SpECTRE
v2023.05.16
|
Functionality for parallelization. More...
Namespaces | |
namespace | InboxInserters |
Structs that have insert_into_inbox methods for commonly used cases. | |
Classes | |
struct | ArrayIndex |
The array index used for indexing Chare Arrays, mostly an implementation detail. More... | |
class | Callback |
An abstract base class, whose derived class holds a function that can be invoked at a later time. The function is intended to be invoked only once. More... | |
struct | get_action_list_from_phase_dep_action_list |
(Lazy) metafunction to get the action list from a PhaseActions More... | |
struct | get_mutable_cache_tag |
struct | get_parallel_component_from_proxy |
Retrieve a parallel component from its proxy. More... | |
struct | get_parallel_component_from_proxy< Proxy< ParallelComponent, Ts... > > |
class | GlobalCache |
A Charm++ chare that caches constant data once per Charm++ node or non-constant data once per Charm++ core. More... | |
struct | has_pup_member |
Check if T has a pup member function. More... | |
struct | is_array |
Check if T is a SpECTRE Array. More... | |
struct | is_array_element_proxy |
Check if T is a Charm++ proxy for an array chare element (from indexing into the array chare) More... | |
struct | is_array_proxy |
Check if T is a Charm++ proxy for an array chare (the entire array) More... | |
struct | is_bound_array |
Check if T is a ParallelComponent for a Charm++ bound array. More... | |
struct | is_bound_array< T, std::void_t< typename T::bind_to > > |
struct | is_group |
Check if T is a SpECTRE Group. More... | |
struct | is_group_proxy |
Check if T is a Charm++ proxy for a group chare. More... | |
struct | is_node_group_proxy |
Check if T is a Charm++ proxy for a node group chare. More... | |
struct | is_nodegroup |
Check if T is a SpECTRE Nodegroup. More... | |
struct | is_pupable |
Check if type T has operator| defined for Charm++ serialization. More... | |
struct | is_singleton |
Check if T is a SpECTRE Singleton. More... | |
class | Main |
The main function of a Charm++ executable. See the Parallelization documentation for an overview of Metavariables, Phases, and parallel components. More... | |
struct | MutableCacheTag |
class | MutableGlobalCache |
A Charm++ chare that caches mutable data once per Charm++ core. More... | |
class | NodeLock |
A typesafe wrapper for a lock for synchronization of shared resources on a given node, with safe creation, destruction, and serialization. More... | |
struct | NoSection |
Can be used instead of a Parallel::Section when no section is desired. More... | |
class | PerformAlgorithmCallback |
Wraps a call to perform_algorithm. More... | |
struct | PhaseActions |
List of all the actions to be executed in the specified phase. More... | |
struct | ReductionData< ReductionDatum< Ts, InvokeCombines, InvokeFinals, InvokeFinalExtraArgsIndices >... > |
Used for reducing a possibly heterogeneous collection of types in a single reduction call. More... | |
struct | ReductionDatum |
The data to be reduced, and invokables to be called whenever two reduction messages are combined and after the reduction has been completed. More... | |
struct | ResourceInfo |
Holds resource info for all singletons and for avoiding placing array elements/singletons on the global proc 0. More... | |
struct | Section |
A subset of chares in a parallel component. More... | |
class | SimpleActionCallback |
Wraps a call to a simple action and its arguments. Can be invoked only once. More... | |
class | SimpleActionCallback< SimpleAction, Proxy > |
Wraps a call to a simple action without arguments. More... | |
struct | SingletonInfoHolder |
Holds resource info for a single singleton component. More... | |
struct | SingletonPack |
struct | SingletonPack< tmpl::list< ParallelComponents... > > |
Holds options for a group of singleton components. More... | |
Typedefs | |
using | iterable_action_return_t = std::tuple< AlgorithmExecution, std::optional< size_t > > |
Return type for iterable actions. The std::optional can be used to specify the index of the action to be called next in the phase dependent action list. Passing std::nullopt will execute the next action in the list. | |
using | ArrayIndex1D = ArrayIndex< CkIndex1D > |
using | ArrayIndex2D = ArrayIndex< CkIndex2D > |
using | ArrayIndex3D = ArrayIndex< CkIndex3D > |
using | ArrayIndex4D = ArrayIndex< CkIndex4D > |
using | ArrayIndex5D = ArrayIndex< CkIndex5D > |
using | ArrayIndex6D = ArrayIndex< CkIndex6D > |
template<class ActionsList > | |
