representing the remote localities currently available to this application.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return an empty vector otherwise.

See

hpx::find_here(), hpx::find_locality()

Parameters

• type: [in] The type of the components for which the function should return the available remote localities.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

naming::id_type find_locality(components::component_type type, error_code &ec = throws)

Return the global id representing an arbitrary locality which supports the given component type.

The function find_locality() can be used to retrieve the global id of an arbitrary locality currently available to this application which supports the creation of instances of the given component type.

Note

Generally, the id of a locality can be used for instance to create new instances of components and to invoke plain actions (global functions).

Return

The global id representing an arbitrary locality currently available to this application which supports the creation of instances of the given component type. If no locality supporting the given component type is currently available, this function will return hpx::naming::invalid_id.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function will return meaningful results only if called from an HPX-thread. It will return hpx::naming::invalid_id otherwise.

See

hpx::find_here(), hpx::find_all_localities()

Parameters

• type: [in] The type of the components for which the function should return any available locality.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

naming::id_type get_colocation_id(launch::sync_policy, naming::id_type const &id, error_code &ec = throws)

Return the id of the locality where the object referenced by the given id is currently located on.

The function hpx::get_colocation_id() returns the id of the locality where the given object is currently located.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

See

hpx::get_colocation_id()

Parameters

• id: [in] The id of the object to locate.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

lcos::future<naming::id_type> get_colocation_id(naming::id_type const &id)

Asynchronously return the id of the locality where the object referenced by the given id is currently located on.

See

hpx::get_colocation_id(launch::sync_policy)

Parameters

• id: [in] The id of the object to locate.

template<typename Component>
hpx::future<std::shared_ptr<Component>> get_ptr(naming::id_type const &id)

Returns a future referring to the pointer to the underlying memory of a component.

The function hpx::get_ptr can be used to extract a future referring to the pointer to the underlying memory of a given component.

Return

This function returns a future representing the pointer to the underlying memory for the component instance with the given id.

Note

This function will successfully return the requested result only if the given component is currently located on the calling locality. Otherwise the function will raise an error.

Note

The component instance the returned pointer refers to can not be migrated as long as there is at least one copy of the returned shared_ptr alive.

Parameters

• id: [in] The global id of the component for which the pointer to the underlying memory should be retrieved.

Template Parameters

• The: only template parameter has to be the type of the server side component.

template<typename Derived, typename Stub>
hpx::future<std::shared_ptr<typename components::client_base<Derived, Stub>::server_component_type>> get_ptr(components::client_base<Derived, Stub> const &c)

Returns a future referring to the pointer to the underlying memory of a component.

The function hpx::get_ptr can be used to extract a future referring to the pointer to the underlying memory of a given component.

Return

This function returns a future representing the pointer to the underlying memory for the component instance with the given id.

Note

This function will successfully return the requested result only if the given component is currently located on the calling locality. Otherwise the function will raise an error.

Note

The component instance the returned pointer refers to can not be migrated as long as there is at least one copy of the returned shared_ptr alive.

Parameters

• c: [in] A client side representation of the component for which the pointer to the underlying memory should be retrieved.

template<typename Component>
std::shared_ptr<Component> get_ptr(launch::sync_policy p, naming::id_type const &id, error_code &ec = throws)

Returns the pointer to the underlying memory of a component.

The function hpx::get_ptr_sync can be used to extract the pointer to the underlying memory of a given component.

Return

This function returns the pointer to the underlying memory for the component instance with the given id.

Note

This function will successfully return the requested result only if the given component is currently located on the requesting locality. Otherwise the function will raise and error.

Note

The component instance the returned pointer refers to can not be migrated as long as there is at least one copy of the returned shared_ptr alive.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• p: [in] The parameter p represents a placeholder type to turn make the call synchronous.
• id: [in] The global id of the component for which the pointer to the underlying memory should be retrieved.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

Template Parameters

• The: only template parameter has to be the type of the server side component.

template<typename Derived, typename Stub>
std::shared_ptr<typename components::client_base<Derived, Stub>::server_component_type> get_ptr(launch::sync_policy p, components::client_base<Derived, Stub> const &c, error_code &ec = throws)

Returns the pointer to the underlying memory of a component.

The function hpx::get_ptr_sync can be used to extract the pointer to the underlying memory of a given component.

Return

This function returns the pointer to the underlying memory for the component instance with the given id.

Note

This function will successfully return the requested result only if the given component is currently located on the requesting locality. Otherwise the function will raise and error.

Note

The component instance the returned pointer refers to can not be migrated as long as there is at least one copy of the returned shared_ptr alive.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• p: [in] The parameter p represents a placeholder type to turn make the call synchronous.
• c: [in] A client side representation of the component for which the pointer to the underlying memory should be retrieved.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::uint32_t get_locality_id(error_code &ec = throws)

Return the number of the locality this function is being called from.

This function returns the id of the current locality.

Note

The returned value is zero based and its maximum value is smaller than the overall number of localities the current application is running on (as returned by get_num_localities()).

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

This function needs to be executed on a HPX-thread. It will fail otherwise (it will return -1).

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::string get_locality_name()

Return the name of the locality this function is called on.

This function returns the name for the locality on which this function is called.

Return

This function returns the name for the locality on which the function is called. The name is retrieved from the underlying networking layer and may be different for different parcelports.

See

future<std::string> get_locality_name(naming::id_type const& id)

future<std::string> get_locality_name(naming::id_type const &id)

Return the name of the referenced locality.

This function returns a future referring to the name for the locality of the given id.

Return

This function returns the name for the locality of the given id. The name is retrieved from the underlying networking layer and may be different for different parcel ports.

See

std::string get_locality_name()

Parameters

• id: [in] The global id of the locality for which the name should be retrieved

std::uint32_t get_initial_num_localities()

Return the number of localities which were registered at startup for the running application.

The function get_initial_num_localities returns the number of localities which were connected to the console at application startup.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

See

hpx::find_all_localities, hpx::get_num_localities

lcos::future<std::uint32_t> get_num_localities()

Asynchronously return the number of localities which are currently registered for the running application.

The function get_num_localities asynchronously returns the number of localities currently connected to the console. The returned future represents the actual result.

Note

This function will return meaningful results only if called from an HPX-thread. It will return 0 otherwise.

See

hpx::find_all_localities, hpx::get_num_localities

std::uint32_t get_num_localities(launch::sync_policy, error_code &ec = throws)

Return the number of localities which are currently registered for the running application.

The function get_num_localities returns the number of localities currently connected to the console.

Note

This function will return meaningful results only if called from an HPX-thread. It will return 0 otherwise.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

See

hpx::find_all_localities, hpx::get_num_localities

Parameters

• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

lcos::future<std::uint32_t> get_num_localities(components::component_type t)

Asynchronously return the number of localities which are currently registered for the running application.

The function get_num_localities asynchronously returns the number of localities currently connected to the console which support the creation of the given component type. The returned future represents the actual result.

Note

This function will return meaningful results only if called from an HPX-thread. It will return 0 otherwise.

See

hpx::find_all_localities, hpx::get_num_localities

Parameters

• t: The component type for which the number of connected localities should be retrieved.

std::uint32_t get_num_localities(launch::sync_policy, components::component_type t, error_code &ec = throws)

Synchronously return the number of localities which are currently registered for the running application.

The function get_num_localities returns the number of localities currently connected to the console which support the creation of the given component type. The returned future represents the actual result.

Note

This function will return meaningful results only if called from an HPX-thread. It will return 0 otherwise.

See

hpx::find_all_localities, hpx::get_num_localities

Parameters

• t: The component type for which the number of connected localities should be retrieved.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::size_t get_os_thread_count()

Return the number of OS-threads running in the runtime instance the current HPX-thread is associated with.

std::size_t get_os_thread_count(threads::executor const &exec)

Return the number of worker OS- threads used by the given executor to execute HPX threads.

This function returns the number of cores used to execute HPX threads for the given executor. If the function is called while no HPX runtime system is active, it will return zero. If the executor is not valid, this function will fall back to retrieving the number of OS threads used by HPX.

Parameters

• exec: [in] The executor to be used.

std::string get_thread_name()

Return the name of the calling thread.

This function returns the name of the calling thread. This name uniquely identifies the thread in the context of HPX. If the function is called while no HPX runtime system is active, the result will be “<unknown>”.

std::size_t get_worker_thread_num()

Return the number of the current OS-thread running in the runtime instance the current HPX-thread is executed with.

This function returns the zero based index of the OS-thread which executes the current HPX-thread.

Note

The returned value is zero based and its maximum value is smaller than the overall number of OS-threads executed (as returned by get_os_thread_count().

Note

This function needs to be executed on a HPX-thread. It will fail otherwise (it will return -1).

std::size_t get_worker_thread_num(error_code &ec)

Return the number of the current OS-thread running in the runtime instance the current HPX-thread is executed with.

This function returns the zero based index of the OS-thread which executes the current HPX-thread.

Note

The returned value is zero based and its maximum value is smaller than the overall number of OS-threads executed (as returned by get_os_thread_count(). It will return -1 if the current thread is not a known thread or if the runtime is not in running state.

Note

This function needs to be executed on a HPX-thread. It will fail otherwise (it will return -1).

Parameters

• ec: [in,out] this represents the error status on exit.

void report_error(std::size_t num_thread, std::exception_ptr const &e)

The function report_error reports the given exception to the console.

void report_error(std::exception_ptr const &e)

The function report_error reports the given exception to the console.

char const *get_runtime_mode_name(runtime_mode state)

Get the readable string representing the name of the given runtime_mode constant.

runtime_mode get_runtime_mode_from_name(std::string const &mode)

Get the internal representation (runtime_mode constant) from the readable string representing the name.

parcel_write_handler_type set_parcel_write_handler(parcel_write_handler_type const &f)

Set the default parcel write handler which is invoked once a parcel has been sent if no explicit write handler was specified.

Return

The function returns the parcel write handler which was installed before this function was called.

Note

If no parcel handler function is registered by the user the system will call a default parcel handler function which is not performing any actions. However, this default function will terminate the application in case of any errors detected during preparing or sending the parcel.

Parameters

• f: The new parcel write handler to use from this point on

void register_pre_shutdown_function(shutdown_function_type f)

Add a function to be executed by a HPX thread during hpx::finalize() but guaranteed before any shutdown function is executed (system-wide)

Any of the functions registered with register_pre_shutdown_function are guaranteed to be executed by an HPX thread during the execution of hpx::finalize() before any of the registered shutdown functions are executed (see: hpx::register_shutdown_function()).

Note

If this function is called while the pre-shutdown functions are being executed, or after that point, it will raise a invalid_status exception.

See

hpx::register_shutdown_function()

Parameters

• f: [in] The function to be registered to run by an HPX thread as a pre-shutdown function.

void register_shutdown_function(shutdown_function_type f)

Add a function to be executed by a HPX thread during hpx::finalize() but guaranteed after any pre-shutdown function is executed (system-wide)

Any of the functions registered with register_shutdown_function are guaranteed to be executed by an HPX thread during the execution of hpx::finalize() after any of the registered pre-shutdown functions are executed (see: hpx::register_pre_shutdown_function()).

Note

If this function is called while the shutdown functions are being executed, or after that point, it will raise a invalid_status exception.

See

hpx::register_pre_shutdown_function()

Parameters

• f: [in] The function to be registered to run by an HPX thread as a shutdown function.

void register_pre_startup_function(startup_function_type f)

Add a function to be executed by a HPX thread before hpx_main but guaranteed before any startup function is executed (system-wide).

Any of the functions registered with register_pre_startup_function are guaranteed to be executed by an HPX thread before any of the registered startup functions are executed (see hpx::register_startup_function()).

This function is one of the few API functions which can be called before the runtime system has been fully initialized. It will automatically stage the provided startup function to the runtime system during its initialization (if necessary).

Note

If this function is called while the pre-startup functions are being executed or after that point, it will raise a invalid_status exception.

Parameters

• f: [in] The function to be registered to run by an HPX thread as a pre-startup function.

See

hpx::register_startup_function()

void register_startup_function(startup_function_type f)

Add a function to be executed by a HPX thread before hpx_main but guaranteed after any pre-startup function is executed (system-wide).

Any of the functions registered with register_startup_function are guaranteed to be executed by an HPX thread after any of the registered pre-startup functions are executed (see: hpx::register_pre_startup_function()), but before hpx_main is being called.

This function is one of the few API functions which can be called before the runtime system has been fully initialized. It will automatically stage the provided startup function to the runtime system during its initialization (if necessary).

Note

If this function is called while the startup functions are being executed or after that point, it will raise a invalid_status exception.

Parameters

• f: [in] The function to be registered to run by an HPX thread as a startup function.

See

hpx::register_pre_startup_function()

void trigger_lco_event(naming::id_type const &id, naming::address &&addr, bool move_credits = true)

Trigger the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should be triggered.
• addr: [in] This represents the addr of the LCO which should be triggered.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void trigger_lco_event(naming::id_type const &id, bool move_credits = true)

Trigger the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should be triggered.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void trigger_lco_event(naming::id_type const &id, naming::address &&addr, naming::id_type const &cont, bool move_credits = true)

Trigger the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should be triggered.
• addr: [in] This represents the addr of the LCO which should be triggered.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void trigger_lco_event(naming::id_type const &id, naming::id_type const &cont, bool move_credits = true)

Trigger the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should be triggered.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Result>
void set_lco_value(naming::id_type const &id, naming::address &&addr, Result &&t, bool move_credits = true)

Set the result value for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the given value.
• addr: [in] This represents the addr of the LCO which should be triggered.
• t: [in] This is the value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Result>
std::enable_if<!std::is_same<typename util::decay<Result>::type, naming::address>::value>::type set_lco_value(naming::id_type const &id, Result &&t, bool move_credits = true)

Set the result value for the (managed) LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the given value.
• t: [in] This is the value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Result>
std::enable_if<!std::is_same<typename util::decay<Result>::type, naming::address>::value>::type set_lco_value_unmanaged(naming::id_type const &id, Result &&t, bool move_credits = true)

Set the result value for the (unmanaged) LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the given value.
• t: [in] This is the value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Result>
void set_lco_value(naming::id_type const &id, naming::address &&addr, Result &&t, naming::id_type const &cont, bool move_credits = true)

Set the result value for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the given value.
• addr: [in] This represents the addr of the LCO which should be triggered.
• t: [in] This is the value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Result>
std::enable_if<!std::is_same<typename util::decay<Result>::type, naming::address>::value>::type set_lco_value(naming::id_type const &id, Result &&t, naming::id_type const &cont, bool move_credits = true)

Set the result value for the (managed) LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the given value.
• t: [in] This is the value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Result>
std::enable_if<!std::is_same<typename util::decay<Result>::type, naming::address>::value>::type set_lco_value_unmanaged(naming::id_type const &id, Result &&t, naming::id_type const &cont, bool move_credits = true)

Set the result value for the (unmanaged) LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the given value.
• t: [in] This is the value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, naming::address &&addr, std::exception_ptr const &e, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• addr: [in] This represents the addr of the LCO which should be triggered.
• e: [in] This is the error value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, naming::address &&addr, std::exception_ptr &&e, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• addr: [in] This represents the addr of the LCO which should be triggered.
• e: [in] This is the error value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, std::exception_ptr const &e, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• e: [in] This is the error value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, std::exception_ptr &&e, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• e: [in] This is the error value which should be sent to the LCO.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, naming::address &&addr, std::exception_ptr const &e, naming::id_type const &cont, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• addr: [in] This represents the addr of the LCO which should be triggered.
• e: [in] This is the error value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, naming::address &&addr, std::exception_ptr &&e, naming::id_type const &cont, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• addr: [in] This represents the addr of the LCO which should be triggered.
• e: [in] This is the error value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, std::exception_ptr const &e, naming::id_type const &cont, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• e: [in] This is the error value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

void set_lco_error(naming::id_type const &id, std::exception_ptr &&e, naming::id_type const &cont, bool move_credits = true)

Set the error state for the LCO referenced by the given id.

Parameters

• id: [in] This represents the id of the LCO which should receive the error value.
• e: [in] This is the error value which should be sent to the LCO.
• cont: [in] This represents the LCO to trigger after completion.
• move_credits: [in] If this is set to true then it is ok to send all credits in id along with the generated message. The default value is true.

template<typename Component, typename ... Ts><unspecified> hpx::new_(id_type const & locality, Ts &&... vs)

Create one or more new instances of the given Component type on the specified locality.

This function creates one or more new instances of the given Component type on the specified locality and returns a future object for the global address which can be used to reference the new component instance.

Note

This function requires to specify an explicit template argument which will define what type of component(s) to create, for instance:

hpx::future<hpx::id_type> f =
   hpx::new_<some_component>(hpx::find_here(), ...);
hpx::id_type id = f.get();

Return

The function returns different types depending on its use:

• If the explicit template argument Component represents a component type (traits::is_component<Component>::value evaluates to true), the function will return an hpx::future object instance which can be used to retrieve the global address of the newly created component.
• If the explicit template argument Component represents a client side object (traits::is_client<Component>::value evaluates to true), the function will return a new instance of that type which can be used to refer to the newly created component instance.

Parameters

• locality: [in] The global address of the locality where the new instance should be created on.
• vs: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the constructor of the created component instance.

template<typename Component, typename ... Ts><unspecified> hpx::local_new(Ts &&... vs)

Create one new instance of the given Component type on the current locality.

This function creates one new instance of the given Component type on the current locality and returns a future object for the global address which can be used to reference the new component instance.

Note

This function requires to specify an explicit template argument which will define what type of component(s) to create, for instance:

hpx::future<hpx::id_type> f =
   hpx::local_new<some_component>(...);
hpx::id_type id = f.get();

Return

The function returns different types depending on its use:

• If the explicit template argument Component represents a component type (traits::is_component<Component>::value evaluates to true), the function will return an hpx::future object instance which can be used to retrieve the global address of the newly created component. If the first argument is hpx::launch::sync the function will directly return an hpx::id_type.
• If the explicit template argument Component represents a client side object (traits::is_client<Component>::value evaluates to true), the function will return a new instance of that type which can be used to refer to the newly created component instance.

Note

The difference of this funtion to hpx::new_ is that it can be used in cases where the supplied arguments are non-copyable and non-movable. All operations are guaranteed to be local only.

Parameters

• vs: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the constructor of the created component instance.

template<typename Component, typename ... Ts><unspecified> hpx::new_(id_type const & locality, std::size_t count, Ts &&... vs)

Create multiple new instances of the given Component type on the specified locality.

This function creates multiple new instances of the given Component type on the specified locality and returns a future object for the global address which can be used to reference the new component instance.

Note

This function requires to specify an explicit template argument which will define what type of component(s) to create, for instance:

hpx::future<std::vector<hpx::id_type> > f =
   hpx::new_<some_component[]>(hpx::find_here(), 10, ...);
hpx::id_type id = f.get();

Return

The function returns different types depending on its use:

• If the explicit template argument Component represents an array of a component type (i.e. Component[], where traits::is_component<Component>::value evaluates to true), the function will return an hpx::future object instance which holds a std::vector<hpx::id_type>, where each of the items in this vector is a global address of one of the newly created components.
• If the explicit template argument Component represents an array of a client side object type (i.e. Component[], where traits::is_client<Component>::value evaluates to true), the function will return an hpx::future object instance which holds a std::vector<hpx::id_type>, where each of the items in this vector is a client side instance of the given type, each representing one of the newly created components.

Parameters

• locality: [in] The global address of the locality where the new instance should be created on.
• count: [in] The number of component instances to create
• vs: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the constructor of the created component instance.

template<typename Component, typename DistPolicy, typename ... Ts><unspecified> hpx::new_(DistPolicy const & policy, Ts &&... vs)

Create one or more new instances of the given Component type based on the given distribution policy.

This function creates one or more new instances of the given Component type on the localities defined by the given distribution policy and returns a future object for global address which can be used to reference the new component instance(s).

Note

This function requires to specify an explicit template argument which will define what type of component(s) to create, for instance:

hpx::future<hpx::id_type> f =
   hpx::new_<some_component>(hpx::default_layout, ...);
hpx::id_type id = f.get();

Return

The function returns different types depending on its use:

• If the explicit template argument Component represents a component type (traits::is_component<Component>::value evaluates to true), the function will return an hpx::future object instance which can be used to retrieve the global address of the newly created component.
• If the explicit template argument Component represents a client side object (traits::is_client<Component>::value evaluates to true), the function will return a new instance of that type which can be used to refer to the newly created component instance.

Parameters

• policy: [in] The distribution policy used to decide where to place the newly created.
• vs: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the constructor of the created component instance.

template<typename Component, typename DistPolicy, typename ... Ts><unspecified> hpx::new_(DistPolicy const & policy, std::size_t count, Ts &&... vs)

Create multiple new instances of the given Component type on the localities as defined by the given distribution policy.

This function creates multiple new instances of the given Component type on the localities defined by the given distribution policy and returns a future object for the global address which can be used to reference the new component instance.

Note

This function requires to specify an explicit template argument which will define what type of component(s) to create, for instance:

hpx::future<std::vector<hpx::id_type> > f =
   hpx::new_<some_component[]>(hpx::default_layout, 10, ...);
hpx::id_type id = f.get();

Return

The function returns different types depending on its use:

• If the explicit template argument Component represents an array of a component type (i.e. Component[], where traits::is_component<Component>::value evaluates to true), the function will return an hpx::future object instance which holds a std::vector<hpx::id_type>, where each of the items in this vector is a global address of one of the newly created components.
• If the explicit template argument Component represents an array of a client side object type (i.e. Component[], where traits::is_client<Component>::value evaluates to true), the function will return an hpx::future object instance which holds a std::vector<hpx::id_type>, where each of the items in this vector is a client side instance of the given type, each representing one of the newly created components.

Parameters

• policy: [in] The distribution policy used to decide where to place the newly created.
• count: [in] The number of component instances to create
• vs: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the constructor of the created component instance.

template<typename ...Ts>
tuple<future<Ts>...> split_future(future<tuple<Ts...>> &&f)

The function split_future is an operator allowing to split a given future of a sequence of values (any tuple, std::pair, or std::array) into an equivalent container of futures where each future represents one of the values from the original future. In some sense this function provides the inverse operation of when_all.

Return

Returns an equivalent container (same container type as passed as the argument) of futures, where each future refers to the corresponding value in the input parameter. All of the returned futures become ready once the input future has become ready. If the input future is exceptional, all output futures will be exceptional as well.

Note

The following cases are special:

tuple<future<void> > split_future(future<tuple<> > && f);
array<future<void>, 1> split_future(future<array<T, 0> > && f);

here the returned futures are directly representing the futures which were passed to the function.

Parameters

• f: [in] A future holding an arbitrary sequence of values stored in a tuple-like container. This facility supports hpx::util::tuple<>, std::pair<T1, T2>, and std::array<T, N>

template<typename T>
std::vector<future<T>> split_future(future<std::vector<T>> &&f, std::size_t size)

The function split_future is an operator allowing to split a given future of a sequence of values (any std::vector) into a std::vector of futures where each future represents one of the values from the original std::vector. In some sense this function provides the inverse operation of when_all.

Return

Returns a std::vector of futures, where each future refers to the corresponding value in the input parameter. All of the returned futures become ready once the input future has become ready. If the input future is exceptional, all output futures will be exceptional as well.

Parameters

• f: [in] A future holding an arbitrary sequence of values stored in a std::vector.
• size: [in] The number of elements the vector will hold once the input future has become ready

template<typename InputIter>
void wait_all(InputIter first, InputIter last)

The function wait_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns after they finished executing.

Note

The function wait_all returns after all futures have become ready. All input futures are still valid after wait_all returns.

Parameters

• first: The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_all should wait.
• last: The iterator pointing to the last element of a sequence of future or shared_future objects for which wait_all should wait.

template<typename R>
void wait_all(std::vector<future<R>> &&futures)

The function wait_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns after they finished executing.

Note

The function wait_all returns after all futures have become ready. All input futures are still valid after wait_all returns.

Parameters

• futures: A vector or array holding an arbitrary amount of future or shared_future objects for which wait_all should wait.

template<typename R, std::size_t N>
void wait_all(std::array<future<R>, N> &&futures)

The function wait_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns after they finished executing.

Note

The function wait_all returns after all futures have become ready. All input futures are still valid after wait_all returns.

Parameters

• futures: A vector or array holding an arbitrary amount of future or shared_future objects for which wait_all should wait.

template<typename ...T>
void wait_all(T&&... futures)

The function wait_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns after they finished executing.

Note

The function wait_all returns after all futures have become ready. All input futures are still valid after wait_all returns.

Parameters

• futures: An arbitrary number of future or shared_future objects, possibly holding different types for which wait_all should wait.

template<typename InputIter>
InputIter wait_all_n(InputIter begin, std::size_t count)

The function wait_all_n is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns after they finished executing.

Return

The function wait_all_n will return an iterator referring to the first element in the input sequence after the last processed element.

Note

The function wait_all_n returns after all futures have become ready. All input futures are still valid after wait_all_n returns.

Parameters

• begin: The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_all_n should wait.
• count: The number of elements in the sequence starting at first.

template<typename InputIter, typename Container = vector<future<typename std::iterator_traits<InputIter>::value_type>>>
future<Container> when_all(InputIter first, InputIter last)

The function when_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after they finished executing.

Return

Returns a future holding the same list of futures as has been passed to when_all.

• future<Container<future<R>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Note

Calling this version of when_all where first == last, returns a future with an empty container that is immediately ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_all will not throw an exception, but the futures held in the output collection may.

Parameters

• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_all should wait.
• last: [in] The iterator pointing to the last element of a sequence of future or shared_future objects for which when_all should wait.

template<typename Range>
future<Range> when_all(Range &&values)

The function when_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after they finished executing.

Return

Returns a future holding the same list of futures as has been passed to when_all.

• future<Container<future<R>>>: If the input cardinality is unknown at compile time and the futures are all of the same type.

Note

Calling this version of when_all where the input container is empty, returns a future with an empty container that is immediately ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_all will not throw an exception, but the futures held in the output collection may.

Parameters

• values: [in] A range holding an arbitrary amount of future or shared_future objects for which when_all should wait.

template<typename ...T>
future<tuple<future<T>...>> when_all(T&&... futures)

The function when_all is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after they finished executing.

Return

Returns a future holding the same list of futures as has been passed to when_all.

• future<tuple<future<T0>, future<T1>, future<T2>…>>: If inputs are fixed in number and are of heterogeneous types. The inputs can be any arbitrary number of future objects.
• future<tuple<>> if when_all is called with zero arguments. The returned future will be initially ready.

Note

Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_all will not throw an exception, but the futures held in the output collection may.

Parameters

• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which when_all should wait.

template<typename InputIter, typename Container = vector<future<typename std::iterator_traits<InputIter>::value_type>>>
future<Container> when_all_n(InputIter begin, std::size_t count)

The function when_all_n is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after they finished executing.

Return

Returns a future holding the same list of futures as has been passed to when_all_n.

• future<Container<future<R>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output vector will be the same as given by the input iterator.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

None of the futures in the input sequence are invalidated.

Parameters

• begin: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_all_n should wait.
• count: [in] The number of elements in the sequence starting at first.

Exceptions

• This: function will throw errors which are encountered while setting up the requested operation only. Errors encountered while executing the operations delivering the results to be stored in the futures are reported through the futures themselves.

template<typename InputIter>
void wait_any(InputIter first, InputIter last, error_code &ec = throws)

The function wait_any is a non-deterministic choice operator. It OR-composes all future objects given and returns after one future of that list finishes execution.

Note

The function wait_any returns after at least one future has become ready. All input futures are still valid after wait_any returns.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

None of the futures in the input sequence are invalidated.

Parameters

• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_any should wait.
• last: [in] The iterator pointing to the last element of a sequence of future or shared_future objects for which wait_any should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename R>
void wait_any(std::vector<future<R>> &futures, error_code &ec = throws)

The function wait_any is a non-deterministic choice operator. It OR-composes all future objects given and returns after one future of that list finishes execution.

Note

The function wait_any returns after at least one future has become ready. All input futures are still valid after wait_any returns.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

None of the futures in the input sequence are invalidated.

Parameters

• futures: [in] A vector holding an arbitrary amount of future or shared_future objects for which wait_any should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename R, std:;size_t N>void hpx::wait_any(std::array< future< R >, N > & futures, error_code & ec = throws)

The function wait_any is a non-deterministic choice operator. It OR-composes all future objects given and returns after one future of that list finishes execution.

Note

The function wait_any returns after at least one future has become ready. All input futures are still valid after wait_any returns.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

None of the futures in the input sequence are invalidated.

Parameters

• futures: [in] Amn array holding an arbitrary amount of future or shared_future objects for which wait_any should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename ...T>
void wait_any(error_code &ec, T&&... futures)

The function wait_any is a non-deterministic choice operator. It OR-composes all future objects given and returns after one future of that list finishes execution.

Note

The function wait_any returns after at least one future has become ready. All input futures are still valid after wait_any returns.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

None of the futures in the input sequence are invalidated.

Parameters

• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which wait_any should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename ...T>
void wait_any(T&&... futures)

The function wait_any is a non-deterministic choice operator. It OR-composes all future objects given and returns after one future of that list finishes execution.

Note

The function wait_any returns after at least one future has become ready. All input futures are still valid after wait_any returns.

Note

None of the futures in the input sequence are invalidated.

Parameters

• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which wait_any should wait.

template<typename InputIter>
InputIter wait_any_n(InputIter first, std::size_t count, error_code &ec = throws)

The function wait_any_n is a non-deterministic choice operator. It OR-composes all future objects given and returns after one future of that list finishes execution.

Note

The function wait_any_n returns after at least one future has become ready. All input futures are still valid after wait_any_n returns.

Return

The function wait_all_n will return an iterator referring to the first element in the input sequence after the last processed element.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Note

None of the futures in the input sequence are invalidated.

Parameters

• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_any_n should wait.
• count: [in] The number of elements in the sequence starting at first.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename InputIter, typename Container = vector<future<typename std::iterator_traits<InputIter>::value_type>>>
future<when_any_result<Container>> when_any(InputIter first, InputIter last)

The function when_any is a non-deterministic choice operator. It OR-composes all future objects given and returns a new future object representing the same list of futures after one future of that list finishes execution.

Return

Returns a when_any_result holding the same list of futures as has been passed to when_any and an index pointing to a ready future.

• future<when_any_result<Container<future<R>>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Parameters

• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_any should wait.
• last: [in] The iterator pointing to the last element of a sequence of future or shared_future objects for which when_any should wait.

template<typename Range>
future<when_any_result<Range>> when_any(Range &values)

The function when_any is a non-deterministic choice operator. It OR-composes all future objects given and returns a new future object representing the same list of futures after one future of that list finishes execution.

Return

Returns a when_any_result holding the same list of futures as has been passed to when_any and an index pointing to a ready future.

• future<when_any_result<Container<future<R>>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Parameters

• values: [in] A range holding an arbitrary amount of futures or shared_future objects for which when_any should wait.

template<typename ...T>
future<when_any_result<tuple<future<T>...>>> when_any(T&&... futures)

The function when_any is a non-deterministic choice operator. It OR-composes all future objects given and returns a new future object representing the same list of futures after one future of that list finishes execution.

Return

Returns a when_any_result holding the same list of futures as has been passed to when_any and an index pointing to a ready future..

• future<when_any_result<tuple<future<T0>, future<T1>…>>>: If inputs are fixed in number and are of heterogeneous types. The inputs can be any arbitrary number of future objects.
• future<when_any_result<tuple<>>> if when_any is called with zero arguments. The returned future will be initially ready.

Parameters

• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which when_any should wait.

template<typename InputIter, typename Container = vector<future<typename std::iterator_traits<InputIter>::value_type>>>
future<when_any_result<Container>> when_any_n(InputIter first, std::size_t count)

The function when_any_n is a non-deterministic choice operator. It OR-composes all future objects given and returns a new future object representing the same list of futures after one future of that list finishes execution.

Return

Returns a when_any_result holding the same list of futures as has been passed to when_any and an index pointing to a ready future.

• future<when_any_result<Container<future<R>>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Note

None of the futures in the input sequence are invalidated.

Parameters

• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_any_n should wait.
• count: [in] The number of elements in the sequence starting at first.

template<typename InputIter>
future<vector<future<typename std::iterator_traits<InputIter>::value_type>>> wait_some(std::size_t n, Iterator first, Iterator last, error_code &ec = throws)

The function wait_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The future returned by the function wait_some becomes ready when at least n argument futures have become ready.

Return

Returns a future holding the same list of futures as has been passed to wait_some.

• future<vector<future<R>>>: If the input cardinality is unknown at compile time and the futures are all of the same type.

Note

Calling this version of wait_some where first == last, returns a future with an empty vector that is immediately ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by wait_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_all should wait.
• last: [in] The iterator pointing to the last element of a sequence of future or shared_future objects for which when_all should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename R>
void wait_some(std::size_t n, std::vector<future<R>> &&futures, error_code &ec = throws)

The function wait_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The function wait_all returns after n futures have become ready. All input futures are still valid after wait_all returns.

Note

Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by wait_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• futures: [in] A vector holding an arbitrary amount of future or shared_future objects for which wait_some should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename R, std::size_t N>
void wait_some(std::size_t n, std::array<future<R>, N> &&futures, error_code &ec = throws)

The function wait_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The function wait_all returns after n futures have become ready. All input futures are still valid after wait_all returns.

Note

Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by wait_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• futures: [in] An array holding an arbitrary amount of future or shared_future objects for which wait_some should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename ...T>
void wait_some(std::size_t n, T&&... futures, error_code &ec = throws)

The function wait_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The function wait_all returns after n futures have become ready. All input futures are still valid after wait_all returns.

Note

Calling this version of wait_some where first == last, returns a future with an empty vector that is immediately ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by wait_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which wait_some should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename InputIter>
InputIter wait_some_n(std::size_t n, Iterator first, std::size_t count, error_code &ec = throws)

The function wait_some_n is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The function wait_all returns after n futures have become ready. All input futures are still valid after wait_all returns.

Return

This function returns an Iterator referring to the first element after the last processed input element.

Note

Calling this version of wait_some_n where count == 0, returns a future with the same elements as the arguments that is immediately ready. Possibly none of the futures in that vector are ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by wait_some_n will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_all should wait.
• count: [in] The number of elements in the sequence starting at first.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename InputIter, typename Container = vector<future<typename std::iterator_traits<InputIter>::value_type>>>
future<when_some_result<Container>> when_some(std::size_t n, Iterator first, Iterator last, error_code &ec = throws)

The function when_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The future returned by the function when_some becomes ready when at least n argument futures have become ready.

Return

Returns a when_some_result holding the same list of futures as has been passed to when_some and indices pointing to ready futures.

• future<when_some_result<Container<future<R>>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Note

Calling this version of when_some where first == last, returns a future with an empty container that is immediately ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_all should wait.
• last: [in] The iterator pointing to the last element of a sequence of future or shared_future objects for which when_all should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename Range>
future<when_some_result<Range>> when_some(std::size_t n, Range &&futures, error_code &ec = throws)

The function when_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The future returned by the function when_some becomes ready when at least n argument futures have become ready.

Return

Returns a when_some_result holding the same list of futures as has been passed to when_some and indices pointing to ready futures.

• future<when_some_result<Container<future<R>>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Note

Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• futures: [in] A container holding an arbitrary amount of future or shared_future objects for which when_some should wait.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename ...T>
future<when_some_result<tuple<future<T>...>>> when_some(std::size_t n, error_code &ec, T&&... futures)

The function when_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The future returned by the function when_some becomes ready when at least n argument futures have become ready.

Return

Returns a when_some_result holding the same list of futures as has been passed to when_some and an index pointing to a ready future..

• future<when_some_result<tuple<future<T0>, future<T1>…>>>: If inputs are fixed in number and are of heterogeneous types. The inputs can be any arbitrary number of future objects.
• future<when_some_result<tuple<>>> if when_some is called with zero arguments. The returned future will be initially ready.

Note

Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.
• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which when_some should wait.

template<typename ...T>
future<when_some_result<tuple<future<T>...>>> when_some(std::size_t n, T&&... futures)

The function when_some is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The future returned by the function when_some becomes ready when at least n argument futures have become ready.

Return

Returns a when_some_result holding the same list of futures as has been passed to when_some and an index pointing to a ready future..

• future<when_some_result<tuple<future<T0>, future<T1>…>>>: If inputs are fixed in number and are of heterogeneous types. The inputs can be any arbitrary number of future objects.
• future<when_some_result<tuple<>>> if when_some is called with zero arguments. The returned future will be initially ready.

Note

Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_some will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• futures: [in] An arbitrary number of future or shared_future objects, possibly holding different types for which when_some should wait.

template<typename InputIter, typename Container = vector<future<typename std::iterator_traits<InputIter>::value_type>>>
future<when_some_result<Container>> when_some_n(std::size_t n, Iterator first, std::size_t count, error_code &ec = throws)

The function when_some_n is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns a new future object representing the same list of futures after n of them finished executing.

Note

The future returned by the function when_some_n becomes ready when at least n argument futures have become ready.

Return

Returns a when_some_result holding the same list of futures as has been passed to when_some and indices pointing to ready futures.

• future<when_some_result<Container<future<R>>>>: If the input cardinality is unknown at compile time and the futures are all of the same type. The order of the futures in the output container will be the same as given by the input iterator.

Note

Calling this version of when_some_n where count == 0, returns a future with the same elements as the arguments that is immediately ready. Possibly none of the futures in that container are ready. Each future and shared_future is waited upon and then copied into the collection of the output (returned) future, maintaining the order of the futures in the input collection. The future returned by when_some_n will not throw an exception, but the futures held in the output collection may.

Parameters

• n: [in] The number of futures out of the arguments which have to become ready in order for the returned future to get ready.
• first: [in] The iterator pointing to the first element of a sequence of future or shared_future objects for which when_all should wait.
• count: [in] The number of elements in the sequence starting at first.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

template<typename F, typename Future>
void wait_each(F &&f, std::vector<Future> &&futures)

The function wait_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns after they finished executing. Additionally, the supplied function is called for each of the passed futures as soon as the future has become ready. wait_each returns after all futures have been become ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• futures: A vector holding an arbitrary amount of future or shared_future objects for which wait_each should wait.

template<typename F, typename Iterator>
void wait_each(F &&f, Iterator begin, Iterator end)

The function wait_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns after they finished executing. Additionally, the supplied function is called for each of the passed futures as soon as the future has become ready. wait_each returns after all futures have been become ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• begin: The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_each should wait.
• end: The iterator pointing to the last element of a sequence of future or shared_future objects for which wait_each should wait.

template<typename F, typename ...T>
void wait_each(F &&f, T&&... futures)

The function wait_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns after they finished executing. Additionally, the supplied function is called for each of the passed futures as soon as the future has become ready. wait_each returns after all futures have been become ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• futures: An arbitrary number of future or shared_future objects, possibly holding different types for which wait_each should wait.

template<typename F, typename Iterator>
void wait_each_n(F &&f, Iterator begin, std::size_t count)

The function wait_each is an operator allowing to join on the result of all given futures. It AND-composes all future objects given and returns after they finished executing. Additionally, the supplied function is called for each of the passed futures as soon as the future has become ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• begin: The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_each_n should wait.
• count: The number of elements in the sequence starting at first.

template<typename F, typename Future>
future<void> when_each(F &&f, std::vector<Future> &&futures)

The function when_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns a new future object representing the event of all those futures having finished executing. It also calls the supplied callback for each of the futures which becomes ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Return

Returns a future representing the event of all input futures being ready.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• futures: A vector holding an arbitrary amount of future or shared_future objects for which wait_each should wait.

template<typename F, typename Iterator>
future<Iterator> when_each(F &&f, Iterator begin, Iterator end)

The function when_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns a new future object representing the event of all those futures having finished executing. It also calls the supplied callback for each of the futures which becomes ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Return

Returns a future representing the event of all input futures being ready.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• begin: The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_each should wait.
• end: The iterator pointing to the last element of a sequence of future or shared_future objects for which wait_each should wait.

template<typename F, typename ...Ts>
future<void> when_each(F &&f, Ts&&... futures)

The function when_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns a new future object representing the event of all those futures having finished executing. It also calls the supplied callback for each of the futures which becomes ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Return

Returns a future representing the event of all input futures being ready.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• futures: An arbitrary number of future or shared_future objects, possibly holding different types for which wait_each should wait.

template<typename F, typename Iterator>
future<Iterator> when_each_n(F &&f, Iterator begin, std::size_t count)

The function when_each is an operator allowing to join on the results of all given futures. It AND-composes all future objects given and returns a new future object representing the event of all those futures having finished executing. It also calls the supplied callback for each of the futures which becomes ready.

