]> https://gitweb.dealii.org/ - dealii-svn.git/commitdiff
Update documentation in a number of places.
authorbangerth <bangerth@0785d39b-7218-0410-832d-ea1e28bc413d>
Wed, 22 Jun 2011 12:44:28 +0000 (12:44 +0000)
committerbangerth <bangerth@0785d39b-7218-0410-832d-ea1e28bc413d>
Wed, 22 Jun 2011 12:44:28 +0000 (12:44 +0000)
git-svn-id: https://svn.dealii.org/trunk@23850 0785d39b-7218-0410-832d-ea1e28bc413d

deal.II/include/deal.II/base/parallel.h

index 7387e3a630053b65d3f53f0d9ee41acbb0e27746..9c977004c89485a11c9dfb486f0c2eb9b90c962d 100644 (file)
@@ -145,13 +145,21 @@ namespace parallel
                                    * std::transform does. The difference is
                                    * that the function can run in parallel
                                    * when deal.II is configured to use
-                                   * multiple threads. In that case, the last
-                                   * argument denotes the minimum number each
-                                   * thread can work on; the number must be
+                                   * multiple threads.
+                                   * 
+                                   * If running in parallel, the iterator range
+                                   * is split into several chunks that are
+                                   * each packaged up as a task and given to
+                                   * the Threading Building Blocks scheduler
+                                   * to work on as compute resources are
+                                   * available. The function returns once all
+                                   * chunks have been worked on. The last
+                                   * argument denotes the minimum number of
+                                   * elements of the iterator range per
+                                   * task; the number must be
                                    * large enough to amortize the startup
-                                   * cost of new threads, and small enough to
-                                   * ensure that (i) threads will be started
-                                   * at all, and (ii) threads can be
+                                   * cost of new tasks, and small enough to
+                                   * ensure that tasks can be
                                    * reasonably load balanced.
                                    *
                                    * For a discussion of the kind of
@@ -201,13 +209,21 @@ namespace parallel
                                    * std::transform does. The difference is
                                    * that the function can run in parallel
                                    * when deal.II is configured to use
-                                   * multiple threads. In that case, the last
-                                   * argument denotes the minimum number each
-                                   * thread can work on; the number must be
+                                   * multiple threads.
+                                   * 
+                                   * If running in parallel, the iterator range
+                                   * is split into several chunks that are
+                                   * each packaged up as a task and given to
+                                   * the Threading Building Blocks scheduler
+                                   * to work on as compute resources are
+                                   * available. The function returns once all
+                                   * chunks have been worked on. The last
+                                   * argument denotes the minimum number of
+                                   * elements of the iterator range per
+                                   * task; the number must be
                                    * large enough to amortize the startup
-                                   * cost of new threads, and small enough to
-                                   * ensure that (i) threads will be started
-                                   * at all, and (ii) threads can be
+                                   * cost of new tasks, and small enough to
+                                   * ensure that tasks can be
                                    * reasonably load balanced.
                                    *
                                    * For a discussion of the kind of
@@ -261,13 +277,21 @@ namespace parallel
                                    * std::transform does. The difference is
                                    * that the function can run in parallel
                                    * when deal.II is configured to use
-                                   * multiple threads. In that case, the last
-                                   * argument denotes the minimum number each
-                                   * thread can work on; the number must be
+                                   * multiple threads.
+                                   * 
+                                   * If running in parallel, the iterator range
+                                   * is split into several chunks that are
+                                   * each packaged up as a task and given to
+                                   * the Threading Building Blocks scheduler
+                                   * to work on as compute resources are
+                                   * available. The function returns once all
+                                   * chunks have been worked on. The last
+                                   * argument denotes the minimum number of
+                                   * elements of the iterator range per
+                                   * task; the number must be
                                    * large enough to amortize the startup
-                                   * cost of new threads, and small enough to
-                                   * ensure that (i) threads will be started
-                                   * at all, and (ii) threads can be
+                                   * cost of new tasks, and small enough to
+                                   * ensure that tasks can be
                                    * reasonably load balanced.
                                    *
                                    * For a discussion of the kind of
@@ -372,7 +396,7 @@ namespace parallel
                                    *     parallel::apply_to_subranges
                                    *        (0, A.n_rows(),
                                    *         std_cxx1x::bind (&mat_vec_on_subranges,
-                                   *                          _1, _2,
+                                   *                          std_cxx1x::_1, std_cxx1x::_2,
                                    *                          std_cxx1x::cref(A),
                                    *                          std_cxx1x::cref(x),
                                    *                          std_cxx1x::ref(y)),
@@ -589,7 +613,7 @@ namespace parallel
                                    *       (parallel::accumulate_from_subranges<double>
                                    *        (0, A.n_rows(),
                                    *         std_cxx1x::bind (&mat_norm_sqr_on_subranges,
-                                   *                          _1, _2,
+                                   *                          std_cxx1x::_1, std_cxx1x::_2,
                                    *                          std_cxx1x::cref(A),
                                    *                          std_cxx1x::cref(x)),
                                    *         50);
@@ -633,7 +657,7 @@ namespace parallel
                                    * available CPU resources, the result may
                                    * also be $a+c+b$ or any other
                                    * permutation; because floating point
-                                   * addition does not commute (as oppose, of
+                                   * addition does not commute (as opposed, of
                                    * course, to addition of real %numbers),
                                    * the result of invoking this function
                                    * several times may differ on the order of

In the beginning the Universe was created. This has made a lot of people very angry and has been widely regarded as a bad move.

Douglas Adams


Typeset in Trocchi and Trocchi Bold Sans Serif.