using | get_inbox_tags = tmpl::remove_duplicates< tmpl::join< tmpl::transform< ActionsList, detail::get_inbox_tags_from_action< tmpl::_1 > > > > |
Given a list of Actions, get a list of the unique inbox tags. | |
template<class ActionsList > | |
using | get_const_global_cache_tags_from_actions = tmpl::remove_duplicates< tmpl::join< tmpl::transform< ActionsList, detail::get_const_global_cache_tags_from_parallel_struct< tmpl::_1 > > > > |
Given a list of Actions, get a list of the unique tags specified in the actions' const_global_cache_tags aliases. | |
template<class ActionsList > | |
using | get_mutable_global_cache_tags_from_actions = tmpl::remove_duplicates< tmpl::join< tmpl::transform< ActionsList, detail::get_mutable_global_cache_tags_from_parallel_struct< tmpl::_1 > > > > |
Given a list of Actions, get a list of the unique tags specified in the actions' mutable_global_cache_tags aliases. | |
template<typename Metavariables > | |
using | get_const_global_cache_tags = tmpl::remove_duplicates< tmpl::flatten< tmpl::list< typename detail::get_const_global_cache_tags_from_parallel_struct< Metavariables >::type, tmpl::transform< typename Metavariables::component_list, detail::get_const_global_cache_tags_from_parallel_struct< tmpl::_1 > >, tmpl::transform< typename Metavariables::component_list, detail::get_const_global_cache_tags_from_pdal< tmpl::_1 > > > > > |
Given the metavariables, get a list of the unique tags that will specify the items in the GlobalCache. | |
template<typename Metavariables > | |
using | get_mutable_global_cache_tags = tmpl::remove_duplicates< tmpl::flatten< tmpl::list< typename detail::get_mutable_global_cache_tags_from_parallel_struct< Metavariables >::type, tmpl::transform< typename Metavariables::component_list, detail::get_mutable_global_cache_tags_from_parallel_struct< tmpl::_1 > >, tmpl::transform< typename Metavariables::component_list, detail::get_mutable_global_cache_tags_from_pdal< tmpl::_1 > > > > > |
Given the metavariables, get a list of the unique tags that will specify the mutable items in the GlobalCache. | |
template<typename Metavariables > | |
using | get_mutable_global_cache_tag_storage = tmpl::transform< get_mutable_global_cache_tags< Metavariables >, get_mutable_cache_tag< tmpl::_1 > > |
template<typename PhaseDepActionList > | |
using | get_initialization_actions_list = tmpl::flatten< tmpl::transform< PhaseDepActionList, detail::get_initialization_actions_list< tmpl::_1 > > > |
Given the phase dependent action list, return the list of actions in the Initialization phase (or an empty list if the Initialization phase is absent from the phase dependent action list) | |
template<typename InitializationActionsList > | |
using | get_simple_tags_from_options = tmpl::remove_duplicates< tmpl::flatten< tmpl::transform< InitializationActionsList, detail::get_simple_tags_from_options_from_action< tmpl::_1 > > > > |
Given a list of initialization actions, returns a list of the unique simple_tags_from_options for all the actions. | |
template<typename SimpleTagsList , typename Metavariables > | |
using | get_option_tags = tmpl::remove_duplicates< tmpl::flatten< tmpl::transform< SimpleTagsList, tmpl::bind< detail::get_option_tags_from_simple_tag< Metavariables >::template f, tmpl::_1 > > > > |
Given a list of simple tags, returns a list of the unique option tags required to construct them. | |
Enumerations | |
enum class | AlgorithmExecution { Continue , Retry , Pause , Halt } |
The possible options for altering the current execution of the algorithm, used in the return type of iterable actions. More... | |
enum class | ExitCode : int { Complete = 0 , Abort = 1 , ContinueFromCheckpoint = 2 } |
Exit code of an executable. More... | |
enum class | Phase { Phase::AdjustDomain , Phase::CheckDomain , Phase::Cleanup , Phase::EvaluateAmrCriteria , Phase::Evolve , Phase::Execute , Phase::Exit , Phase::ImportInitialData , Phase::Initialization , Phase::InitializeInitialDataDependentQuantities , Phase::InitializeTimeStepperHistory , Phase::LoadBalancing , Phase::PostFailureCleanup , Phase::Register , Phase::RegisterWithElementDataReader , Phase::Solve , Phase::Testing , Phase::WriteCheckpoint } |
The possible phases of an executable. More... | |
Functions | |
template<typename Metavariables , typename... Tags, typename... OptionTags> | |
tuples::TaggedTuple< Tags... > | create_from_options (const tuples::TaggedTuple< OptionTags... > &options, tmpl::list< Tags... >) |
Given a list of tags and a tagged tuple containing items created from input options, return a tagged tuple of items constructed by calls to create_from_options for each tag in the list. | |