Note

This function consumes the futures as they are passed on to the supplied function. The callback should take one or two parameters, namely either a future to be processed or a type that std::size_t is implicitly convertible to as the first parameter and the future as the second parameter. The first parameter will correspond to the index of the current future in the collection.

Return

Returns a future holding the iterator pointing to the first element after the last one.

Parameters

• f: The function which will be called for each of the input futures once the future has become ready.
• begin: The iterator pointing to the first element of a sequence of future or shared_future objects for which wait_each_n should wait.
• count: The number of elements in the sequence starting at first.

Variables

error_code throws

Predefined error_code object used as “throw on error” tag.

The predefined hpx::error_code object hpx::throws is supplied for use as a “throw on error” tag.

Functions that specify an argument in the form ‘error_code& ec=throws’ (with appropriate namespace qualifiers), have the following error handling semantics:

If &ec != &throws and an error occurred: ec.value() returns the implementation specific error number for the particular error that occurred and ec.category() returns the error_category for ec.value().

If &ec != &throws and an error did not occur, ec.clear().

If an error occurs and &ec == &throws, the function throws an exception of type hpx::exception or of a type derived from it. The exception’s get_errorcode() member function returns a reference to an hpx::error_code object with the behavior as specified above.

namespace actions

namespace applier

Functions

applier &get_applier()

The function get_applier returns a reference to the (thread specific) applier instance.

applier *get_applier_ptr()

The function get_applier returns a pointer to the (thread specific) applier instance. The returned pointer is NULL if the current thread is not known to HPX or if the runtime system is not active.

namespace components

Functions

template<typename Component>
future<naming::id_type> migrate_from_storage(naming::id_type const &to_resurrect, naming::id_type const &target = naming::invalid_id)

Migrate the component with the given id from the specified target storage (resurrect the object)

The function migrate_from_storage<Component> will migrate the component referenced by to_resurrect from the storage facility specified where the object is currently stored on. It returns a future referring to the migrated component instance. The component instance is resurrected on the locality specified by target_locality.

Return

A future representing the global id of the migrated component instance. This should be the same as to_resurrect.

Parameters

• to_resurrect: [in] The global id of the component to migrate.
• target: [in] The optional locality to resurrect the object on. By default the object is resurrected on the locality it was located on last.

Template Parameters

• The: only template argument specifies the component type of the component to migrate from the given storage facility.

template<typename Component>
future<naming::id_type> migrate_to_storage(naming::id_type const &to_migrate, naming::id_type const &target_storage)

Migrate the component with the given id to the specified target storage

The function migrate_to_storage<Component> will migrate the component referenced by to_migrate to the storage facility specified with target_storage. It returns a future referring to the migrated component instance.

Return

A future representing the global id of the migrated component instance. This should be the same as migrate_to.

Parameters

• to_migrate: [in] The global id of the component to migrate.
• target_storage: [in] The id of the storage facility to migrate this object to.

Template Parameters

• The: only template argument specifies the component type of the component to migrate to the given storage facility.

template<typename Derived, typename Stub>
Derived migrate_to_storage(client_base<Derived, Stub> const &to_migrate, hpx::components::component_storage const &target_storage)

Migrate the given component to the specified target storage

The function migrate_to_storage will migrate the component referenced by to_migrate to the storage facility specified with target_storage. It returns a future referring to the migrated component instance.

Return

A client side representation of representing of the migrated component instance. This should be the same as migrate_to.

Parameters

• to_migrate: [in] The client side representation of the component to migrate.
• target_storage: [in] The id of the storage facility to migrate this object to.

template<typename Component>
future<naming::id_type> copy(naming::id_type const &to_copy)

Copy given component to the specified target locality.

The function copy<Component> will create a copy of the component referenced by to_copy on the locality specified with target_locality. It returns a future referring to the newly created component instance.

Return

A future representing the global id of the newly (copied) component instance.

Note

The new component instance is created on the locality of the component instance which is to be copied.

Parameters

• to_copy: [in] The global id of the component to copy

Template Parameters

• The: only template argument specifies the component type to create.

template<typename Component>
future<naming::id_type> copy(naming::id_type const &to_copy, naming::id_type const &target_locality)

Copy given component to the specified target locality.

The function copy<Component> will create a copy of the component referenced by to_copy on the locality specified with target_locality. It returns a future referring to the newly created component instance.

Return

A future representing the global id of the newly (copied) component instance.

Parameters

• to_copy: [in] The global id of the component to copy
• target_locality: [in ] The locality where the copy should be created.

Template Parameters

• The: only template argument specifies the component type to create.

template<typename Derived, typename Stub>
Derived copy(client_base<Derived, Stub> const &to_copy, naming::id_type const &target_locality = naming::invalid_id)

Copy given component to the specified target locality.

The function copy will create a copy of the component referenced by the client side object to_copy on the locality specified with target_locality. It returns a new client side object future referring to the newly created component instance.

Return

A future representing the global id of the newly (copied) component instance.

Note

If the second argument is omitted (or is invalid_id) the new component instance is created on the locality of the component instance which is to be copied.

Parameters

• to_copy: [in] The client side object representing the component to copy
• target_locality: [in, optional] The locality where the copy should be created (default is same locality as source).

Template Parameters

• The: only template argument specifies the component type to create.

template<typename Component, typename DistPolicy>
future<naming::id_type> migrate(naming::id_type const &to_migrate, DistPolicy const &policy)

Migrate the given component to the specified target locality

The function migrate<Component> will migrate the component referenced by to_migrate to the locality specified with target_locality. It returns a future referring to the migrated component instance.

Return

A future representing the global id of the migrated component instance. This should be the same as migrate_to.

Parameters

• to_migrate: [in] The client side representation of the component to migrate.
• policy: [in] A distribution policy which will be used to determine the locality to migrate this object to.

Template Parameters

• Component: Specifies the component type of the component to migrate.
• DistPolicy: Specifies the distribution policy to use to determine the destination locality.

template<typename Derived, typename Stub, typename DistPolicy>
Derived migrate(client_base<Derived, Stub> const &to_migrate, DistPolicy const &policy)

Migrate the given component to the specified target locality

The function migrate<Component> will migrate the component referenced by to_migrate to the locality specified with target_locality. It returns a future referring to the migrated component instance.

Return

A future representing the global id of the migrated component instance. This should be the same as migrate_to.

Parameters

• to_migrate: [in] The client side representation of the component to migrate.
• policy: [in] A distribution policy which will be used to determine the locality to migrate this object to.

Template Parameters

• Derived: Specifies the component type of the component to migrate.
• DistPolicy: Specifies the distribution policy to use to determine the destination locality.

template<typename Component>
future<naming::id_type> migrate(naming::id_type const &to_migrate, naming::id_type const &target_locality)

Migrate the component with the given id to the specified target locality

The function migrate<Component> will migrate the component referenced by to_migrate to the locality specified with target_locality. It returns a future referring to the migrated component instance.

Return

A future representing the global id of the migrated component instance. This should be the same as migrate_to.

Parameters

• to_migrate: [in] The global id of the component to migrate.
• target_locality: [in] The locality where the component should be migrated to.

Template Parameters

• Component: Specifies the component type of the component to migrate.

template<typename Derived, typename Stub>
Derived migrate(client_base<Derived, Stub> const &to_migrate, naming::id_type const &target_locality)

Migrate the given component to the specified target locality

The function migrate<Component> will migrate the component referenced by to_migrate to the locality specified with target_locality. It returns a future referring to the migrated component instance.

Return

A client side representation of representing of the migrated component instance. This should be the same as migrate_to.

Parameters

• to_migrate: [in] The client side representation of the component to migrate.
• target_locality: [in] The id of the locality to migrate this object to.

Template Parameters

• Derived: Specifies the component type of the component to migrate.

Variables

char const *const default_binpacking_counter_name = "/runtime{locality/total}/count/component@"

binpacking_distribution_policy const binpacked

A predefined instance of the binpacking distribution_policy. It will represent the local locality and will place all items to create here.

colocating_distribution_policy const colocated

A predefined instance of the co-locating distribution_policy. It will represent the local locality and will place all items to create here.

default_distribution_policy const default_layout = {}

A predefined instance of the default distribution_policy. It will represent the local locality and will place all items to create here.

namespace lcos

Functions

template<typename Action, typename ArgN, ...>hpx::future<std::vector<decltype(Action(hpx::id_type, ArgN, ...))> > hpx::lcos::broadcast(std::vector< hpx::id_type > const & ids, ArgN argN, ...)

Perform a distributed broadcast operation.

The function hpx::lcos::broadcast performs a distributed broadcast operation resulting in action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The given action is invoked asynchronously on all given identifiers, and the arguments ArgN are passed along to those invocations.

Return

This function returns a future representing the result of the overall reduction operation.

Note

If decltype(Action(…)) is void, then the result of this function is future<void>.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• argN: [in] Any number of arbitrary arguments (passed by const reference) which will be forwarded to the action invocation.

template<typename Action, typename ArgN, ...>void hpx::lcos::broadcast_apply(std::vector< hpx::id_type > const & ids, ArgN argN, ...)

Perform an asynchronous (fire&forget) distributed broadcast operation.

The function hpx::lcos::broadcast_apply performs an asynchronous (fire&forget) distributed broadcast operation resulting in action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The given action is invoked asynchronously on all given identifiers, and the arguments ArgN are passed along to those invocations.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• argN: [in] Any number of arbitrary arguments (passed by const reference) which will be forwarded to the action invocation.

template<typename Action, typename ArgN, ...>hpx::future<std::vector<decltype(Action(hpx::id_type, ArgN, ..., std::size_t))> > hpx::lcos::broadcast_with_index(std::vector< hpx::id_type > const & ids, ArgN argN, ...)

Perform a distributed broadcast operation.

The function hpx::lcos::broadcast_with_index performs a distributed broadcast operation resulting in action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The given action is invoked asynchronously on all given identifiers, and the arguments ArgN are passed along to those invocations.

The function passes the index of the global identifier in the given list of identifiers as the last argument to the action.

Return

This function returns a future representing the result of the overall reduction operation.

Note

If decltype(Action(…)) is void, then the result of this function is future<void>.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• argN: [in] Any number of arbitrary arguments (passed by const reference) which will be forwarded to the action invocation.

template<typename Action, typename ArgN, ...>void hpx::lcos::broadcast_apply_with_index(std::vector< hpx::id_type > const & ids, ArgN argN, ...)

Perform an asynchronous (fire&forget) distributed broadcast operation.

The function hpx::lcos::broadcast_apply_with_index performs an asynchronous (fire&forget) distributed broadcast operation resulting in action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The given action is invoked asynchronously on all given identifiers, and the arguments ArgN are passed along to those invocations.

The function passes the index of the global identifier in the given list of identifiers as the last argument to the action.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• argN: [in] Any number of arbitrary arguments (passed by const reference) which will be forwarded to the action invocation.

template<typename Action, typename FoldOp, typename Init, typename ArgN, ...>hpx::future<decltype(Action(hpx::id_type, ArgN, ...))> hpx::lcos::fold(std::vector< hpx::id_type > const & ids, FoldOp && fold_op, Init && init, ArgN argN, ...)

Perform a distributed fold operation.

The function hpx::lcos::fold performs a distributed folding operation over results returned from action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

Note

The type of the initial value must be convertible to the result type returned from the invoked action.

Return

This function returns a future representing the result of the overall folding operation.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• fold_op: [in] A binary function expecting two results as returned from the action invocations. The function (or function object) is expected to return the result of the folding operation performed on its arguments.
• init: [in] The initial value to be used for the folding operation
• argN: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the action invocation.

template<typename Action, typename FoldOp, typename Init, typename ArgN, ...>hpx::future<decltype(Action(hpx::id_type, ArgN, ..., std::size_t))> hpx::lcos::fold_with_index(std::vector< hpx::id_type > const & ids, FoldOp && fold_op, Init && init, ArgN argN, ...)

Perform a distributed folding operation.

The function hpx::lcos::fold_with_index performs a distributed folding operation over results returned from action invocations on a given set of global identifiers. The action can be either plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The function passes the index of the global identifier in the given list of identifiers as the last argument to the action.

Note

The type of the initial value must be convertible to the result type returned from the invoked action.

Return

This function returns a future representing the result of the overall folding operation.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• fold_op: [in] A binary function expecting two results as returned from the action invocations. The function (or function object) is expected to return the result of the folding operation performed on its arguments.
• init: [in] The initial value to be used for the folding operation
• argN: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the action invocation.

template<typename Action, typename FoldOp, typename Init, typename ArgN, ...>hpx::future<decltype(Action(hpx::id_type, ArgN, ...))> hpx::lcos::inverse_fold(std::vector< hpx::id_type > const & ids, FoldOp && fold_op, Init && init, ArgN argN, ...)

Perform a distributed inverse folding operation.

The function hpx::lcos::inverse_fold performs an inverse distributed folding operation over results returned from action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

Note

The type of the initial value must be convertible to the result type returned from the invoked action.

Return

This function returns a future representing the result of the overall folding operation.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• fold_op: [in] A binary function expecting two results as returned from the action invocations. The function (or function object) is expected to return the result of the folding operation performed on its arguments.
• init: [in] The initial value to be used for the folding operation
• argN: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the action invocation.

template<typename Action, typename FoldOp, typename Init, typename ArgN, ...>hpx::future<decltype(Action(hpx::id_type, ArgN, ..., std::size_t))> hpx::lcos::inverse_fold_with_index(std::vector< hpx::id_type > const & ids, FoldOp && fold_op, Init && init, ArgN argN, ...)

Perform a distributed inverse folding operation.

The function hpx::lcos::inverse_fold_with_index performs an inverse distributed folding operation over results returned from action invocations on a given set of global identifiers. The action can be either plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The function passes the index of the global identifier in the given list of identifiers as the last argument to the action.

Note

The type of the initial value must be convertible to the result type returned from the invoked action.

Return

This function returns a future representing the result of the overall folding operation.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• fold_op: [in] A binary function expecting two results as returned from the action invocations. The function (or function object) is expected to return the result of the folding operation performed on its arguments.
• init: [in] The initial value to be used for the folding operation
• argN: [in] Any number of arbitrary arguments (passed by value, by const reference or by rvalue reference) which will be forwarded to the action invocation.

template<typename T>
hpx::future<std::vector<T>> gather_here(char const *basename, hpx::future<T> result, std::size_t num_sites = std::size_t(-1), std::size_t generation = std::size_t(-1), std::size_t this_site = std::size_t(-1))

Gather a set of values from different call sites

This function receives a set of values from all call sites operating on the given base name.

Note

Each gather operation has to be accompanied with a unique usage of the HPX_REGISTER_GATHER macro to define the necessary internal facilities used by gather_here and gather_there

Return

This function returns a future holding a vector with all gathered values. It will become ready once the gather operation has been completed.

Parameters

• basename: The base name identifying the gather operation
• result: A future referring to the value to transmit to the central gather point from this call site.
• num_sites: The number of participating sites (default: all localities).
• generation: The generational counter identifying the sequence number of the gather operation performed on the given base name. This is optional and needs to be supplied only if the gather operation on the given base name has to be performed more than once.
• this_site: The sequence number of this invocation (usually the locality id). This value is optional and defaults to whatever hpx::get_locality_id() returns.

template<typename T>
hpx::future<void> gather_there(char const *basename, hpx::future<T> result, std::size_t generation = std::size_t(-1), std::size_t root_site = 0, std::size_t this_site = std::size_t(-1))

Gather a given value at the given call site

This function transmits the value given by result to a central gather site (where the corresponding gather_here is executed)

Note

Each gather operation has to be accompanied with a unique usage of the HPX_REGISTER_GATHER macro to define the necessary internal facilities used by gather_here and gather_there

Return

This function returns a future which will become ready once the gather operation has been completed.

Parameters

• basename: The base name identifying the gather operation
• result: A future referring to the value to transmit to the central gather point from this call site.
• generation: The generational counter identifying the sequence number of the gather operation performed on the given base name. This is optional and needs to be supplied only if the gather operation on the given base name has to be performed more than once.
• root_site: The sequence number of the central gather point (usually the locality id). This value is optional and defaults to 0.
• this_site: The sequence number of this invocation (usually the locality id). This value is optional and defaults to whatever hpx::get_locality_id() returns.

template<typename T>
hpx::future<std::vector<typename std::decay<T>::type>> gather_here(char const *basename, T &&result, std::size_t num_sites = std::size_t(-1), std::size_t generation = std::size_t(-1), std::size_t this_site = std::size_t(-1))

Gather a set of values from different call sites

This function receives a set of values from all call sites operating on the given base name.

Note

Each gather operation has to be accompanied with a unique usage of the HPX_REGISTER_GATHER macro to define the necessary internal facilities used by gather_here and gather_there

Return

This function returns a future holding a vector with all gathered values. It will become ready once the gather operation has been completed.

Parameters

• basename: The base name identifying the gather operation
• result: The value to transmit to the central gather point from this call site.
• num_sites: The number of participating sites (default: all localities).
• generation: The generational counter identifying the sequence number of the gather operation performed on the given base name. This is optional and needs to be supplied only if the gather operation on the given base name has to be performed more than once.
• this_site: The sequence number of this invocation (usually the locality id). This value is optional and defaults to whatever hpx::get_locality_id() returns.

template<typename T>
hpx::future<void> gather_there(char const *basename, T &&result, std::size_t generation = std::size_t(-1), std::size_t root_site = 0, std::size_t this_site = std::size_t(-1))

Gather a given value at the given call site

This function transmits the value given by result to a central gather site (where the corresponding gather_here is executed)

Note

Each gather operation has to be accompanied with a unique usage of the HPX_REGISTER_GATHER macro to define the necessary internal facilities used by gather_here and gather_there

Return

This function returns a future which will become ready once the gather operation has been completed.

Parameters

• basename: The base name identifying the gather operation
• result: The value to transmit to the central gather point from this call site.
• generation: The generational counter identifying the sequence number of the gather operation performed on the given base name. This is optional and needs to be supplied only if the gather operation on the given base name has to be performed more than once.
• root_site: The sequence number of the central gather point (usually the locality id). This value is optional and defaults to 0.
• this_site: The sequence number of this invocation (usually the locality id). This value is optional and defaults to whatever hpx::get_locality_id() returns.

template<typename Action, typename ReduceOp, typename ArgN, ...>hpx::future<decltype(Action(hpx::id_type, ArgN, ...))> hpx::lcos::reduce(std::vector< hpx::id_type > const & ids, ReduceOp && reduce_op, ArgN argN, ...)

Perform a distributed reduction operation.

The function hpx::lcos::reduce performs a distributed reduction operation over results returned from action invocations on a given set of global identifiers. The action can be either a plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

Return

This function returns a future representing the result of the overall reduction operation.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• reduce_op: [in] A binary function expecting two results as returned from the action invocations. The function (or function object) is expected to return the result of the reduction operation performed on its arguments.
• argN: [in] Any number of arbitrary arguments (passed by by const reference) which will be forwarded to the action invocation.

template<typename Action, typename ReduceOp, typename ArgN, ...>hpx::future<decltype(Action(hpx::id_type, ArgN, ..., std::size_t))> hpx::lcos::reduce_with_index(std::vector< hpx::id_type > const & ids, ReduceOp && reduce_op, ArgN argN, ...)

Perform a distributed reduction operation.

The function hpx::lcos::reduce_with_index performs a distributed reduction operation over results returned from action invocations on a given set of global identifiers. The action can be either plain action (in which case the global identifiers have to refer to localities) or a component action (in which case the global identifiers have to refer to instances of a component type which exposes the action.

The function passes the index of the global identifier in the given list of identifiers as the last argument to the action.

Return

This function returns a future representing the result of the overall reduction operation.

Parameters

• ids: [in] A list of global identifiers identifying the target objects for which the given action will be invoked.
• reduce_op: [in] A binary function expecting two results as returned from the action invocations. The function (or function object) is expected to return the result of the reduction operation performed on its arguments.
• argN: [in] Any number of arbitrary arguments (passed by by const reference) which will be forwarded to the action invocation.

namespace naming

Functions

id_type unmanaged(id_type const &id)

The helper function hpx::unmanaged can be used to generate a global identifier which does not participate in the automatic garbage collection.

Return

This function returns a new global id referencing the same object as the parameter id. The only difference is that the returned global identifier does not participate in the automatic garbage collection.

Note

This function allows to apply certain optimizations to the process of memory management in HPX. It however requires the user to take full responsibility for keeping the referenced objects alive long enough.

Parameters

• id: [in] The id to generated the unmanaged global id from This parameter can be itself a managed or a unmanaged global id.

namespace parallel

namespace execution

Typedefs

using parallel_executor = parallel_policy_executor<hpx::launch>

using service_executor = threads::executors::service_executor

A service_executor exposes one of the predefined HPX thread pools through an executor interface.

Note

All tasks executed by one of these executors will run on one of the OS-threads dedicated for the given thread pool. The tasks will not run as HPX-threads.

using io_pool_executor = threads::executors::io_pool_executor

A io_pool_executor exposes the predefined HPX IO thread pool through an executor interface.

Note

All tasks executed by one of these executors will run on one of the OS-threads dedicated for the IO thread pool. The tasks will not run as HPX-threads.

using parcel_pool_executor = threads::executors::parcel_pool_executor

A io_pool_executor exposes the predefined HPX parcel thread pool through an executor interface.

Note

All tasks executed by one of these executors will run on one of the OS-threads dedicated for the parcel thread pool. The tasks will not run as HPX-threads.

using timer_pool_executor = threads::executors::timer_pool_executor

A io_pool_executor exposes the predefined HPX timer thread pool through an executor interface.

Note

All tasks executed by one of these executors will run on one of the OS-threads dedicated for the timer thread pool. The tasks will not run as HPX-threads.

using main_pool_executor = threads::executors::main_pool_executor

A io_pool_executor exposes the predefined HPX main thread pool through an executor interface.

Note

All tasks executed by one of these executors will run on one of the OS-threads dedicated for the main thread pool. The tasks will not run as HPX-threads.

using local_priority_queue_executor = threads::executors::local_priority_queue_executor

Creates a new local_priority_queue_executor

Parameters

• max_punits: [in] The maximum number of processing units to associate with the newly created executor.
• min_punits: [in] The minimum number of processing units to associate with the newly created executor (default: 1).

Variables

task_policy_tag HPX_CONSTEXPR_OR_CONST hpx::parallel::execution::task

Default sequential execution policy object.

HPX_STATIC_CONSTEXPR sequenced_policy hpx::parallel::execution::seq

Default sequential execution policy object.

HPX_STATIC_CONSTEXPR parallel_policy hpx::parallel::execution::par

Default parallel execution policy object.

HPX_STATIC_CONSTEXPR parallel_unsequenced_policy hpx::parallel::execution::par_unseq

Default vector execution policy object.

namespace [anonymous]

namespace [anonymous]

namespace v1

Functions

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type adjacent_difference(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest)

Assigns each value in the range given by result its corresponding element in the range [first, last] and the one preceding it except *result, which is assigned *first

The difference operations in the parallel

adjacent_difference invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly (last - first) - 1 application of the binary operator and (last - first) assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the input range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the output range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the range the algorithm will be applied to.
• dest: Refers to the beginning of the sequence of elements the results will be assigned to.

The difference operations in the parallel adjacent_difference invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

This overload of

adjacent_find is available if the user decides to provide their algorithm their own binary predicate op.

Return

The adjacent_difference algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The adjacent_find algorithm returns an iterator to the last element in the output range.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Op>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type adjacent_difference(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, Op &&op)

Assigns each value in the range given by result its corresponding element in the range [first, last] and the one preceding it except *result, which is assigned *first

The difference operations in the parallel

adjacent_difference invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly (last - first) - 1 application of the binary operator and (last - first) assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the input range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the output range (deduced). This iterator type must meet the requirements of an forward iterator.
• Op: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of adjacent_difference requires Op to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the range the algorithm will be applied to.
• dest: Refers to the beginning of the sequence of elements the results will be assigned to.
• op: The binary operator which returns the difference of elements. The signature should be equivalent to the following:

  bool op(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 must be such that objects of type FwdIter1 can be dereferenced and then implicitly converted to the dereferenced type of dest.

The difference operations in the parallel adjacent_difference invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The adjacent_difference algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The adjacent_find algorithm returns an iterator to the last element in the output range.

template<typename ExPolicy, typename FwdIter, typename Pred = detail::equal_to>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter>::type>::type adjacent_find(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&op = Pred())

Searches the range [first, last) for two consecutive identical elements. This version uses the given binary predicate op

The comparison operations in the parallel

adjacent_find invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly the smaller of (result - first) + 1 and (last - first) - 1 application of the predicate where result is the value returned

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of adjacent_find requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the range the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1 .

The comparison operations in the parallel adjacent_find invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

This overload of

adjacent_find is available if the user decides to provide their algorithm their own binary predicate op.

Return

The adjacent_find algorithm returns a hpx::future<InIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns InIter otherwise. The adjacent_find algorithm returns an iterator to the first of the identical elements. If no such elements are found, last is returned.

template<typename ExPolicy, typename FwdIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type none_of(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f, Proj &&proj = Proj())

Checks if unary predicate f returns true for no elements in the range [first, last).

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last - first applications of the predicate f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of none_of requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The none_of algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The none_of algorithm returns true if the unary predicate f returns true for no elements in the range, false otherwise. It returns true if the range is empty.

template<typename ExPolicy, typename FwdIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type any_of(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f, Proj &&proj = Proj())

Checks if unary predicate f returns true for at least one element in the range [first, last).

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last - first applications of the predicate f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of any_of requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The any_of algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The any_of algorithm returns true if the unary predicate f returns true for at least one element in the range, false otherwise. It returns false if the range is empty.

template<typename ExPolicy, typename FwdIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type all_of(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f, Proj &&proj = Proj())

Checks if unary predicate f returns true for all elements in the range [first, last).

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last - first applications of the predicate f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of all_of requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The all_of algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The all_of algorithm returns true if the unary predicate f returns true for all elements in the range, false otherwise. It returns true if the range is empty.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest)

Copies the elements in the range, defined by [first, last), to another range beginning at dest.

The assignments in the parallel

copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename Size, typename FwdIter2>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type copy_n(ExPolicy &&policy, FwdIter1 first, Size count, FwdIter2 dest)

Copies the elements in the range [first, first + count), starting from first and proceeding to first + count - 1., to another range beginning at dest.

The assignments in the parallel

copy_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, if count > 0, no assignments otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel copy_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The copy_n algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type copy_if(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, F &&f, Proj &&proj = Proj())

Copies the elements in the range, defined by [first, last), to another range beginning at dest. Copies only the elements for which the predicate f returns true. The order of the elements that are not removed is preserved.

The assignments in the parallel

copy_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The copy_if algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIterB, typename FwdIterE, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename std::iterator_traits<FwdIterB>::difference_type>::type count(ExPolicy &&policy, FwdIterB first, FwdIterE last, T const &value, Proj &&proj = Proj())

Returns the number of elements in the range [first, last) satisfying a specific criteria. This version counts the elements that are equal to the given value.

The comparisons in the parallel

count algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first comparisons.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the comparisons.
• FwdIterB: The type of the source begin iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIterE: The type of the source end iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to search for (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• value: The value to search for.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

Note

The comparisons in the parallel count algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The count algorithm returns a hpx::future<difference_type> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by std::iterator_traits<FwdIterB>::difference_type. The count algorithm returns the number of elements satisfying the given criteria.

template<typename ExPolicy, typename FwdIterB, typename FwdIterE, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename std::iterator_traits<FwdIterB>::difference_type>::type count_if(ExPolicy &&policy, FwdIterB first, FwdIterE last, F &&f, Proj &&proj = Proj())

Returns the number of elements in the range [first, last) satisfying a specific criteria. This version counts elements for which predicate f returns true.

Note

Complexity: Performs exactly last - first applications of the predicate.

Note

The assignments in the parallel count_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

The assignments in the parallel count_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The count_if algorithm returns hpx::future<difference_type> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by std::iterator_traits<FwdIterB>::difference_type. The count algorithm returns the number of elements satisfying the given criteria.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the comparisons.
• FwdIterB: The type of the source begin iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIterE: The type of the source end iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of count_if requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIterB can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

template<typename ExPolicy, typename FwdIter>
util::detail::algorithm_result<ExPolicy>::type destroy(ExPolicy &&policy, FwdIter first, FwdIter last)

Destroys objects of type typename iterator_traits<ForwardIt>::value_type in the range [first, last).

The operations in the parallel

destroy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first operations.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.

The operations in the parallel destroy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The destroy algorithm returns a hpx::future<void>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename ExPolicy, typename FwdIter, typename Size>
util::detail::algorithm_result<ExPolicy, FwdIter>::type destroy_n(ExPolicy &&policy, FwdIter first, Size count)

Destroys objects of type typename iterator_traits<ForwardIt>::value_type in the range [first, first + count).

The operations in the parallel

destroy_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count operations, if count > 0, no assignments otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Size: The type of the argument specifying the number of elements to apply this algorithm to.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.

The operations in the parallel destroy_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The destroy_n algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The destroy_n algorithm returns the iterator to the element in the source range, one past the last element constructed.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, bool>::type>::type equal(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, Pred &&op = Pred())

Returns true if the range [first1, last1) is equal to the range [first2, last2), and false otherwise.

The comparison operations in the parallel

equal algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most min(last1 - first1, last2 - first2) applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of equal requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively

The comparison operations in the parallel equal algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The two ranges are considered equal if, for every iterator i in the range [first1,last1), *i equals *(first2 + (i - first1)). This overload of equal uses operator== to determine if two elements are equal.

Return

The equal algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The equal algorithm returns true if the elements in the two ranges are equal, otherwise it returns false. If the length of the range [first1, last1) does not equal the length of the range [first2, last2), it returns false.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, bool>::type>::type equal(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, Pred &&op = Pred())

Returns true if the range [first1, last1) is equal to the range starting at first2, and false otherwise.

The comparison operations in the parallel

equal algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last1 - first1 applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of equal requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively

The comparison operations in the parallel equal algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The two ranges are considered equal if, for every iterator i in the range [first1,last1), *i equals *(first2 + (i - first1)). This overload of equal uses operator== to determine if two elements are equal.

Return

The equal algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The equal algorithm returns true if the elements in the two ranges are equal, otherwise it returns false.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T, typename Op>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type exclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, T init, Op &&op)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(binary_op, init, *first, …, *(first + (i - result) - 1)).

The reduce operations in the parallel

exclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be used as initial (and intermediate) values (deduced).
• Op: The type of the binary function object used for the reduction operation.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• init: The initial value for the generalized sum.
• op: Specifies the function (or function object) which will be invoked for each of the values of the input sequence. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Ret must be such that an object of a type as given by the input sequence can be implicitly converted to any of those types.

The reduce operations in the parallel exclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

exclusive_scan and inclusive_scan is that inclusive_scan includes the ith input element in the ith sum. If op is not mathematically associative, the behavior of inclusive_scan may be non-deterministic.

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The exclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aN) is defined as:

• a1 when N is 1
• op(GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aK), GENERALIZED_NONCOMMUTATIVE_SUM(op, aM, …, aN)) where 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type exclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, T init)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(+, init, *first, …, *(first + (i - result) - 1))

The reduce operations in the parallel

exclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate std::plus<T>.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be used as initial (and intermediate) values (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• init: The initial value for the generalized sum.

The reduce operations in the parallel exclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

exclusive_scan and inclusive_scan is that inclusive_scan includes the ith input element in the ith sum.

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The exclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(+, a1, …, aN) is defined as:

• a1 when N is 1
• GENERALIZED_NONCOMMUTATIVE_SUM(+, a1, …, aK)
  • GENERALIZED_NONCOMMUTATIVE_SUM(+, aM, …, aN) where 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIter, typename T>
util::detail::algorithm_result<ExPolicy>::type fill(ExPolicy &&policy, FwdIter first, FwdIter last, T value)

Assigns the given value to the elements in the range [first, last).

The comparisons in the parallel

fill algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be assigned (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• value: The value to be assigned.

The comparisons in the parallel fill algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The fill algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by void.

template<typename ExPolicy, typename FwdIter, typename Size, typename T>
util::detail::algorithm_result<ExPolicy, FwdIter>::type fill_n(ExPolicy &&policy, FwdIter first, Size count, T value)

Assigns the given value value to the first count elements in the range beginning at first if count > 0. Does nothing otherwise.

The comparisons in the parallel

fill_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, for count > 0.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an output iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• T: The type of the value to be assigned (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• value: The value to be assigned.

The comparisons in the parallel fill_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The fill_n algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by void.

template<typename ExPolicy, typename FwdIter, typename T>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter>::type>::type find(ExPolicy &&policy, FwdIter first, FwdIter last, T const &val)

Returns the first element in the range [first, last) that is equal to value

The comparison operations in the parallel

find algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last - first applications of the operator==().

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to find (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• val: the value to compare the elements to

The comparison operations in the parallel find algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The find algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The find algorithm returns the first element in the range [first,last) that is equal to val. If no such element in the range of [first,last) is equal to val, then the algorithm returns last.

template<typename ExPolicy, typename FwdIter, typename F>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter>::type>::type find_if(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f)

Returns the first element in the range [first, last) for which predicate f returns true

The comparison operations in the parallel

find_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• f: The unary predicate which returns true for the required element. The signature of the predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type.

The comparison operations in the parallel find_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The find_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The find_if algorithm returns the first element in the range [first,last) that satisfies the predicate f. If no such element exists that satisfies the predicate f, the algorithm returns last.

template<typename ExPolicy, typename FwdIter, typename F>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter>::type>::type find_if_not(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f)

Returns the first element in the range [first, last) for which predicate f returns false

The comparison operations in the parallel

find_if_not algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• f: The unary predicate which returns false for the required element. The signature of the predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type.

The comparison operations in the parallel find_if_not algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The find_if_not algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The find_if_not algorithm returns the first element in the range [first, last) that does not satisfy the predicate f. If no such element exists that does not satisfy the predicate f, the algorithm returns last.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to, typename Proj = util::projection_identity>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter1>::type>::type find_end(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, Pred &&op = Pred(), Proj &&proj = Proj())

Returns the last subsequence of elements [first2, last2) found in the range [first, last) using the given predicate f to compare elements.

The comparison operations in the parallel

find_end algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most S*(N-S+1) comparisons where S = distance(first2, last2) and N = distance(first1, last1).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of replace requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of type dereferenced FwdIter1 and dereferenced FwdIter2.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements the algorithm will be searching for.
• last2: Refers to the end of the sequence of elements of the algorithm will be searching for.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively.
• proj: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter1 and dereferenced FwdIter2 as a projection operation before the function f is invoked.

The comparison operations in the parallel find_end algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

This overload of

find_end is available if the user decides to provide the algorithm their own predicate f.

Return

The find_end algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The find_end algorithm returns an iterator to the beginning of the last subsequence [first2, last2) in range [first, last). If the length of the subsequence [first2, last2) is greater than the length of the range [first1, last1), last1 is returned. Additionally if the size of the subsequence is empty or no subsequence is found, last1 is also returned.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter1>::type>::type find_first_of(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 s_first, FwdIter2 s_last, Pred &&op = Pred(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Searches the range [first, last) for any elements in the range [s_first, s_last). Uses binary predicate p to compare elements

The comparison operations in the parallel

find_first_of algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most (S*N) comparisons where S = distance(s_first, s_last) and N = distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of equal requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj1: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of type dereferenced FwdIter1.
• Proj2: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of type dereferenced FwdIter2.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• s_first: Refers to the beginning of the sequence of elements the algorithm will be searching for.
• s_last: Refers to the end of the sequence of elements of the algorithm will be searching for.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively.
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter1 as a projection operation before the function op is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter2 as a projection operation before the function op is invoked.

The comparison operations in the parallel find_first_of algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The find_first_of algorithm returns a hpx::future<FwdIter1> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter1 otherwise. The find_first_of algorithm returns an iterator to the first element in the range [first, last) that is equal to an element from the range [s_first, s_last). If the length of the subsequence [s_first, s_last) is greater than the length of the range [first, last), last is returned. Additionally if the size of the subsequence is empty or no subsequence is found, last is also returned. This overload of find_end is available if the user decides to provide the algorithm their own predicate f.

template<typename ExPolicy, typename FwdIter, typename Size, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type for_each_n(ExPolicy &&policy, FwdIter first, Size count, F &&f, Proj &&proj = Proj())

Applies f to the result of dereferencing every iterator in the range [first, first + count), starting from first and proceeding to first + count - 1.

If

f returns a result, the result is ignored.

Note

Complexity: Applies f exactly count times.

If the type of first satisfies the requirements of a mutable iterator, f may apply non-constant functions through the dereferenced iterator.

Unlike its sequential form, the parallel overload of for_each does not return a copy of its Function parameter, since parallelization may not permit efficient state accumulation.

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of for_each requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  <ignored> pred(const Type &a);
  

  The signature does not need to have const&. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate f is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The for_each_n algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. It returns first + count for non-negative values of count and first for negative values.

template<typename ExPolicy, typename FwdIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type for_each(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f, Proj &&proj = Proj())

Applies f to the result of dereferencing every iterator in the range [first, last).

If

f returns a result, the result is ignored.

Note

Complexity: Applies f exactly last - first times.

If the type of first satisfies the requirements of a mutable iterator, f may apply non-constant functions through the dereferenced iterator.

Unlike its sequential form, the parallel overload of for_each does not return a copy of its Function parameter, since parallelization may not permit efficient state accumulation.

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of for_each requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  <ignored> pred(const Type &a);
  

  The signature does not need to have const&. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate f is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The for_each algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. It returns last.

template<typename ExPolicy, typename FwdIter, typename F>
util::detail::algorithm_result<ExPolicy, FwdIter>::type generate(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f)

Assign each element in range [first, last) a value generated by the given function object f

The assignments in the parallel

generate algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly distance(first, last) invocations of f and assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: generator function that will be called. signature of function should be equivalent to the following:

  Ret fun();
  

  The type

  Ret must be such that an object of type FwdIter can be dereferenced and assigned a value of type Ret.

The assignments in the parallel generate algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. It returns last.

template<typename ExPolicy, typename FwdIter, typename Size, typename F>
util::detail::algorithm_result<ExPolicy, FwdIter>::type generate_n(ExPolicy &&policy, FwdIter first, Size count, F &&f)

Assigns each element in range [first, first+count) a value generated by the given function object g.

