| /oneTBB/doc/main/tbb_userguide/design_patterns/ |
| H A D | Non-Preemptive_Priorities.rst | 13 Choose the next work item to do, based on priorities.
26 to choose work based on some kind of priority relationship.
35 - Given multiple work items, there is a rule for which item should
51 Put the work in a shared work pile. Decouple tasks from specific
52 work, so that task execution chooses the actual piece of work to be
94 // Abstract base class for a prioritized piece of work.
98 // Derived class defines the actual work.
119 collection of work and fires off tasks through the ``oneapi::tbb::task_group::run`` interface
120 and then choose a work from the collection:
167 scheduler to choose when to do a work item, not which work item to do.
|
| H A D | Local_Serializer.rst | 36 waiting at the lock when they could be doing useful work
46 Sequence the work items using a FIFO (first-in first-out structure).
101 * - No higher priority work is ready to run.
111 first relevant constraint between pieces of work.
129 further details of the work.
135 // Abstract base class for a prioritized piece of work.
139 // Derived class defines the actual work.
225 - A queue of WorkItem waiting for prior work to complete.
228 - A count of queued or in-flight work.
238 that no other work is in flight and thus the work should be moved
[all …]
|
| H A D | GUI_Thread.rst | 40 - The GUI thread needs to offload work onto other threads without
41 waiting for the work to complete.
45 dedicated to doing the offloaded work.
64 The GUI thread offloads the work by firing off a task to do it using
66 When finished, the task posts an event to the GUI thread to indicate that the work is done.
89 // Event posted from enqueued task when it finishes its work.
138 #. The GUI thread calls ``LongRunningWork``, which hands off the work
|
| H A D | Agglomeration.rst | 14 or communication swamps the useful work.
59 distribute work evenly across processors.
118 For recursive computations, most of the work is towards the leaves,
|
| /oneTBB/ |
| H A D | third-party-programs.txt | 168 …work means to copy from or adapt all or part of the work in a fashion requiring copyright permissi…
170 A "covered work" means either the unmodified Program or a work based on the Program.
179 …ce code" for a work means the preferred form of the work for making modifications to it. "Object c…
185 …work in object code form means all the source code needed to generate, install, and (for an execut…
189 The Corresponding Source for a work in source code form is that same work.
201 …work, you waive any legal power to forbid circumvention of technological measures to the extent su…
211 a) The work must carry prominent notices stating that you modified it, and giving a relevant date.
214 …work has interactive user interfaces, each must display Appropriate Legal Notices; however, if the…
276 …authorizes use under this License of the Program or a work on which the Program is based. The work…
286 …work, and grant a patent license to some of the parties receiving the covered work authorizing the…
[all …]
|
| H A D | LICENSE.txt | 35 "Work" shall mean the work of authorship, whether in Source or
37 copyright notice that is included in or attached to the work
40 "Derivative Works" shall mean any work, whether in Source or Object
43 represent, as a whole, an original work of authorship. For the purposes
48 "Contribution" shall mean any work of authorship, including
161 work stoppage, computer failure or malfunction, or any and all
178 APPENDIX: How to apply the Apache License to your work.
180 To apply the Apache License to your work, attach the following
|
| /oneTBB/python/tbb/ |
| H A D | test.py | 74 def work(mseconds): function
138 assert list(pool.imap(work, range(10, 3, -1), chunksize=4)) == list(map(
142 assert sorted(pool.imap_unordered(work, range(10, 3, -1))) == sorted(map(
182 result = pool.imap_unordered_async(work, range(2, -10, -1), callback=cb)
191 result = pool.imap_async(work, range(2, -10, -1), callback=cb)
|
| /oneTBB/doc/main/tbb_userguide/Migration_Guide/ |
| H A D | Task_API.rst | 59 method. With this new approach, you work with functors in a C++-standard way by implementing ``void
89 Adding more work during task execution
94 In TBB, cases when the amount of work is not known in advance and the work needs to be added during
112 // ... do some work for item m_i ...
140 In oneTBB ``tbb::parallel_do`` interface was removed. Instead, the functionality of adding new work
157 // ... do some work for item i ...
183 // ... do some work for item i ...