template<typename ParallelComponent , typename SectionIdTag , typename DbTagsList > | |
auto & | get_section (const gsl::not_null< db::DataBox< DbTagsList > * > box) |
Retrieve the section that the element belongs to, or Parallel::no_section() if SectionIdTag is void . More... | |
template<typename T , typename DistribObject > | |
T | number_of_procs (const DistribObject &distributed_object) |
Number of processing elements. | |
template<typename T , typename DistribObject > | |
T | my_proc (const DistribObject &distributed_object) |
Index of my processing element. | |
template<typename T , typename DistribObject > | |
T | number_of_nodes (const DistribObject &distributed_object) |
Number of nodes. | |
template<typename T , typename DistribObject > | |
T | my_node (const DistribObject &distributed_object) |
Index of my node. | |
template<typename T , typename R , typename DistribObject > | |
T | procs_on_node (const R node_index, const DistribObject &distributed_object) |
Number of processing elements on the given node. | |
template<typename T , typename DistribObject > | |
T | my_local_rank (const DistribObject &distributed_object) |
The local index of my processing element on my node. This is in the interval 0, ..., procs_on_node(my_node()) - 1. | |
template<typename T , typename R , typename DistribObject > | |
T | first_proc_on_node (const R node_index, const DistribObject &distributed_object) |
Index of first processing element on the given node. | |
template<typename T , typename R , typename DistribObject > | |
T | node_of (const R proc_index, const DistribObject &distributed_object) |
Index of the node for the given processing element. | |
template<typename T , typename R , typename DistribObject > | |
T | local_rank_of (const R proc_index, const DistribObject &distributed_object) |
The local index for the given processing element on its node. | |
template<typename ReceiveTag , typename Proxy , typename ReceiveDataType > | |
void | receive_data (Proxy &&proxy, typename ReceiveTag::temporal_id temporal_id, ReceiveDataType &&receive_data, const bool enable_if_disabled=false) |
Send the data args... to the algorithm running on proxy , and tag the message with the identifier temporal_id . More... | |
template<typename ReceiveTag , typename Proxy , typename MessageType > | |
void | receive_data (Proxy &&proxy, MessageType *message) |
Send a pointer message to the algorithm running on proxy . More... | |
template<typename Action , typename Proxy , typename... Args> | |
decltype(auto) | local_synchronous_action (Proxy &&proxy, Args &&... args) |
Invoke a local synchronous action on proxy | |
template<typename Proxy > | |
auto * | local (Proxy &&proxy) |
Wrapper for calling Charm++'s .ckLocal() on a proxy. More... | |
template<typename Proxy > | |
auto * | local_branch (Proxy &&proxy) |
Wrapper for calling Charm++'s .ckLocalBranch() on a proxy. More... | |
std::vector< Phase > | known_phases () |
std::ostream & | operator<< (std::ostream &os, const Phase &phase) |
Output operator for a Phase. | |
template<typename... Args> | |
void | printf (const std::string &format, Args &&... args) |
Print an atomic message to stdout with C printf usage. More... | |
template<typename... Args> | |
void | printf_error (const std::string &format, Args &&... args) |
Print an atomic message to stderr with C printf usage. More... | |
template<class Action , class SenderProxy , class TargetProxy , class... Ts, class SectionType = NoSection> | |
void | contribute_to_reduction (ReductionData< Ts... > reduction_data, const SenderProxy &sender_component, const TargetProxy &target_component, const gsl::not_null< SectionType * > section=&no_section()) |
Perform a reduction from the sender_component (typically your own parallel component) to the target_component , performing the Action upon receiving the reduction. More... | |
template<typename ParallelComponent > | |
bool | operator== (const SingletonInfoHolder< ParallelComponent > &lhs, const SingletonInfoHolder< ParallelComponent > &rhs) |
template<typename ParallelComponent > | |
bool | operator!= (const SingletonInfoHolder< ParallelComponent > &lhs, const SingletonInfoHolder< ParallelComponent > &rhs) |
template<typename... Components> | |
bool | operator== (const SingletonPack< tmpl::list< Components... > > &lhs, const SingletonPack< tmpl::list< Components... > > &rhs) |
template<typename... Components> | |