The assignments in the parallel

generate_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly count invocations of f and assignments, for count > 0.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements in the sequence the algorithm will be applied to.
• f: Refers to the generator function object that will be called. The signature of the function should be equivalent to

  Ret fun();
  

  The type

  Ret must be such that an object of type OutputIt can be dereferenced and assigned a value of type Ret.

The assignments in the parallel generate_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. It returns last.

template<typename ExPolicy, typename RandIter, typename Comp = detail::less, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type is_heap(ExPolicy &&policy, RandIter first, RandIter last, Comp &&comp = Comp(), Proj &&proj = Proj())

Returns whether the range is max heap. That is, true if the range is max heap, false otherwise. The function uses the given comparison function object comp (defaults to using operator<()).

comp has to induce a strict weak ordering on the values.

Note

Complexity: Performs at most N applications of the comparison comp, at most 2 * N applications of the projection proj, where N = last - first.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• RandIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• Comp: The type of the function/function object to use (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_heap algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The is_heap algorithm returns whether the range is max heap. That is, true if the range is max heap, false otherwise.

template<typename ExPolicy, typename RandIter, typename Comp = detail::less, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, RandIter>::type is_heap_until(ExPolicy &&policy, RandIter first, RandIter last, Comp &&comp = Comp(), Proj &&proj = Proj())

Returns the upper bound of the largest range beginning at first which is a max heap. That is, the last iterator it for which range [first, it) is a max heap. The function uses the given comparison function object comp (defaults to using operator<()).

comp has to induce a strict weak ordering on the values.

Note

Complexity: Performs at most N applications of the comparison comp, at most 2 * N applications of the projection proj, where N = last - first.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• RandIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• Comp: The type of the function/function object to use (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_heap_until algorithm returns a hpx::future<RandIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns RandIter otherwise. The is_heap_until algorithm returns the upper bound of the largest range beginning at first which is a max heap. That is, the last iterator it for which range [first, it) is a max heap.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, bool>::type>::type includes(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, Pred &&op = Pred())

Returns true if every element from the sorted range [first2, last2) is found within the sorted range [first1, last1). Also returns true if [first2, last2) is empty. The version expects both ranges to be sorted with the user supplied binary predicate f.

The comparison operations in the parallel

includes algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

At most 2*(N1+N2-1) comparisons, where N1 = std::distance(first1, last1) and N2 = std::distance(first2, last2).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of includes requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as includes. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively

The comparison operations in the parallel includes algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The includes algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The includes algorithm returns true every element from the sorted range [first2, last2) is found within the sorted range [first1, last1). Also returns true if [first2, last2) is empty.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Op, typename T>
util::detail::algorithm_result<ExPolicy, FwdIter2>::type inclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, Op &&op, T init)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(op, init, *first, …, *(first + (i - result))).

The reduce operations in the parallel

inclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be used as initial (and intermediate) values (deduced).
• Op: The type of the binary function object used for the reduction operation.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• init: The initial value for the generalized sum.
• op: Specifies the function (or function object) which will be invoked for each of the values of the input sequence. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Ret must be such that an object of a type as given by the input sequence can be implicitly converted to any of those types.

The reduce operations in the parallel inclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

exclusive_scan and inclusive_scan is that inclusive_scan includes the ith input element in the ith sum. If op is not mathematically associative, the behavior of inclusive_scan may be non-deterministic.

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The inclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aN) is defined as:

• a1 when N is 1
• op(GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aK), GENERALIZED_NONCOMMUTATIVE_SUM(op, aM, …, aN)) where 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Op>
util::detail::algorithm_result<ExPolicy, FwdIter2>::type inclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, Op &&op)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(op, *first, …, *(first + (i - result))).

The reduce operations in the parallel

inclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• Op: The type of the binary function object used for the reduction operation.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• op: Specifies the function (or function object) which will be invoked for each of the values of the input sequence. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Ret must be such that an object of a type as given by the input sequence can be implicitly converted to any of those types.

The reduce operations in the parallel inclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

exclusive_scan and inclusive_scan is that inclusive_scan includes the ith input element in the ith sum.

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The inclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(+, a1, …, aN) is defined as:

• a1 when N is 1
• GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aK)
  • GENERALIZED_NONCOMMUTATIVE_SUM(+, aM, …, aN) where 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type inclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest)

Assigns through each iterator i in [result, result + (last - first)) the value of gENERALIZED_NONCOMMUTATIVE_SUM(+, *first, …, *(first + (i - result))).

The reduce operations in the parallel

inclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The reduce operations in the parallel inclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

exclusive_scan and inclusive_scan is that inclusive_scan includes the ith input element in the ith sum.

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The inclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(+, a1, …, aN) is defined as:

• a1 when N is 1
• GENERALIZED_NONCOMMUTATIVE_SUM(+, a1, …, aK)
  • GENERALIZED_NONCOMMUTATIVE_SUM(+, aM, …, aN) where 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIter, typename Pred>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, bool>::type>::type is_partitioned(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&pred)

Determines if the range [first, last) is partitioned.

The predicate operations in the parallel

is_partitioned algorithm invoked with an execution policy object of type sequenced_policy executes in sequential order in the calling thread.

Note

Complexity: at most (N) predicate evaluations where N = distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the This iterator type must meet the requirements of a forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of that the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of that the algorithm will be applied to.
• pred: Refers to the binary predicate which returns true if the first argument should be treated as less than the second argument. The signature of the function should be equivalent to

  bool pred(const Type &a, const Type &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type must be such that objects of types FwdIter can be dereferenced and then implicitly converted to Type.

The comparison operations in the parallel is_partitioned algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_partitioned algorithm returns a hpx::future<bool> if the execution policy is of type task_execution_policy and returns bool otherwise. The is_partitioned algorithm returns true if each element in the sequence for which pred returns true precedes those for which pred returns false. Otherwise is_partitioned returns false. If the range [first, last) contains less than two elements, the function is always true.

template<typename ExPolicy, typename FwdIter, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, bool>::type>::type is_sorted(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&pred = Pred())

Determines if the range [first, last) is sorted. Uses pred to compare elements.

The comparison operations in the parallel

is_sorted algorithm invoked with an execution policy object of type sequenced_policy executes in sequential order in the calling thread.

Note

Complexity: at most (N+S-1) comparisons where N = distance(first, last). S = number of partitions

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the This iterator type must meet the requirements of a forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of is_sorted requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of that the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of that the algorithm will be applied to.
• pred: Refers to the binary predicate which returns true if the first argument should be treated as less than the second argument. The signature of the function should be equivalent to

  bool pred(const Type &a, const Type &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type must be such that objects of types FwdIter can be dereferenced and then implicitly converted to Type.

The comparison operations in the parallel is_sorted algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_sorted algorithm returns a hpx::future<bool> if the execution policy is of type task_execution_policy and returns bool otherwise. The is_sorted algorithm returns a bool if each element in the sequence [first, last) satisfies the predicate passed. If the range [first, last) contains less than two elements, the function always returns true.

template<typename ExPolicy, typename FwdIter, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter>::type>::type is_sorted_until(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&pred = Pred())

Returns the first element in the range [first, last) that is not sorted. Uses a predicate to compare elements or the less than operator.

The comparison operations in the parallel

is_sorted_until algorithm invoked with an execution policy object of type sequenced_policy executes in sequential order in the calling thread.

Note

Complexity: at most (N+S-1) comparisons where N = distance(first, last). S = number of partitions

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the This iterator type must meet the requirements of a forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of is_sorted_until requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of that the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of that the algorithm will be applied to.
• pred: Refers to the binary predicate which returns true if the first argument should be treated as less than the second argument. The signature of the function should be equivalent to

  bool pred(const Type &a, const Type &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type must be such that objects of types FwdIter can be dereferenced and then implicitly converted to Type.

The comparison operations in the parallel is_sorted_until algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_sorted_until algorithm returns a hpx::future<FwdIter> if the execution policy is of type task_execution_policy and returns FwdIter otherwise. The is_sorted_until algorithm returns the first unsorted element. If the sequence has less than two elements or the sequence is sorted, last is returned.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, bool>::type>::type lexicographical_compare(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, Pred &&pred = Pred())

Checks if the first range [first1, last1) is lexicographically less than the second range [first2, last2). uses a provided predicate to compare elements.

The comparison operations in the parallel

lexicographical_compare algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most 2 * min(N1, N2) applications of the comparison operation, where N1 = std::distance(first1, last) and N2 = std::distance(first2, last2).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of lexicographical_compare requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• pred: Refers to the comparison function that the first and second ranges will be applied to

The comparison operations in the parallel lexicographical_compare algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

Lexicographical comparison is an operation with the following properties

• Two ranges are compared element by element
• The first mismatching element defines which range is lexicographically less or greater than the other
• If one range is a prefix of another, the shorter range is lexicographically less than the other
• If two ranges have equivalent elements and are of the same length, then the ranges are lexicographically equal
• An empty range is lexicographically less than any non-empty range
• Two empty ranges are lexicographically equal

Return

The lexicographically_compare algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The lexicographically_compare algorithm returns true if the first range is lexicographically less, otherwise it returns false. range [first2, last2), it returns false.

template<typename ExPolicy, typename RandIter1, typename RandIter2, typename RandIter3, typename Comp = detail::less, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in1(RandIter1), tag::in2(RandIter2), tag::out(RandIter3)>>::type merge(ExPolicy &&policy, RandIter1 first1, RandIter1 last1, RandIter2 first2, RandIter2 last2, RandIter3 dest, Comp &&comp = Comp(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Merges two sorted ranges [first1, last1) and [first2, last2) into one sorted range beginning at dest. The order of equivalent elements in the each of original two ranges is preserved. For equivalent elements in the original two ranges, the elements from the first range precede the elements from the second range. The destination range cannot overlap with either of the input ranges.

The assignments in the parallel

merge algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs O(std::distance(first1, last1) + std::distance(first2, last2)) applications of the comparison comp and the each projection.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• RandIter1: The type of the source iterators used (deduced) representing the first sorted range. This iterator type must meet the requirements of an random access iterator.
• RandIter2: The type of the source iterators used (deduced) representing the second sorted range. This iterator type must meet the requirements of an random access iterator.
• RandIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an random access iterator.
• Comp: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of merge requires Comp to meet the requirements of CopyConstructible. This defaults to std::less<>
• Proj1: The type of an optional projection function to be used for elements of the first range. This defaults to util::projection_identity
• Proj2: The type of an optional projection function to be used for elements of the second range. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the first range of elements the algorithm will be applied to.
• last1: Refers to the end of the first range of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second range of elements the algorithm will be applied to.
• last2: Refers to the end of the second range of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• comp: comp is a callable object which returns true if the first argument is less than the second, and false otherwise. The signature of this comparison should be equivalent to:

  bool comp(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types RandIter1 and RandIter2 can be dereferenced and then implicitly converted to both Type1 and Type2
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of the first range as a projection operation before the actual comparison comp is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of the second range as a projection operation before the actual comparison comp is invoked.

The assignments in the parallel merge algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The merge algorithm returns a hpx::future<tagged_tuple<tag::in1(RandIter1), tag::in2(RandIter2), tag::out(RandIter3)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_tuple<tag::in1(RandIter1), tag::in2(RandIter2), tag::out(RandIter3)> otherwise. The merge algorithm returns the tuple of the source iterator last1, the source iterator last2, the destination iterator to the end of the dest range.

template<typename ExPolicy, typename RandIter, typename Comp = detail::less, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, RandIter>::type inplace_merge(ExPolicy &&policy, RandIter first, RandIter middle, RandIter last, Comp &&comp = Comp(), Proj &&proj = Proj())

Merges two consecutive sorted ranges [first, middle) and [middle, last) into one sorted range [first, last). The order of equivalent elements in the each of original two ranges is preserved. For equivalent elements in the original two ranges, the elements from the first range precede the elements from the second range.

The assignments in the parallel

inplace_merge algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs O(std::distance(first, last)) applications of the comparison comp and the each projection.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• RandIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an random access iterator.
• Comp: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of inplace_merge requires Comp to meet the requirements of CopyConstructible. This defaults to std::less<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the first sorted range the algorithm will be applied to.
• middle: Refers to the end of the first sorted range and the beginning of the second sorted range the algorithm will be applied to.
• last: Refers to the end of the second sorted range the algorithm will be applied to.
• comp: comp is a callable object which returns true if the first argument is less than the second, and false otherwise. The signature of this comparison should be equivalent to:

  bool comp(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types RandIter can be dereferenced and then implicitly converted to both Type1 and Type2
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel inplace_merge algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The inplace_merge algorithm returns a hpx::future<RandIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns RandIter otherwise. The inplace_merge algorithm returns the source iterator last

template<typename ExPolicy, typename FwdIter, typename Proj = util::projection_identity, typename F = detail::less>
util::detail::algorithm_result<ExPolicy, FwdIter>::type min_element(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f = F(), Proj &&proj = Proj())

Finds the smallest element in the range [first, last) using the given comparison function f.

The comparisons in the parallel

min_element algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly max(N-1, 0) comparisons, where N = std::distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of min_element requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: The binary predicate which returns true if the the left argument is less than the right element. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The comparisons in the parallel min_element algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The min_element algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The min_element algorithm returns the iterator to the smallest element in the range [first, last). If several elements in the range are equivalent to the smallest element, returns the iterator to the first such element. Returns last if the range is empty.

template<typename ExPolicy, typename FwdIter, typename Proj = util::projection_identity, typename F = detail::less>
util::detail::algorithm_result<ExPolicy, FwdIter>::type max_element(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f = F(), Proj &&proj = Proj())

Finds the greatest element in the range [first, last) using the given comparison function f.

The comparisons in the parallel

max_element algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly max(N-1, 0) comparisons, where N = std::distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of max_element requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: The binary predicate which returns true if the This argument is optional and defaults to std::less. the left argument is less than the right element. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The comparisons in the parallel max_element algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The max_element algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The max_element algorithm returns the iterator to the smallest element in the range [first, last). If several elements in the range are equivalent to the smallest element, returns the iterator to the first such element. Returns last if the range is empty.

template<typename ExPolicy, typename FwdIter, typename Proj = util::projection_identity, typename F = detail::less>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::min(FwdIter), tag::max(FwdIter)>>::type minmax_element(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f = F(), Proj &&proj = Proj())

Finds the greatest element in the range [first, last) using the given comparison function f.

The comparisons in the parallel

minmax_element algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most max(floor(3/2*(N-1)), 0) applications of the predicate, where N = std::distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of minmax_element requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: The binary predicate which returns true if the the left argument is less than the right element. This argument is optional and defaults to std::less. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The comparisons in the parallel minmax_element algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The minmax_element algorithm returns a hpx::future<tagged_pair<tag::min(FwdIter), tag::max(FwdIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::min(FwdIter), tag::max(FwdIter)> otherwise. The minmax_element algorithm returns a pair consisting of an iterator to the smallest element as the first element and an iterator to the greatest element as the second. Returns std::make_pair(first, first) if the range is empty. If several elements are equivalent to the smallest element, the iterator to the first such element is returned. If several elements are equivalent to the largest element, the iterator to the last such element is returned.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, std::pair<FwdIter1, FwdIter2>>::type>::type mismatch(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, Pred &&op = Pred())

Returns true if the range [first1, last1) is mismatch to the range [first2, last2), and false otherwise.

The comparison operations in the parallel

mismatch algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most min(last1 - first1, last2 - first2) applications of the predicate f. If FwdIter1 and FwdIter2 meet the requirements of RandomAccessIterator and (last1 - first1) != (last2 - first2) then no applications of the predicate f are made.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of mismatch requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as mismatch. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively

The comparison operations in the parallel mismatch algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The two ranges are considered mismatch if, for every iterator i in the range [first1,last1), *i mismatchs *(first2 + (i - first1)). This overload of mismatch uses operator== to determine if two elements are mismatch.

Return

The mismatch algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The mismatch algorithm returns true if the elements in the two ranges are mismatch, otherwise it returns false. If the length of the range [first1, last1) does not mismatch the length of the range [first2, last2), it returns false.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, std::pair<FwdIter1, FwdIter2>>::type>::type mismatch(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, Pred &&op = Pred())

Returns std::pair with iterators to the first two non-equivalent elements.

The comparison operations in the parallel

mismatch algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most last1 - first1 applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of mismatch requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as mismatch. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively

The comparison operations in the parallel mismatch algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The mismatch algorithm returns a hpx::future<std::pair<FwdIter1, FwdIter2> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns std::pair<FwdIter1, FwdIter2> otherwise. The mismatch algorithm returns the first mismatching pair of elements from two ranges: one defined by [first1, last1) and another defined by [first2, last2).

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type move(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest)

Moves the elements in the range [first, last), to another range beginning at dest. After this operation the elements in the moved-from range will still contain valid values of the appropriate type, but not necessarily the same values as before the move.

The move assignments in the parallel

move algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first move assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the move assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The move assignments in the parallel move algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The move algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The move algorithm returns the pair of the input iterator last and the output iterator to the element in the destination range, one past the last element moved.

template<typename ExPolicy, typename BidirIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, BidirIter>::type stable_partition(ExPolicy &&policy, BidirIter first, BidirIter last, F &&f, Proj &&proj = Proj())

Permutes the elements in the range [first, last) such that there exists an iterator i such that for every iterator j in the range [first, i) INVOKE(f, INVOKE (proj, *j)) != false, and for every iterator k in the range [i, last), INVOKE(f, INVOKE (proj, *k)) == false

The invocations of

f in the parallel stable_partition algorithm invoked with an execution policy object of type sequenced_policy executes in sequential order in the calling thread.

Note

Complexity: At most (last - first) * log(last - first) swaps, but only linear number of swaps if there is enough extra memory. Exactly last - first applications of the predicate and projection.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• BidirIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Unary predicate which returns true if the element should be ordered before other elements. Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool fun(const Type &a);
  

  The signature does not need to have const&. The type

  Type must be such that an object of type BidirIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel stable_partition algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The stable_partition algorithm returns an iterator i such that for every iterator j in the range [first, i), f(*j) != false INVOKE(f, INVOKE(proj, *j)) != false, and for every iterator k in the range [i, last), f(*k) == false INVOKE(f, INVOKE (proj, *k)) == false. The relative order of the elements in both groups is preserved. If the execution policy is of type parallel_task_policy the algorithm returns a future<> referring to this iterator.

template<typename ExPolicy, typename FwdIter, typename Pred, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type partition(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&pred, Proj &&proj = Proj())

Reorders the elements in the range [first, last) in such a way that all elements for which the predicate pred returns true precede the elements for which the predicate pred returns false. Relative order of the elements is not preserved.

The assignments in the parallel

partition algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most 2 * (last - first) swaps. Exactly last - first applications of the predicate and projection.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of partition requires Pred to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is an unary predicate for partitioning the source iterators. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type InIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel partition algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The partition algorithm returns a hpx::future<FwdIter> if the execution policy is of type parallel_task_policy and returns FwdIter otherwise. The partition algorithm returns the iterator to the first element of the second group.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename Pred, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in(FwdIter1), tag::out1(FwdIter2), tag::out2(FwdIter3)>>::type partition_copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest_true, FwdIter3 dest_false, Pred &&pred, Proj &&proj = Proj())

Copies the elements in the range, defined by [first, last), to two different ranges depending on the value returned by the predicate pred. The elements, that satisfy the predicate pred, are copied to the range beginning at dest_true. The rest of the elements are copied to the range beginning at dest_false. The order of the elements is preserved.

The assignments in the parallel

partition_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range for the elements that satisfy the predicate pred (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range for the elements that don’t satisfy the predicate pred (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of partition_copy requires Pred to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest_true: Refers to the beginning of the destination range for the elements that satisfy the predicate pred.
• dest_false: Refers to the beginning of the destination range for the elements that don’t satisfy the predicate pred.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is an unary predicate for partitioning the source iterators. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel partition_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The partition_copy algorithm returns a hpx::future<tagged_tuple<tag::in(InIter), tag::out1(OutIter1), tag::out2(OutIter2)> > if the execution policy is of type parallel_task_policy and returns tagged_tuple<tag::in(InIter), tag::out1(OutIter1), tag::out2(OutIter2)> otherwise. The partition_copy algorithm returns the tuple of the source iterator last, the destination iterator to the end of the dest_true range, and the destination iterator to the end of the dest_false range.

template<typename ExPolicy, typename FwdIterB, typename FwdIterE, typename T, typename F>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, T>::type>::type reduce(ExPolicy &&policy, FwdIterB first, FwdIterE last, T init, F &&f)

Returns GENERALIZED_SUM(f, init, *first, …, *(first + (last - first) - 1)).

The reduce operations in the parallel

reduce algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIterB: The type of the source begin iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIterE: The type of the source end iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.
• T: The type of the value to be used as initial (and intermediate) values (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&. The types

  Type1 Ret must be such that an object of type FwdIterB can be dereferenced and then implicitly converted to any of those types.
• init: The initial value for the generalized sum.

The reduce operations in the parallel copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

reduce and accumulate is that the behavior of reduce may be non-deterministic for non-associative or non-commutative binary predicate.

Return

The reduce algorithm returns a hpx::future<T> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns T otherwise. The reduce algorithm returns the result of the generalized sum over the elements given by the input range [first, last).

Note

GENERALIZED_SUM(op, a1, …, aN) is defined as follows:

• a1 when N is 1
• op(GENERALIZED_SUM(op, b1, …, bK), GENERALIZED_SUM(op, bM, …, bN)), where:
  • b1, …, bN may be any permutation of a1, …, aN and
  • 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIterB, typename FwdIterE, typename T>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, T>::type>::type reduce(ExPolicy &&policy, FwdIterB first, FwdIterE last, T init)

Returns GENERALIZED_SUM(+, init, *first, …, *(first + (last - first) - 1)).

The reduce operations in the parallel

reduce algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the operator+().

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIterB: The type of the source begin iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIterE: The type of the source end iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be used as initial (and intermediate) values (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• init: The initial value for the generalized sum.

The reduce operations in the parallel copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

reduce and accumulate is that the behavior of reduce may be non-deterministic for non-associative or non-commutative binary predicate.

Return

The reduce algorithm returns a hpx::future<T> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns T otherwise. The reduce algorithm returns the result of the generalized sum (applying operator+()) over the elements given by the input range [first, last).

Note

GENERALIZED_SUM(+, a1, …, aN) is defined as follows:

• a1 when N is 1
• op(GENERALIZED_SUM(+, b1, …, bK), GENERALIZED_SUM(+, bM, …, bN)), where:
  • b1, …, bN may be any permutation of a1, …, aN and
  • 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIterB, typename FwdIterE>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, typename std::iterator_traits<FwdIterB>::value_type>::type>::type reduce(ExPolicy &&policy, FwdIterB first, FwdIterE last)

Returns GENERALIZED_SUM(+, T(), *first, …, *(first + (last - first) - 1)).

The reduce operations in the parallel

reduce algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the operator+().

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIterB: The type of the source begin iterator used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIterE: The type of the source end iterator used (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.

The reduce operations in the parallel copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

reduce and accumulate is that the behavior of reduce may be non-deterministic for non-associative or non-commutative binary predicate.

Return

The reduce algorithm returns a hpx::future<T> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns T otherwise (where T is the value_type of FwdIterB). The reduce algorithm returns the result of the generalized sum (applying operator+()) over the elements given by the input range [first, last).

Note

The type of the initial value (and the result type) T is determined from the value_type of the used FwdIterB.

Note

GENERALIZED_SUM(+, a1, …, aN) is defined as follows:

• a1 when N is 1
• op(GENERALIZED_SUM(+, b1, …, bK), GENERALIZED_SUM(+, bM, …, bN)), where:
  • b1, …, bN may be any permutation of a1, …, aN and
  • 1 < K+1 = M <= N.

template<typename ExPolicy, typename RanIter, typename RanIter2, typename FwdIter1, typename FwdIter2, typename Compare = std::equal_to<typename std::iterator_traits<RanIter>::value_type>, typename Func = std::plus<typename std::iterator_traits<RanIter2>::value_type>>
util::detail::algorithm_result<ExPolicy, std::pair<FwdIter1, FwdIter2>>::type reduce_by_key(ExPolicy &&policy, RanIter key_first, RanIter key_last, RanIter2 values_first, FwdIter1 keys_output, FwdIter2 values_output, Compare &&comp = Compare(), Func &&func = Func())

Reduce by Key performs an inclusive scan reduction operation on elements supplied in key/value pairs. The algorithm produces a single output value for each set of equal consecutive keys in [key_first, key_last). the value being the GENERALIZED_NONCOMMUTATIVE_SUM(op, init, *first, …, *(first + (i - result))). for the run of consecutive matching keys. The number of keys supplied must match the number of values.

comp has to induce a strict weak ordering on the values.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• RanIter: The type of the key iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• RanIter2: The type of the value iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• FwdIter1: The type of the iterator representing the destination key range (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination value range (deduced). This iterator type must meet the requirements of an forward iterator.
• Compare: The type of the optional function/function object to use to compare keys (deduced). Assumed to be std::equal_to otherwise.
• Func: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• key_first: Refers to the beginning of the sequence of key elements the algorithm will be applied to.
• key_last: Refers to the end of the sequence of key elements the algorithm will be applied to.
• values_first: Refers to the beginning of the sequence of value elements the algorithm will be applied to.
• keys_output: Refers to the start output location for the keys produced by the algorithm.
• values_output: Refers to the start output location for the values produced by the algorithm.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• func: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&. The types

  Type1 Ret must be such that an object of type FwdIter can be dereferenced and then implicitly converted to any of those types.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The reduce_by_key algorithm returns a hpx::future<pair<Iter1,Iter2>> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns pair<Iter1,Iter2> otherwise.

template<typename ExPolicy, typename FwdIter, typename Pred, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type remove_if(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&pred, Proj &&proj = Proj())

Removes all elements satisfying specific criteria from the range [first, last) and returns a past-the-end iterator for the new end of the range. This version removes all elements for which predicate pred returns true.

The assignments in the parallel

remove_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate pred and the projection proj.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of remove_if requires Pred to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The remove_if algorithm returns the iterator to the new end of the range.

template<typename ExPolicy, typename FwdIter, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type remove(ExPolicy &&policy, FwdIter first, FwdIter last, T const &value, Proj &&proj = Proj())

Removes all elements satisfying specific criteria from the range [first, last) and returns a past-the-end iterator for the new end of the range. This version removes all elements that are equal to value.

The assignments in the parallel

remove algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the operator==() and the projection proj.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to remove (deduced). This value type must meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• value: Specifies the value of elements to remove.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The remove algorithm returns the iterator to the new end of the range.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type remove_copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, T const &val, Proj &&proj = Proj())

Copies the elements in the range, defined by [first, last), to another range beginning at dest. Copies only the elements for which the comparison operator returns false when compare to val. The order of the elements that are not removed is preserved.

Effects: Copies all the elements referred to by the iterator it in the range [first,last) for which the following corresponding conditions do not hold: INVOKE(proj, *it) == value

The assignments in the parallel

remove_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type that the result of dereferencing FwdIter1 is compared to.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• val: Value to be removed.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove_copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type remove_copy_if(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, F &&f, Proj &&proj = Proj())

Copies the elements in the range, defined by [first, last), to another range beginning at dest. Copies only the elements for which the predicate f returns false. The order of the elements that are not removed is preserved.

Effects: Copies all the elements referred to by the iterator it in the range [first,last) for which the following corresponding conditions do not hold: INVOKE(pred, INVOKE(proj, *it)) != false.

The assignments in the parallel

remove_copy_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the elements to be removed. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove_copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove_copy_if algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter, typename T1, typename T2, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type replace(ExPolicy &&policy, FwdIter first, FwdIter last, T1 const &old_value, T2 const &new_value, Proj &&proj = Proj())

Replaces all elements satisfying specific criteria with new_value in the range [first, last).

Effects: Substitutes elements referred by the iterator it in the range [first, last) with new_value, when the following corresponding conditions hold: INVOKE(proj, *it) == old_value

The assignments in the parallel

replace algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• T1: The type of the old value to replace (deduced).
• T2: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• old_value: Refers to the old value of the elements to replace.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise. It returns last.

template<typename ExPolicy, typename FwdIter, typename F, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type replace_if(ExPolicy &&policy, FwdIter first, FwdIter last, F &&f, T const &new_value, Proj &&proj = Proj())

Replaces all elements satisfying specific criteria (for which predicate f returns true) with new_value in the range [first, last).

Effects: Substitutes elements referred by the iterator it in the range [first, last) with new_value, when the following corresponding conditions hold: INVOKE(f, INVOKE(proj, *it)) != false

The assignments in the parallel

replace_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible. (deduced).
• T: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the elements which need to replaced. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. It returns last.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T1, typename T2, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type replace_copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, T1 const &old_value, T2 const &new_value, Proj &&proj = Proj())

Copies the all elements from the range [first, last) to another range beginning at dest replacing all elements satisfying a specific criteria with new_value.

Effects: Assigns to every iterator it in the range [result, result + (last - first)) either new_value or *(first + (it - result)) depending on whether the following corresponding condition holds: INVOKE(proj, *(first + (i - result))) == old_value

The assignments in the parallel

replace_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• T1: The type of the old value to replace (deduced).
• T2: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• old_value: Refers to the old value of the elements to replace.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename F, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type replace_copy_if(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, F &&f, T const &new_value, Proj &&proj = Proj())

Copies the all elements from the range [first, last) to another range beginning at dest replacing all elements satisfying a specific criteria with new_value.

Effects: Assigns to every iterator it in the range [result, result + (last - first)) either new_value or *(first + (it - result)) depending on whether the following corresponding condition holds: INVOKE(f, INVOKE(proj, *(first + (i - result)))) != false

The assignments in the parallel

replace_copy_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible. (deduced).
• T: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the elements which need to replaced. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace_copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_copy_if algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The replace_copy_if algorithm returns the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename BidirIter>
util::detail::algorithm_result<ExPolicy, BidirIter>::type reverse(ExPolicy &&policy, BidirIter first, BidirIter last)

Reverses the order of the elements in the range [first, last). Behaves as if applying std::iter_swap to every pair of iterators first+i, (last-i) - 1 for each non-negative i < (last-first)/2.

The assignments in the parallel

reverse algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Linear in the distance between first and last.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• BidirIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an bidirectional iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.

The assignments in the parallel reverse algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The reverse algorithm returns a hpx::future<BidirIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns BidirIter otherwise. It returns last.

template<typename ExPolicy, typename BidirIter, typename FwdIter>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(BidirIter), tag::out(FwdIter)>>::type reverse_copy(ExPolicy &&policy, BidirIter first, BidirIter last, FwdIter dest_first)

Copies the elements from the range [first, last) to another range beginning at dest_first in such a way that the elements in the new range are in reverse order. Behaves as if by executing the assignment *(dest_first + (last - first) - 1 - i) = *(first + i) once for each non-negative i < (last - first) If the source and destination ranges (that is, [first, last) and [dest_first, dest_first+(last-first)) respectively) overlap, the behavior is undefined.

The assignments in the parallel

reverse_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• BidirIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an bidirectional iterator.
• FwdIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest_first: Refers to the begin of the destination range.

The assignments in the parallel reverse_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The reverse_copy algorithm returns a hpx::future<tagged_pair<tag::in(BidirIter), tag::out(FwdIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(BidirIter), tag::out(FwdIter)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::begin(FwdIter), tag::end(FwdIter)>>::type rotate(ExPolicy &&policy, FwdIter first, FwdIter new_first, FwdIter last)

Performs a left rotation on a range of elements. Specifically, rotate swaps the elements in the range [first, last) in such a way that the element new_first becomes the first element of the new range and new_first - 1 becomes the last element.

The assignments in the parallel

rotate algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Linear in the distance between first and last.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• new_first: Refers to the element that should appear at the beginning of the rotated range.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.

The assignments in the parallel rotate algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The type of dereferenced FwdIter must meet the requirements of MoveAssignable and MoveConstructible.

Return

The rotate algorithm returns a hpx::future<tagged_pair<tag::begin(FwdIter), tag::end(FwdIter)> > if the execution policy is of type parallel_task_policy and returns tagged_pair<tag::begin(FwdIter), tag::end(FwdIter)> otherwise. The rotate algorithm returns the iterator equal to pair(first + (last - new_first), last).

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type rotate_copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 new_first, FwdIter1 last, FwdIter2 dest_first)

Copies the elements from the range [first, last), to another range beginning at dest_first in such a way, that the element new_first becomes the first element of the new range and new_first - 1 becomes the last element.

The assignments in the parallel

rotate_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an bidirectional iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• new_first: Refers to the element that should appear at the beginning of the rotated range.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest_first: Refers to the begin of the destination range.

The assignments in the parallel rotate_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The rotate_copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The rotate_copy algorithm returns the output iterator to the element past the last element copied.

template<typename ExPolicy, typename FwdIter, typename FwdIter2, typename Pred = detail::equal_to, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type search(ExPolicy &&policy, FwdIter first, FwdIter last, FwdIter2 s_first, FwdIter2 s_last, Pred &&op = Pred(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Searches the range [first, last) for any elements in the range [s_first, s_last). Uses a provided predicate to compare elements.

The comparison operations in the parallel

search algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most (S*N) comparisons where S = distance(s_first, s_last) and N = distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an input iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of adjacent_find requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj1: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of type dereferenced FwdIter.
• Proj2: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of type dereferenced FwdIter2.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• s_first: Refers to the beginning of the sequence of elements the algorithm will be searching for.
• s_last: Refers to the end of the sequence of elements of the algorithm will be searching for.
• op: Refers to the binary predicate which returns true if the elements should be treated as equal. the signature of the function should be equivalent to

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter1 as a projection operation before the actual predicate is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter2 as a projection operation before the actual predicate is invoked.

The comparison operations in the parallel search algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The search algorithm returns a hpx::future<FwdIter> if the execution policy is of type task_execution_policy and returns FwdIter otherwise. The search algorithm returns an iterator to the beginning of the first subsequence [s_first, s_last) in range [first, last). If the length of the subsequence [s_first, s_last) is greater than the length of the range [first, last), last is returned. Additionally if the size of the subsequence is empty first is returned. If no subsequence is found, last is returned.

template<typename ExPolicy, typename FwdIter, typename FwdIter2, typename Pred = detail::equal_to, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type search_n(ExPolicy &&policy, FwdIter first, std::size_t count, FwdIter2 s_first, FwdIter2 s_last, Pred &&op = Pred(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Searches the range [first, last) for any elements in the range [s_first, s_last). Uses a provided predicate to compare elements.

The comparison operations in the parallel

search_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most (S*N) comparisons where S = distance(s_first, s_last) and N = count.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used for the first range (deduced). This iterator type must meet the requirements of an input iterator.
• FwdIter2: The type of the source iterators used for the second range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of adjacent_find requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• count: Refers to the range of elements of the first range the algorithm will be applied to.
• s_first: Refers to the beginning of the sequence of elements the algorithm will be searching for.
• s_last: Refers to the end of the sequence of elements of the algorithm will be searching for.
• op: Refers to the binary predicate which returns true if the elements should be treated as equal. the signature of the function should be equivalent to

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter1 as a projection operation before the actual predicate is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced FwdIter2 as a projection operation before the actual predicate is invoked.

The comparison operations in the parallel search_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The search_n algorithm returns a hpx::future<FwdIter> if the execution policy is of type task_execution_policy and returns FwdIter otherwise. The search_n algorithm returns an iterator to the beginning of the last subsequence [s_first, s_last) in range [first, first+count). If the length of the subsequence [s_first, s_last) is greater than the length of the range [first, first+count), first is returned. Additionally if the size of the subsequence is empty or no subsequence is found, first is also returned.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter3>::type>::type set_difference(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, FwdIter3 dest, Pred &&op = Pred())

Constructs a sorted range beginning at dest consisting of all elements present in the range [first1, last1) and not present in the range [first2, last2). This algorithm expects both input ranges to be sorted with the given binary predicate f.

Equivalent elements are treated individually, that is, if some element is found

m times in [first1, last1) and n times in [first2, last2), it will be copied to dest exactly std::max(m-n, 0) times. The resulting range cannot overlap with either of the input ranges.

Note

Complexity: At most 2*(N1 + N2 - 1) comparisons, where N1 is the length of the first sequence and N2 is the length of the second sequence.

The resulting range cannot overlap with either of the input ranges.

The application of function objects in parallel algorithm invoked with a sequential execution policy object execute in sequential order in the calling thread (

sequenced_policy) or in a single new thread spawned from the current thread (for sequenced_task_policy).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter1: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of set_difference requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type InIter can be dereferenced and then implicitly converted to Type1

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The set_difference algorithm returns a hpx::future<FwdIter3> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter3 otherwise. The set_difference algorithm returns the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter3>::type>::type set_intersection(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, FwdIter3 dest, Pred &&op = Pred())

Constructs a sorted range beginning at dest consisting of all elements present in both sorted ranges [first1, last1) and [first2, last2). This algorithm expects both input ranges to be sorted with the given binary predicate f.

If some element is found

m times in [first1, last1) and n times in [first2, last2), the first std::min(m, n) elements will be copied from the first range to the destination range. The order of equivalent elements is preserved. The resulting range cannot overlap with either of the input ranges.

Note

Complexity: At most 2*(N1 + N2 - 1) comparisons, where N1 is the length of the first sequence and N2 is the length of the second sequence.

The resulting range cannot overlap with either of the input ranges.

The application of function objects in parallel algorithm invoked with a sequential execution policy object execute in sequential order in the calling thread (

sequenced_policy) or in a single new thread spawned from the current thread (for sequenced_task_policy).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter1: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of set_intersection requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type InIter can be dereferenced and then implicitly converted to Type1

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The set_intersection algorithm returns a hpx::future<FwdIter3> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter3 otherwise. The set_intersection algorithm returns the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter3>::type>::type set_symmetric_difference(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, FwdIter3 dest, Pred &&op = Pred())

Constructs a sorted range beginning at dest consisting of all elements present in either of the sorted ranges [first1, last1) and [first2, last2), but not in both of them are copied to the range beginning at dest. The resulting range is also sorted. This algorithm expects both input ranges to be sorted with the given binary predicate f.

If some element is found

m times in [first1, last1) and n times in [first2, last2), it will be copied to dest exactly std::abs(m-n) times. If m>n, then the last m-n of those elements are copied from [first1,last1), otherwise the last n-m elements are copied from [first2,last2). The resulting range cannot overlap with either of the input ranges.

Note

Complexity: At most 2*(N1 + N2 - 1) comparisons, where N1 is the length of the first sequence and N2 is the length of the second sequence.

The resulting range cannot overlap with either of the input ranges.

The application of function objects in parallel algorithm invoked with a sequential execution policy object execute in sequential order in the calling thread (

sequenced_policy) or in a single new thread spawned from the current thread (for sequenced_task_policy).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter1: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of set_symmetric_difference requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type InIter can be dereferenced and then implicitly converted to Type1

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The set_symmetric_difference algorithm returns a hpx::future<FwdIter3> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter3 otherwise. The set_symmetric_difference algorithm returns the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename Pred = detail::less>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter3>::type>::type set_union(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, FwdIter3 dest, Pred &&op = Pred())

Constructs a sorted range beginning at dest consisting of all elements present in one or both sorted ranges [first1, last1) and [first2, last2). This algorithm expects both input ranges to be sorted with the given binary predicate f.