210 // do some work processing m_shared_data
242 // do some work processing m_shared_data
345 // some work to do ...
[all …]
|
| /oneTBB/doc/main/tbb_userguide/ |
| H A D | Cook_Until_Done_parallel_do.rst | 65 iterators for sequential programs work correctly. This convenience makes
66 ``parallel_for_each`` unscalable, because the fetching of work is
71 There are two ways that ``parallel_for_each`` can acquire work scalably.
79 work by calling ``feeder.add(item)``. For example, suppose processing
|
| H A D | work_isolation.rst | 28 oneapi::tbb::parallel_for( 0, N2, []( int j ) { /* Some work */ } );
53 oneapi::tbb::parallel_for( 0, N2, []( int j ) { /* Some work */ } );
82 oneapi::tbb::parallel_for( 0, N2, []( int j ) { /* Some work */ } );
88 However, using a separate arena for work isolation is not always
133 oneapi::tbb::parallel_for( 0, N2, []( int j ) { /* Some work */ } );
|
| H A D | Controlling_Chunking_os.rst | 82 useful work as the gray area inside a brown border that represents
87 of useful work depends upon the grainsize, not on the number of grains.
129 computation over a million indices. There is little work per iteration.
144 scheduling overhead, not useful work. An increase in grainsize brings a
157 work than an iteration of an inner loop.
|
| H A D | Guiding_Task_Scheduler_Execution.rst | 39 a different NUMA node. To reduce this overhead, the work may be divided among several ``task_arena``
68 The example shows how to set the most performant core type as preferable for work execution:
94 /*parallel work*/
107 /*parallel work*/
|
| H A D | Timing.rst | 26 ... do some work ...
28 printf("work took %g seconds\n",(t1-t0).seconds());
|
| H A D | Throughput_of_pipeline.rst | 17 keep the sequential filters fast, and when possible, shift work to the
31 useful work. Making windows too large may cause them to spill out of
|
| H A D | Mapping_Nodes2Tasks.rst | 17 from the work pool.
47 used to execute tasks from the oneTBB work pool while it is waiting.
|
| H A D | Flow_Graph_resource_tips.rst | 8 your graph, or control the maximum number of tasks in the work pool.
|
| H A D | More_on_HashCompare.rst | 8 ``concurrent_hash_map`` work for your own types.
47 might not work. You could trivially meet this requirement by always
|
| H A D | How_Task_Scheduler_Works.rst | 8 it was designed to work efficiently for fork-join parallelism with lots of forks.
47 which causes the depth-first execution until the thread runs out of work.
|
| H A D | Automatic_Chunking.rst | 7 A parallel loop construct incurs overhead cost for every chunk of work
|
| H A D | When_Not_to_Use_Queues.rst | 30 do other work until a value shows up. It also tries to keep items hot in
|
| H A D | Dependence_Graph.rst | 111 from the oneTBB work pool.
133 oneTBB work pool while it is waiting.
|
| /oneTBB/test/tbb/ |
| H A D | test_async_node.cpp | 355 work_type work = {input, &gateway}; in submit() local
356 my_work_queue.push( work ); in submit()
361 work_type work; in process() local
362 if( is_active && my_work_queue.try_pop( work ) ) { in process()
366 wrapper_helper<output_type, output_type>::copy_value(work.input, output); in process()
367 wrapper_helper<output_type, output_type>::check(work.input, output); in process()
368 work.gateway->try_put(output); in process()
371 work.gateway->release_wait(); in process()
|
| H A D | test_flow_graph_priorities.cpp | 268 work_type work; in operator ()() local
269 while (activity->my_queue.try_pop(work)) { in operator ()()
271 work.gateway->try_put(work.input); in operator ()()
272 work.gateway->release_wait(); in operator ()()
281 work_type work = { input, gateway }; in submit() local
283 my_queue.push(work); in submit()
|
| /oneTBB/doc/main/intro/ |
| H A D | Benefits.rst | 39 particular goal of parallelizing computationally intensive work,
53 multiple threads to work on different parts of a collection.
|
| /oneTBB/doc/GSG/ |
| H A D | intro.rst | 11 to be executed simultaneously. Using threads can make your program work faster and more efficiently.
|