bool | operator!= (const SingletonPack< tmpl::list< Components... > > &lhs, const SingletonPack< tmpl::list< Components... > > &rhs) |
template<typename Metavars > | |
bool | operator== (const ResourceInfo< Metavars > &lhs, const ResourceInfo< Metavars > &rhs) |
template<typename Metavars > | |
bool | operator!= (const ResourceInfo< Metavars > &lhs, const ResourceInfo< Metavars > &rhs) |
template<typename ParallelComponentTag , typename Metavariables > | |
auto | get_parallel_component (GlobalCache< Metavariables > &cache) -> Parallel::proxy_from_parallel_component< GlobalCache_detail::get_component_if_mocked< typename Metavariables::component_list, ParallelComponentTag > > & |
Access the Charm++ proxy associated with a ParallelComponent. More... | |
template<typename ParallelComponentTag , typename Metavariables > | |
auto | get_parallel_component (const GlobalCache< Metavariables > &cache) -> const Parallel::proxy_from_parallel_component< GlobalCache_detail::get_component_if_mocked< typename Metavariables::component_list, ParallelComponentTag > > & |
Access the Charm++ proxy associated with a ParallelComponent. More... | |
template<typename GlobalCacheTag , typename Metavariables > | |
auto | get (const GlobalCache< Metavariables > &cache) -> const GlobalCache_detail::type_for_get< GlobalCacheTag, Metavariables > & |
Access data in the cache. More... | |
template<typename GlobalCacheTag , typename Function , typename Metavariables > | |
bool | mutable_cache_item_is_ready (GlobalCache< Metavariables > &cache, const Function &function) |
Returns whether the object identified by GlobalCacheTag is ready to be accessed by get . More... | |
template<typename GlobalCacheTag , typename Function , typename Metavariables , typename... Args> | |
void | mutate (GlobalCache< Metavariables > &cache, Args &&... args) |
Mutates non-const data in the cache, by calling Function::apply() More... | |
template<typename Action , typename Proxy > | |
void | simple_action (Proxy &&proxy) |
Invoke a simple action on proxy | |
template<typename Action , typename Proxy , typename Arg0 , typename... Args> | |
void | simple_action (Proxy &&proxy, Arg0 &&arg0, Args &&... args) |
Invoke a simple action on proxy | |
template<typename Action , typename Proxy > | |
void | threaded_action (Proxy &&proxy) |
Invoke a threaded action on proxy , where the proxy must be a nodegroup. | |
template<typename Action , typename Proxy , typename Arg0 , typename... Args> | |
void | threaded_action (Proxy &&proxy, Arg0 &&arg0, Args &&... args) |
Invoke a threaded action on proxy , where the proxy must be a nodegroup. | |
template<typename SenderComponent , typename ArrayIndex , typename Metavariables , class... Ts> | |
void | contribute_to_phase_change_reduction (tuples::TaggedTuple< Ts... > data_for_reduction, Parallel::GlobalCache< Metavariables > &cache, const ArrayIndex &array_index) |
Send data from a parallel component to the Main chare for making phase-change decisions. More... | |
template<typename SenderComponent , typename Metavariables , class... Ts> | |
void | contribute_to_phase_change_reduction (tuples::TaggedTuple< Ts... > data_for_reduction, Parallel::GlobalCache< Metavariables > &cache) |
Send data from a parallel component to the Main chare for making phase-change decisions. More... | |
NoSection & | no_section () |
Variables | |
template<typename Metavariables , typename Tag > | |
constexpr bool | is_in_const_global_cache |
Check whether a tag is retrievable from the const portion of the global cache. More... | |
template<typename Metavariables , typename Tag > | |
constexpr bool | is_in_mutable_global_cache |
Check whether a tag is retrievable from the mutable portion of the global cache. More... | |
template<typename Metavariables , typename Tag > | |
constexpr bool | is_in_global_cache |
Check whether a tag is retrievable from the global cache. More... | |
template<typename T > | |
constexpr bool | is_array_v = is_array<T>::value |
template<typename T > | |
constexpr bool | is_singleton_v = is_singleton<T>::value |
template<typename T > | |
constexpr bool | is_group_v = is_group<T>::value |
template<typename T > | |
constexpr bool | is_nodegroup_v = is_nodegroup<T>::value |
template<typename T > | |
using | has_pup_member_t = typename has_pup_member< T >::type |
template<typename T > | |
constexpr bool | has_pup_member_v = has_pup_member<T>::value |
template<typename T > | |
using | is_pupable_t = typename is_pupable< T >::type |
template<typename T > | |
constexpr bool | is_pupable_v = is_pupable<T>::value |
Functionality for parallelization.