If some element is found

m times in [first1, last1) and n times in [first2, last2), then all m elements will be copied from [first1, last1) to dest, preserving order, and then exactly std::max(n-m, 0) elements will be copied from [first2, last2) to dest, also preserving order.

Note

Complexity: At most 2*(N1 + N2 - 1) comparisons, where N1 is the length of the first sequence and N2 is the length of the second sequence.

The resulting range cannot overlap with either of the input ranges.

The application of function objects in parallel algorithm invoked with a sequential execution policy object execute in sequential order in the calling thread (

sequenced_policy) or in a single new thread spawned from the current thread (for sequenced_task_policy).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• FwdIter1: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators used (deduced) representing the first sequence. This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• Op: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of set_union requires Pred to meet the requirements of CopyConstructible. This defaults to std::less<>

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the sequence of elements of the first range the algorithm will be applied to.
• last1: Refers to the end of the sequence of elements of the first range the algorithm will be applied to.
• first2: Refers to the beginning of the sequence of elements of the second range the algorithm will be applied to.
• last2: Refers to the end of the sequence of elements of the second range the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type InIter can be dereferenced and then implicitly converted to Type1

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The set_union algorithm returns a hpx::future<FwdIter3> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter3 otherwise. The set_union algorithm returns the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename RandomIt, typename Proj = util::projection_identity, typename Compare = detail::less>
util::detail::algorithm_result<ExPolicy, RandomIt>::type sort(ExPolicy &&policy, RandomIt first, RandomIt last, Compare &&comp = Compare(), Proj &&proj = Proj())

Sorts the elements in the range [first, last) in ascending order. The order of equal elements is not guaranteed to be preserved. The function uses the given comparison function object comp (defaults to using operator<()).

A sequence is sorted with respect to a comparator

comp and a projection proj if for every iterator i pointing to the sequence and every non-negative integer n such that i + n is a valid iterator pointing to an element of the sequence, and INVOKE(comp, INVOKE(proj, *(i + n)), INVOKE(proj, *i)) == false.

Note

Complexity: O(Nlog(N)), where N = std::distance(first, last) comparisons.

comp has to induce a strict weak ordering on the values.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• Iter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• Comp: The type of the function/function object to use (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• proj: Specifies the function (or function object) which will be invoked for each pair of elements as a projection operation before the actual predicate comp is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The sort algorithm returns a hpx::future<RandomIt> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns RandomIt otherwise. The algorithm returns an iterator pointing to the first element after the last element in the input sequence.

template<typename ExPolicy, typename KeyIter, typename ValueIter, typename Compare = detail::less>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in1(KeyIter), tag::in2(ValueIter)>>::type sort_by_key(ExPolicy &&policy, KeyIter key_first, KeyIter key_last, ValueIter value_first, Compare &&comp = Compare())

Sorts one range of data using keys supplied in another range. The key elements in the range [key_first, key_last) are sorted in ascending order with the corresponding elements in the value range moved to follow the sorted order. The algorithm is not stable, the order of equal elements is not guaranteed to be preserved. The function uses the given comparison function object comp (defaults to using operator<()).

A sequence is sorted with respect to a comparator

comp and a projection proj if for every iterator i pointing to the sequence and every non-negative integer n such that i + n is a valid iterator pointing to an element of the sequence, and INVOKE(comp, INVOKE(proj, *(i + n)), INVOKE(proj, *i)) == false.

Note

Complexity: O(Nlog(N)), where N = std::distance(first, last) comparisons.

comp has to induce a strict weak ordering on the values.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• KeyIter: The type of the key iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• ValueIter: The type of the value iterators used (deduced). This iterator type must meet the requirements of a random access iterator.
• Comp: The type of the function/function object to use (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• key_first: Refers to the beginning of the sequence of key elements the algorithm will be applied to.
• key_last: Refers to the end of the sequence of key elements the algorithm will be applied to.
• value_first: Refers to the beginning of the sequence of value elements the algorithm will be applied to, the range of elements must match [key_first, key_last)
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The sort_by-key algorithm returns a hpx::future<tagged_pair<tag::in1(KeyIter>, tag::in2(ValueIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns otherwise. The algorithm returns a pair holding an iterator pointing to the first element after the last element in the input key sequence and an iterator pointing to the first element after the last element in the input value sequence.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type swap_ranges(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2)

Exchanges elements between range [first1, last1) and another range starting at first2.

The swap operations in the parallel

swap_ranges algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Linear in the distance between first1 and last1

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the swap operations.
• FwdIter1: The type of the first range of iterators to swap (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the second range of iterators to swap (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the first sequence of elements the algorithm will be applied to.
• last1: Refers to the end of the first sequence of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second sequence of elements the algorithm will be applied to.

The swap operations in the parallel swap_ranges algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The swap_ranges algorithm returns a hpx::future<FwdIter2> if the execution policy is of type parallel_task_policy and returns FwdIter2 otherwise. The swap_ranges algorithm returns iterator to the element past the last element exchanged in the range beginning with first2.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type transform(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, F &&f, Proj &&proj = Proj())

Applies the given function f to the range [first, last) and stores the result in another range, beginning at dest.

The invocations of

f in the parallel transform algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly last - first applications of f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type &a);
  

  The signature does not need to have const&. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type. The type Ret must be such that an object of type FwdIter2 can be dereferenced and assigned a value of type Ret.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel transform algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The transform algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The transform algorithm returns a tuple holding an iterator referring to the first element after the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename F, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in1(FwdIter1), tag::in2(FwdIter2), tag::out(FwdIter3)>>::type transform(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter3 dest, F &&f, Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Applies the given function f to pairs of elements from two ranges: one defined by [first1, last1) and the other beginning at first2, and stores the result in another range, beginning at dest.

The invocations of

f in the parallel transform algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly last - first applications of f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• FwdIter1: The type of the source iterators for the first range used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators for the second range used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj1: The type of an optional projection function to be used for elements of the first sequence. This defaults to util::projection_identity
• Proj2: The type of an optional projection function to be used for elements of the second sequence. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the first sequence of elements the algorithm will be applied to.
• last1: Refers to the end of the first sequence of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively. The type Ret must be such that an object of type FwdIter3 can be dereferenced and assigned a value of type Ret.
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of the first sequence as a projection operation before the actual predicate f is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of the second sequence as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel transform algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The transform algorithm returns a hpx::future<tagged_tuple<tag::in1(FwdIter1), tag::in2(FwdIter2), tag::out(FwdIter3)> > if the execution policy is of type parallel_task_policy and returns tagged_tuple<tag::in1(FwdIter1), tag::in2(FwdIter2), tag::out(FwdIter3)> otherwise. The transform algorithm returns a tuple holding an iterator referring to the first element after the first input sequence, an iterator referring to the first element after the second input sequence, and the output iterator referring to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename FwdIter3, typename F, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in1(FwdIter1), tag::in2(FwdIter2), tag::out(FwdIter3)>>::type transform(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, FwdIter2 last2, FwdIter3 dest, F &&f, Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Applies the given function f to pairs of elements from two ranges: one defined by [first1, last1) and the other beginning at first2, and stores the result in another range, beginning at dest.

The invocations of

f in the parallel transform algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly min(last2-first2, last1-first1) applications of f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• FwdIter1: The type of the source iterators for the first range used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the source iterators for the second range used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj1: The type of an optional projection function to be used for elements of the first sequence. This defaults to util::projection_identity
• Proj2: The type of an optional projection function to be used for elements of the second sequence. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the first sequence of elements the algorithm will be applied to.
• last1: Refers to the end of the first sequence of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second sequence of elements the algorithm will be applied to.
• last2: Refers to the end of the second sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively. The type Ret must be such that an object of type FwdIter3 can be dereferenced and assigned a value of type Ret.
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of the first sequence as a projection operation before the actual predicate f is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of the second sequence as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel transform algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The algorithm will invoke the binary predicate until it reaches the end of the shorter of the two given input sequences

Return

The transform algorithm returns a hpx::future<tagged_tuple<tag::in1(FwdIter1), tag::in2(FwdIter2), tag::out(FwdIter3)> > if the execution policy is of type parallel_task_policy and returns tagged_tuple<tag::in1(FwdIter1), tag::in2(FwdIter2), tag::out(FwdIter3)> otherwise. The transform algorithm returns a tuple holding an iterator referring to the first element after the first input sequence, an iterator referring to the first element after the second input sequence, and the output iterator referring to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T, typename Op, typename Conv>
util::detail::algorithm_result<ExPolicy, FwdIter2>::type transform_exclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, T init, Op &&op, Conv &&conv)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(binary_op, init, conv(*first), …, conv(*(first + (i - result) - 1))).

The reduce operations in the parallel

transform_exclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicates op and conv.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• Conv: The type of the unary function object used for the conversion operation.
• T: The type of the value to be used as initial (and intermediate) values (deduced).
• Op: The type of the binary function object used for the reduction operation.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• conv: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is a unary predicate. The signature of this predicate should be equivalent to:

  R fun(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type. The type R must be such that an object of this type can be implicitly converted to T.
• init: The initial value for the generalized sum.
• op: Specifies the function (or function object) which will be invoked for each of the values of the input sequence. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Ret must be such that an object of a type as given by the input sequence can be implicitly converted to any of those types.

The reduce operations in the parallel transform_exclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Neither

conv nor op shall invalidate iterators or subranges, or modify elements in the ranges [first,last) or [result,result + (last - first)).

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The transform_exclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aN) is defined as:

• a1 when N is 1
• op(GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aK), GENERALIZED_NONCOMMUTATIVE_SUM(op, aM, …, aN) where 1 < K+1 = M <= N.

The behavior of transform_exclusive_scan may be non-deterministic for a non-associative predicate.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Op, typename Conv, typename T>
util::detail::algorithm_result<ExPolicy, FwdIter2>::type transform_inclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, Op &&op, Conv &&conv, T init)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(op, init, conv(*first), …, conv(*(first + (i - result)))).

The reduce operations in the parallel

transform_inclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• Conv: The type of the unary function object used for the conversion operation.
• T: The type of the value to be used as initial (and intermediate) values (deduced).
• Op: The type of the binary function object used for the reduction operation.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• conv: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is a unary predicate. The signature of this predicate should be equivalent to:

  R fun(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type. The type R must be such that an object of this type can be implicitly converted to T.
• init: The initial value for the generalized sum.
• op: Specifies the function (or function object) which will be invoked for each of the values of the input sequence. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Ret must be such that an object of a type as given by the input sequence can be implicitly converted to any of those types.

The reduce operations in the parallel transform_inclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Neither

conv nor op shall invalidate iterators or subranges, or modify elements in the ranges [first,last) or [result,result + (last - first)).

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The transform_inclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aN) is defined as:

• a1 when N is 1
• op(GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aK), GENERALIZED_NONCOMMUTATIVE_SUM(op, aM, …, aN)) where 1 < K+1 = M <= N.

The difference between exclusive_scan and transform_inclusive_scan is that transform_inclusive_scan includes the ith input element in the ith sum. If op is not mathematically associative, the behavior of transform_inclusive_scan may be non-deterministic.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Conv, typename Op>
util::detail::algorithm_result<ExPolicy, FwdIter2>::type transform_inclusive_scan(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, Op &&op, Conv &&conv)

Assigns through each iterator i in [result, result + (last - first)) the value of GENERALIZED_NONCOMMUTATIVE_SUM(op, conv(*first), …, conv(*(first + (i - result)))).

The reduce operations in the parallel

transform_inclusive_scan algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• Conv: The type of the unary function object used for the conversion operation.
• T: The type of the value to be used as initial (and intermediate) values (deduced).
• Op: The type of the binary function object used for the reduction operation.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• conv: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is a unary predicate. The signature of this predicate should be equivalent to:

  R fun(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type. The type R must be such that an object of this type can be implicitly converted to T.
• op: Specifies the function (or function object) which will be invoked for each of the values of the input sequence. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Ret must be such that an object of a type as given by the input sequence can be implicitly converted to any of those types.

The reduce operations in the parallel transform_inclusive_scan algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Neither

conv nor op shall invalidate iterators or subranges, or modify elements in the ranges [first,last) or [result,result + (last - first)).

Return

The copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The transform_inclusive_scan algorithm returns the output iterator to the element in the destination range, one past the last element copied.

Note

GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aN) is defined as:

• a1 when N is 1
• op(GENERALIZED_NONCOMMUTATIVE_SUM(op, a1, …, aK), GENERALIZED_NONCOMMUTATIVE_SUM(op, aM, …, aN)) where 1 < K+1 = M <= N.

The difference between exclusive_scan and transform_inclusive_scan is that transform_inclusive_scan includes the ith input element in the ith sum.

template<typename ExPolicy, typename FwdIter, typename T, typename Reduce, typename Convert>
util::detail::algorithm_result<ExPolicy, T>::type transform_reduce(ExPolicy &&policy, FwdIter first, FwdIter last, T init, Reduce &&red_op, Convert &&conv_op)

Returns GENERALIZED_SUM(red_op, init, conv_op(*first), …, conv_op(*(first + (last - first) - 1))).

The reduce operations in the parallel

transform_reduce algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicates red_op and conv_op.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.
• T: The type of the value to be used as initial (and intermediate) values (deduced).
• Reduce: The type of the binary function object used for the reduction operation.
• Convert: The type of the unary function object used to transform the elements of the input sequence before invoking the reduce function.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• conv_op: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is a unary predicate. The signature of this predicate should be equivalent to:

  R fun(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type. The type R must be such that an object of this type can be implicitly converted to T.
• init: The initial value for the generalized sum.
• red_op: Specifies the function (or function object) which will be invoked for each of the values returned from the invocation of conv_op. This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1, Type2, and Ret must be such that an object of a type as returned from conv_op can be implicitly converted to any of those types.

The reduce operations in the parallel transform_reduce algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

The difference between

transform_reduce and accumulate is that the behavior of transform_reduce may be non-deterministic for non-associative or non-commutative binary predicate.

Return

The transform_reduce algorithm returns a hpx::future<T> if the execution policy is of type parallel_task_policy and returns T otherwise. The transform_reduce algorithm returns the result of the generalized sum over the values returned from conv_op when applied to the elements given by the input range [first, last).

Note

GENERALIZED_SUM(op, a1, …, aN) is defined as follows:

• a1 when N is 1
• op(GENERALIZED_SUM(op, b1, …, bK), GENERALIZED_SUM(op, bM, …, bN)), where:
  • b1, …, bN may be any permutation of a1, …, aN and
  • 1 < K+1 = M <= N.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T>
util::detail::algorithm_result<ExPolicy, T>::type transform_reduce(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, T init)

Returns the result of accumulating init with the inner products of the pairs formed by the elements of two ranges starting at first1 and first2.

The operations in the parallel

transform_reduce algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op2.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the first source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the second source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be used as return) values (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the first sequence of elements the result will be calculated with.
• last1: Refers to the end of the first sequence of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second sequence of elements the result will be calculated with.
• init: The initial value for the sum.

The operations in the parallel transform_reduce algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The transform_reduce algorithm returns a hpx::future<T> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns T otherwise.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename T, typename Reduce, typename Convert>
util::detail::algorithm_result<ExPolicy, T>::type transform_reduce(ExPolicy &&policy, FwdIter1 first1, FwdIter1 last1, FwdIter2 first2, T init, Reduce &&red_op, Convert &&conv_op)

Returns the result of accumulating init with the inner products of the pairs formed by the elements of two ranges starting at first1 and first2.

The operations in the parallel

transform_reduce algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: O(last - first) applications of the predicate op2.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the first source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the second source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be used as return) values (deduced).
• Reduce: The type of the binary function object used for the multiplication operation.
• Convert: The type of the unary function object used to transform the elements of the input sequence before invoking the reduce function.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first1: Refers to the beginning of the first sequence of elements the result will be calculated with.
• last1: Refers to the end of the first sequence of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second sequence of elements the result will be calculated with.
• init: The initial value for the sum.
• red_op: Specifies the function (or function object) which will be invoked for the initial value and each of the return values of op2. This is a binary predicate. The signature of this predicate should be equivalent to should be equivalent to:

  Ret fun(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Ret must be such that it can be implicitly converted to a type of T.
• conv_op: Specifies the function (or function object) which will be invoked for each of the input values of the sequence. This is a binary predicate. The signature of this predicate should be equivalent to

  Ret fun(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Ret must be such that it can be implicitly converted to an object for the second argument type of op1.

The operations in the parallel transform_reduce algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The transform_reduce algorithm returns a hpx::future<T> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns T otherwise.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type uninitialized_copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest)

Copies the elements in the range, defined by [first, last), to an uninitialized memory area beginning at dest. If an exception is thrown during the copy operation, the function has no effects.

The assignments in the parallel

uninitialized_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of a forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel uninitialized_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_copy algorithm returns a hpx::future<FwdIter2>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The uninitialized_copy algorithm returns the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter1, typename Size, typename FwdIter2>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy, FwdIter2>::type>::type uninitialized_copy_n(ExPolicy &&policy, FwdIter1 first, Size count, FwdIter2 dest)

Copies the elements in the range [first, first + count), starting from first and proceeding to first + count - 1., to another range beginning at dest. If an exception is thrown during the copy operation, the function has no effects.

The assignments in the parallel

uninitialized_copy_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, if count > 0, no assignments otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an input iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of a forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel uninitialized_copy_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_copy_n algorithm returns a hpx::future<FwdIter2> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The uninitialized_copy_n algorithm returns the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename FwdIter>
util::detail::algorithm_result<ExPolicy>::type uninitialized_default_construct(ExPolicy &&policy, FwdIter first, FwdIter last)

Constructs objects of type typename iterator_traits<ForwardIt>::value_type in the uninitialized storage designated by the range [first, last) by default-initialization. If an exception is thrown during the initialization, the function has no effects.

The assignments in the parallel

uninitialized_default_construct algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.

The assignments in the parallel uninitialized_default_construct algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_default_construct algorithm returns a hpx::future<void>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename ExPolicy, typename FwdIter, typename Size>
util::detail::algorithm_result<ExPolicy, FwdIter>::type uninitialized_default_construct_n(ExPolicy &&policy, FwdIter first, Size count)

Constructs objects of type typename iterator_traits<ForwardIt>::value_type in the uninitialized storage designated by the range [first, first + count) by default-initialization. If an exception is thrown during the initialization, the function has no effects.

The assignments in the parallel

uninitialized_default_construct_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, if count > 0, no assignments otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Size: The type of the argument specifying the number of elements to apply f to.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.

The assignments in the parallel uninitialized_default_construct_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_default_construct_n algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The uninitialized_default_construct_n algorithm returns the iterator to the element in the source range, one past the last element constructed.

template<typename ExPolicy, typename FwdIter, typename T>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy>::type>::type uninitialized_fill(ExPolicy &&policy, FwdIter first, FwdIter last, T const &value)

Copies the given value to an uninitialized memory area, defined by the range [first, last). If an exception is thrown during the initialization, the function has no effects.

The initializations in the parallel

uninitialized_fill algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Linear in the distance between first and last

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• T: The type of the value to be assigned (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• value: The value to be assigned.

The initializations in the parallel uninitialized_fill algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_fill algorithm returns a hpx::future<void>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns nothing otherwise.

template<typename ExPolicy, typename FwdIter, typename Size, typename T>
std::enable_if<execution::is_execution_policy<ExPolicy>::value, typename util::detail::algorithm_result<ExPolicy>::type>::type uninitialized_fill_n(ExPolicy &&policy, FwdIter first, Size count, T const &value)

Copies the given value value to the first count elements in an uninitialized memory area beginning at first. If an exception is thrown during the initialization, the function has no effects.

The initializations in the parallel

uninitialized_fill_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, if count > 0, no assignments otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of a forward iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• T: The type of the value to be assigned (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• value: The value to be assigned.

The initializations in the parallel uninitialized_fill_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_fill_n algorithm returns a hpx::future<void>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns nothing otherwise.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2>
util::detail::algorithm_result<ExPolicy, FwdIter2>::type uninitialized_move(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest)

Moves the elements in the range, defined by [first, last), to an uninitialized memory area beginning at dest. If an exception is thrown during the initialization, some objects in [first, last) are left in a valid but unspecified state.

The assignments in the parallel

uninitialized_move algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first move operations.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of a forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel uninitialized_move algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_move algorithm returns a hpx::future<FwdIter2>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter2 otherwise. The uninitialized_move algorithm returns the output iterator to the element in the destination range, one past the last element moved.

template<typename ExPolicy, typename FwdIter1, typename Size, typename FwdIter2>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type uninitialized_move_n(ExPolicy &&policy, FwdIter1 first, Size count, FwdIter2 dest)

Moves the elements in the range [first, first + count), starting from first and proceeding to first + count - 1., to another range beginning at dest. If an exception is thrown during the initialization, some objects in [first, first + count) are left in a valid but unspecified state.

The assignments in the parallel

uninitialized_move_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count movements, if count > 0, no move operations otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of a forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel uninitialized_move_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_move_n algorithm returns a hpx::future<std::pair<FwdIter1, FwdIter2>> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns std::pair<FwdIter1, FwdIter2> otherwise. The uninitialized_move_n algorithm returns the pair of the input iterator to the element past in the source range and an output iterator to the element in the destination range, one past the last element moved.

template<typename ExPolicy, typename FwdIter>
util::detail::algorithm_result<ExPolicy>::type uninitialized_value_construct(ExPolicy &&policy, FwdIter first, FwdIter last)

Constructs objects of type typename iterator_traits<ForwardIt>::value_type in the uninitialized storage designated by the range [first, last) by default-initialization. If an exception is thrown during the initialization, the function has no effects.

The assignments in the parallel

uninitialized_value_construct algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.

The assignments in the parallel uninitialized_value_construct algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_value_construct algorithm returns a hpx::future<void>, if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename ExPolicy, typename FwdIter, typename Size>
util::detail::algorithm_result<ExPolicy, FwdIter>::type uninitialized_value_construct_n(ExPolicy &&policy, FwdIter first, Size count)

Constructs objects of type typename iterator_traits<ForwardIt>::value_type in the uninitialized storage designated by the range [first, first + count) by default-initialization. If an exception is thrown during the initialization, the function has no effects.

The assignments in the parallel

uninitialized_value_construct_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, if count > 0, no assignments otherwise.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Size: The type of the argument specifying the number of elements to apply f to.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.

The assignments in the parallel uninitialized_value_construct_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The uninitialized_value_construct_n algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The uninitialized_value_construct_n algorithm returns the iterator to the element in the source range, one past the last element constructed.

template<typename ExPolicy, typename FwdIter, typename Pred = detail::equal_to, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, FwdIter>::type unique(ExPolicy &&policy, FwdIter first, FwdIter last, Pred &&pred = Pred(), Proj &&proj = Proj())

Eliminates all but the first element from every consecutive group of equivalent elements from the range [first, last) and returns a past-the-end iterator for the new logical end of the range.

The assignments in the parallel

unique algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first - 1 applications of the predicate pred and no more than twice as many applications of the projection proj.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of unique requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is an binary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter can be dereferenced and then implicitly converted to both Type1 and Type2
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel unique algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The unique algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The unique algorithm returns the iterator to the new end of the range.

template<typename ExPolicy, typename FwdIter1, typename FwdIter2, typename Pred = detail::equal_to, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)>>::type unique_copy(ExPolicy &&policy, FwdIter1 first, FwdIter1 last, FwdIter2 dest, Pred &&pred = Pred(), Proj &&proj = Proj())

Copies the elements from the range [first, last), to another range beginning at dest in such a way that there are no consecutive equal elements. Only the first element of each group of equal elements is copied.

The assignments in the parallel

unique_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first - 1 applications of the predicate pred and no more than twice as many applications of the projection proj

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• FwdIter1: The type of the source iterators used (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of unique_copy requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is an binary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a, const Type &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel unique_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The unique_copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The unique_copy algorithm returns the pair of the source iterator to last, and the destination iterator to the end of the dest range.

template<typename ExPolicy, typename Rng, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type none_of(ExPolicy &&policy, Rng &&rng, F &&f, Proj &&proj = Proj())

Checks if unary predicate f returns true for no elements in the range rng.

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most std::distance(begin(rng), end(rng)) applications of the predicate f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of none_of requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The none_of algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The none_of algorithm returns true if the unary predicate f returns true for no elements in the range, false otherwise. It returns true if the range is empty.

template<typename ExPolicy, typename Rng, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type any_of(ExPolicy &&policy, Rng &&rng, F &&f, Proj &&proj = Proj())

Checks if unary predicate f returns true for at least one element in the range rng.

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most std::distance(begin(rng), end(rng)) applications of the predicate f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of none_of requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The any_of algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The any_of algorithm returns true if the unary predicate f returns true for at least one element in the range, false otherwise. It returns false if the range is empty.

template<typename ExPolicy, typename Rng, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type all_of(ExPolicy &&policy, Rng &&rng, F &&f, Proj &&proj = Proj())

Checks if unary predicate f returns true for all elements in the range rng.

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most std::distance(begin(rng), end(rng)) applications of the predicate f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of none_of requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The all_of algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The all_of algorithm returns true if the unary predicate f returns true for all elements in the range, false otherwise. It returns true if the range is empty.

template<typename ExPolicy, typename Rng, typename OutIter>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_traits<Rng>::iterator_type), tag::out(OutIter)>>::type copy(ExPolicy &&policy, Rng &&rng, OutIter dest)

Copies the elements in the range rng to another range beginning at dest.

The assignments in the parallel

copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly std::distance(begin(rng), end(rng)) assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.

The assignments in the parallel copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The copy algorithm returns a hpx::future<tagged_pair<tag::in(iterator_t<Rng>), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(iterator_t<Rng>), tag::out(FwdIter2)> otherwise. The copy algorithm returns the pair of the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename OutIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_traits<Rng>::iterator_type), tag::out(OutIter)>>::type copy_if(ExPolicy &&policy, Rng &&rng, OutIter dest, F &&f, Proj &&proj = Proj())

Copies the elements in the range rng to another range beginning at dest. Copies only the elements for which the predicate f returns true. The order of the elements that are not removed is preserved.

The assignments in the parallel

copy_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than std::distance(begin(rng), end(rng)) assignments, exactly std::distance(begin(rng), end(rng)) applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type InIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The copy_if algorithm returns a hpx::future<tagged_pair<tag::in(iterator_t<Rng>), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(iterator_t<Rng>), tag::out(FwdIter2)> otherwise. The copy_if algorithm returns the pair of the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename std::iterator_traits<typename hpx::traits::range_traits<Rng>::iterator_type>::difference_type>::type count(ExPolicy &&policy, Rng &&rng, T const &value, Proj &&proj = Proj())

Returns the number of elements in the range [first, last) satisfying a specific criteria. This version counts the elements that are equal to the given value.

The comparisons in the parallel

count algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first comparisons.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the comparisons.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• T: The type of the value to search for (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• value: The value to search for.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

Note

The comparisons in the parallel count algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The count algorithm returns a hpx::future<difference_type> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by std::iterator_traits<FwdIter>::difference_type. The count algorithm returns the number of elements satisfying the given criteria.

template<typename ExPolicy, typename Rng, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename std::iterator_traits<typename hpx::traits::range_traits<Rng>::iterator_type>::difference_type>::type count_if(ExPolicy &&policy, Rng &&rng, F &&f, Proj &&proj = Proj())

Returns the number of elements in the range [first, last) satisfying a specific criteria. This version counts elements for which predicate f returns true.

Note

Complexity: Performs exactly last - first applications of the predicate.

Note

The assignments in the parallel count_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

The assignments in the parallel count_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The count_if algorithm returns hpx::future<difference_type> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by std::iterator_traits<FwdIter>::difference_type. The count algorithm returns the number of elements satisfying the given criteria.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the comparisons.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of count_if requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

template<typename ExPolicy, typename Rng, typename T>
util::detail::algorithm_result<ExPolicy>::type fill(ExPolicy &&policy, Rng &&rng, T value)

Assigns the given value to the elements in the range [first, last).

The comparisons in the parallel

fill algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• T: The type of the value to be assigned (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• value: The value to be assigned.

The comparisons in the parallel fill algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The fill algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by void.

template<typename ExPolicy, typename Rng, typename Size, typename T>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_traits<Rng>::iterator_type>::type fill_n(ExPolicy &&policy, Rng &rng, Size count, T value)

Assigns the given value value to the first count elements in the range beginning at first if count > 0. Does nothing otherwise.

The comparisons in the parallel

fill_n algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly count assignments, for count > 0.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Size: The type of the argument specifying the number of elements to apply f to.
• T: The type of the value to be assigned (deduced).

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• count: Refers to the number of elements starting at first the algorithm will be applied to.
• value: The value to be assigned.

The comparisons in the parallel fill_n algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The fill_n algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns difference_type otherwise (where difference_type is defined by void.

template<typename ExPolicy, typename Rng, typename Rng2, typename Pred = detail::equal_to, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type find_end(ExPolicy &&policy, Rng &&rng, Rng2 &&rng2, Pred &&op = Pred(), Proj &&proj = Proj())

Returns the last subsequence of elements rng2 found in the range rng using the given predicate f to compare elements.

The comparison operations in the parallel

find_end algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most S*(N-S+1) comparisons where S = distance(begin(rng2), end(rng2)) and N = distance(begin(rng), end(rng)).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the first source range (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• Rng2: The type of the second source range (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of replace requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the first sequence of elements the algorithm will be applied to.
• rng2: Refers to the second sequence of elements the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types iterator_t<Rng> and iterator_t<Rng2> can be dereferenced and then implicitly converted to Type1 and Type2 respectively.
• proj: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced iterator_t<Rng> and dereferenced iterator_t<Rng2> as a projection operation before the function op is invoked.

The comparison operations in the parallel find_end algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

This overload of

find_end is available if the user decides to provide the algorithm their own predicate op.

Return

The find_end algorithm returns a hpx::future<iterator_t<Rng> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns iterator_t<Rng> otherwise. The find_end algorithm returns an iterator to the beginning of the last subsequence rng2 in range rng. If the length of the subsequence rng2 is greater than the length of the range rng, end(rng) is returned. Additionally if the size of the subsequence is empty or no subsequence is found, end(rng) is also returned.

template<typename ExPolicy, typename Rng1, typename Rng2, typename Pred = detail::equal_to, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng1>::type>::type find_first_of(ExPolicy &&policy, Rng1 &&rng1, Rng2 &&rng2, Pred &&op = Pred(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Searches the range rng1 for any elements in the range rng2. Uses binary predicate p to compare elements

The comparison operations in the parallel

find_first_of algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most (S*N) comparisons where S = distance(begin(rng2), end(rng2)) and N = distance(begin(rng1), end(rng1)).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng1: The type of the first source range (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• Rng2: The type of the second source range (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of replace requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj1: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements in rng1.
• Proj2: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements in rng2.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng1: Refers to the first sequence of elements the algorithm will be applied to.
• rng2: Refers to the second sequence of elements the algorithm will be applied to.
• op: The binary predicate which returns true if the elements should be treated as equal. The signature should be equivalent to the following:

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types iterator_t<Rng1> and iterator_t<Rng2> can be dereferenced and then implicitly converted to Type1 and Type2 respectively.
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced iterator_t<Rng1> before the function op is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of type dereferenced iterator_t<Rng2> before the function op is invoked.

The comparison operations in the parallel find_first_of algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

This overload of

find_first_of is available if the user decides to provide the algorithm their own predicate op.

Return

The find_end algorithm returns a hpx::future<iterator_t<Rng1> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns iterator_t<Rng1> otherwise. The find_first_of algorithm returns an iterator to the first element in the range rng1 that is equal to an element from the range rng2. If the length of the subsequence rng2 is greater than the length of the range rng1, end(rng1) is returned. Additionally if the size of the subsequence is empty or no subsequence is found, end(rng1) is also returned.

template<typename ExPolicy, typename Rng, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type for_each(ExPolicy &&policy, Rng &&rng, F &&f, Proj &&proj = Proj())

Applies f to the result of dereferencing every iterator in the given range rng.

If

f returns a result, the result is ignored.

Note

Complexity: Applies f exactly size(rng) times.

If the type of first satisfies the requirements of a mutable iterator, f may apply non-constant functions through the dereferenced iterator.

Unlike its sequential form, the parallel overload of for_each does not return a copy of its Function parameter, since parallelization may not permit efficient state accumulation.

The application of function objects in parallel algorithm invoked with an execution policy object of type

sequenced_policy execute in sequential order in the calling thread.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of for_each requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). The signature of this predicate should be equivalent to:

  <ignored> pred(const Type &a);
  

  The signature does not need to have const&. The type

  Type must be such that an object of type InIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The for_each algorithm returns a hpx::future<InIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns InIter otherwise. It returns last.

template<typename ExPolicy, typename Rng, typename F>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type generate(ExPolicy &&policy, Rng &&rng, F &&f)

Assign each element in range [first, last) a value generated by the given function object f

The assignments in the parallel

generate algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly distance(first, last) invocations of f and assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: generator function that will be called. signature of function should be equivalent to the following:

  Ret fun();
  

  The type

  Ret must be such that an object of type FwdIter can be dereferenced and assigned a value of type Ret.

The assignments in the parallel generate algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. It returns last.

template<typename ExPolicy, typename Rng, typename Comp = detail::less, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, bool>::type is_heap(ExPolicy &&policy, Rng &&rng, Comp &&comp = Comp(), Proj &&proj = Proj())

Returns whether the range is max heap. That is, true if the range is max heap, false otherwise. The function uses the given comparison function object comp (defaults to using operator<()).

comp has to induce a strict weak ordering on the values.

Note

Complexity: Performs at most N applications of the comparison comp, at most 2 * N applications of the projection proj, where N = last - first.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an random access iterator.
• Comp: The type of the function/function object to use (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_heap algorithm returns a hpx::future<bool> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns bool otherwise. The is_heap algorithm returns whether the range is max heap. That is, true if the range is max heap, false otherwise.

template<typename ExPolicy, typename Rng, typename Comp = detail::less, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type is_heap_until(ExPolicy &&policy, Rng &&rng, Comp &&comp = Comp(), Proj &&proj = Proj())

Returns the upper bound of the largest range beginning at first which is a max heap. That is, the last iterator it for which range [first, it) is a max heap. The function uses the given comparison function object comp (defaults to using operator<()).

comp has to induce a strict weak ordering on the values.

Note

Complexity: Performs at most N applications of the comparison comp, at most 2 * N applications of the projection proj, where N = last - first.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an random access iterator.
• Comp: The type of the function/function object to use (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The is_heap_until algorithm returns a hpx::future<RandIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns RandIter otherwise. The is_heap_until algorithm returns the upper bound of the largest range beginning at first which is a max heap. That is, the last iterator it for which range [first, it) is a max heap.

template<typename ExPolicy, typename Rng1, typename Rng2, typename RandIter3, typename Comp = detail::less, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in1(typename hpx::traits::range_iterator<Rng1>::type), tag::in2(typename hpx::traits::range_iterator<Rng2>::type), tag::out(RandIter3)>>::type merge(ExPolicy &&policy, Rng1 &&rng1, Rng2 &&rng2, RandIter3 dest, Comp &&comp = Comp(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Merges two sorted ranges [first1, last1) and [first2, last2) into one sorted range beginning at dest. The order of equivalent elements in the each of original two ranges is preserved. For equivalent elements in the original two ranges, the elements from the first range precede the elements from the second range. The destination range cannot overlap with either of the input ranges.

The assignments in the parallel

merge algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs O(std::distance(first1, last1) + std::distance(first2, last2)) applications of the comparison comp and the each projection.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng1: The type of the first source range used (deduced). The iterators extracted from this range type must meet the requirements of an random access iterator.
• Rng2: The type of the second source range used (deduced). The iterators extracted from this range type must meet the requirements of an random access iterator.
• RandIter3: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an random access iterator.
• Comp: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of merge requires Comp to meet the requirements of CopyConstructible. This defaults to std::less<>
• Proj1: The type of an optional projection function to be used for elements of the first range. This defaults to util::projection_identity
• Proj2: The type of an optional projection function to be used for elements of the second range. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng1: Refers to the first range of elements the algorithm will be applied to.
• rng2: Refers to the second range of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• comp: comp is a callable object which returns true if the first argument is less than the second, and false otherwise. The signature of this comparison should be equivalent to:

  bool comp(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types RandIter1 and RandIter2 can be dereferenced and then implicitly converted to both Type1 and Type2
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of the first range as a projection operation before the actual comparison comp is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of the second range as a projection operation before the actual comparison comp is invoked.

The assignments in the parallel merge algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The merge algorithm returns a hpx::future<tagged_tuple<tag::in1(RandIter1), tag::in2(RandIter2), tag::out(RandIter3)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_tuple<tag::in1(RandIter1), tag::in2(RandIter2), tag::out(RandIter3)> otherwise. The merge algorithm returns the tuple of the source iterator last1, the source iterator last2, the destination iterator to the end of the dest range.

template<typename ExPolicy, typename Rng, typename RandIter, typename Comp = detail::less, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, RandIter>::type inplace_merge(ExPolicy &&policy, Rng &&rng, RandIter middle, Comp &&comp = Comp(), Proj &&proj = Proj())

Merges two consecutive sorted ranges [first, middle) and [middle, last) into one sorted range [first, last). The order of equivalent elements in the each of original two ranges is preserved. For equivalent elements in the original two ranges, the elements from the first range precede the elements from the second range.

The assignments in the parallel

inplace_merge algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs O(std::distance(first, last)) applications of the comparison comp and the each projection.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an random access iterator.
• RandIter: The type of the source iterators used (deduced). This iterator type must meet the requirements of an random access iterator.
• Comp: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of inplace_merge requires Comp to meet the requirements of CopyConstructible. This defaults to std::less<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the range of elements the algorithm will be applied to.
• middle: Refers to the end of the first sorted range and the beginning of the second sorted range the algorithm will be applied to.
• comp: comp is a callable object which returns true if the first argument is less than the second, and false otherwise. The signature of this comparison should be equivalent to:

  bool comp(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types RandIter can be dereferenced and then implicitly converted to both Type1 and Type2
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel inplace_merge algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The inplace_merge algorithm returns a hpx::future<RandIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns RandIter otherwise. The inplace_merge algorithm returns the source iterator last

template<typename ExPolicy, typename Rng, typename Proj = util::projection_identity, typename F = detail::less>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_traits<Rng>::iterator_type>::type min_element(ExPolicy &&policy, Rng &&rng, F &&f = F(), Proj &&proj = Proj())

Finds the smallest element in the range [first, last) using the given comparison function f.

The comparisons in the parallel

min_element algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly max(N-1, 0) comparisons, where N = std::distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of min_element requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: The binary predicate which returns true if the the left argument is less than the right element. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The comparisons in the parallel min_element algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The min_element algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The min_element algorithm returns the iterator to the smallest element in the range [first, last). If several elements in the range are equivalent to the smallest element, returns the iterator to the first such element. Returns last if the range is empty.

template<typename ExPolicy, typename Rng, typename Proj = util::projection_identity, typename F = detail::less>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_traits<Rng>::iterator_type>::type max_element(ExPolicy &&policy, Rng &&rng, F &&f = F(), Proj &&proj = Proj())

Finds the greatest element in the range [first, last) using the given comparison function f.

The comparisons in the parallel

max_element algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly max(N-1, 0) comparisons, where N = std::distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of max_element requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: The binary predicate which returns true if the This argument is optional and defaults to std::less. the left argument is less than the right element. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The comparisons in the parallel max_element algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The max_element algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The max_element algorithm returns the iterator to the smallest element in the range [first, last). If several elements in the range are equivalent to the smallest element, returns the iterator to the first such element. Returns last if the range is empty.

template<typename ExPolicy, typename Rng, typename Proj = util::projection_identity, typename F = detail::less>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::min(typename hpx::traits::range_traits<Rng>::iterator_type), tag::max(typename hpx::traits::range_traits<Rng>::iterator_type)>>::type minmax_element(ExPolicy &&policy, Rng &&rng, F &&f = F(), Proj &&proj = Proj())

Finds the greatest element in the range [first, last) using the given comparison function f.

The comparisons in the parallel

minmax_element algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: At most max(floor(3/2*(N-1)), 0) applications of the predicate, where N = std::distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of minmax_element requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: The binary predicate which returns true if the the left argument is less than the right element. This argument is optional and defaults to std::less. The signature of the predicate function should be equivalent to the following:

  bool pred(const Type1 &a, const Type1 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The type

  Type1 must be such that objects of type FwdIter can be dereferenced and then implicitly converted to Type1.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The comparisons in the parallel minmax_element algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The minmax_element algorithm returns a hpx::future<tagged_pair<tag::min(FwdIter), tag::max(FwdIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::min(FwdIter), tag::max(FwdIter)> otherwise. The minmax_element algorithm returns a pair consisting of an iterator to the smallest element as the first element and an iterator to the greatest element as the second. Returns std::make_pair(first, first) if the range is empty. If several elements are equivalent to the smallest element, the iterator to the first such element is returned. If several elements are equivalent to the largest element, the iterator to the last such element is returned.

template<typename ExPolicy, typename Rng, typename Pred, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type partition(ExPolicy &&policy, Rng &&rng, Pred &&pred, Proj &&proj = Proj())

Reorders the elements in the range rng in such a way that all elements for which the predicate pred returns true precede the elements for which the predicate pred returns false. Relative order of the elements is not preserved.

The assignments in the parallel

partition algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs at most 2 * N swaps, exactly N applications of the predicate and projection, where N = std::distance(begin(rng), end(rng)).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of partition requires Pred to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by the range rng. This is an unary predicate for partitioning the source iterators. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel partition algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The partition algorithm returns a hpx::future<FwdIter> if the execution policy is of type parallel_task_policy and returns FwdIter otherwise. The partition algorithm returns the iterator to the first element of the second group.

template<typename ExPolicy, typename Rng, typename FwdIter2, typename FwdIter3, typename Pred, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in(typename hpx::traits::range_iterator<Rng>::type), tag::out1(FwdIter2), tag::out2(FwdIter3)>>::type partition_copy(ExPolicy &&policy, Rng &&rng, FwdIter2 dest_true, FwdIter3 dest_false, Pred &&pred, Proj &&proj = Proj())

Copies the elements in the range rng, to two different ranges depending on the value returned by the predicate pred. The elements, that satisfy the predicate pred, are copied to the range beginning at dest_true. The rest of the elements are copied to the range beginning at dest_false. The order of the elements is preserved.

The assignments in the parallel

partition_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than N assignments, exactly N applications of the predicate pred, where N = std::distance(begin(rng), end(rng)).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range for the elements that satisfy the predicate pred (deduced). This iterator type must meet the requirements of an forward iterator.
• FwdIter3: The type of the iterator representing the destination range for the elements that don’t satisfy the predicate pred (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of partition_copy requires Pred to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest_true: Refers to the beginning of the destination range for the elements that satisfy the predicate pred.
• dest_false: Refers to the beginning of the destination range for the elements that don’t satisfy the predicate pred.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by the range rng. This is an unary predicate for partitioning the source iterators. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel partition_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The partition_copy algorithm returns a hpx::future<tagged_tuple<tag::in(InIter), tag::out1(OutIter1), tag::out2(OutIter2)> > if the execution policy is of type parallel_task_policy and returns tagged_tuple<tag::in(InIter), tag::out1(OutIter1), tag::out2(OutIter2)> otherwise. The partition_copy algorithm returns the tuple of the source iterator last, the destination iterator to the end of the dest_true range, and the destination iterator to the end of the dest_false range.

template<typename ExPolicy, typename Rng, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type remove(ExPolicy &&policy, Rng &&rng, T const &value, Proj &&proj = Proj())

Removes all elements satisfying specific criteria from the range [first, last) and returns a past-the-end iterator for the new end of the range. This version removes all elements that are equal to value.

The assignments in the parallel

remove algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the operator==() and the projection proj.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• T: The type of the value to remove (deduced). This value type must meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• value: Specifies the value of elements to remove.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The remove algorithm returns the iterator to the new end of the range.

template<typename ExPolicy, typename Rng, typename Pred, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type remove_if(ExPolicy &&policy, Rng &&rng, Pred &&pred, Proj &&proj = Proj())

Removes all elements satisfying specific criteria from the range [first, last) and returns a past-the-end iterator for the new end of the range. This version removes all elements for which predicate pred returns true.

The assignments in the parallel

remove_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate pred and the projection proj.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of remove_if requires Pred to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The remove_if algorithm returns the iterator to the new end of the range.

template<typename ExPolicy, typename Rng, typename OutIter, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_traits<Rng>::iterator_type), tag::out(OutIter)>>::type remove_copy(ExPolicy &&policy, Rng &&rng, OutIter dest, T const &val, Proj &&proj = Proj())

Copies the elements in the range, defined by [first, last), to another range beginning at dest. Copies only the elements for which the comparison operator returns false when compare to val. The order of the elements that are not removed is preserved.

Effects: Copies all the elements referred to by the iterator it in the range [first,last) for which the following corresponding conditions do not hold: INVOKE(proj, *it) == value

The assignments in the parallel

remove_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• T: The type that the result of dereferencing InIter is compared to.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• val: Value to be removed.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove_copy algorithm returns a hpx::future<tagged_pair<tag::in(InIter), tag::out(OutIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(InIter), tag::out(OutIter)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename OutIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_traits<Rng>::iterator_type), tag::out(OutIter)>>::type remove_copy_if(ExPolicy &&policy, Rng &&rng, OutIter dest, F &&f, Proj &&proj = Proj())

Copies the elements in the range, defined by [first, last), to another range beginning at dest. Copies only the elements for which the predicate f returns false. The order of the elements that are not removed is preserved.

Effects: Copies all the elements referred to by the iterator it in the range [first,last) for which the following corresponding conditions do not hold: INVOKE(pred, INVOKE(proj, *it)) != false.

The assignments in the parallel

remove_copy_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than last - first assignments, exactly last - first applications of the predicate f.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of copy_if requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the elements to be removed. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type InIter can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel remove_copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The remove_copy_if algorithm returns a hpx::future<tagged_pair<tag::in(InIter), tag::out(OutIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(InIter), tag::out(OutIter)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename T1, typename T2, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_traits<Rng>::iterator_type>::type replace(ExPolicy &&policy, Rng &&rng, T1 const &old_value, T2 const &new_value, Proj &&proj = Proj())

Replaces all elements satisfying specific criteria with new_value in the range [first, last).

Effects: Substitutes elements referred by the iterator it in the range [first,last) with new_value, when the following corresponding conditions hold: INVOKE(proj, *i) == old_value

Note

Complexity: Performs exactly last - first assignments.

The assignments in the parallel

replace algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• T1: The type of the old value to replace (deduced).
• T2: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• old_value: Refers to the old value of the elements to replace.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename ExPolicy, typename Rng, typename F, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_traits<Rng>::iterator_type>::type replace_if(ExPolicy &&policy, Rng &&rng, F &&f, T const &new_value, Proj &&proj = Proj())

Replaces all elements satisfying specific criteria (for which predicate f returns true) with new_value in the range [first, last).

Effects: Substitutes elements referred by the iterator it in the range [first, last) with new_value, when the following corresponding conditions hold: INVOKE(f, INVOKE(proj, *it)) != false

Note

Complexity: Performs exactly last - first applications of the predicate.

The assignments in the parallel

replace_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible. (deduced).
• T: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the elements which need to replaced. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_if algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise. It returns last.

template<typename ExPolicy, typename Rng, typename OutIter, typename T1, typename T2, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_traits<Rng>::iterator_type), tag::out(OutIter)>>::type replace_copy(ExPolicy &&policy, Rng &&rng, OutIter dest, T1 const &old_value, T2 const &new_value, Proj &&proj = Proj())

Copies the all elements from the range [first, last) to another range beginning at dest replacing all elements satisfying a specific criteria with new_value.

Effects: Assigns to every iterator it in the range [result, result + (last - first)) either new_value or *(first + (it - result)) depending on whether the following corresponding condition holds: INVOKE(proj, *(first + (i - result))) == old_value

The assignments in the parallel

replace_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• T1: The type of the old value to replace (deduced).
• T2: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• old_value: Refers to the old value of the elements to replace.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_copy algorithm returns a hpx::future<tagged_pair<tag::in(InIter), tag::out(OutIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(InIter), tag::out(OutIter)> otherwise. The copy algorithm returns the pair of the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename OutIter, typename F, typename T, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_traits<Rng>::iterator_type), tag::out(OutIter)>>::type replace_copy_if(ExPolicy &&policy, Rng &&rng, OutIter dest, F &&f, T const &new_value, Proj &&proj = Proj())

Copies the all elements from the range [first, last) to another range beginning at dest replacing all elements satisfying a specific criteria with new_value.

Effects: Assigns to every iterator it in the range [result, result + (last - first)) either new_value or *(first + (it - result)) depending on whether the following corresponding condition holds: INVOKE(f, INVOKE(proj, *(first + (i - result)))) != false

The assignments in the parallel

replace_copy_if algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first applications of the predicate.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of equal requires F to meet the requirements of CopyConstructible. (deduced).
• T: The type of the new values to replace (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate which returns true for the elements which need to replaced. The signature of this predicate should be equivalent to:

  bool pred(const Type &a);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter can be dereferenced and then implicitly converted to Type.
• new_value: Refers to the new value to use as the replacement.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel replace_copy_if algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The replace_copy_if algorithm returns a hpx::future<tagged_pair<tag::in(InIter), tag::out(OutIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(InIter), tag::out(OutIter)> otherwise. The replace_copy_if algorithm returns the input iterator last and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type reverse(ExPolicy &&policy, Rng &&rng)

Reverses the order of the elements in the range [first, last). Behaves as if applying std::iter_swap to every pair of iterators first+i, (last-i) - 1 for each non-negative i < (last-first)/2.

The assignments in the parallel

reverse algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Linear in the distance between first and last.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of a bidirectional iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.

The assignments in the parallel reverse algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The reverse algorithm returns a hpx::future<BidirIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns BidirIter otherwise. It returns last.

template<typename ExPolicy, typename Rng, typename OutIter>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_iterator<Rng>::type), tag::out(OutIter)>>::type reverse_copy(ExPolicy &&policy, Rng &&rng, OutIter dest_first)

Copies the elements from the range [first, last) to another range beginning at dest_first in such a way that the elements in the new range are in reverse order. Behaves as if by executing the assignment *(dest_first + (last - first) - 1 - i) = *(first + i) once for each non-negative i < (last - first) If the source and destination ranges (that is, [first, last) and [dest_first, dest_first+(last-first)) respectively) overlap, the behavior is undefined.

The assignments in the parallel

reverse_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of a bidirectional iterator.
• OutputIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest_first: Refers to the begin of the destination range.

The assignments in the parallel reverse_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The reverse_copy algorithm returns a hpx::future<tagged_pair<tag::in(BidirIter), tag::out(OutIter)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(BidirIter), tag::out(OutIter)> otherwise. The copy algorithm returns the pair of the input iterator forwarded to the first element after the last in the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::begin(typename hpx::traits::range_iterator<Rng>::type), tag::end(typename hpx::traits::range_iterator<Rng>::type)>>::type rotate(ExPolicy &&policy, Rng &&rng, typename hpx::traits::range_iterator<Rng>::type middle)

Performs a left rotation on a range of elements. Specifically, rotate swaps the elements in the range [first, last) in such a way that the element new_first becomes the first element of the new range and new_first - 1 becomes the last element.

The assignments in the parallel

rotate algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Linear in the distance between first and last.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• middle: Refers to the element that should appear at the beginning of the rotated range.

The assignments in the parallel rotate algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The type of dereferenced FwdIter must meet the requirements of MoveAssignable and MoveConstructible.

Return

The rotate algorithm returns a hpx::future<tagged_pair<tag::begin(FwdIter), tag::end(FwdIter)> > if the execution policy is of type parallel_task_policy and returns tagged_pair<tag::begin(FwdIter), tag::end(FwdIter)> otherwise. The rotate algorithm returns the iterator equal to pair(first + (last - new_first), last).

template<typename ExPolicy, typename Rng, typename OutIter>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_iterator<Rng>::type), tag::out(OutIter)>>::type rotate_copy(ExPolicy &&policy, Rng &&rng, typename hpx::traits::range_iterator<Rng>::type middle, OutIter dest_first)

Copies the elements from the range [first, last), to another range beginning at dest_first in such a way, that the element new_first becomes the first element of the new range and new_first - 1 becomes the last element.

The assignments in the parallel

rotate_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs exactly last - first assignments.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of a forward iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• middle: Refers to the element that should appear at the beginning of the rotated range.
• dest_first: Refers to the begin of the destination range.

The assignments in the parallel rotate_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The rotate_copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter), tag::out(OutIter)> > if the execution policy is of type parallel_task_policy and returns tagged_pair<tag::in(FwdIter), tag::out(OutIter)> otherwise. The rotate_copy algorithm returns the output iterator to the element past the last element copied.

template<typename ExPolicy, typename Rng1, typename Rng2, typename Pred = detail::equal_to, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng1>::type>::type search(ExPolicy &&policy, Rng1 &&rng1, Rng2 &&rng2, Pred &&op = Pred(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Searches the range [first, last) for any elements in the range [s_first, s_last). Uses a provided predicate to compare elements.

The comparison operations in the parallel

search algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most (S*N) comparisons where S = distance(s_first, s_last) and N = distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng1: The type of the examine range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Rng2: The type of the search range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of adjacent_find requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj1: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of Rng1.
• Proj2: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of Rng2.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng1: Refers to the sequence of elements the algorithm will be examining.
• rng2: Refers to the sequence of elements the algorithm will be searching for.
• op: Refers to the binary predicate which returns true if the elements should be treated as equal. the signature of the function should be equivalent to

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of rng1 as a projection operation before the actual predicate is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of rng2 as a projection operation before the actual predicate is invoked.

The comparison operations in the parallel search algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The search algorithm returns a hpx::future<FwdIter> if the execution policy is of type task_execution_policy and returns FwdIter otherwise. The search algorithm returns an iterator to the beginning of the first subsequence [s_first, s_last) in range [first, last). If the length of the subsequence [s_first, s_last) is greater than the length of the range [first, last), last is returned. Additionally if the size of the subsequence is empty first is returned. If no subsequence is found, last is returned.

template<typename ExPolicy, typename Rng1, typename Rng2, typename Pred = detail::equal_to, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng1>::type>::type search_n(ExPolicy &&policy, Rng1 &&rng1, std::size_t count, Rng2 &&rng2, Pred &&op = Pred(), Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Searches the range [first, last) for any elements in the range [s_first, s_last). Uses a provided predicate to compare elements.

The comparison operations in the parallel

search algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: at most (S*N) comparisons where S = distance(s_first, s_last) and N = distance(first, last).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng1: The type of the examine range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Rng2: The type of the search range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Pred: The type of an optional function/function object to use. Unlike its sequential form, the parallel overload of adjacent_find requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj1: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of Rng1.
• Proj2: The type of an optional projection function. This defaults to util::projection_identity and is applied to the elements of Rng2.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng1: Refers to the sequence of elements the algorithm will be examining.
• count: The number of elements to apply the algorithm on.
• rng2: Refers to the sequence of elements the algorithm will be searching for.
• op: Refers to the binary predicate which returns true if the elements should be treated as equal. the signature of the function should be equivalent to

  bool pred(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const &, but the function must not modify the objects passed to it. The types

  Type1 and Type2 must be such that objects of types FwdIter1 and FwdIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of rng1 as a projection operation before the actual predicate is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of rng2 as a projection operation before the actual predicate is invoked.

The comparison operations in the parallel search algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The search algorithm returns a hpx::future<FwdIter> if the execution policy is of type task_execution_policy and returns FwdIter otherwise. The search algorithm returns an iterator to the beginning of the first subsequence [s_first, s_last) in range [first, last). If the length of the subsequence [s_first, s_last) is greater than the length of the range [first, last), last is returned. Additionally if the size of the subsequence is empty first is returned. If no subsequence is found, last is returned.

template<typename ExPolicy, typename Rng, typename Proj = util::projection_identity, typename Compare = detail::less>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type sort(ExPolicy &&policy, Rng &&rng, Compare &&comp = Compare(), Proj &&proj = Proj())

Sorts the elements in the range rng in ascending order. The order of equal elements is not guaranteed to be preserved. The function uses the given comparison function object comp (defaults to using operator<()).

A sequence is sorted with respect to a comparator

comp and a projection proj if for every iterator i pointing to the sequence and every non-negative integer n such that i + n is a valid iterator pointing to an element of the sequence, and INVOKE(comp, INVOKE(proj, *(i + n)), INVOKE(proj, *i)) == false.

Note

Complexity: O(Nlog(N)), where N = std::distance(begin(rng), end(rng)) comparisons.

comp has to induce a strict weak ordering on the values.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Comp: The type of the function/function object to use (deduced).
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• comp: comp is a callable object. The return value of the INVOKE operation applied to an object of type Comp, when contextually converted to bool, yields true if the first argument of the call is less than the second, and false otherwise. It is assumed that comp will not apply any non-constant function through the dereferenced iterator.
• proj: Specifies the function (or function object) which will be invoked for each pair of elements as a projection operation before the actual predicate comp is invoked.

The application of function objects in parallel algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

The application of function objects in parallel algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The sort algorithm returns a hpx::future<Iter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns Iter otherwise. It returns last.

template<typename ExPolicy, typename Rng, typename OutIter, typename F, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_iterator<Rng>::type), tag::out(OutIter)>>::type transform(ExPolicy &&policy, Rng &&rng, OutIter dest, F &&f, Proj &&proj = Proj())

Applies the given function f to the given range rng and stores the result in another range, beginning at dest.

The invocations of

f in the parallel transform algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly size(rng) applications of f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is an unary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type &a);
  

  The signature does not need to have const&. The type

  Type must be such that an object of type InIter can be dereferenced and then implicitly converted to Type. The type Ret must be such that an object of type OutIter can be dereferenced and assigned a value of type Ret.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel transform algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The transform algorithm returns a hpx::future<tagged_pair<tag::in(InIter), tag::out(OutIter)> > if the execution policy is of type parallel_task_policy and returns tagged_pair<tag::in(InIter), tag::out(OutIter)> otherwise. The transform algorithm returns a tuple holding an iterator referring to the first element after the input sequence and the output iterator to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename InIter2, typename OutIter, typename F, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in1(typename hpx::traits::range_iterator<Rng>::type), tag::in2(InIter2), tag::out(OutIter)>>::type transform(ExPolicy &&policy, Rng &&rng, InIter2 first2, OutIter dest, F &&f, Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Applies the given function f to pairs of elements from two ranges: one defined by rng and the other beginning at first2, and stores the result in another range, beginning at dest.

The invocations of

f in the parallel transform algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly size(rng) applications of f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• InIter2: The type of the source iterators for the second range used (deduced). This iterator type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj1: The type of an optional projection function to be used for elements of the first sequence. This defaults to util::projection_identity
• Proj2: The type of an optional projection function to be used for elements of the second sequence. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• first2: Refers to the beginning of the second sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&. The types

  Type1 and Type2 must be such that objects of types InIter1 and InIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively. The type Ret must be such that an object of type OutIter can be dereferenced and assigned a value of type Ret.
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of the first sequence as a projection operation before the actual predicate f is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of the second sequence as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel transform algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The transform algorithm returns a hpx::future<tagged_tuple<tag::in1(InIter1), tag::in2(InIter2), tag::out(OutIter)> > if the execution policy is of type parallel_task_policy and returns tagged_tuple<tag::in1(InIter1), tag::in2(InIter2), tag::out(OutIter)> otherwise. The transform algorithm returns a tuple holding an iterator referring to the first element after the first input sequence, an iterator referring to the first element after the second input sequence, and the output iterator referring to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng1, typename Rng2, typename OutIter, typename F, typename Proj1 = util::projection_identity, typename Proj2 = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_tuple<tag::in1(typename hpx::traits::range_iterator<Rng1>::type), tag::in2(typename hpx::traits::range_iterator<Rng2>::type), tag::out(OutIter)>>::type transform(ExPolicy &&policy, Rng1 &&rng1, Rng2 &&rng2, OutIter dest, F &&f, Proj1 &&proj1 = Proj1(), Proj2 &&proj2 = Proj2())

Applies the given function f to pairs of elements from two ranges: one defined by [first1, last1) and the other beginning at first2, and stores the result in another range, beginning at dest.

The invocations of

f in the parallel transform algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Exactly min(last2-first2, last1-first1) applications of f

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the invocations of f.
• Rng1: The type of the first source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• Rng2: The type of the second source range used (deduced). The iterators extracted from this range type must meet the requirements of an input iterator.
• OutIter: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an output iterator.
• F: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of transform requires F to meet the requirements of CopyConstructible.
• Proj1: The type of an optional projection function to be used for elements of the first sequence. This defaults to util::projection_identity
• Proj2: The type of an optional projection function to be used for elements of the second sequence. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng1: Refers to the first sequence of elements the algorithm will be applied to.
• rng2: Refers to the second sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• f: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last).This is a binary predicate. The signature of this predicate should be equivalent to:

  Ret fun(const Type1 &a, const Type2 &b);
  

  The signature does not need to have const&. The types

  Type1 and Type2 must be such that objects of types InIter1 and InIter2 can be dereferenced and then implicitly converted to Type1 and Type2 respectively. The type Ret must be such that an object of type OutIter can be dereferenced and assigned a value of type Ret.
• proj1: Specifies the function (or function object) which will be invoked for each of the elements of the first sequence as a projection operation before the actual predicate f is invoked.
• proj2: Specifies the function (or function object) which will be invoked for each of the elements of the second sequence as a projection operation before the actual predicate f is invoked.

The invocations of f in the parallel transform algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Note

The algorithm will invoke the binary predicate until it reaches the end of the shorter of the two given input sequences

Return

The transform algorithm returns a hpx::future<tagged_tuple<tag::in1(InIter1), tag::in2(InIter2), tag::out(OutIter)> > if the execution policy is of type parallel_task_policy and returns tagged_tuple<tag::in1(InIter1), tag::in2(InIter2), tag::out(OutIter)> otherwise. The transform algorithm returns a tuple holding an iterator referring to the first element r the first input sequence, an iterator referring to the first element after the second input sequence, and the output iterator referring to the element in the destination range, one past the last element copied.

template<typename ExPolicy, typename Rng, typename Pred = detail::equal_to, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, typename hpx::traits::range_iterator<Rng>::type>::type unique(ExPolicy &&policy, Rng &&rng, Pred &&pred = Pred(), Proj &&proj = Proj())

Eliminates all but the first element from every consecutive group of equivalent elements from the range rng and returns a past-the-end iterator for the new logical end of the range.

The assignments in the parallel

unique algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than N assignments, exactly N - 1 applications of the predicate pred and no more than twice as many applications of the projection proj, where N = std::distance(begin(rng), end(rng)).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of unique requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last). This is an binary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a, const Type &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel unique algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The unique algorithm returns a hpx::future<FwdIter> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns FwdIter otherwise. The unique algorithm returns the iterator to the new end of the range.

template<typename ExPolicy, typename Rng, typename FwdIter2, typename Pred = detail::equal_to, typename Proj = util::projection_identity>
util::detail::algorithm_result<ExPolicy, hpx::util::tagged_pair<tag::in(typename hpx::traits::range_iterator<Rng>::type), tag::out(FwdIter2)>>::type unique_copy(ExPolicy &&policy, Rng &&rng, FwdIter2 dest, Pred &&pred = Pred(), Proj &&proj = Proj())

Copies the elements from the range rng, to another range beginning at dest in such a way that there are no consecutive equal elements. Only the first element of each group of equal elements is copied.

The assignments in the parallel

unique_copy algorithm invoked with an execution policy object of type sequenced_policy execute in sequential order in the calling thread.

Note

Complexity: Performs not more than N assignments, exactly N - 1 applications of the predicate pred, where N = std::distance(begin(rng), end(rng)).

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it executes the assignments.
• Rng: The type of the source range used (deduced). The iterators extracted from this range type must meet the requirements of an forward iterator.
• FwdIter2: The type of the iterator representing the destination range (deduced). This iterator type must meet the requirements of an forward iterator.
• Pred: The type of the function/function object to use (deduced). Unlike its sequential form, the parallel overload of unique_copy requires Pred to meet the requirements of CopyConstructible. This defaults to std::equal_to<>
• Proj: The type of an optional projection function. This defaults to util::projection_identity

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• rng: Refers to the sequence of elements the algorithm will be applied to.
• dest: Refers to the beginning of the destination range.
• pred: Specifies the function (or function object) which will be invoked for each of the elements in the sequence specified by the range rng. This is an binary predicate which returns true for the required elements. The signature of this predicate should be equivalent to:

  bool pred(const Type &a, const Type &b);
  

  The signature does not need to have const&, but the function must not modify the objects passed to it. The type

  Type must be such that an object of type FwdIter1 can be dereferenced and then implicitly converted to Type.
• proj: Specifies the function (or function object) which will be invoked for each of the elements as a projection operation before the actual predicate is invoked.

The assignments in the parallel unique_copy algorithm invoked with an execution policy object of type parallel_policy or parallel_task_policy are permitted to execute in an unordered fashion in unspecified threads, and indeterminately sequenced within each thread.

Return

The unique_copy algorithm returns a hpx::future<tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> > if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns tagged_pair<tag::in(FwdIter1), tag::out(FwdIter2)> otherwise. The unique_copy algorithm returns the pair of the source iterator to last, and the destination iterator to the end of the dest range.

namespace v2

Functions

template<typename ExPolicy, typename F>
util::detail::algorithm_result<ExPolicy>::type define_task_block(ExPolicy &&policy, F &&f)

Constructs a task_block, tr, using the given execution policy policy,and invokes the expression f(tr) on the user-provided object, f.

Postcondition: All tasks spawned from

f have finished execution. A call to define_task_block may return on a different thread than that on which it was called.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the task block may be parallelized.
• F: The type of the user defined function to invoke inside the define_task_block (deduced). F shall be MoveConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• f: The user defined function to invoke inside the task block. Given an lvalue tr of type task_block, the expression, (void)f(tr), shall be well-formed.

Note

It is expected (but not mandated) that f will (directly or indirectly) call tr.run(callable_object).

Exceptions

• An: exception_list, as specified in Exception Handling.

template<typename F>
void define_task_block(F &&f)

Constructs a task_block, tr, and invokes the expression f(tr) on the user-provided object, f. This version uses parallel_policy for task scheduling.

Postcondition: All tasks spawned from

f have finished execution. A call to define_task_block may return on a different thread than that on which it was called.

Template Parameters

• F: The type of the user defined function to invoke inside the define_task_block (deduced). F shall be MoveConstructible.

Parameters

• f: The user defined function to invoke inside the task block. Given an lvalue tr of type task_block, the expression, (void)f(tr), shall be well-formed.

Note

It is expected (but not mandated) that f will (directly or indirectly) call tr.run(callable_object).

Exceptions

• An: exception_list, as specified in Exception Handling.

template<typename ExPolicy, typename F>
util::detail::algorithm_result<ExPolicy>::type define_task_block_restore_thread(ExPolicy &&policy, F &&f)

Constructs a task_block, tr, and invokes the expression f(tr) on the user-provided object, f.

Postcondition: All tasks spawned from

f have finished execution. A call to define_task_block_restore_thread always returns on the same thread as that on which it was called.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the task block may be parallelized.
• F: The type of the user defined function to invoke inside the define_task_block (deduced). F shall be MoveConstructible.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• f: The user defined function to invoke inside the define_task_block. Given an lvalue tr of type task_block, the expression, (void)f(tr), shall be well-formed.

Exceptions

• An: exception_list, as specified in Exception Handling.

Note

It is expected (but not mandated) that f will (directly or indirectly) call tr.run(callable_object).

template<typename F>
void define_task_block_restore_thread(F &&f)

Constructs a task_block, tr, and invokes the expression f(tr) on the user-provided object, f. This version uses parallel_policy for task scheduling.

Postcondition: All tasks spawned from

f have finished execution. A call to define_task_block_restore_thread always returns on the same thread as that on which it was called.

Template Parameters

• F: The type of the user defined function to invoke inside the define_task_block (deduced). F shall be MoveConstructible.

Parameters

• f: The user defined function to invoke inside the define_task_block. Given an lvalue tr of type task_block, the expression, (void)f(tr), shall be well-formed.

Exceptions

• An: exception_list, as specified in Exception Handling.

Note

It is expected (but not mandated) that f will (directly or indirectly) call tr.run(callable_object).

template<typename ExPolicy, typename I, typename ...Args>
util::detail::algorithm_result<ExPolicy>::type for_loop(ExPolicy &&policy, typename std::decay<I>::type first, I last, Args&&... args)

The for_loop implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

Return

The for_loop algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename I, typename ...Args>
void for_loop(typename std::decay<I>::type first, I last, Args&&... args)

The for_loop implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

The execution of for_loop without specifying an execution policy is equivalent to specifying parallel::execution::seq as the execution policy.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

template<typename ExPolicy, typename I, typename S, typename... Args, &&std::is_integral< S >::value>util::detail::algorithm_result<ExPolicy>::type hpx::parallel::v2::for_loop_strided(ExPolicy && policy, typename std::decay< I >::type first, I last, S stride, Args &&... args)

The for_loop_strided implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• S: The type of the stride variable. This should be an integral type.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• stride: Refers to the stride of the iteration steps. This shall have non-zero value and shall be negative only if I has integral type or meets the requirements of a bidirectional iterator.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

Return

The for_loop_strided algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename I, typename S, typename... Args, &&std::is_integral< S >::value>void hpx::parallel::v2::for_loop_strided(typename std::decay< I >::type first, I last, S stride, Args &&... args)

The for_loop_strided implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

The execution of for_loop without specifying an execution policy is equivalent to specifying parallel::execution::seq as the execution policy.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• S: The type of the stride variable. This should be an integral type.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• last: Refers to the end of the sequence of elements the algorithm will be applied to.
• stride: Refers to the stride of the iteration steps. This shall have non-zero value and shall be negative only if I has integral type or meets the requirements of a bidirectional iterator.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

template<typename ExPolicy, typename I, typename Size, typename... Args, &&std::is_integral< Size >::value>util::detail::algorithm_result<ExPolicy>::type hpx::parallel::v2::for_loop_n(ExPolicy && policy, I first, Size size, Args &&... args)

The for_loop_n implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• Size: The type of a non-negative integral value specifying the number of items to iterate over.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• size: Refers to the number of items the algorithm will be applied to.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

Return

The for_loop_n algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename I, typename Size, typename... Args, &&std::is_integral< Size >::value>void hpx::parallel::v2::for_loop_n(I first, Size size, Args &&... args)

The for_loop implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

The execution of for_loop without specifying an execution policy is equivalent to specifying parallel::execution::seq as the execution policy.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• Size: The type of a non-negative integral value specifying the number of items to iterate over.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• size: Refers to the number of items the algorithm will be applied to.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

template<typename ExPolicy, typename I, typename Size, typename S, typename... Args, &&std::is_integral< Size >::value &&std::is_integral< S >::value>util::detail::algorithm_result<ExPolicy>::type hpx::parallel::v2::for_loop_n_strided(ExPolicy && policy, I first, Size size, S stride, Args &&... args)

The for_loop_n_strided implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• ExPolicy: The type of the execution policy to use (deduced). It describes the manner in which the execution of the algorithm may be parallelized and the manner in which it applies user-provided function objects.
• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• Size: The type of a non-negative integral value specifying the number of items to iterate over.
• S: The type of the stride variable. This should be an integral type.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• policy: The execution policy to use for the scheduling of the iterations.
• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• size: Refers to the number of items the algorithm will be applied to.
• stride: Refers to the stride of the iteration steps. This shall have non-zero value and shall be negative only if I has integral type or meets the requirements of a bidirectional iterator.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

Return

The for_loop_n_strided algorithm returns a hpx::future<void> if the execution policy is of type sequenced_task_policy or parallel_task_policy and returns void otherwise.

template<typename I, typename Size, typename S, typename... Args, &&std::is_integral< Size >::value &&std::is_integral< S >::value>void hpx::parallel::v2::for_loop_n_strided(I first, Size size, S stride, Args &&... args)

The for_loop_n_strided implements loop functionality over a range specified by integral or iterator bounds. For the iterator case, these algorithms resemble for_each from the Parallelism TS, but leave to the programmer when and if to dereference the iterator.

The execution of for_loop without specifying an execution policy is equivalent to specifying parallel::execution::seq as the execution policy.

Requires:

I shall be an integral type or meet the requirements of an input iterator type. The args parameter pack shall have at least one element, comprising objects returned by invocations of reduction and/or induction function templates followed by exactly one element invocable element-access function, f. f shall meet the requirements of MoveConstructible.

Template Parameters

• I: The type of the iteration variable. This could be an (forward) iterator type or an integral type.
• Size: The type of a non-negative integral value specifying the number of items to iterate over.
• S: The type of the stride variable. This should be an integral type.
• Args: A parameter pack, it’s last element is a function object to be invoked for each iteration, the others have to be either conforming to the induction or reduction concept.

Parameters

• first: Refers to the beginning of the sequence of elements the algorithm will be applied to.
• size: Refers to the number of items the algorithm will be applied to.
• stride: Refers to the stride of the iteration steps. This shall have non-zero value and shall be negative only if I has integral type or meets the requirements of a bidirectional iterator.
• args: The last element of this parameter pack is the function (object) to invoke, while the remaining elements of the parameter pack are instances of either induction or reduction objects. The function (or function object) which will be invoked for each of the elements in the sequence specified by [first, last) should expose a signature equivalent to:

  <ignored> pred(I const& a, ...);
  

  The signature does not need to have const&. It will receive the current value of the iteration variable and one argument for each of the induction or reduction objects passed to the algorithms, representing their current values.

Effects: Applies f to each element in the input sequence, with additional arguments corresponding to the reductions and inductions in the args parameter pack. The length of the input sequence is last - first.

The first element in the input sequence is specified by first. Each subsequent element is generated by incrementing the previous element.

Along with an element from the input sequence, for each member of the

args parameter pack excluding f, an additional argument is passed to each application of f as follows:

Note

As described in the C++ standard, arithmetic on non-random-access iterators is performed using advance and distance.

Note

The order of the elements of the input sequence is important for determining ordinal position of an application of f, even though the applications themselves may be unordered.

If the pack member is an object returned by a call to a reduction function listed in section, then the additional argument is a reference to a view of that reduction object. If the pack member is an object returned by a call to induction, then the additional argument is the induction value for that induction object corresponding to the position of the application of f in the input sequence.

Complexity: Applies f exactly once for each element of the input sequence.

Remarks: If f returns a result, the result is ignored.

template<typename T>
detail::induction_stride_helper<T> induction(T &&value, std::size_t stride)

The function template returns an induction object of unspecified type having a value type and encapsulating an initial value value of that type and, optionally, a stride.

For each element in the input range, a looping algorithm over input sequence S computes an induction value from an induction variable and ordinal position p within S by the formula i + p * stride if a stride was specified or i + p otherwise. This induction value is passed to the element access function.

If the value argument to induction is a non-const lvalue, then that lvalue becomes the live-out object for the returned induction object. For each induction object that has a live-out object, the looping algorithm assigns the value of i + n * stride to the live-out object upon return, where n is the number of elements in the input range.

Return

This returns an induction object with value type T, initial value value, and (if specified) stride stride. If T is an lvalue of non-const type, value is used as the live-out object for the induction object; otherwise there is no live-out object.

Template Parameters

• T: The value type to be used by the induction object.

Parameters

• value: [in] The initial value to use for the induction object
• stride: [in] The (optional) stride to use for the induction object (default: 1)

template<typename T, typename Op>
detail::reduction_helper<T, typename std::decay<Op>::type> reduction(T &var, T const &identity, Op &&combiner)

The function template returns a reduction object of unspecified type having a value type and encapsulating an identity value for the reduction, a combiner function object, and a live-out object from which the initial value is obtained and into which the final value is stored.

A parallel algorithm uses reduction objects by allocating an unspecified number of instances, called views, of the reduction’s value type. Each view is initialized with the reduction object’s identity value, except that the live-out object (which was initialized by the caller) comprises one of the views. The algorithm passes a reference to a view to each application of an element-access function, ensuring that no two concurrently-executing invocations share the same view. A view can be shared between two applications that do not execute concurrently, but initialization is performed only once per view.

Modifications to the view by the application of element access functions accumulate as partial results. At some point before the algorithm returns, the partial results are combined, two at a time, using the reduction object’s combiner operation until a single value remains, which is then assigned back to the live-out object.

T shall meet the requirements of CopyConstructible and MoveAssignable. The expression var = combiner(var, var) shall be well formed.

Template Parameters

• T: The value type to be used by the induction object.
• Op: The type of the binary function (object) used to perform the reduction operation.

Parameters

• var: [in,out] The life-out value to use for the reduction object. This will hold the reduced value after the algorithm is finished executing.
• identity: [in] The identity value to use for the reduction operation.
• combiner: [in] The binary function (object) used to perform a pairwise reduction on the elements.

Note

In order to produce useful results, modifications to the view should be limited to commutative operations closely related to the combiner operation. For example if the combiner is plus<T>, incrementing the view would be consistent with the combiner but doubling it or assigning to it would not.

Return

This returns a reduction object of unspecified type having a value type of T. When the return value is used by an algorithm, the reference to var is used as the live-out object, new views are initialized to a copy of identity, and views are combined by invoking the copy of combiner, passing it the two views to be combined.

namespace performance_counters

Functions

counter_status install_counter_type(std::string const &name, hpx::util::function_nonser<std::int64_t(bool)> const &counter_value, std::string const &helptext = "", std::string const &uom = "", error_code &ec = throws, )

Install a new generic performance counter type in a way, which will uninstall it automatically during shutdown.

The function install_counter_type will register a new generic counter type based on the provided function. The counter type will be automatically unregistered during system shutdown. Any consumer querying any instance of this this counter type will cause the provided function to be called and the returned value to be exposed as the counter value.

The counter type is registered such that there can be one counter instance per locality. The expected naming scheme for the counter instances is: '/objectname{locality#<*>/total}/countername' where ‘<*>’ is a zero based integer identifying the locality the counter is created on.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Return

If successful, this function returns status_valid_data, otherwise it will either throw an exception or return an error_code from the enum counter_status (also, see note related to parameter ec).

Note

The counter type registry is a locality based service. You will have to register each counter type on every locality where a corresponding performance counter will be created.

Parameters

• name: [in] The global virtual name of the counter type. This name is expected to have the format /objectname/countername.
• counter_value: [in] The function to call whenever the counter value is requested by a consumer.
• helptext: [in, optional] A longer descriptive text shown to the user to explain the nature of the counters created from this type.
• uom: [in] The unit of measure for the new performance counter type.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

counter_status install_counter_type(std::string const &name, hpx::util::function_nonser<std::vector<std::int64_t>(bool)> const &counter_value, std::string const &helptext = "", std::string const &uom = "", error_code &ec = throws, )

Install a new generic performance counter type returning an array of values in a way, that will uninstall it automatically during shutdown.