The functions in namespace Parallel
that return information on nodes and cores are templated on DistribObject. Actions should use these functions rather than the raw charm++ versions (in the sys namespace in Utilities/System/ParallelInfo.hpp) so that the mocking framework will see the mocked cores and nodes.
using Parallel::has_pup_member_t = typedef typename has_pup_member<T>::type |
using Parallel::is_pupable_t = typedef typename is_pupable<T>::type |
|
strong |
The possible options for altering the current execution of the algorithm, used in the return type of iterable actions.
|
strong |
Exit code of an executable.
Enumerator | |
---|---|
Complete | Program is complete. |
Abort | Program aborted because of an error. |
ContinueFromCheckpoint | Program is incomplete and should be continued from the last checkpoint. |
void Parallel::contribute_to_phase_change_reduction | ( | tuples::TaggedTuple< Ts... > | data_for_reduction, |
Parallel::GlobalCache< Metavariables > & | cache | ||
) |
Send data from a parallel component to the Main chare for making phase-change decisions.
This function is distinct from Parallel::contribute_to_reduction
because this function contributes reduction data to the Main chare via the entry method phase_change_reduction
. This must be done because the entry method must alter member data specific to the Main chare, and the Main chare cannot execute actions like most SpECTRE parallel components. For all cases other than sending phase-change decision data to the Main chare, you should use Parallel::contribute_to_reduction
.
void Parallel::contribute_to_phase_change_reduction | ( | tuples::TaggedTuple< Ts... > | data_for_reduction, |
Parallel::GlobalCache< Metavariables > & | cache, | ||
const ArrayIndex & | array_index | ||
) |
Send data from a parallel component to the Main chare for making phase-change decisions.
This function is distinct from Parallel::contribute_to_reduction
because this function contributes reduction data to the Main chare via the entry method phase_change_reduction
. This must be done because the entry method must alter member data specific to the Main chare, and the Main chare cannot execute actions like most SpECTRE parallel components. For all cases other than sending phase-change decision data to the Main chare, you should use Parallel::contribute_to_reduction
.
auto & Parallel::get_section | ( | const gsl::not_null< db::DataBox< DbTagsList > * > | box | ) |
Retrieve the section that the element belongs to, or Parallel::no_section()
if SectionIdTag
is void
.
This function is useful to support sections in parallel algorithms. Specify the SectionIdTag
template parameter to retrieve the associated section, or set it to void
when the parallel algorithm runs over all elements of the parallel component. See Parallel::Section
for details on sections.
Only call this function on elements that are part of a section. In case not all elements are part of a section with the SectionIdTag
, make sure to skip those elements before calling this function.
auto * Parallel::local | ( | Proxy && | proxy | ) |
Wrapper for calling Charm++'s .ckLocal()
on a proxy.
The Proxy must be to a Charm++ array chare (implementing a singleton design pattern using a single-element array) or to a Charm++ array element chare (i.e., the proxy obtained by indexing into a Charm++ array chare).
The function returns a pointer to the chare if it exists on the local processor, and NULL if it does not. See the Charm++ documentation. It is the responsibility of the user to check the result pointer is valid.
auto * Parallel::local_branch | ( | Proxy && | proxy | ) |
Wrapper for calling Charm++'s .ckLocalBranch()
on a proxy.
The Proxy must be to a Charm++ group chare or nodegroup chare.
The function returns a pointer to the local group/nodegroup chare.
|
constexpr |
|
constexpr |