The function install_counter_type will register a new generic counter type that returns an array of values based on the provided function. The counter type will be automatically unregistered during system shutdown. Any consumer querying any instance of this this counter type will cause the provided function to be called and the returned array value to be exposed as the counter value.

The counter type is registered such that there can be one counter instance per locality. The expected naming scheme for the counter instances is: '/objectname{locality#<*>/total}/countername' where ‘<*>’ is a zero based integer identifying the locality the counter is created on.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Return

If successful, this function returns status_valid_data, otherwise it will either throw an exception or return an error_code from the enum counter_status (also, see note related to parameter ec).

Note

The counter type registry is a locality based service. You will have to register each counter type on every locality where a corresponding performance counter will be created.

Parameters

• name: [in] The global virtual name of the counter type. This name is expected to have the format /objectname/countername.
• counter_value: [in] The function to call whenever the counter value (array of values) is requested by a consumer.
• helptext: [in, optional] A longer descriptive text shown to the user to explain the nature of the counters created from this type.
• uom: [in] The unit of measure for the new performance counter type.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void install_counter_type(std::string const &name, counter_type type, error_code &ec = throws)

Install a new performance counter type in a way, which will uninstall it automatically during shutdown.

The function install_counter_type will register a new counter type based on the provided counter_type_info. The counter type will be automatically unregistered during system shutdown.

Return

If successful, this function returns status_valid_data, otherwise it will either throw an exception or return an error_code from the enum counter_status (also, see note related to parameter ec).

Note

The counter type registry is a locality based service. You will have to register each counter type on every locality where a corresponding performance counter will be created.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• name: [in] The global virtual name of the counter type. This name is expected to have the format /objectname/countername.
• type: [in] The type of the counters of this counter_type.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

counter_status install_counter_type(std::string const &name, counter_type type, std::string const &helptext, std::string const &uom = "", std::uint32_t version = HPX_PERFORMANCE_COUNTER_V1, error_code &ec = throws)

Install a new performance counter type in a way, which will uninstall it automatically during shutdown.

The function install_counter_type will register a new counter type based on the provided counter_type_info. The counter type will be automatically unregistered during system shutdown.

Return

If successful, this function returns status_valid_data, otherwise it will either throw an exception or return an error_code from the enum counter_status (also, see note related to parameter ec).

Note

The counter type registry is a locality based service. You will have to register each counter type on every locality where a corresponding performance counter will be created.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• name: [in] The global virtual name of the counter type. This name is expected to have the format /objectname/countername.
• type: [in] The type of the counters of this counter_type.
• helptext: [in] A longer descriptive text shown to the user to explain the nature of the counters created from this type.
• uom: [in] The unit of measure for the new performance counter type.
• version: [in] The version of the counter type. This is currently expected to be set to HPX_PERFORMANCE_COUNTER_V1.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

counter_status install_counter_type(std::string const &name, counter_type type, std::string const &helptext, create_counter_func const &create_counter, discover_counters_func const &discover_counters, std::uint32_t version = HPX_PERFORMANCE_COUNTER_V1, std::string const &uom = "", error_code &ec = throws)

Install a new generic performance counter type in a way, which will uninstall it automatically during shutdown.

The function install_counter_type will register a new generic counter type based on the provided counter_type_info. The counter type will be automatically unregistered during system shutdown.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Return

If successful, this function returns status_valid_data, otherwise it will either throw an exception or return an error_code from the enum counter_status (also, see note related to parameter ec).

Note

The counter type registry is a locality based service. You will have to register each counter type on every locality where a corresponding performance counter will be created.

Parameters

• name: [in] The global virtual name of the counter type. This name is expected to have the format /objectname/countername.
• type: [in] The type of the counters of this counter_type.
• helptext: [in] A longer descriptive text shown to the user to explain the nature of the counters created from this type.
• version: [in] The version of the counter type. This is currently expected to be set to HPX_PERFORMANCE_COUNTER_V1.
• create_counter: [in] The function which will be called to create a new instance of this counter type.
• discover_counters: [in] The function will be called to discover counter instances which can be created.
• uom: [in] The unit of measure of the counter type (default: “”)
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

namespace resource

Typedefs

using scheduler_function = util::function_nonser<std::unique_ptr<hpx::threads::thread_pool_base>(hpx::threads::policies::callback_notifier&, std::size_t, std::size_t, std::size_t, std::string const&)>

Enums

enum partitioner_mode

This enumeration describes the modes available when creating a resource partitioner.

Values:

mode_default = 0

Default mode.

mode_allow_oversubscription = 1

Allow processing units to be oversubscribed, i.e. multiple worker threads to share a single processing unit.

mode_allow_dynamic_pools = 2

Allow worker threads to be added and removed from thread pools.

enum scheduling_policy

This enumeration lists the available scheduling policies (or schedulers) when creating thread pools.

Values:

user_defined = -2

unspecified = -1

local = 0

local_priority_fifo = 1

local_priority_lifo = 2

static_ = 3

static_priority = 4

abp_priority_fifo = 5

abp_priority_lifo = 6

shared_priority = 7

Functions

detail::partitioner &get_partitioner()

May be used anywhere in code and returns a reference to the single, global resource partitioner.

bool is_partitioner_valid()

Returns true if the resource partitioner has been initialized. Returns false otherwise.

namespace this_thread

Functions

threads::thread_state_ex_enum suspend(threads::thread_state_enum state, threads::thread_id_type const &id, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to the thread state passed as the parameter.

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_state_ex_enum suspend(threads::thread_state_enum state = threads::pending, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to the thread state passed as the parameter.

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_state_ex_enum suspend(util::steady_time_point const &abs_time, threads::thread_id_type const &id, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to suspended and schedules a wakeup for this threads at the given time.

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_state_ex_enum suspend(util::steady_time_point const &abs_time, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to suspended and schedules a wakeup for this threads at the given time.

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_state_ex_enum suspend(util::steady_duration const &rel_time, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to suspended and schedules a wakeup for this threads after the given duration.

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_state_ex_enum suspend(util::steady_duration const &rel_time, threads::thread_id_type const &id, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to suspended and schedules a wakeup for this threads after the given duration.

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_state_ex_enum suspend(std::uint64_t ms, util::thread_description const &description = util::thread_description("this_thread::suspend"), error_code &ec = throws)

The function suspend will return control to the thread manager (suspends the current thread). It sets the new state of this thread to suspended and schedules a wakeup for this threads after the given time (specified in milliseconds).

Note

Must be called from within a HPX-thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::executors::current_executor get_executor(error_code &ec = throws)

Returns a reference to the executor which was used to create the current thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_pool_base *get_pool(error_code &ec = throws)

Returns a pointer to the pool that was used to run the current thread

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

namespace threads

Enums

enum thread_state_enum

The thread_state_enum enumerator encodes the current state of a thread instance

Values:

unknown = 0

active = 1

thread is currently active (running, has resources)

pending = 2

thread is pending (ready to run, but no hardware resource available)

suspended = 3

thread has been suspended (waiting for synchronization event, but still known and under control of the thread-manager)

depleted = 4

thread has been depleted (deeply suspended, it is not known to the thread-manager)

terminated = 5

thread has been stopped an may be garbage collected

staged = 6

this is not a real thread state, but allows to reference staged task descriptions, which eventually will be converted into thread objects

pending_do_not_schedule = 7

pending_boost = 8

enum thread_priority

This enumeration lists all possible thread-priorities for HPX threads.

Values:

thread_priority_unknown = -1

thread_priority_default = 0

Will assign the priority of the task to the default (normal) priority.

thread_priority_low = 1

Task goes onto a special low priority queue and will not be executed until all high/normal priority tasks are done, even if they are added after the low priority task.

thread_priority_normal = 2

Task will be executed when it is taken from the normal priority queue, this is usually a first in-first-out ordering of tasks (depending on scheduler choice). This is the default priority.

thread_priority_high_recursive = 3

The task is a high priority task and any child tasks spawned by this task will be made high priority as well - unless they are specifically flagged as non default priority.

thread_priority_boost = 4

Same as thread_priority_high except that the thread will fall back to thread_priority_normal if resumed after being suspended.

thread_priority_high = 5

Task goes onto a special high priority queue and will be executed before normal/low priority tasks are taken (some schedulers modify the behavior slightly and the documentation for those should be consulted).

enum thread_state_ex_enum

The thread_state_ex_enum enumerator encodes the reason why a thread is being restarted

Values:

wait_unknown = 0

wait_signaled = 1

The thread has been signaled.

wait_timeout = 2

The thread has been reactivated after a timeout.

wait_terminate = 3

The thread needs to be terminated.

wait_abort = 4

The thread needs to be aborted.

enum thread_stacksize

A thread_stacksize references any of the possible stack-sizes for HPX threads.

Values:

thread_stacksize_unknown = -1

thread_stacksize_small = 1

use small stack size

thread_stacksize_medium = 2

use medium sized stack size

thread_stacksize_large = 3

use large stack size

thread_stacksize_huge = 4

use very large stack size

thread_stacksize_current = 5

use size of current thread’s stack

thread_stacksize_default = thread_stacksize_small

use default stack size

thread_stacksize_minimal = thread_stacksize_small

use minimally stack size

thread_stacksize_maximal = thread_stacksize_huge

use maximally stack size

enum thread_schedule_hint_mode

The type of hint given when creating new tasks.

Values:

thread_schedule_hint_mode_none = 0

thread_schedule_hint_mode_thread = 1

thread_schedule_hint_mode_numa = 2

Functions

char const *get_thread_state_name(thread_state_enum state)

Get the readable string representing the name of the given thread_state constant.

char const *get_thread_priority_name(thread_priority priority)

Get the readable string representing the name of the given thread_priority constant.

char const *get_thread_state_ex_name(thread_state_ex_enum state)

Get the readable string representing the name of the given thread_state_ex_enum constant.

char const *get_thread_state_name(thread_state state)

Get the readable string representing the name of the given thread_state constant.

char const *get_stack_size_name(std::ptrdiff_t size)

Get the readable string representing the given stack size constant.

thread_self &get_self()

The function get_self returns a reference to the (OS thread specific) self reference to the current HPX thread.

thread_self *get_self_ptr()

The function get_self_ptr returns a pointer to the (OS thread specific) self reference to the current HPX thread.

thread_self_impl_type *get_ctx_ptr()

The function get_ctx_ptr returns a pointer to the internal data associated with each coroutine.

thread_self *get_self_ptr_checked(error_code &ec = throws)

The function get_self_ptr_checked returns a pointer to the (OS thread specific) self reference to the current HPX thread.

thread_id_type get_self_id()

The function get_self_id returns the HPX thread id of the current thread (or zero if the current thread is not a HPX thread).

thread_id_type get_parent_id()

The function get_parent_id returns the HPX thread id of the current thread’s parent (or zero if the current thread is not a HPX thread).

Note

This function will return a meaningful value only if the code was compiled with HPX_HAVE_THREAD_PARENT_REFERENCE being defined.

std::size_t get_parent_phase()

The function get_parent_phase returns the HPX phase of the current thread’s parent (or zero if the current thread is not a HPX thread).

Note

This function will return a meaningful value only if the code was compiled with HPX_HAVE_THREAD_PARENT_REFERENCE being defined.

std::size_t get_self_stacksize()

The function get_self_stacksize returns the stack size of the current thread (or zero if the current thread is not a HPX thread).

std::uint32_t get_parent_locality_id()

The function get_parent_locality_id returns the id of the locality of the current thread’s parent (or zero if the current thread is not a HPX thread).

Note

This function will return a meaningful value only if the code was compiled with HPX_HAVE_THREAD_PARENT_REFERENCE being defined.

std::uint64_t get_self_component_id()

The function get_self_component_id returns the lva of the component the current thread is acting on

Note

This function will return a meaningful value only if the code was compiled with HPX_HAVE_THREAD_TARGET_ADDRESS being defined.

std::int64_t get_thread_count(thread_state_enum state = unknown)

The function get_thread_count returns the number of currently known threads.

Note

If state == unknown this function will not only return the number of currently existing threads, but will add the number of registered task descriptions (which have not been converted into threads yet).

Parameters

• state: [in] This specifies the thread-state for which the number of threads should be retrieved.

std::int64_t get_thread_count(thread_priority priority, thread_state_enum state = unknown)

The function get_thread_count returns the number of currently known threads.

Note

If state == unknown this function will not only return the number of currently existing threads, but will add the number of registered task descriptions (which have not been converted into threads yet).

Parameters

• priority: [in] This specifies the thread-priority for which the number of threads should be retrieved.
• state: [in] This specifies the thread-state for which the number of threads should be retrieved.

bool enumerate_threads(util::function_nonser<bool(thread_id_type)> const &f, thread_state_enum state = unknown, )

The function enumerate_threads will invoke the given function f for each thread with a matching thread state.

Parameters

• f: [in] The function which should be called for each matching thread. Returning ‘false’ from this function will stop the enumeration process.
• state: [in] This specifies the thread-state for which the threads should be enumerated.

thread_state set_thread_state(thread_id_type const &id, thread_state_enum state = pending, thread_state_ex_enum stateex = wait_signaled, thread_priority priority = thread_priority_normal, bool retry_on_active = true, hpx::error_code &ec = throws)

Set the thread state of the thread referenced by the thread_id id.

Note

If the thread referenced by the parameter id is in thread_state::active state this function schedules a new thread which will set the state of the thread as soon as its not active anymore. The function returns thread_state::active in this case.

Return

This function returns the previous state of the thread referenced by the id parameter. It will return one of the values as defined by the thread_state enumeration. If the thread is not known to the thread-manager the return value will be thread_state::unknown.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread the state should be modified for.
• state: [in] The new state to be set for the thread referenced by the id parameter.
• stateex: [in] The new extended state to be set for the thread referenced by the id parameter.
• priority:
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

thread_id_type set_thread_state(thread_id_type const &id, util::steady_time_point const &abs_time, std::atomic<bool> *started, thread_state_enum state = pending, thread_state_ex_enum stateex = wait_timeout, thread_priority priority = thread_priority_normal, bool retry_on_active = true, error_code &ec = throws)

Set the thread state of the thread referenced by the thread_id id.

Set a timer to set the state of the given thread to the given new value after it expired (at the given time)

Return

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread the state should be modified for.
• abs_time: [in] Absolute point in time for the new thread to be run
• started: [in,out] A helper variable allowing to track the state of the timer helper thread
• state: [in] The new state to be set for the thread referenced by the id parameter.
• stateex: [in] The new extended state to be set for the thread referenced by the id parameter.
• priority:
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

thread_id_type set_thread_state(thread_id_type const &id, util::steady_time_point const &abs_time, thread_state_enum state = pending, thread_state_ex_enum stateex = wait_timeout, thread_priority priority = thread_priority_normal, bool retry_on_active = true, error_code& = throws)

thread_id_type set_thread_state(thread_id_type const &id, util::steady_duration const &rel_time, thread_state_enum state = pending, thread_state_ex_enum stateex = wait_timeout, thread_priority priority = thread_priority_normal, bool retry_on_active = true, error_code &ec = throws)

Set the thread state of the thread referenced by the thread_id id.

Set a timer to set the state of the given thread to the given new value after it expired (after the given duration)

Return

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread the state should be modified for.
• rel_time: [in] Time duration after which the new thread should be run
• state: [in] The new state to be set for the thread referenced by the id parameter.
• stateex: [in] The new extended state to be set for the thread referenced by the id parameter.
• priority:
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

util::thread_description get_thread_description(thread_id_type const &id, error_code &ec = throws)

The function get_thread_description is part of the thread related API allows to query the description of one of the threads known to the thread-manager.

Return

This function returns the description of the thread referenced by the id parameter. If the thread is not known to the thread-manager the return value will be the string “<unknown>”.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread being queried.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

util::thread_description set_thread_description(thread_id_type const &id, util::thread_description const &desc = util::thread_description(), error_code &ec = throws)

util::thread_description get_thread_lco_description(thread_id_type const &id, error_code &ec = throws)

util::thread_description set_thread_lco_description(thread_id_type const &id, util::thread_description const &desc = util::thread_description(), error_code &ec = throws)

thread_state get_thread_state(thread_id_type const &id, error_code &ec = throws)

The function get_thread_backtrace is part of the thread related API allows to query the currently stored thread back trace (which is captured during thread suspension).

Return

This function returns the currently captured stack back trace of the thread referenced by the id parameter. If the thread is not known to the thread-manager the return value will be the zero.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception. The function get_thread_state is part of the thread related API. It queries the state of one of the threads known to the thread-manager.

Return

This function returns the thread state of the thread referenced by the id parameter. If the thread is not known to the thread-manager the return value will be terminated.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread being queried.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

Parameters

• id: [in] The thread id of the thread the state should be modified for.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::size_t get_thread_phase(thread_id_type const &id, error_code &ec = throws)

The function get_thread_phase is part of the thread related API. It queries the phase of one of the threads known to the thread-manager.

Return

This function returns the thread phase of the thread referenced by the id parameter. If the thread is not known to the thread-manager the return value will be ~0.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread the phase should be modified for.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::size_t get_numa_node_number()

bool get_thread_interruption_enabled(thread_id_type const &id, error_code &ec = throws)

Returns whether the given thread can be interrupted at this point.

Return

This function returns true if the given thread can be interrupted at this point in time. It will return false otherwise.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread which should be queried.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

bool set_thread_interruption_enabled(thread_id_type const &id, bool enable, error_code &ec = throws)

Set whether the given thread can be interrupted at this point.

Return

This function returns the previous value of whether the given thread could have been interrupted.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread which should receive the new value.
• enable: [in] This value will determine the new interruption enabled status for the given thread.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

bool get_thread_interruption_requested(thread_id_type const &id, error_code &ec = throws)

Returns whether the given thread has been flagged for interruption.

Return

This function returns true if the given thread was flagged for interruption. It will return false otherwise.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread which should be queried.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void interrupt_thread(thread_id_type const &id, bool flag, error_code &ec = throws)

Flag the given thread for interruption.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread which should be interrupted.
• flag: [in] The flag encodes whether the thread should be interrupted (if it is true), or ‘uninterrupted’ (if it is false).
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

void interrupt_thread(thread_id_type const &id, error_code &ec = throws)

void interruption_point(thread_id_type const &id, error_code &ec = throws)

Interrupt the current thread at this point if it was canceled. This will throw a thread_interrupted exception, which will cancel the thread.

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread which should be interrupted.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

threads::thread_priority get_thread_priority(thread_id_type const &id, error_code &ec = throws)

Return priority of the given thread

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread whose priority is queried.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

std::ptrdiff_t get_stack_size(thread_id_type const &id, error_code &ec = throws)

Return stack size of the given thread

Note

As long as ec is not pre-initialized to hpx::throws this function doesn’t throw but returns the result code using the parameter ec. Otherwise it throws an instance of hpx::exception.

Parameters

• id: [in] The thread id of the thread whose priority is queried.
• ec: [in,out] this represents the error status on exit, if this is pre-initialized to hpx::throws the function will throw on error instead.

threads::executors::current_executor get_executor(thread_id_type const &id, error_code &ec = throws)

Returns a reference to the executor which was used to create the given thread.

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

threads::thread_pool_base *get_pool(thread_id_type const &id, error_code &ec = throws)

Returns a pointer to the pool that was used to run the current thread

Exceptions

• If: &ec != &throws, never throws, but will set ec to an appropriate value when an error occurs. Otherwise, this function will throw an hpx::exception with an error code of hpx::yield_aborted if it is signaled with wait_aborted. If called outside of a HPX-thread, this function will throw an hpx::exception with an error code of hpx::null_thread_id. If this function is called while the thread-manager is not running, it will throw an hpx::exception with an error code of hpx::invalid_status.

namespace policies

Enums

enum scheduler_mode

This enumeration describes the possible modes of a scheduler.

Values:

nothing_special = 0

can be used to disable all other options.

As the name suggests, this option

do_background_work = 0x1

The scheduler will periodically call a provided callback function from a special HPX thread to enable performing background-work, for instance driving networking progress or garbage-collect AGAS.

reduce_thread_priority = 0x02

os-thread driving the scheduler will be reduced below normal.

The kernel priority of the

delay_exit = 0x04

The scheduler will wait for some unspecified amount of time before exiting the scheduling loop while being terminated to make sure no other work is being scheduled during processing the shutdown request.

Some schedulers have the capability to act as ‘embedded’ schedulers. In this case it needs to periodically invoke a provided callback into the outer scheduler more frequently than normal. This option enables this behavior.

enable_elasticity = 0x10

This option allows for the scheduler to dynamically increase and reduce the number of processing units it runs on. Setting this value not succeed for schedulers that do not support this functionality.

enable_stealing = 0x20

schedulers to explicitly disable thread stealing

This option allows for certain

enable_idle_backoff = 0x40

schedulers to explicitly disable exponential idle-back off

This option allows for certain

default_mode = do_background_work | reduce_thread_priority | delay_exit | enable_stealing | enable_idle_backoff

This option represents the default mode.

all_flags = do_background_work | reduce_thread_priority | delay_exit | fast_idle_mode | enable_elasticity | enable_stealing | enable_idle_backoff

namespace traits

namespace util

Functions

std::ostream &operator<<(std::ostream &ost, checkpoint const &ckp)

Operator<< Overload

This overload is the main way to write data from a checkpoint to an object such as a file. Inside the function, the size of the checkpoint will be written to the stream before the checkpoint’s data. The operator>> overload uses this to read the correct number of bytes. Be mindful of this additional write and read when you use different facilities to write out or read in data to a checkpoint!

Parameters

• ost: Output stream to write to.
• ckp: Checkpoint to copy from.

Return

Operator<< returns the ostream object.

std::istream &operator>>(std::istream &ist, checkpoint &ckp)

Operator>> Overload

This overload is the main way to read in data from an object such as a file to a checkpoint. It is important to note that inside the function, the first variable to be read is the size of the checkpoint. This size variable is written to the stream before the checkpoint’s data in the operator<< overload. Be mindful of this additional read and write when you use different facilities to read in or write out data from a checkpoint!

Parameters

• ist: Input stream to write from.
• ckp: Checkpoint to write to.

Return

Operator>> returns the ostream object.

template<typename T, typename ...Ts, typename U = typename std::enable_if<!hpx::traits::is_launch_policy<T>::value && !std::is_same<typename std::decay<T>::type, checkpoint>::value>::type>
hpx::future<checkpoint> save_checkpoint(T &&t, Ts&&... ts)

Save_checkpoint

Save_checkpoint takes any number of objects which a user may wish to store and returns a future to a checkpoint object. Additionally the function can take a policy as a first object which changes its behavior depending on the policy passed to it. Most notably, if a sync policy is used save_checkpoint will simply return a checkpoint object.

Template Parameters

• T: Containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• Ts: More containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• U: This parameter is used to make sure that T is not a launch policy or a checkpoint. This forces the compiler to choose the correct overload.

Parameters

• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

Return

Save_checkpoint returns a future to a checkpoint with one exception: if you pass hpx::launch::sync as the first argument. In this case save_checkpoint will simply return a checkpoint.

template<typename T, typename ...Ts>
hpx::future<checkpoint> save_checkpoint(checkpoint &&c, T &&t, Ts&&... ts)

Save_checkpoint - Take a pre-initialized checkpoint

Save_checkpoint takes any number of objects which a user may wish to store and returns a future to a checkpoint object. Additionally the function can take a policy as a first object which changes its behavior depending on the policy passed to it. Most notably, if a sync policy is used save_checkpoint will simply return a checkpoint object.

Template Parameters

• T: Containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• Ts: More containers passed to save_checkpoint to be serialized and placed into a checkpoint object.

Parameters

• c: Takes a pre-initialized checkpoint to copy data into.
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

Return

Save_checkpoint returns a future to a checkpoint with one exception: if you pass hpx::launch::sync as the first argument. In this case save_checkpoint will simply return a checkpoint.

template<typename T, typename ...Ts>
hpx::future<checkpoint> save_checkpoint(hpx::launch p, T &&t, Ts&&... ts)

Save_checkpoint - Policy overload

Save_checkpoint takes any number of objects which a user may wish to store and returns a future to a checkpoint object. Additionally the function can take a policy as a first object which changes its behavior depending on the policy passed to it. Most notably, if a sync policy is used save_checkpoint will simply return a checkpoint object.

Template Parameters

• T: Containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• Ts: More containers passed to save_checkpoint to be serialized and placed into a checkpoint object.

Parameters

• p: Takes an HPX launch policy. Allows the user to change the way the function is launched i.e. async, sync, etc.
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

Return

Save_checkpoint returns a future to a checkpoint with one exception: if you pass hpx::launch::sync as the first argument. In this case save_checkpoint will simply return a checkpoint.

template<typename T, typename ...Ts>
hpx::future<checkpoint> save_checkpoint(hpx::launch p, checkpoint &&c, T &&t, Ts&&... ts)

Save_checkpoint - Policy overload & pre-initialized checkpoint

Save_checkpoint takes any number of objects which a user may wish to store and returns a future to a checkpoint object. Additionally the function can take a policy as a first object which changes its behavior depending on the policy passed to it. Most notably, if a sync policy is used save_checkpoint will simply return a checkpoint object.

Template Parameters

• T: Containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• Ts: More containers passed to save_checkpoint to be serialized and placed into a checkpoint object.

Parameters

• p: Takes an HPX launch policy. Allows the user to change the way the function is launched i.e. async, sync, etc.
• c: Takes a pre-initialized checkpoint to copy data into.
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

Return

Save_checkpoint returns a future to a checkpoint with one exception: if you pass hpx::launch::sync as the first argument. In this case save_checkpoint will simply return a checkpoint.

template<typename T, typename ...Ts, typename U = typename std::enable_if<!std::is_same<typename std::decay<T>::type, checkpoint>::value>::type>
checkpoint save_checkpoint(hpx::launch::sync_policy sync_p, T &&t, Ts&&... ts)

Save_checkpoint - Sync_policy overload

Save_checkpoint takes any number of objects which a user may wish to store and returns a future to a checkpoint object. Additionally the function can take a policy as a first object which changes its behavior depending on the policy passed to it. Most notably, if a sync policy is used save_checkpoint will simply return a checkpoint object.

Template Parameters

• T: Containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• Ts: More containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• U: This parameter is used to make sure that T is not a checkpoint. This forces the compiler to choose the correct overload.

Parameters

• sync_p: hpx::launch::sync_policy
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

Return

Save_checkpoint which is passed hpx::launch::sync_policy will return a checkpoint which contains the serialized values checkpoint.

template<typename T, typename ...Ts>
checkpoint save_checkpoint(hpx::launch::sync_policy sync_p, checkpoint &&c, T &&t, Ts&&... ts)

Save_checkpoint - Sync_policy overload & pre-init. checkpoint

Save_checkpoint takes any number of objects which a user may wish to store and returns a future to a checkpoint object. Additionally the function can take a policy as a first object which changes its behavior depending on the policy passed to it. Most notably, if a sync policy is used save_checkpoint will simply return a checkpoint object.

Template Parameters

• T: Containers passed to save_checkpoint to be serialized and placed into a checkpoint object.
• Ts: More containers passed to save_checkpoint to be serialized and placed into a checkpoint object.

Parameters

• sync_p: hpx::launch::sync_policy
• c: Takes a pre-initialized checkpoint to copy data into.
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

Return

Save_checkpoint which is passed hpx::launch::sync_policy will return a checkpoint which contains the serialized values checkpoint.

template<typename T, typename ...Ts>
void restore_checkpoint(checkpoint const &c, T &t, Ts&... ts)

Resurrect

Restore_checkpoint takes a checkpoint object as a first argument and the containers which will be filled from the byte stream (in the same order as they were placed in save_checkpoint).

Return

Restore_checkpoint returns void.

Template Parameters

• T: A container to restore.
• Ts: Other containers to restore. Containers must be in the same order that they were inserted into the checkpoint.

Parameters

• c: The checkpoint to restore.
• t: A container to restore.
• ts: Other containers to restore Containers must be in the same order that they were inserted into the checkpoint.

void attach_debugger()

Tries to break an attached debugger, if not supported a loop is invoked which gives enough time to attach a debugger manually.

template<typename F, typename ... Ts>HPX_HOST_DEVICE util::invoke_result<F, Ts...>::type hpx::util::invoke(F && f, Ts &&... vs)

Invokes the given callable object f with the content of the argument pack vs

Return

The result of the callable object when it’s called with the given argument types.

Note

This function is similar to std::invoke (C++17)

Parameters

• f: Requires to be a callable object. If f is a member function pointer, the first argument in the pack will be treated as the callee (this object).
• vs: An arbitrary pack of arguments

Exceptions

• std::exception: like objects thrown by call to object f with the argument types vs.

template<typename R, typename F, typename ... Ts>HPX_HOST_DEVICE R hpx::util::invoke_r(F && f, Ts &&... vs)

Invokes the given callable object f with the content of the argument pack vs

Return

The result of the callable object when it’s called with the given argument types.

Note

This function is similar to std::invoke (C++17)

Parameters

• f: Requires to be a callable object. If f is a member function pointer, the first argument in the pack will be treated as the callee (this object).
• vs: An arbitrary pack of arguments

Exceptions

• std::exception: like objects thrown by call to object f with the argument types vs.

Template Parameters

• R: The result type of the function when it’s called with the content of the given argument types vs.

template<typename F, typename Tuple>HPX_HOST_DEVICE detail::invoke_fused_result<F, Tuple>::type hpx::util::invoke_fused(F && f, Tuple && t)

Invokes the given callable object f with the content of the sequenced type t (tuples, pairs)

Return

The result of the callable object when it’s called with the content of the given sequenced type.

Note

This function is similar to std::apply (C++17)

Parameters

• f: Must be a callable object. If f is a member function pointer, the first argument in the sequenced type will be treated as the callee (this object).
• t: A type which is content accessible through a call to hpx::util::get.

Exceptions

• std::exception: like objects thrown by call to object f with the arguments contained in the sequenceable type t.

template<typename R, typename F, typename Tuple>HPX_HOST_DEVICE R hpx::util::invoke_fused_r(F && f, Tuple && t)

Invokes the given callable object f with the content of the sequenced type t (tuples, pairs)

Return

The result of the callable object when it’s called with the content of the given sequenced type.

Note

This function is similar to std::apply (C++17)

Parameters

• f: Must be a callable object. If f is a member function pointer, the first argument in the sequenced type will be treated as the callee (this object).
• t: A type which is content accessible through a call to hpx::util::get.

Exceptions

• std::exception: like objects thrown by call to object f with the arguments contained in the sequenceable type t.

Template Parameters

• R: The result type of the function when it’s called with the content of the given sequenced type.

template<typename Mapper, typename... T><unspecified> hpx::util::map_pack(Mapper && mapper, T &&... pack)

Maps the pack with the given mapper.

This function tries to visit all plain elements which may be wrapped in:

• homogeneous containers (std::vector, std::list)
• heterogenous containers (hpx::tuple, std::pair, std::array) and re-assembles the pack with the result of the mapper. Mapping from one type to a different one is supported.

Elements that aren’t accepted by the mapper are routed through and preserved through the hierarchy.

// Maps all integers to floats
map_pack([](int value) {
    return float(value);
},
1, hpx::util::make_tuple(2, std::vector<int>{3, 4}), 5);

Return

The mapped element or in case the pack contains multiple elements, the pack is wrapped into a hpx::tuple.

Exceptions

• std::exception: like objects which are thrown by an invocation to the mapper.

Parameters

• mapper: A callable object, which accept an arbitrary type and maps it to another type or the same one.
• pack: An arbitrary variadic pack which may contain any type.

template<typename Visitor, typename ...T>
auto traverse_pack_async(Visitor &&visitor, T&&... pack)

Traverses the pack with the given visitor in an asynchronous way.

This function works in the same way as traverse_pack, however, we are able to suspend and continue the traversal at later time. Thus we require a visitor callable object which provides three operator() overloads as depicted by the code sample below:

struct my_async_visitor
{
    template <typename T>
    bool operator()(async_traverse_visit_tag, T&& element)
    {
        return true;
    }

    template <typename T, typename N>
    void operator()(async_traverse_detach_tag, T&& element, N&& next)
    {
    }

    template <typename T>
    void operator()(async_traverse_complete_tag, T&& pack)
    {
    }
};

See

traverse_pack for a detailed description about the traversal behavior and capabilities.

Return

A boost::intrusive_ptr that references an instance of the given visitor object.

Parameters

• visitor: A visitor object which provides the three operator() overloads that were described above. Additionally the visitor must be compatible for referencing it from a boost::intrusive_ptr. The visitor should must have a virtual destructor!
• pack: The arbitrary parameter pack which is traversed asynchronously. Nested objects inside containers and tuple like types are traversed recursively.

template<typename Allocator, typename Visitor, typename ...T>
auto traverse_pack_async_allocator(Allocator const &alloc, Visitor &&visitor, T&&... pack)

Traverses the pack with the given visitor in an asynchronous way.

This function works in the same way as traverse_pack, however, we are able to suspend and continue the traversal at later time. Thus we require a visitor callable object which provides three operator() overloads as depicted by the code sample below:

struct my_async_visitor
{
    template <typename T>
    bool operator()(async_traverse_visit_tag, T&& element)
    {
        return true;
    }

    template <typename T, typename N>
    void operator()(async_traverse_detach_tag, T&& element, N&& next)
    {
    }

    template <typename T>
    void operator()(async_traverse_complete_tag, T&& pack)
    {
    }
};

See

traverse_pack for a detailed description about the traversal behavior and capabilities.

Return

A boost::intrusive_ptr that references an instance of the given visitor object.

Parameters

• visitor: A visitor object which provides the three operator() overloads that were described above. Additionally the visitor must be compatible for referencing it from a boost::intrusive_ptr. The visitor should must have a virtual destructor!
• pack: The arbitrary parameter pack which is traversed asynchronously. Nested objects inside containers and tuple like types are traversed recursively.
• alloc: Allocator instance to use to create the traversal frame.

template<typename ...Args>
auto unwrap(Args&&... args)

A helper function for retrieving the actual result of any hpx::lcos::future like type which is wrapped in an arbitrary way.

Unwraps the given pack of arguments, so that any hpx::lcos::future object is replaced by its future result type in the argument pack:

• hpx::future<int> -> int
• hpx::future<std::vector<float>> -> std::vector<float>
• std::vector<future<float>> -> std::vector<float>

The function is capable of unwrapping hpx::lcos::future like objects that are wrapped inside any container or tuple like type, see hpx::util::map_pack() for a detailed description about which surrounding types are supported. Non hpx::lcos::future like types are permitted as arguments and passed through.

// Single arguments
int i1 = hpx:util::unwrap(hpx::lcos::make_ready_future(0));

// Multiple arguments
hpx::tuple<int, int> i2 =
    hpx:util::unwrap(hpx::lcos::make_ready_future(1),
                     hpx::lcos::make_ready_future(2));

Note

This function unwraps the given arguments until the first traversed nested hpx::lcos::future which corresponds to an unwrapping depth of one. See hpx::util::unwrap_n() for a function which unwraps the given arguments to a particular depth or hpx::util::unwrap_all() that unwraps all future like objects recursively which are contained in the arguments.

Return

Depending on the count of arguments this function returns a hpx::util::tuple containing the unwrapped arguments if multiple arguments are given. In case the function is called with a single argument, the argument is unwrapped and returned.

Parameters

• args: the arguments that are unwrapped which may contain any arbitrary future or non future type.

Exceptions

• std::exception: like objects in case any of the given wrapped hpx::lcos::future objects were resolved through an exception. See hpx::lcos::future::get() for details.

template<std::size_t Depth, typename ...Args>
auto unwrap_n(Args&&... args)

An alterntive version of hpx::util::unwrap(), which unwraps the given arguments to a certain depth of hpx::lcos::future like objects.

See unwrap for a detailed description.

Template Parameters

• Depth: The count of hpx::lcos::future like objects which are unwrapped maximally.

template<typename ...Args>
auto unwrap_all(Args&&... args)

An alterntive version of hpx::util::unwrap(), which unwraps the given arguments recursively so that all contained hpx::lcos::future like objects are replaced by their actual value.

See hpx::util::unwrap() for a detailed description.

template<typename T>
auto unwrapping(T &&callable)

Returns a callable object which unwraps its arguments upon invocation using the hpx::util::unwrap() function and then passes the result to the given callable object.

auto callable = hpx::util::unwrapping([](int left, int right) {
    return left + right;
});

int i1 = callable(hpx::lcos::make_ready_future(1),
                  hpx::lcos::make_ready_future(2));

See hpx::util::unwrap() for a detailed description.

Parameters

• callable: the callable object which which is called with the result of the corresponding unwrap function.

template<std::size_t Depth, typename T>
auto unwrapping_n(T &&callable)

Returns a callable object which unwraps its arguments upon invocation using the hpx::util::unwrap_n() function and then passes the result to the given callable object.

See hpx::util::unwrapping() for a detailed description.

template<typename T>
auto unwrapping_all(T &&callable)

Returns a callable object which unwraps its arguments upon invocation using the hpx::util::unwrap_all() function and then passes the result to the given callable object.

See hpx::util::unwrapping() for a detailed description.

namespace functional

file migrate_from_storage.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/error_code.hpp>#include <hpx/lcos/detail/future_data.hpp>#include <hpx/lcos/local/detail/condition_variable.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/util/steady_clock.hpp>#include <boost/intrusive/slist.hpp>#include <cstddef>#include <mutex>#include <utility>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <type_traits>#include <hpx/runtime/threads/coroutines/detail/get_stack_pointer.hpp>#include <limits>#include <hpx/runtime/threads/thread_executor.hpp>#include <hpx/runtime/get_os_thread_count.hpp>#include <hpx/runtime/threads/cpu_mask.hpp>#include <hpx/util/assert.hpp>#include <climits>#include <cstdint>#include <string>#include <hpx/runtime/threads/policies/scheduler_mode.hpp>#include <hpx/runtime/threads/topology.hpp>#include <hpx/compat/thread.hpp>#include <thread>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/naming_fwd.hpp>#include <hpx/runtime/resource/partitioner_fwd.hpp>#include <hpx/runtime/threads/policies/callback_notifier.hpp>#include <hpx/runtime/threads_fwd.hpp>#include <hpx/util/function.hpp>#include <exception>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>#include <memory>#include <hpx/util/spinlock.hpp>#include <hpx/util/itt_notify.hpp>#include <hpx/util/register_locks.hpp>#include <boost/smart_ptr/detail/spinlock.hpp>#include <hpx/util/static.hpp>#include <hpx/compat/mutex.hpp>#include <iosfwd>#include <vector>#include <hwloc.h>#include <hpx/util/atomic_count.hpp>#include <atomic>#include <hpx/util/thread_description.hpp>#include <hpx/util/unique_function.hpp>#include <boost/intrusive_ptr.hpp>#include <chrono>#include <hpx/runtime/threads/thread_helpers.hpp>#include <hpx/throw_exception.hpp>#include <hpx/traits/future_access.hpp>#include <hpx/traits/future_traits.hpp>#include <hpx/traits/is_future.hpp>#include <boost/ref.hpp>#include <functional>#include <hpx/traits/get_remote_result.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/assert_owns_lock.hpp>#include <hpx/traits/has_member_xxx.hpp>#include <hpx/util/bind.hpp>#include <hpx/traits/get_function_address.hpp>#include <hpx/traits/get_function_annotation.hpp>#include <hpx/traits/is_action.hpp>#include <hpx/traits/is_bind_expression.hpp>#include <hpx/traits/is_placeholder.hpp>#include <boost/bind/arg.hpp>#include <hpx/util/decay.hpp>#include <hpx/util/detail/pack.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/invoke_fused.hpp>#include <hpx/util/result_of.hpp>#include <hpx/util/tuple.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/util/one_shot.hpp>#include <hpx/util/unused.hpp>#include <boost/container/small_vector.hpp>#include <hpx/lcos/detail/future_traits.hpp>#include <hpx/util/always_void.hpp>#include <iterator>#include <hpx/lcos_fwd.hpp>#include <hpx/traits/is_component.hpp>#include <hpx/traits/promise_local_result.hpp>#include <hpx/traits/promise_remote_result.hpp>#include <hpx/runtime/actions/continuation_fwd.hpp>#include <hpx/runtime/serialization/detail/polymorphic_nonintrusive_factory.hpp>#include <hpx/preprocessor/stringize.hpp>#include <hpx/preprocessor/strip_parens.hpp>#include <hpx/runtime/serialization/detail/non_default_constructible.hpp>#include <hpx/traits/needs_automatic_registration.hpp>#include <hpx/traits/polymorphic_traits.hpp>#include <hpx/traits/has_xxx.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/util/debug/demangle_helper.hpp>#include <hpx/util/jenkins_hash.hpp>#include <typeinfo>#include <unordered_map>#include <hpx/traits/acquire_shared_state.hpp>#include <hpx/util/range.hpp>#include <hpx/traits/detail/reserve.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/traits/is_future_range.hpp>#include <algorithm>#include <hpx/traits/concepts.hpp>#include <hpx/traits/future_then_result.hpp>#include <hpx/util/identity.hpp>#include <hpx/util/lazy_conditional.hpp>#include <hpx/traits/is_executor.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_launch_policy.hpp>#include <hpx/traits/executor_traits.hpp>#include <hpx/util/detected.hpp>#include <hpx/util/allocator_deleter.hpp>#include <hpx/util/internal_allocator.hpp>#include <hpx/util/lazy_enable_if.hpp>#include <hpx/util/serialize_exception.hpp>#include <hpx/lcos/local/packaged_continuation.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/execution_fwd.hpp>#include <hpx/parallel/executors/fused_bulk_execute.hpp>#include <hpx/util/deferred_call.hpp>#include <hpx/exception_list.hpp>#include <hpx/exception.hpp>#include <boost/system/error_code.hpp>#include <list>#include <hpx/lcos/dataflow.hpp>#include <hpx/lcos/detail/future_transforms.hpp>#include <hpx/traits/acquire_future.hpp>#include <array>#include <hpx/runtime/get_worker_thread_num.hpp>#include <hpx/traits/extract_action.hpp>#include <hpx/util/pack_traversal_async.hpp>#include <hpx/util/detail/pack_traversal_async_impl.hpp>#include <hpx/util/detail/container_category.hpp>#include <hpx/traits/is_tuple_like.hpp>#include <hpx/parallel/executors/parallel_executor.hpp>#include <hpx/async_launch_policy_dispatch.hpp>#include <hpx/lcos/async_fwd.hpp>#include <hpx/lcos/local/futures_factory.hpp>#include <hpx/lcos/local/latch.hpp>#include <hpx/util/cache_aligned_data.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/traits/is_iterator.hpp>#include <boost/iterator/iterator_categories.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <cstdlib>#include <hpx/parallel/executors/post_policy_dispatch.hpp>#include <hpx/parallel/executors/static_chunk_size.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/runtime/serialization/access.hpp>#include <hpx/runtime/serialization/brace_initializable_fwd.hpp>#include <hpx/traits/brace_initializable_traits.hpp>#include <hpx/runtime/serialization/input_archive.hpp>#include <hpx/runtime/serialization/basic_archive.hpp>#include <iostream>#include <map>#include <hpx/runtime/serialization/detail/raw_ptr.hpp>#include <hpx/runtime/serialization/detail/pointer.hpp>#include <hpx/runtime/serialization/detail/polymorphic_id_factory.hpp>#include <hpx/runtime/serialization/detail/polymorphic_intrusive_factory.hpp>#include <hpx/runtime/serialization/string.hpp>#include <hpx/runtime/serialization/input_container.hpp>#include <hpx/runtime/serialization/binary_filter.hpp>#include <hpx/runtime/serialization/container.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/util/spinlock_pool.hpp>#include <hpx/util/fibhash.hpp>#include <boost/version.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/traits/is_bitwise_serializable.hpp>#include <hpx/util/detail/yield_k.hpp>#include <hpx/runtime/naming/id_type_impl.hpp>#include <hpx/runtime/serialization/serialization_chunk.hpp>#include <cstring>#include <hpx/traits/serialization_access_data.hpp>#include <boost/cstdint.hpp>#include <boost/predef/other/endian.h>#include <hpx/runtime/serialization/output_archive.hpp>#include <hpx/runtime/serialization/output_container.hpp>#include <hpx/runtime/serialization/detail/polymorphic_nonintrusive_factory_impl.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <hpx/parallel/executors/execution_parameters_fwd.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/util/detail/unwrap_impl.hpp>#include <hpx/util/pack_traversal.hpp>#include <hpx/util/detail/pack_traversal_impl.hpp>#include <hpx/lcos/wait_all.hpp>#include <hpx/traits/detail/wrap_int.hpp>#include <stdexcept>#include <hpx/components/component_storage/server/migrate_from_storage.hpp>#include <hpx/runtime/components/runtime_support.hpp>#include <hpx/runtime/applier/applier.hpp>#include <hpx/runtime/agas_fwd.hpp>#include <hpx/runtime/applier_fwd.hpp>#include <hpx/runtime/components/component_type.hpp>#include <hpx/preprocessor/expand.hpp>#include <hpx/preprocessor/nargs.hpp>#include <hpx/traits/component_type_database.hpp>#include <hpx/util_fwd.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/runtime/parcelset/parcel.hpp>#include <hpx/runtime/actions_fwd.hpp>#include <hpx/runtime/parcelset_fwd.hpp>#include <hpx/util/thread_specific_ptr.hpp>#include <boost/thread/tss.hpp>#include <hpx/runtime/components/stubs/runtime_support.hpp>#include <hpx/async.hpp>#include <hpx/lcos/async.hpp>#include <hpx/lcos/detail/async_implementations.hpp>#include <hpx/lcos/detail/async_implementations_fwd.hpp>#include <hpx/lcos/packaged_action.hpp>#include <hpx/lcos/promise.hpp>#include <hpx/lcos/detail/promise_base.hpp>#include <hpx/lcos/detail/promise_lco.hpp>#include <hpx/lcos/base_lco_with_value.hpp>#include <hpx/lcos/base_lco.hpp>#include <hpx/runtime/actions/basic_action.hpp>#include <hpx/lcos/sync_fwd.hpp>#include <hpx/runtime/actions/action_support.hpp>#include <hpx/runtime/components/pinned_ptr.hpp>#include <hpx/runtime/get_lva.hpp>#include <hpx/runtime/components_fwd.hpp>#include <hpx/traits/managed_component_policies.hpp>#include <hpx/traits/action_decorate_function.hpp>#include <hpx/traits/component_pin_support.hpp>#include <hpx/runtime/serialization/base_object.hpp>#include <hpx/runtime/threads/thread_init_data.hpp>#include <hpx/traits/action_remote_result.hpp>#include <hpx/runtime/actions/basic_action_fwd.hpp>#include <hpx/runtime/actions/continuation.hpp>#include <hpx/runtime/actions/action_priority.hpp>#include <hpx/traits/action_priority.hpp>#include <hpx/runtime/actions/trigger.hpp>#include <hpx/runtime/agas/interface.hpp>#include <boost/dynamic_bitset.hpp>#include <hpx/runtime/trigger_lco.hpp>#include <hpx/runtime/applier/detail/apply_implementations_fwd.hpp>#include <hpx/traits/is_continuation.hpp>#include <hpx/util/logging.hpp>#include <hpx/runtime/actions/detail/action_factory.hpp>#include <hpx/runtime/actions/detail/invocation_count_registry.hpp>#include <hpx/performance_counters/counters_fwd.hpp>#include <hpx/runtime/actions/preassigned_action_id.hpp>#include <hpx/runtime/actions/transfer_action.hpp>#include <hpx/runtime/actions/transfer_base_action.hpp>#include <hpx/runtime/actions/base_action.hpp>#include <hpx/runtime/threads/thread_id_type.hpp>#include <hpx/runtime/serialization/unique_ptr.hpp>#include <hpx/traits/action_does_termination_detection.hpp>#include <hpx/traits/action_message_handler.hpp>#include <hpx/traits/action_schedule_thread.hpp>#include <hpx/traits/action_serialization_filter.hpp>#include <hpx/traits/action_stacksize.hpp>#include <hpx/traits/action_was_object_migrated.hpp>#include <hpx/util/get_and_reset_value.hpp>#include <hpx/runtime/applier/apply_helper.hpp>#include <hpx/runtime_fwd.hpp>#include <hpx/runtime/basename_registration_fwd.hpp>#include <hpx/components_fwd.hpp>#include <hpx/runtime/components/make_client.hpp>#include <hpx/traits/is_client.hpp>#include <hpx/runtime/config_entry.hpp>#include <hpx/runtime/find_localities.hpp>#include <hpx/runtime/get_colocation_id.hpp>#include <hpx/runtime/get_locality_id.hpp>#include <hpx/runtime/get_locality_name.hpp>#include <hpx/runtime/get_num_localities.hpp>#include <hpx/runtime/get_thread_name.hpp>#include <hpx/runtime/report_error.hpp>#include <hpx/runtime/runtime_fwd.hpp>#include <hpx/runtime/runtime_mode.hpp>#include <hpx/runtime/set_parcel_write_handler.hpp>#include <hpx/runtime/shutdown_function.hpp>#include <hpx/runtime/startup_function.hpp>#include <hpx/state.hpp>#include <hpx/traits/action_continuation.hpp>#include <hpx/traits/action_decorate_continuation.hpp>#include <hpx/traits/action_select_direct_execution.hpp>#include <hpx/runtime/parcelset/detail/per_action_data_counter_registry.hpp>#include <hpx/runtime/actions/transfer_continuation_action.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <boost/utility/string_ref.hpp>#include <sstream>#include <hpx/runtime/actions/component_action.hpp>#include <hpx/runtime/components/server/managed_component_base.hpp>#include <hpx/runtime/components/server/create_component_fwd.hpp>#include <hpx/runtime/components/server/component_heap.hpp>#include <hpx/util/reinitializable_static.hpp>#include <hpx/util/bind_front.hpp>#include <hpx/util/static_reinit.hpp>#include <hpx/runtime/components/server/wrapper_heap.hpp>#include <hpx/util/generate_unique_ids.hpp>#include <hpx/util/wrapper_heap_base.hpp>#include <new>#include <hpx/runtime/components/server/wrapper_heap_list.hpp>#include <hpx/util/one_size_heap_list.hpp>#include <hpx/util/unlock_guard.hpp>#include <hpx/plugins/parcel/coalescing_message_handler_registration.hpp>#include <hpx/runtime/components/server/component_base.hpp>#include <hpx/util/ini.hpp>#include <boost/lexical_cast.hpp>#include <hpx/lcos/local/promise.hpp>#include <boost/utility/swap.hpp>#include <hpx/runtime/applier/apply.hpp>#include <hpx/runtime/applier/detail/apply_implementations.hpp>#include <hpx/traits/action_is_target_valid.hpp>#include <hpx/traits/component_supports_migration.hpp>#include <hpx/util/format.hpp>#include <cctype>#include <cstdio>#include <ostream>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <hpx/lcos/detail/async_colocated_fwd.hpp>#include <hpx/runtime/naming/unmanaged.hpp>#include <hpx/runtime/parcelset/detail/parcel_await.hpp>#include <hpx/runtime/parcelset/put_parcel.hpp>#include <hpx/runtime.hpp>#include <hpx/performance_counters/counters.hpp>#include <hpx/runtime/parcelset/locality.hpp>#include <hpx/runtime/serialization/map.hpp>#include <hpx/runtime/thread_hooks.hpp>#include <hpx/util/runtime_configuration.hpp>#include <hpx/runtime/components/static_factory_data.hpp>#include <hpx/util/plugin/export_plugin.hpp>#include <hpx/util/plugin/abstract_factory.hpp>#include <hpx/util/plugin/virtual_constructor.hpp>#include <hpx/util/plugin/config.hpp>#include <boost/any.hpp>#include <boost/shared_ptr.hpp>#include <hpx/util/plugin/concrete_factory.hpp>#include <hpx/util/plugin/plugin_wrapper.hpp>#include <boost/algorithm/string/case_conv.hpp>#include <hpx/util/plugin/dll.hpp>#include <hpx/util/plugin/detail/dll_dlopen.hpp>#include <boost/filesystem/convenience.hpp>#include <boost/filesystem/path.hpp>#include <link.h>#include <dlfcn.h>#include <limits.h>#include <hpx/plugins/plugin_registry_base.hpp>#include <hpx/util/plugin.hpp>#include <hpx/util/plugin/plugin_factory.hpp>#include <boost/filesystem.hpp>#include <set>#include <hpx/runtime/naming/split_gid.hpp>#include <hpx/runtime/parcelset/parcelhandler.hpp>#include <hpx/runtime/parcelset/parcelport.hpp>#include <hpx/performance_counters/parcels/data_point.hpp>#include <hpx/performance_counters/parcels/gatherer.hpp>#include <hpx/lcos/local/no_mutex.hpp>#include <hpx/runtime/parcelset/detail/per_action_data_counter.hpp>#include <hpx/util/high_resolution_timer.hpp>#include <hpx/util/high_resolution_clock.hpp>#include <hpx/plugins/parcelport_factory_base.hpp>#include <hpx/traits/component_type_is_compatible.hpp>#include <hpx/traits/is_valid_action.hpp>#include <hpx/runtime/applier/apply_callback.hpp>#include <boost/asio/error.hpp>#include <hpx/runtime/threads/thread.hpp>#include <hpx/lcos/sync.hpp>#include <hpx/lcos/detail/sync_implementations.hpp>#include <hpx/lcos/detail/sync_implementations_fwd.hpp>#include <hpx/lcos/async_continue.hpp>#include <hpx/lcos/async_continue_fwd.hpp>#include <hpx/util/bind_action.hpp>#include <hpx/runtime/actions/manage_object_action.hpp>#include <hpx/runtime/serialization/array.hpp>#include <boost/array.hpp>#include <hpx/runtime/serialization/serialize_buffer.hpp>#include <hpx/traits/supports_streaming_with_any.hpp>#include <boost/shared_array.hpp>#include <hpx/runtime/components/server/runtime_support.hpp>#include <hpx/compat/condition_variable.hpp>#include <condition_variable>#include <hpx/lcos/local/condition_variable.hpp>#include <hpx/lcos/local/mutex.hpp>#include <hpx/plugins/plugin_factory_base.hpp>#include <hpx/runtime/components/server/create_component.hpp>#include <hpx/runtime/find_here.hpp>#include <boost/program_options/options_description.hpp>#include <hpx/runtime/serialization/vector.hpp>#include <hpx/runtime/serialization/detail/serialize_collection.hpp>#include <hpx/runtime/components/server/migrate_component.hpp>#include <hpx/runtime/actions/plain_action.hpp>#include <hpx/runtime/get_ptr.hpp>#include <hpx/runtime/agas/gva.hpp>#include <boost/io/ios_state.hpp>#include <hpx/components/component_storage/export_definitions.hpp>#include <hpx/config/export_definitions.hpp>#include <hpx/components/component_storage/server/component_storage.hpp>#include <hpx/components/containers/unordered/unordered_map.hpp>#include <hpx/runtime/components/copy_component.hpp>#include <hpx/lcos/detail/async_colocated.hpp>#include <hpx/runtime/agas/primary_namespace.hpp>#include <hpx/runtime/agas/server/primary_namespace.hpp>#include <hpx/runtime/components/server/fixed_component_base.hpp>#include <hpx/runtime/applier/bind_naming_wrappers.hpp>#include <hpx/util/functional/colocated_helpers.hpp>#include <hpx/runtime/components/server/copy_component.hpp>#include <hpx/runtime/components/new.hpp>#include <hpx/runtime/components/default_distribution_policy.hpp>#include <hpx/runtime/serialization/shared_ptr.hpp>#include <hpx/runtime/components/server/distributed_metadata_base.hpp>#include <hpx/runtime/components/server/simple_component_base.hpp>#include <hpx/runtime/components/server/component.hpp>#include <hpx/traits/component_heap_type.hpp>#include <hpx/runtime/serialization/unordered_map.hpp>#include <hpx/components/containers/container_distribution_policy.hpp>#include <hpx/components/containers/unordered/partition_unordered_map_component.hpp>#include <hpx/lcos/reduce.hpp>#include <hpx/runtime/components/component_factory.hpp>#include <hpx/runtime/components/server/locking_hook.hpp>#include <hpx/runtime/threads/coroutines/coroutine.hpp>#include <hpx/runtime/threads/coroutines/coroutine_fwd.hpp>#include <hpx/runtime/threads/coroutines/detail/coroutine_accessor.hpp>#include <hpx/runtime/threads/coroutines/detail/coroutine_impl.hpp>#include <hpx/runtime/threads/coroutines/detail/context_base.hpp>#include <hpx/runtime/threads/coroutines/detail/context_impl.hpp>#include <hpx/runtime/threads/coroutines/detail/swap_context.hpp>#include <hpx/runtime/threads/coroutines/detail/tss.hpp>#include <hpx/runtime/threads/coroutines/detail/coroutine_self.hpp>#include <tuple>#include <hpx/components/containers/unordered/unordered_map_segmented_iterator.hpp>#include <hpx/util/iterator_adaptor.hpp>#include <hpx/util/iterator_facade.hpp>#include <boost/integer.hpp>

file migrate_to_storage.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/traits/is_component.hpp>#include <hpx/components/component_storage/component_storage.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/components/component_storage/server/component_storage.hpp>#include <cstddef>#include <vector>#include <hpx/components/component_storage/server/migrate_to_storage.hpp>#include <hpx/throw_exception.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/components/component_storage/export_definitions.hpp>#include <cstdint>#include <memory>#include <utility>#include <type_traits>

file error.hpp

#include <hpx/config.hpp>#include <boost/system/error_code.hpp>#include <string>

file error_code.hpp

#include <hpx/config.hpp>#include <hpx/error.hpp>#include <hpx/exception_fwd.hpp>#include <boost/system/error_code.hpp>#include <exception>#include <stdexcept>#include <string>#include <hpx/throw_exception.hpp>

file exception.hpp

#include <hpx/config.hpp>#include <hpx/error.hpp>#include <hpx/error_code.hpp>#include <hpx/error.hpp>#include <hpx/exception_fwd.hpp>#include <boost/system/error_code.hpp>#include <exception>#include <stdexcept>#include <string>#include <hpx/throw_exception.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/exception_info.hpp>#include <hpx/error_code.hpp>#include <hpx/util/detail/pack.hpp>#include <cstddef>#include <type_traits>#include <hpx/util/tuple.hpp>#include <hpx/runtime/serialization/detail/non_default_constructible.hpp>#include <memory>#include <hpx/traits/is_bitwise_serializable.hpp>#include <hpx/util/decay.hpp>#include <boost/ref.hpp>#include </hpx/build/docs/hpx/util/functional>#include <utility>#include <boost/array.hpp>#include <array>#include <algorithm>#include <typeinfo>#include <hpx/runtime/naming_fwd.hpp>#include <hpx/runtime/agas_fwd.hpp>#include <hpx/util/function.hpp>#include <cstdint>#include <boost/system/system_error.hpp>#include <hpx/config/warnings_prefix.hpp>#include <hpx/throw_exception.hpp>#include <hpx/config/warnings_suffix.hpp>

file exception_fwd.hpp

#include <hpx/config.hpp>#include <hpx/error.hpp>#include <hpx/throw_exception.hpp>

file exception_list.hpp

#include <hpx/config.hpp>#include <hpx/exception.hpp>#include <hpx/error.hpp>#include <hpx/error_code.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/exception_info.hpp>#include <hpx/runtime/naming_fwd.hpp>#include <boost/system/error_code.hpp>#include <boost/system/system_error.hpp>#include <cstddef>#include <cstdint>#include <exception>#include <string>#include <hpx/config/warnings_prefix.hpp>#include <hpx/throw_exception.hpp>#include <hpx/config/warnings_suffix.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/runtime/threads/thread_helpers.hpp>#include <hpx/runtime/threads_fwd.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <hpx/runtime/threads/coroutines/coroutine_fwd.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/runtime/threads/detail/combined_tagged_state.hpp>#include <hpx/util/assert.hpp>#include <hpx/runtime/threads/thread_id_type.hpp>#include <hpx/config/constexpr.hpp>#include <hpx/config/export_definitions.hpp>#include <functional>#include <iosfwd>#include <hpx/util_fwd.hpp>#include <hpx/util/function.hpp>#include <hpx/util/unique_function.hpp>#include <utility>#include <memory>#include <hpx/runtime/thread_pool_helpers.hpp>#include <hpx/runtime/threads/policies/scheduler_mode.hpp>#include <hpx/util/register_locks.hpp>#include <hpx/traits/has_member_xxx.hpp>#include <hpx/preprocessor/cat.hpp>#include <type_traits>#include <hpx/util/steady_clock.hpp>#include <chrono>#include <hpx/util/thread_description.hpp>#include <hpx/runtime/actions/basic_action_fwd.hpp>#include <hpx/runtime/actions/preassigned_action_id.hpp>#include <hpx/traits/get_function_address.hpp>#include <hpx/traits/get_function_annotation.hpp>#include <hpx/traits/is_action.hpp>#include <hpx/util/always_void.hpp>#include <hpx/util/decay.hpp>#include <atomic>#include <hpx/util/detail/yield_k.hpp>#include <sched.h>#include <time.h>#include <hpx/util/itt_notify.hpp>#include <boost/smart_ptr/detail/spinlock.hpp>#include <list>#include <mutex>

file hpx_finalize.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>

file hpx_init.hpp

#include <hpx/config.hpp>#include <hpx/config/attributes.hpp>#include <hpx/config/defines.hpp>#include <hpx/config/compiler_specific.hpp>#include <hpx/config/branch_hints.hpp>#include <hpx/config/compiler_fence.hpp>#include <hpx/config/compiler_native_tls.hpp>#include <ciso646>#include <hpx/config/constexpr.hpp>#include <hpx/config/debug.hpp>#include <hpx/config/emulate_deleted.hpp>#include <hpx/config/export_definitions.hpp>#include <hpx/config/forceinline.hpp>#include <hpx/config/lambda_capture.hpp>#include <utility>#include <hpx/config/manual_profiling.hpp>#include <hpx/config/threads_stack.hpp>#include <hpx/config/version.hpp>#include <hpx/config/weak_symbol.hpp>#include <boost/version.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/preprocessor/stringize.hpp>#include <hpx/hpx_finalize.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/error.hpp>#include <boost/system/error_code.hpp>#include <string>#include <hpx/throw_exception.hpp>#include <hpx/preprocessor/expand.hpp>#include <hpx/preprocessor/nargs.hpp>#include <boost/current_function.hpp>#include <exception>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>#include <hpx/hpx_suspend.hpp>#include <hpx/runtime/runtime_mode.hpp>#include <hpx/runtime/shutdown_function.hpp>#include <hpx/util/unique_function.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <hpx/preprocessor/strip_parens.hpp>#include <type_traits>#include <hpx/traits/get_function_address.hpp>#include <cstddef>#include <memory>#include <hpx/traits/get_function_annotation.hpp>#include <hpx/util/itt_notify.hpp>#include <cstdint>#include <cstring>#include <hpx/traits/is_callable.hpp>#include <hpx/util/always_void.hpp>#include <hpx/util/result_of.hpp>#include <boost/ref.hpp>#include <hpx/util/detail/basic_function.hpp>#include <hpx/util/assert.hpp>#include <assert.h>#include <cstdlib>#include <iostream>#include <hpx/util/detail/empty_function.hpp>#include <hpx/util/detail/vtable/vtable.hpp>#include <hpx/util/detail/vtable/function_vtable.hpp>#include <hpx/util/detail/vtable/callable_vtable.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/void_guard.hpp>#include </hpx/build/docs/hpx/util/functional>#include <hpx/util/detail/vtable/copyable_vtable.hpp>#include <new>#include <hpx/util/detail/vtable/serializable_function_vtable.hpp>#include <hpx/runtime/serialization/detail/polymorphic_intrusive_factory.hpp>#include <hpx/util/debug/demangle_helper.hpp>#include <typeinfo>#include <hpx/util/jenkins_hash.hpp>#include <random>#include <unordered_map>#include <hpx/util/detail/function_registration.hpp>#include <hpx/util/detail/vtable/serializable_vtable.hpp>#include <hpx/util_fwd.hpp>#include <hpx/runtime/startup_function.hpp>#include <hpx/util/function.hpp>#include <boost/program_options/options_description.hpp>#include <boost/program_options/variables_map.hpp>#include <vector>

file hpx_start.hpp

#include <hpx/config.hpp>#include <hpx/hpx_finalize.hpp>#include <hpx/runtime/runtime_mode.hpp>#include <hpx/runtime/shutdown_function.hpp>#include <hpx/runtime/startup_function.hpp>#include <hpx/util/function.hpp>#include <boost/program_options/options_description.hpp>#include <boost/program_options/variables_map.hpp>#include <cstddef>#include <string>#include <vector>

file hpx_suspend.hpp

#include <hpx/exception_fwd.hpp>

file barrier.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/components/server/managed_component_base.hpp>#include <hpx/runtime/launch_policy.hpp>#include <boost/intrusive_ptr.hpp>#include <cstddef>#include <string>#include <utility>#include <vector>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>

file broadcast.hpp

file fold.hpp

file gather.hpp

Defines

HPX_REGISTER_GATHER_DECLARATION(type, name)

Declare a gather object named name for a given data type type.

The macro HPX_REGISTER_GATHER_DECLARATION can be used to declare all facilities necessary for a (possibly remote) gather operation.

The parameter type specifies for which data type the gather operations should be enabled.

The (optional) parameter name should be a unique C-style identifier which will be internally used to identify a particular gather operation. If this defaults to <type>_gather if not specified.

Note

The macro HPX_REGISTER_GATHER_DECLARATION can be used with 1 or 2 arguments. The second argument is optional and defaults to <type>_gather.

HPX_REGISTER_GATHER(type, name)

Define a gather object named name for a given data type type.

The macro HPX_REGISTER_GATHER can be used to define all facilities necessary for a (possibly remote) gather operation.

The parameter type specifies for which data type the gather operations should be enabled.

The (optional) parameter name should be a unique C-style identifier which will be internally used to identify a particular gather operation. If this defaults to <type>_gather if not specified.

Note

The macro HPX_REGISTER_GATHER can be used with 1 or 2 arguments. The second argument is optional and defaults to <type>_gather.

file split_future.hpp

file wait_all.hpp

file wait_any.hpp

file wait_each.hpp

file wait_some.hpp

file when_all.hpp

file when_any.hpp

file when_each.hpp

file when_some.hpp

file algorithm.hpp

#include <hpx/config.hpp>#include <algorithm>#include <hpx/parallel/algorithms/adjacent_find.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <hpx/throw_exception.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/decay.hpp>#include <type_traits>#include <utility>#include <hpx/parallel/exception_list.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/datapar/execution_policy.hpp>#include <hpx/parallel/execution_policy_fwd.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <hpx/parallel/executors/parallel_executor.hpp>#include <hpx/parallel/executors/rebind_executor.hpp>#include <hpx/parallel/executors/sequenced_executor.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/executor_traits.hpp>#include <hpx/traits/is_execution_policy.hpp>#include <hpx/traits/is_executor.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <hpx/traits/is_launch_policy.hpp>#include <memory>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/util/detail/scoped_executor_parameters.hpp>#include <hpx/util/tuple.hpp>#include <string>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/cancellation_token.hpp>#include <atomic>#include <functional>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/result_of.hpp>#include <cstddef>#include <iterator>#include <vector>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/dataflow.hpp>#include <hpx/lcos/dataflow.hpp>#include <hpx/exception_list.hpp>#include <hpx/lcos/wait_all.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/util/detail/chunk_size.hpp>#include <hpx/util/iterator_range.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/executors/execution_information.hpp>#include <hpx/runtime/threads/topology.hpp>#include <hpx/traits/detail/wrap_int.hpp>#include <hpx/parallel/executors/execution_information_fwd.hpp>#include <hpx/parallel/executors/execution_fwd.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <hpx/parallel/util/detail/chunk_size_iterator.hpp>#include <hpx/util/min.hpp>#include <hpx/util/iterator_facade.hpp>#include <hpx/parallel/util/detail/handle_local_exceptions.hpp>#include <hpx/async.hpp>#include <hpx/hpx_finalize.hpp>#include <exception>#include <list>#include <hpx/parallel/util/detail/partitioner_iteration.hpp>#include <hpx/util/invoke_fused.hpp>#include <hpx/parallel/util/detail/select_partitioner.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/tagged.hpp>#include <hpx/util/detail/pack.hpp>#include </hpx/build/docs/hpx/util/functional>#include <hpx/util/zip_iterator.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/util/functional/segmented_iterator_helpers.hpp>#include <hpx/parallel/algorithms/all_any_none.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/vector_pack_load_store.hpp>#include <hpx/parallel/traits/vector_pack_type.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/util/always_void.hpp>#include <hpx/parallel/util/invoke_projected.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/transfer.hpp>#include <hpx/parallel/segmented_algorithms/detail/transfer.hpp>#include <hpx/parallel/segmented_algorithms/detail/dispatch.hpp>#include <hpx/runtime/actions/plain_action.hpp>#include <hpx/runtime/components/colocating_distribution_policy.hpp>#include <hpx/lcos/detail/async_colocated.hpp>#include <hpx/lcos/detail/async_colocated_callback.hpp>#include <hpx/lcos/async_continue_callback.hpp>#include <hpx/lcos/async_callback_fwd.hpp>#include <hpx/lcos/async_fwd.hpp>#include <hpx/runtime/actions/basic_action_fwd.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/lcos/async_continue.hpp>#include <hpx/runtime/applier/apply_callback.hpp>#include <hpx/traits/extract_action.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/traits/promise_local_result.hpp>#include <hpx/traits/promise_remote_result.hpp>#include <hpx/lcos/detail/async_colocated_callback_fwd.hpp>#include <hpx/lcos/detail/async_colocated_fwd.hpp>#include <hpx/runtime/agas/primary_namespace.hpp>#include <hpx/runtime/agas/server/primary_namespace.hpp>#include <hpx/lcos/detail/async_implementations.hpp>#include <hpx/runtime/applier/detail/apply_colocated_callback_fwd.hpp>#include <hpx/runtime/actions/action_support.hpp>#include <hpx/runtime/applier/detail/apply_colocated_fwd.hpp>#include <hpx/traits/is_continuation.hpp>#include <hpx/runtime/applier/detail/apply_implementations.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <hpx/runtime/find_here.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/parallel/util/detail/handle_remote_exceptions.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/transfer.hpp>#include <hpx/traits/pointer_category.hpp>#include <cstring>#include <boost/shared_array.hpp>#include <hpx/parallel/algorithms/count.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/algorithms/detail/distance.hpp>#include <hpx/parallel/traits/vector_pack_count_bits.hpp>#include <hpx/parallel/algorithms/equal.hpp>#include <hpx/parallel/algorithms/fill.hpp>#include <hpx/traits/is_value_proxy.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/identity.hpp>#include <cstdint>#include <hpx/parallel/algorithms/find.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/algorithms/generate.hpp>#include <hpx/parallel/algorithms/includes.hpp>#include <hpx/parallel/algorithms/is_heap.hpp>#include <hpx/parallel/algorithms/is_partitioned.hpp>#include <hpx/parallel/algorithms/is_sorted.hpp>#include <hpx/parallel/algorithms/lexicographical_compare.hpp>#include <hpx/parallel/algorithms/mismatch.hpp>#include <hpx/parallel/algorithms/merge.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/parallel/algorithms/minmax.hpp>#include <hpx/parallel/algorithms/mismatch.hpp>#include <hpx/parallel/algorithms/move.hpp>#include <hpx/parallel/algorithms/partition.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/parallel/algorithms/remove.hpp>#include <hpx/parallel/algorithms/remove_copy.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/replace.hpp>#include <hpx/parallel/algorithms/reverse.hpp>#include <hpx/parallel/algorithms/rotate.hpp>#include <hpx/parallel/algorithms/reverse.hpp>#include <hpx/parallel/algorithms/search.hpp>#include <hpx/parallel/algorithms/set_difference.hpp>#include <hpx/parallel/algorithms/detail/set_operation.hpp>#include <hpx/parallel/algorithms/set_intersection.hpp>#include <hpx/parallel/algorithms/set_symmetric_difference.hpp>#include <hpx/parallel/algorithms/set_union.hpp>#include <hpx/parallel/algorithms/sort.hpp>#include <hpx/parallel/algorithms/swap_ranges.hpp>#include <hpx/parallel/algorithms/unique.hpp>#include <hpx/parallel/algorithms/for_loop.hpp>#include <hpx/parallel/algorithms/for_loop_induction.hpp>#include <cstdlib>#include <hpx/parallel/algorithms/for_loop_reduction.hpp>#include <hpx/runtime/get_os_thread_count.hpp>#include <hpx/runtime/get_worker_thread_num.hpp>#include <hpx/util/cache_aligned_data.hpp>

file adjacent_difference.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/zip_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>

file adjacent_find.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file all_any_none.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/range.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/invoke_projected.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file all_any_none.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/all_any_none.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/util/decay.hpp>#include <hpx/util/result_of.hpp>#include <hpx/parallel/traits/projected.hpp>#include <iterator>#include <type_traits>#include <hpx/parallel/util/projection_identity.hpp>#include <utility>

file copy.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/detail/transfer.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/transfer.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <cstring>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>#include <boost/shared_array.hpp>

file copy.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <type_traits>#include <utility>

file count.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/range.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/distance.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/vector_pack_count_bits.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/invoke_projected.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <functional>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file count.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/count.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file destroy.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>

file equal.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file exclusive_scan.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/util/zip_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/inclusive_scan.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>#include <hpx/parallel/execution_policy.hpp>

file fill.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_value_proxy.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>

file fill.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_execution_policy.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/fill.hpp>#include <type_traits>#include <utility>

file find.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/util/invoke_projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file find.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_execution_policy.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/find.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file for_each.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_value_proxy.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/identity.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <iterator>#include <type_traits>#include <utility>

file for_each.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file for_loop.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/decay.hpp>#include <hpx/util/detail/pack.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tuple.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/for_loop_induction.hpp>#include <hpx/parallel/algorithms/for_loop_reduction.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file for_loop_induction.hpp

#include <hpx/config.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <cstddef>#include <cstdlib>#include <type_traits>#include <utility>

file for_loop_reduction.hpp

#include <hpx/config.hpp>#include <hpx/runtime/get_os_thread_count.hpp>#include <hpx/runtime/get_worker_thread_num.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/decay.hpp>#include <hpx/util/cache_aligned_data.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <boost/shared_array.hpp>#include <cstddef>#include <cstdlib>#include <functional>#include <type_traits>#include <utility>

file generate.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>

file generate.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/generate.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file includes.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/cancellation_token.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file inclusive_scan.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/util/zip_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>

file is_heap.hpp

#include <hpx/config.hpp>#include <hpx/async.hpp>#include <hpx/lcos/future.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <list>#include <vector>#include <type_traits>#include <utility>

file is_heap.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/is_heap.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file is_partitioned.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/cancellation_token.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <functional>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file is_sorted.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/cancellation_token.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <functional>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file lexicographical_compare.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/algorithms/mismatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file merge.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/detail/transfer.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/detail/handle_local_exceptions.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/transfer.hpp>#include <algorithm>#include <cstddef>#include <exception>#include <iterator>#include <list>#include <memory>#include <type_traits>#include <utility>#include <vector>

file merge.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/parallel/algorithms/merge.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <type_traits>#include <utility>

file minmax.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file minmax.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/minmax.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <type_traits>#include <utility>

file mismatch.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file move.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/transfer.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/transfer.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>

file partition.hpp

#include <hpx/config.hpp>#include <hpx/async.hpp>#include <hpx/lcos/dataflow.hpp>#include <hpx/lcos/future.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/exception_list.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/execution_information.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/detail/handle_local_exceptions.hpp>#include <hpx/parallel/util/invoke_projected.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <exception>#include <iterator>#include <list>#include <type_traits>#include <utility>#include <vector>#include <boost/shared_array.hpp>

file partition.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/parallel/algorithms/partition.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <type_traits>#include <utility>

file reduce.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/range.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/algorithms/detail/accumulate.hpp>#include <functional>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/distance.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>

file reduce.hpp

file reduce_by_key.hpp

#include <hpx/config.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/algorithms/inclusive_scan.hpp>#include <hpx/parallel/algorithms/sort.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <hpx/util/range.hpp>#include <hpx/util/transform_iterator.hpp>#include <hpx/util/identity.hpp>#include <hpx/util/iterator_adaptor.hpp>#include <hpx/util/lazy_conditional.hpp>#include <hpx/util/result_of.hpp>#include <iterator>#include <type_traits>#include <hpx/util/tuple.hpp>#include <cstdint>#include <functional>#include <utility>#include <vector>

file remove.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/detail/transfer.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/invoke_projected.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/transfer.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <cstring>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>#include <boost/shared_array.hpp>

file remove.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/remove.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <type_traits>#include <utility>

file remove_copy.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file remove_copy.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/remove_copy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file replace.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file replace.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/replace.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file reverse.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file reverse.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/reverse.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file rotate.hpp

#include <hpx/config.hpp>#include <hpx/dataflow.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/reverse.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/transfer.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file rotate.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/rotate.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file search.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file search.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_execution_policy.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/search.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <cstddef>#include <type_traits>#include <utility>

file set_difference.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/set_operation.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file set_intersection.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/set_operation.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file set_symmetric_difference.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/set_operation.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file set_union.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/algorithms/copy.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/set_operation.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file sort.hpp

#include <hpx/config.hpp>#include <hpx/dataflow.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/decay.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/exception_list.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/execution_information.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <algorithm>#include <cstddef>#include <exception>#include <functional>#include <iterator>#include <list>#include <type_traits>#include <utility>

file sort.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/parallel/algorithms/sort.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <type_traits>#include <utility>

file sort_by_key.hpp

#include <hpx/config.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/tuple.hpp>#include <hpx/parallel/algorithms/sort.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file swap_ranges.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/for_each.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <iterator>#include <type_traits>#include <utility>

file transform.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/util/tuple.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/transform_loop.hpp>#include <hpx/parallel/util/cancellation_token.hpp>#include <hpx/traits/is_execution_policy.hpp>#include <hpx/util/invoke.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <hpx/parallel/util/zip_iterator.hpp>#include <cstdint>

file transform.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/tagged_tuple.hpp>#include <hpx/parallel/algorithms/transform.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tuple.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/transform_loop.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <iterator>#include <type_traits>#include <utility>#include <hpx/parallel/traits/projected_range.hpp>

file transform_exclusive_scan.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/result_of.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/transform_inclusive_scan.hpp>#include <hpx/util/invoke.hpp>#include <hpx/parallel/algorithms/inclusive_scan.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>#include <hpx/parallel/execution_policy.hpp>

file transform_inclusive_scan.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/result_of.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/inclusive_scan.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>

file transform_reduce.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/segmented_iterator_traits.hpp>#include <hpx/util/range.hpp>#include <hpx/util/result_of.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>

file transform_reduce_binary.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/result_of.hpp>#include <hpx/util/zip_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <hpx/util/unused.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <numeric>#include <type_traits>#include <utility>#include <vector>

file uninitialized_copy.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner_with_cleanup.hpp>#include <hpx/dataflow.hpp>#include <hpx/exception_list.hpp>#include <hpx/lcos/wait_all.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <hpx/parallel/util/detail/chunk_size.hpp>#include <hpx/parallel/util/detail/handle_local_exceptions.hpp>#include <hpx/parallel/util/detail/partitioner_iteration.hpp>#include <hpx/parallel/util/detail/scoped_executor_parameters.hpp>#include <hpx/parallel/util/detail/select_partitioner.hpp>#include <hpx/parallel/util/partitioner.hpp>#include <algorithm>#include <cstddef>#include <exception>#include <list>#include <memory>#include <type_traits>#include <utility>#include <vector>#include <hpx/parallel/util/zip_iterator.hpp>#include <iterator>

file uninitialized_default_construct.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner_with_cleanup.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>

file uninitialized_fill.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner_with_cleanup.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>

file uninitialized_move.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner_with_cleanup.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>

file uninitialized_value_construct.hpp

#include <hpx/config.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/void_guard.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/partitioner_with_cleanup.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>

file unique.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/util/unused.hpp>#include <hpx/parallel/algorithms/detail/dispatch.hpp>#include <hpx/parallel/algorithms/detail/is_negative.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/algorithms/detail/transfer.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/util/compare_projected.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <hpx/parallel/util/foreach_partitioner.hpp>#include <hpx/parallel/util/loop.hpp>#include <hpx/parallel/util/projection_identity.hpp>#include <hpx/parallel/util/scan_partitioner.hpp>#include <hpx/parallel/util/transfer.hpp>#include <hpx/parallel/util/zip_iterator.hpp>#include <algorithm>#include <cstddef>#include <cstring>#include <iterator>#include <memory>#include <type_traits>#include <utility>#include <vector>#include <boost/shared_array.hpp>

file unique.hpp

#include <hpx/config.hpp>#include <hpx/traits/concepts.hpp>#include <hpx/traits/is_iterator.hpp>#include <hpx/traits/is_range.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tagged_pair.hpp>#include <hpx/parallel/algorithms/unique.hpp>#include <hpx/parallel/tagspec.hpp>#include <hpx/parallel/traits/projected.hpp>#include <hpx/parallel/traits/projected_range.hpp>#include <type_traits>#include <utility>

file execution_policy.hpp

#include <hpx/config.hpp>#include <hpx/parallel/datapar/execution_policy.hpp>#include <hpx/parallel/execution_policy_fwd.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <hpx/lcos/future.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/preprocessor/stringize.hpp>#include <hpx/runtime/serialization/base_object.hpp>#include <hpx/traits/detail/wrap_int.hpp>#include <hpx/traits/has_member_xxx.hpp>#include <hpx/traits/is_executor.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <hpx/traits/is_launch_policy.hpp>#include <hpx/util/decay.hpp>#include <hpx/util/detail/pack.hpp>#include <hpx/parallel/executors/execution_parameters_fwd.hpp>#include <boost/ref.hpp>#include <cstddef>#include <functional>#include <type_traits>#include <utility>#include <vector>#include <hpx/parallel/executors/parallel_executor.hpp>#include <hpx/parallel/executors/rebind_executor.hpp>#include <hpx/parallel/executors/execution_fwd.hpp>#include <hpx/traits/executor_traits.hpp>#include <hpx/parallel/executors/sequenced_executor.hpp>#include <hpx/async_launch_policy_dispatch.hpp>#include <hpx/runtime/threads/thread_executor.hpp>#include <hpx/sync_launch_policy_dispatch.hpp>#include <hpx/lcos/sync_fwd.hpp>#include <hpx/lcos/local/futures_factory.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/traits/is_action.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/deferred_call.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/exception_list.hpp>#include <hpx/exception_list.hpp>#include <hpx/hpx_finalize.hpp>#include <hpx/util/assert.hpp>#include <exception>#include <iterator>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/is_execution_policy.hpp>#include <memory>

file auto_chunk_size.hpp

#include <hpx/config.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <hpx/util/high_resolution_clock.hpp>#include <hpx/util/steady_clock.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <type_traits>

file dynamic_chunk_size.hpp

#include <hpx/config.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <cstddef>#include <type_traits>

file execution_fwd.hpp

#include <utility>#include <type_traits>#include <hpx/config.hpp>#include <hpx/traits/executor_traits.hpp>

file execution_information_fwd.hpp

#include <hpx/config.hpp>#include <hpx/parallel/executors/execution_fwd.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <hpx/traits/executor_traits.hpp>#include <cstddef>#include <type_traits>#include <utility>

file guided_chunk_size.hpp

#include <hpx/config.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <algorithm>#include <cstddef>#include <type_traits>

file parallel_executor.hpp

#include <hpx/config.hpp>#include <hpx/async_launch_policy_dispatch.hpp>#include <hpx/lcos/future.hpp>#include <hpx/lcos/local/latch.hpp>#include <hpx/parallel/algorithms/detail/predicates.hpp>#include <hpx/parallel/executors/fused_bulk_execute.hpp>#include <hpx/parallel/executors/post_policy_dispatch.hpp>#include <hpx/parallel/executors/static_chunk_size.hpp>#include <hpx/runtime/get_worker_thread_num.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/runtime/threads/thread_helpers.hpp>#include <hpx/traits/future_traits.hpp>#include <hpx/traits/is_executor.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/deferred_call.hpp>#include <hpx/util/internal_allocator.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/one_shot.hpp>#include <hpx/util/range.hpp>#include <hpx/util/unwrap.hpp>#include <algorithm>#include <cstddef>#include <type_traits>#include <utility>#include <vector>

file persistent_auto_chunk_size.hpp

#include <hpx/config.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <hpx/util/high_resolution_clock.hpp>#include <hpx/util/steady_clock.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <type_traits>

file sequenced_executor.hpp

#include <hpx/config.hpp>#include <hpx/async_launch_policy_dispatch.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/threads/thread_executor.hpp>#include <hpx/sync_launch_policy_dispatch.hpp>#include <hpx/traits/is_executor.hpp>#include <hpx/util/deferred_call.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/exception_list.hpp>#include <cstddef>#include <iterator>#include <type_traits>#include <utility>#include <vector>

file service_executors.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/parallel/executors/static_chunk_size.hpp>#include <hpx/parallel/executors/thread_execution.hpp>#include <hpx/lcos/dataflow.hpp>#include <hpx/lcos/local/futures_factory.hpp>#include <hpx/runtime/threads/thread_executor.hpp>#include <hpx/traits/future_access.hpp>#include <hpx/traits/is_launch_policy.hpp>#include <hpx/util/bind.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/deferred_call.hpp>#include <hpx/util/detail/pack.hpp>#include <hpx/util/range.hpp>#include <hpx/util/tuple.hpp>#include <hpx/util/unwrap.hpp>#include <hpx/parallel/executors/execution.hpp>#include <algorithm>#include <type_traits>#include <utility>#include <vector>#include <hpx/runtime/threads/executors/service_executors.hpp>#include <hpx/compat/condition_variable.hpp>#include <hpx/compat/mutex.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/throw_exception.hpp>#include <hpx/util/atomic_count.hpp>#include <hpx/util/steady_clock.hpp>#include <hpx/util/thread_description.hpp>#include <hpx/util/unique_function.hpp>#include <atomic>#include <chrono>#include <cstddef>#include <cstdint>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>#include <hpx/traits/executor_traits.hpp>

file static_chunk_size.hpp

#include <hpx/config.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/traits/is_executor_parameters.hpp>#include <hpx/parallel/executors/execution_parameters_fwd.hpp>#include <cstddef>#include <type_traits>

file thread_pool_executors.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/parallel/executors/execution_parameters.hpp>#include <hpx/parallel/executors/thread_execution.hpp>#include <hpx/parallel/executors/thread_execution_information.hpp>#include <hpx/runtime/get_os_thread_count.hpp>#include <hpx/runtime/threads/policies/scheduler_mode.hpp>#include <hpx/runtime/threads/thread_executor.hpp>#include <hpx/runtime/threads/topology.hpp>#include <hpx/traits/is_launch_policy.hpp>#include <hpx/parallel/executors/execution_information.hpp>#include <cstddef>#include <type_traits>#include <utility>#include <hpx/parallel/executors/thread_timed_execution.hpp>#include <hpx/lcos/local/packaged_task.hpp>#include <hpx/lcos/detail/future_data.hpp>#include <hpx/lcos/local/promise.hpp>#include <hpx/throw_exception.hpp>#include <hpx/traits/is_callable.hpp>#include <hpx/util/annotated_function.hpp>#include <hpx/util/thread_description.hpp>#include <hpx/util/unique_function.hpp>#include <exception>#include <memory>#include <hpx/util/deferred_call.hpp>#include <hpx/util/steady_clock.hpp>#include <hpx/parallel/executors/timed_execution.hpp>#include <hpx/parallel/executors/timed_execution_fwd.hpp>#include <hpx/parallel/executors/execution_fwd.hpp>#include <hpx/parallel/executors/timed_executors.hpp>#include <hpx/runtime/threads/thread.hpp>#include <hpx/traits/detail/wrap_int.hpp>#include <hpx/traits/executor_traits.hpp>#include <hpx/util/bind.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/executors/parallel_executor.hpp>#include <hpx/parallel/executors/sequenced_executor.hpp>#include <chrono>#include <functional>#include <hpx/traits/is_executor.hpp>#include <vector>#include <hpx/runtime/threads/executors/thread_pool_executors.hpp>#include <hpx/lcos/local/counting_semaphore.hpp>#include <hpx/lcos/local/detail/counting_semaphore.hpp>#include <hpx/lcos/local/detail/condition_variable.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/assert_owns_lock.hpp>#include <algorithm>#include <cstdint>#include <mutex>#include <hpx/runtime/resource/detail/partitioner.hpp>#include <hpx/runtime/resource/partitioner.hpp>#include <hpx/runtime/resource/partitioner_fwd.hpp>#include <hpx/runtime/resource/detail/create_partitioner.hpp>#include <hpx/runtime/runtime_mode.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/find_prefix.hpp>#include <hpx/preprocessor/stringize.hpp>#include <string>#include <hpx/util/function.hpp>#include <boost/program_options.hpp>#include <hpx/runtime/threads/cpu_mask.hpp>#include <hpx/runtime/threads/policies/affinity_data.hpp>#include <atomic>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>#include <hpx/util/command_line_handling.hpp>#include <hpx/hpx_init.hpp>#include <hpx/hpx_finalize.hpp>#include <hpx/hpx_suspend.hpp>#include <hpx/runtime/shutdown_function.hpp>#include <hpx/runtime/startup_function.hpp>#include <boost/program_options/options_description.hpp>#include <boost/program_options/variables_map.hpp>#include <hpx/util/manage_config.hpp>#include <hpx/util/safe_lexical_cast.hpp>#include <boost/lexical_cast.hpp>#include <map>#include <hpx/util/runtime_configuration.hpp>#include <hpx/util/tuple.hpp>#include <iosfwd>#include <hpx/runtime/threads/thread_enums.hpp>

file task_block.hpp

#include <hpx/config.hpp>#include <hpx/async.hpp>#include <hpx/exception.hpp>#include <hpx/lcos/dataflow.hpp>#include <hpx/lcos/future.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/lcos/when_all.hpp>#include <hpx/traits/is_future.hpp>#include <hpx/util/bind.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/decay.hpp>#include <hpx/parallel/exception_list.hpp>#include <hpx/parallel/execution_policy.hpp>#include <hpx/parallel/executors/execution.hpp>#include <hpx/parallel/util/detail/algorithm_result.hpp>#include <boost/utility/addressof.hpp>#include <memory>#include <exception>#include <mutex>#include <type_traits>#include <utility>#include <vector>

file manage_counter_type.hpp

#include <hpx/config.hpp>#include <hpx/error_code.hpp>#include <hpx/performance_counters/counters_fwd.hpp>#include <hpx/util/function.hpp>#include <cstddef>#include <cstdint>#include <string>#include <vector>

file basic_action.hpp

#include <hpx/config.hpp>#include <hpx/exception.hpp>#include <hpx/lcos/sync_fwd.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/preprocessor/expand.hpp>#include <hpx/preprocessor/nargs.hpp>#include <hpx/preprocessor/stringize.hpp>#include <hpx/runtime/actions/action_support.hpp>#include <hpx/runtime/actions/basic_action_fwd.hpp>#include <hpx/runtime/actions/continuation.hpp>#include <hpx/runtime/actions/detail/action_factory.hpp>#include <hpx/runtime/actions/detail/invocation_count_registry.hpp>#include <hpx/runtime/actions/preassigned_action_id.hpp>#include <hpx/runtime/actions/transfer_action.hpp>#include <hpx/runtime/actions/transfer_continuation_action.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/runtime/parcelset/detail/per_action_data_counter_registry.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/runtime_fwd.hpp>#include <hpx/traits/action_decorate_function.hpp>#include <hpx/traits/action_priority.hpp>#include <hpx/traits/action_remote_result.hpp>#include <hpx/traits/action_stacksize.hpp>#include <hpx/traits/is_action.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/traits/promise_local_result.hpp>#include <hpx/util/detail/pack.hpp>#include <hpx/util/get_and_reset_value.hpp>#include <hpx/util/invoke_fused.hpp>#include <hpx/util/logging.hpp>#include <hpx/util/tuple.hpp>#include <boost/utility/string_ref.hpp>#include <atomic>#include <cstddef>#include <cstdint>#include <exception>#include <sstream>#include <string>#include <type_traits>#include <utility>

Defines

HPX_REGISTER_ACTION_DECLARATION(...)

Declare the necessary component action boilerplate code.

The macro HPX_REGISTER_ACTION_DECLARATION can be used to declare all the boilerplate code which is required for proper functioning of component actions in the context of HPX.

The parameter action is the type of the action to declare the boilerplate for.

This macro can be invoked with an optional second parameter. This parameter specifies a unique name of the action to be used for serialization purposes. The second parameter has to be specified if the first parameter is not usable as a plain (non-qualified) C++ identifier, i.e. the first parameter contains special characters which cannot be part of a C++ identifier, such as ‘<’, ‘>’, or ‘:’.

namespace app
{
    // Define a simple component exposing one action 'print_greeting'
    class HPX_COMPONENT_EXPORT server
      : public hpx::components::simple_component_base<server>
    {
        void print_greeting ()
        {
            hpx::cout << "Hey, how are you?\n" << hpx::flush;
        }

        // Component actions need to be declared, this also defines the
        // type 'print_greeting_action' representing the action.
        HPX_DEFINE_COMPONENT_ACTION(server,
            print_greeting, print_greeting_action);
    };
}

// Declare boilerplate code required for each of the component actions.
HPX_REGISTER_ACTION_DECLARATION(app::server::print_greeting_action);

Example:

Note

This macro has to be used once for each of the component actions defined using one of the HPX_DEFINE_COMPONENT_ACTION macros. It has to be visible in all translation units using the action, thus it is recommended to place it into the header file defining the component.

HPX_REGISTER_ACTION(...)

Define the necessary component action boilerplate code.

The macro HPX_REGISTER_ACTION can be used to define all the boilerplate code which is required for proper functioning of component actions in the context of HPX.

The parameter action is the type of the action to define the boilerplate for.

This macro can be invoked with an optional second parameter. This parameter specifies a unique name of the action to be used for serialization purposes. The second parameter has to be specified if the first parameter is not usable as a plain (non-qualified) C++ identifier, i.e. the first parameter contains special characters which cannot be part of a C++ identifier, such as ‘<’, ‘>’, or ‘:’.

Note

This macro has to be used once for each of the component actions defined using one of the HPX_DEFINE_COMPONENT_ACTION or HPX_DEFINE_PLAIN_ACTION macros. It has to occur exactly once for each of the actions, thus it is recommended to place it into the source file defining the component.

Note

Only one of the forms of this macro HPX_REGISTER_ACTION or HPX_REGISTER_ACTION_ID should be used for a particular action, never both.

HPX_REGISTER_ACTION_ID(action, actionname, actionid)

Define the necessary component action boilerplate code and assign a predefined unique id to the action.

The macro HPX_REGISTER_ACTION can be used to define all the boilerplate code which is required for proper functioning of component actions in the context of HPX.

The parameter action is the type of the action to define the boilerplate for.

The parameter actionname specifies an unique name of the action to be used for serialization purposes. The second parameter has to be usable as a plain (non-qualified) C++ identifier, it should not contain special characters which cannot be part of a C++ identifier, such as ‘<’, ‘>’, or ‘:’.

The parameter actionid specifies an unique integer value which will be used to represent the action during serialization.

Note

This macro has to be used once for each of the component actions defined using one of the HPX_DEFINE_COMPONENT_ACTION or global actions HPX_DEFINE_PLAIN_ACTION macros. It has to occur exactly once for each of the actions, thus it is recommended to place it into the source file defining the component.

Note

Only one of the forms of this macro HPX_REGISTER_ACTION or HPX_REGISTER_ACTION_ID should be used for a particular action, never both.

file component_action.hpp

#include <hpx/config.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/preprocessor/expand.hpp>#include <hpx/preprocessor/nargs.hpp>#include <hpx/runtime/actions/basic_action.hpp>#include <hpx/runtime/components/pinned_ptr.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/traits/is_future.hpp>#include <boost/utility/string_ref.hpp>#include <cstdlib>#include <sstream>#include <string>#include <type_traits>#include <utility>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>

Defines

HPX_DEFINE_COMPONENT_ACTION(...)

Registers a member function of a component as an action type with HPX.

The macro HPX_DEFINE_COMPONENT_ACTION can be used to register a member function of a component as an action type named action_type.

The parameter component is the type of the component exposing the member function func which should be associated with the newly defined action type. The parameter action_type is the name of the action type to register with HPX.

namespace app
{
    // Define a simple component exposing one action 'print_greeting'
    class HPX_COMPONENT_EXPORT server
      : public hpx::components::simple_component_base<server>
    {
        void print_greeting() const
        {
            hpx::cout << "Hey, how are you?\n" << hpx::flush;
        }

        // Component actions need to be declared, this also defines the
        // type 'print_greeting_action' representing the action.
        HPX_DEFINE_COMPONENT_ACTION(server, print_greeting,
            print_greeting_action);
    };
}

Example:

The first argument must provide the type name of the component the action is defined for.

The second argument must provide the member function name the action should wrap.

The default value for the third argument (the typename of the defined action) is derived from the name of the function (as passed as the second argument) by appending ‘_action’. The third argument can be omitted only if the second argument with an appended suffix ‘_action’ resolves to a valid, unqualified C++ type name.

Note

The macro HPX_DEFINE_COMPONENT_ACTION can be used with 2 or 3 arguments. The third argument is optional.

file plain_action.hpp

#include <hpx/config.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/preprocessor/expand.hpp>#include <hpx/preprocessor/nargs.hpp>#include <hpx/preprocessor/strip_parens.hpp>#include <hpx/runtime/actions/basic_action.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/traits/component_type_database.hpp>#include <hpx/util/assert.hpp>#include <boost/utility/string_ref.hpp>#include <cstdlib>#include <sstream>#include <stdexcept>#include <string>#include <utility>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>

Defines

HPX_DEFINE_PLAIN_ACTION(...)

Defines a plain action type.

namespace app
{
    void some_global_function(double d)
    {
        cout << d;
    }

    // This will define the action type 'app::some_global_action' which
    // represents the function 'app::some_global_function'.
    HPX_DEFINE_PLAIN_ACTION(some_global_function, some_global_action);
}

Example:

Note

Usually this macro will not be used in user code unless the intent is to avoid defining the action_type in global namespace. Normally, the use of the macro HPX_PLAIN_ACTION is recommended.

Note

The macro HPX_DEFINE_PLAIN_ACTION can be used with 1 or 2 arguments. The second argument is optional. The default value for the second argument (the typename of the defined action) is derived from the name of the function (as passed as the first argument) by appending ‘_action’. The second argument can be omitted only if the first argument with an appended suffix ‘_action’ resolves to a valid, unqualified C++ type name.

HPX_DECLARE_PLAIN_ACTION(...)

Declares a plain action type.

HPX_PLAIN_ACTION(...)

Defines a plain action type based on the given function func and registers it with HPX.

The macro HPX_PLAIN_ACTION can be used to define a plain action (e.g. an action encapsulating a global or free function) based on the given function func. It defines the action type name representing the given function. This macro additionally registers the newly define action type with HPX.

The parameter func is a global or free (non-member) function which should be encapsulated into a plain action. The parameter name is the name of the action type defined by this macro.

namespace app
{
    void some_global_function(double d)
    {
        cout << d;
    }
}

// This will define the action type 'some_global_action' which represents
// the function 'app::some_global_function'.
HPX_PLAIN_ACTION(app::some_global_function, some_global_action);

Example:

Note

The macro HPX_PLAIN_ACTION has to be used at global namespace even if the wrapped function is located in some other namespace. The newly defined action type is placed into the global namespace as well.

Note

The macro HPX_PLAIN_ACTION_ID can be used with 1, 2, or 3 arguments. The second and third arguments are optional. The default value for the second argument (the typename of the defined action) is derived from the name of the function (as passed as the first argument) by appending ‘_action’. The second argument can be omitted only if the first argument with an appended suffix ‘_action’ resolves to a valid, unqualified C++ type name. The default value for the third argument is hpx::components::factory_check.

Note

Only one of the forms of this macro HPX_PLAIN_ACTION or HPX_PLAIN_ACTION_ID should be used for a particular action, never both.

HPX_PLAIN_ACTION_ID(func, name, id)

Defines a plain action type based on the given function func and registers it with HPX.

The macro HPX_PLAIN_ACTION_ID can be used to define a plain action (e.g. an action encapsulating a global or free function) based on the given function func. It defines the action type actionname representing the given function. The parameter actionid

The parameter actionid specifies an unique integer value which will be used to represent the action during serialization.

The parameter func is a global or free (non-member) function which should be encapsulated into a plain action. The parameter name is the name of the action type defined by this macro.

The second parameter has to be usable as a plain (non-qualified) C++ identifier, it should not contain special characters which cannot be part of a C++ identifier, such as ‘<’, ‘>’, or ‘:’.

namespace app
{
    void some_global_function(double d)
    {
        cout << d;
    }
}

// This will define the action type 'some_global_action' which represents
// the function 'app::some_global_function'.
HPX_PLAIN_ACTION_ID(app::some_global_function, some_global_action,
  some_unique_id);

Example:

Note

The macro HPX_PLAIN_ACTION_ID has to be used at global namespace even if the wrapped function is located in some other namespace. The newly defined action type is placed into the global namespace as well.

Note

Only one of the forms of this macro HPX_PLAIN_ACTION or HPX_PLAIN_ACTION_ID should be used for a particular action, never both.

file applier_fwd.hpp

#include <hpx/config.hpp>

file basename_registration_fwd.hpp

#include <hpx/config.hpp>#include <hpx/components_fwd.hpp>#include <hpx/lcos_fwd.hpp>#include <hpx/runtime/components/make_client.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <cstddef>#include <string>#include <utility>#include <vector>

file binpacking_distribution_policy.hpp

#include <hpx/config.hpp>#include <hpx/dataflow.hpp>#include <hpx/lcos/future.hpp>#include <hpx/performance_counters/performance_counter.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/util/bind_front.hpp>#include <hpx/performance_counters/counters_fwd.hpp>#include <hpx/performance_counters/stubs/performance_counter.hpp>#include <hpx/performance_counters/server/base_performance_counter.hpp>#include <hpx/lcos/base_lco_with_value.hpp>#include <hpx/performance_counters/counters.hpp>#include <hpx/performance_counters/performance_counter_base.hpp>#include <hpx/runtime/actions/component_action.hpp>#include <hpx/runtime/components/component_type.hpp>#include <hpx/runtime/components/server/component.hpp>#include <hpx/throw_exception.hpp>#include <hpx/util/atomic_count.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <string>#include <utility>#include <vector>#include <hpx/runtime/find_here.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <hpx/runtime/serialization/string.hpp>#include <hpx/runtime/serialization/vector.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/bind_back.hpp>#include <hpx/util/unwrap.hpp>#include <algorithm>#include <cstddef>#include <cstdint>#include <iterator>#include <type_traits>

file colocating_distribution_policy.hpp

#include <hpx/config.hpp>#include <hpx/lcos/detail/async_colocated.hpp>#include <hpx/lcos/detail/async_colocated_callback.hpp>#include <hpx/lcos/detail/async_implementations.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/applier/detail/apply_colocated_callback_fwd.hpp>#include <hpx/runtime/applier/detail/apply_colocated_fwd.hpp>#include <hpx/runtime/applier/detail/apply_implementations.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/find_here.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <hpx/traits/extract_action.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/traits/promise_local_result.hpp>#include <algorithm>#include <cstddef>#include <type_traits>#include <utility>#include <vector>

file component_factory.hpp

Defines

HPX_REGISTER_COMPONENT(type, name, mode)

Define a component factory for a component type.

This macro is used create and to register a minimal component factory for a component type which allows it to be remotely created using the hpx::new_<> function.

This macro can be invoked with one, two or three arguments

Parameters

• type: The type parameter is a (fully decorated) type of the component type for which a factory should be defined.
• name: The name parameter specifies the name to use to register the factory. This should uniquely (system-wide) identify the component type. The name parameter must conform to the C++ identifier rules (without any namespace). If this parameter is not given, the first parameter is used.
• mode: The mode parameter has to be one of the defined enumeration values of the enumeration hpx::components::factory_state_enum. The default for this parameter is hpx::components::factory_enabled.

file copy_component.hpp

#include <hpx/config.hpp>#include <hpx/lcos/async.hpp>#include <hpx/lcos/detail/async_colocated.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/actions/plain_action.hpp>#include <hpx/runtime/components/server/copy_component.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/traits/is_component.hpp>#include <type_traits>

file default_distribution_policy.hpp

#include <hpx/config.hpp>#include <hpx/lcos/dataflow.hpp>#include <hpx/lcos/future.hpp>#include <hpx/lcos/packaged_action.hpp>#include <hpx/runtime/actions/action_support.hpp>#include <hpx/runtime/applier/apply.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/find_here.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <hpx/runtime/serialization/vector.hpp>#include <hpx/runtime/serialization/shared_ptr.hpp>#include <hpx/traits/extract_action.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/traits/promise_local_result.hpp>#include <hpx/util/assert.hpp>#include <algorithm>#include <cstddef>#include <memory>#include <type_traits>#include <utility>#include <vector>

file migrate_component.hpp

#include <hpx/config.hpp>#include <hpx/lcos/async.hpp>#include <hpx/lcos/detail/async_colocated.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/actions/plain_action.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/components/server/migrate_component.hpp>#include <hpx/runtime/components/target_distribution_policy.hpp>#include <hpx/lcos/dataflow.hpp>#include <hpx/lcos/detail/async_implementations_fwd.hpp>#include <hpx/lcos/packaged_action.hpp>#include <hpx/runtime/actions/action_support.hpp>#include <hpx/runtime/agas/interface.hpp>#include <hpx/runtime/applier/detail/apply_implementations_fwd.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <hpx/runtime/find_here.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <hpx/traits/extract_action.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/traits/promise_local_result.hpp>#include <algorithm>#include <cstddef>#include <type_traits>#include <utility>#include <vector>#include <hpx/traits/is_component.hpp>

file new.hpp

#include <hpx/config.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/components/default_distribution_policy.hpp>#include <hpx/runtime/components/server/create_component.hpp>#include <hpx/runtime/components/stubs/stub_base.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/traits/is_client.hpp>#include <hpx/traits/is_component.hpp>#include <hpx/traits/is_distribution_policy.hpp>#include <hpx/util/lazy_enable_if.hpp>#include <algorithm>#include <cstddef>#include <type_traits>#include <utility>#include <vector>

file find_here.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/naming/id_type.hpp>

file find_localities.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/components/component_type.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <vector>

file get_colocation_id.hpp

#include <hpx/exception_fwd.hpp>#include <hpx/lcos_fwd.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/naming/id_type.hpp>

file get_locality_id.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <cstdint>

file get_locality_name.hpp

#include <hpx/config.hpp>#include <hpx/lcos_fwd.hpp>#include <hpx/runtime/naming/id_type.hpp>#include <string>

file get_num_localities.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/lcos_fwd.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/components/component_type.hpp>#include <cstdint>

file get_os_thread_count.hpp

#include <hpx/config.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <cstddef>

file get_ptr.hpp

#include <hpx/config.hpp>#include <hpx/runtime_fwd.hpp>#include <hpx/runtime/agas/gva.hpp>#include <hpx/runtime/components/client_base.hpp>#include <hpx/runtime/components/component_type.hpp>#include <hpx/runtime/get_lva.hpp>#include <hpx/runtime/launch_policy.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/throw_exception.hpp>#include <hpx/traits/component_pin_support.hpp>#include <hpx/traits/component_type_is_compatible.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/bind_back.hpp>#include <memory>

file get_thread_name.hpp

#include <hpx/config.hpp>#include <hpx/util/itt_notify.hpp>#include <string>

file get_worker_thread_num.hpp

#include <hpx/config.hpp>#include <hpx/error_code.hpp>#include <cstddef>

file launch_policy.hpp

#include <hpx/config.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/runtime/serialization/serialization_fwd.hpp>#include <type_traits>#include <utility>

file unmanaged.hpp

#include <hpx/runtime/naming/name.hpp>

file report_error.hpp

#include <hpx/config.hpp>#include <cstddef>#include <exception>

file partitioner.hpp

#include <hpx/config.hpp>#include <hpx/runtime/resource/partitioner_fwd.hpp>#include <hpx/runtime/resource/detail/create_partitioner.hpp>#include <hpx/runtime/runtime_mode.hpp>#include <hpx/runtime/threads/policies/scheduler_mode.hpp>#include <hpx/util/function.hpp>#include <boost/program_options.hpp>#include <cstddef>#include <string>#include <utility>#include <vector>

file partitioner_fwd.hpp

#include <hpx/config.hpp>#include <hpx/runtime/threads/policies/callback_notifier.hpp>#include <hpx/runtime/threads_fwd.hpp>#include <hpx/util/function.hpp>#include <cstddef>#include <memory>#include <string>

file runtime_mode.hpp

#include <hpx/config.hpp>#include <string>

file set_parcel_write_handler.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/parcelset_fwd.hpp>#include <hpx/util/function.hpp>#include <boost/system/error_code.hpp>

file shutdown_function.hpp

#include <hpx/config.hpp>#include <hpx/util/unique_function.hpp>

file startup_function.hpp

#include <hpx/config.hpp>#include <hpx/util/unique_function.hpp>

file scheduler_mode.hpp

file thread_data_fwd.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/threads/coroutines/coroutine_fwd.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/runtime/threads/thread_id_type.hpp>#include <hpx/util_fwd.hpp>#include <hpx/util/function.hpp>#include <hpx/util/unique_function.hpp>#include <cstddef>#include <cstdint>#include <utility>#include <memory>

file thread_enums.hpp

#include <hpx/config.hpp>#include <hpx/runtime/threads/detail/combined_tagged_state.hpp>#include <cstddef>#include <cstdint>

file thread_helpers.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/naming_fwd.hpp>#include <hpx/runtime/threads_fwd.hpp>#include <hpx/runtime/thread_pool_helpers.hpp>#include <hpx/runtime/threads/policies/scheduler_mode.hpp>#include <hpx/runtime/threads/thread_data_fwd.hpp>#include <hpx/runtime/threads/thread_enums.hpp>#include <hpx/util_fwd.hpp>#include <hpx/util/unique_function.hpp>#include <hpx/util/register_locks.hpp>#include <hpx/util/steady_clock.hpp>#include <hpx/util/thread_description.hpp>#include <atomic>#include <chrono>#include <cstddef>#include <cstdint>#include <type_traits>#include <utility>

file thread_pool_base.hpp

#include <hpx/config.hpp>#include <hpx/compat/barrier.hpp>#include <hpx/compat/condition_variable.hpp>#include <hpx/compat/mutex.hpp>#include <climits>#include <cstddef>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>#include <hpx/compat/thread.hpp>#include <hpx/error_code.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/lcos/future.hpp>#include <hpx/lcos/local/no_mutex.hpp>#include <hpx/lcos/local/spinlock.hpp>#include <hpx/runtime/thread_pool_helpers.hpp>#include <hpx/runtime/threads/cpu_mask.hpp>#include <hpx/runtime/threads/policies/affinity_data.hpp>#include <hpx/runtime/threads/policies/callback_notifier.hpp>#include <hpx/runtime/threads/policies/scheduler_mode.hpp>#include <hpx/runtime/threads/thread_executor.hpp>#include <hpx/runtime/threads/thread_init_data.hpp>#include <hpx/runtime/threads/topology.hpp>#include <hpx/state.hpp>#include <hpx/util/steady_clock.hpp>#include <hpx/util_fwd.hpp>#include <cstdint>#include <exception>#include <functional>#include <iosfwd>#include <memory>#include <mutex>#include <string>#include <vector>

file trigger_lco.hpp

#include <hpx/config.hpp>#include <hpx/lcos_fwd.hpp>#include <hpx/runtime/actions/continuation_fwd.hpp>#include <hpx/runtime/actions/action_priority.hpp>#include <hpx/runtime/applier/detail/apply_implementations_fwd.hpp>#include <hpx/runtime/naming/address.hpp>#include <hpx/runtime/naming/name.hpp>#include <hpx/util/assert.hpp>#include <hpx/util/decay.hpp>#include <exception>#include <type_traits>#include <utility>

file runtime_fwd.hpp

#include <hpx/config.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/runtime/basename_registration_fwd.hpp>#include <hpx/runtime/config_entry.hpp>#include <hpx/runtime/find_localities.hpp>#include <hpx/runtime/get_colocation_id.hpp>#include <hpx/runtime/get_locality_id.hpp>#include <hpx/runtime/get_locality_name.hpp>#include <hpx/runtime/get_num_localities.hpp>#include <hpx/runtime/get_os_thread_count.hpp>#include <hpx/runtime/get_thread_name.hpp>#include <hpx/runtime/get_worker_thread_num.hpp>#include <hpx/runtime/naming_fwd.hpp>#include <hpx/runtime/report_error.hpp>#include <hpx/runtime/runtime_fwd.hpp>#include <hpx/runtime/runtime_mode.hpp>#include <hpx/runtime/set_parcel_write_handler.hpp>#include <hpx/runtime/shutdown_function.hpp>#include <hpx/runtime/startup_function.hpp>#include <hpx/util/function.hpp>#include <hpx/util_fwd.hpp>#include <cstddef>#include <cstdint>#include <string>

file throw_exception.hpp

#include <hpx/config.hpp>#include <hpx/error.hpp>#include <hpx/exception_fwd.hpp>#include <hpx/preprocessor/cat.hpp>#include <hpx/preprocessor/expand.hpp>#include <hpx/preprocessor/nargs.hpp>#include <boost/current_function.hpp>#include <boost/system/error_code.hpp>#include <exception>#include <string>#include <hpx/config/warnings_prefix.hpp>#include <hpx/config/warnings_suffix.hpp>

Defines

HPX_THROW_EXCEPTION(errcode, f, msg)

Throw a hpx::exception initialized from the given parameters.

The macro HPX_THROW_EXCEPTION can be used to throw a hpx::exception. The purpose of this macro is to prepend the source file name and line number of the position where the exception is thrown to the error message. Moreover, this associates additional diagnostic information with the exception, such as file name and line number, locality id and thread id, and stack backtrace from the point where the exception was thrown.

The parameter errcode holds the hpx::error code the new exception should encapsulate. The parameter f is expected to hold the name of the function exception is thrown from and the parameter msg holds the error message the new exception should encapsulate.

void raise_exception()
{
    // Throw a hpx::exception initialized from the given parameters.
    // Additionally associate with this exception some detailed
    // diagnostic information about the throw-site.
    HPX_THROW_EXCEPTION(hpx::no_success, "raise_exception", "simulated error");
}

Example:

HPX_THROWS_IF(ec, errcode, f, msg)

Either throw a hpx::exception or initialize hpx::error_code from the given parameters.

The macro HPX_THROWS_IF can be used to either throw a hpx::exception or to initialize a hpx::error_code from the given parameters. If &ec == &hpx::throws, the semantics of this macro are equivalent to HPX_THROW_EXCEPTION. If &ec != &hpx::throws, the hpx::error_code instance ec is initialized instead.

The parameter errcode holds the hpx::error code from which the new exception should be initialized. The parameter f is expected to hold the name of the function exception is thrown from and the parameter msg holds the error message the new exception should encapsulate.

file is_execution_policy.hpp

#include <hpx/config.hpp>#include <hpx/util/decay.hpp>#include <type_traits>

file checkpoint.hpp

#include <hpx/dataflow.hpp>#include <hpx/lcos/future.hpp>#include <hpx/runtime/serialization/serialize.hpp>#include <hpx/runtime/serialization/vector.hpp>#include <cstddef>#include <fstream>#include <iosfwd>#include <sstream>#include <string>#include <type_traits>#include <utility>#include <vector>

This header defines the save_checkpoint and restore_checkpoint functions. These functions are designed to help HPX application developer’s checkpoint their applications. Save_checkpoint serializes one or more objects and saves them as a byte stream. Restore_checkpoint converts the byte stream back into instances of the objects.

file debugging.hpp

#include <hpx/config.hpp>

file invoke.hpp

#include <hpx/config.hpp>#include <hpx/util/result_of.hpp>#include <hpx/util/void_guard.hpp>#include <boost/ref.hpp>#include <functional>#include <type_traits>#include <utility>

Defines

HPX_INVOKE(F, ...)

HPX_INVOKE_R(R, F, ...)

file invoke_fused.hpp

#include <hpx/config.hpp>#include <hpx/util/detail/pack.hpp>#include <hpx/util/invoke.hpp>#include <hpx/util/result_of.hpp>#include <hpx/util/tuple.hpp>#include <hpx/util/void_guard.hpp>#include <cstddef>#include <type_traits>#include <utility>

file pack_traversal.hpp

#include <hpx/util/detail/pack_traversal_impl.hpp>#include <hpx/util/tuple.hpp>#include <type_traits>#include <utility>

file pack_traversal_async.hpp

#include <hpx/util/detail/pack_traversal_async_impl.hpp>#include <utility>

file unwrap.hpp

#include <hpx/config.hpp>#include <hpx/util/detail/unwrap_impl.hpp>#include <cstddef>#include <utility>

file unwrapped.hpp

#include <hpx/config.hpp>

dir /hpx/source/hpx/runtime/actions

dir /hpx/source/hpx/parallel/algorithms

dir /hpx/source/hpx/components/component_storage

dir /hpx/source/hpx/components

dir /hpx/source/hpx/runtime/components

dir /hpx/source/hpx/parallel/container_algorithms

dir /hpx/source/hpx/parallel/executors

dir /hpx/source/hpx

dir /hpx/source/hpx/lcos

dir /hpx/source/hpx/runtime/naming

dir /hpx/source/hpx/parallel

dir /hpx/source/hpx/performance_counters

dir /hpx/source/hpx/runtime/threads/policies

dir /hpx/source/hpx/runtime/resource

dir /hpx/source/hpx/runtime

dir /hpx/source

dir /hpx/source/hpx/runtime/threads

dir /hpx/source/hpx/traits

dir /hpx/source/hpx/util

Modules reference

• preprocessor