Differenze

Queste sono le differenze tra la revisione selezionata e la versione attuale della pagina.

--- roberto.alfieri:pub:openmp [14/06/2017 11:59] – [OpenMP timing] roberto.alfieri
+++ roberto.alfieri:pub:openmp [14/10/2020 18:38] – [Main clauses] roberto.alfieri
@@ Linea 1: / Linea 1: @@
 ====== OpenMP ======
 [[http://openmp.org/ | OpenMP]] adds constructs for shared-memory threading to C, C++ and Fortran
+[[https://computing.llnl.gov/tutorials/openMP/#ProgrammingModel | Memory Model:]]
 {{:roberto.alfieri:pub:shared_mem.png?200|}}
+Fork--Join Model:
+{{:roberto.alfieri:pub:fork_join2.gif?400|}}
 Consists of Compiler directives, runtime routines and environment variables
-Version 4.0 (July 2013) and 4.5 (Nov 2015 ) add support for accelerators (target directives), [[ https://doc.itc.rwth-aachen.de/download/attachments/28344675/SIMD%20Vectorization%20with%20OpenMP.PDF?version=1&modificationDate=1480523704000&api=v2 | vectorization (SIMD directives) ]], [[ https://computing.llnl.gov/tutorials/openMP/ProcessThreadAffinity.pdf | thread affinity ]].
+Version 4.0 (July 2013), 4.5 (Nov 2015 ) and 5.0 (Nov 2018)  add support for accelerators (target directives), vectorization (SIMD directives), thread affinity and cancellation.
+[[https://developers.redhat.com/blog/2019/03/19/whats-new-in-openmp-5-0/ | What's new in OpenMP 5.0 ]]
@@ Linea 14: / Linea 28: @@
 [[https://computing.llnl.gov/tutorials/openMP/ | LLNL]]
--
-[[https://docs.loni.org/wiki/Introduction_to_OpenMP | LONI]]
 -
 [[http://bisqwit.iki.fi/story/howto/openmp/ | Guide into OpenMP: Easy multithreading programming for C++]]
+**[[http://didattica-linux.unipr.it/~alfieri/matdid/HPC/openmp/base/ | Exercises repository ]]**
 === openMP support in the C/C++ Compilers ===
@@ Linea 35: / Linea 49: @@
 INTEL:  OpenMP 4.5 C/C++/Fortran supported in version 17.0 compilers
-  * Compile with -fopenmp (gcc compiler)  or -qopenmp (intel compiler) on Linux
+== How to compile with openMP library ==
+Compile with -fopenmp (gcc compiler)  or -qopenmp (intel compiler) on Linux
 ===== Execution model =====
@@ Linea 126: / Linea 144: @@
   gcc -O2 -fopenmp ex2.c
+  ./a.out
 It is the programmer's responsibility to ensure that multiple threads properly access SHARED variables (such as via CRITICAL sections)
@@ Linea 189: / Linea 209: @@
 A parallel region is a block of code that will be executed by multiple threads. This is the fundamental OpenMP parallel construct.
-Examples: ex1.c ex2.c [[http://www.fis.unipr.it/home/roberto.alfieri/didattica/matdid/prog/omp/esempi/parallel2.c | parallel2.c ]]
+Examples: ex1.c ex2.c parallel-single.c
@@ Linea 211: / Linea 231: @@
   Or use #pragma omp parallel for
-[[http://www.fis.unipr.it/home/roberto.alfieri/didattica/matdid/prog/omp/esempi/ | Examples]]: for.c for2.c
+ Examples: for.c for-schelule.c
@@ Linea 223: / Linea 243: @@
 All other threads on the team skip this section of code.
-[[http://www.fis.unipr.it/home/roberto.alfieri/didattica/matdid/prog/omp/esempi/ | Examples]]: single.c
+Examples: parallel-single.c
 ==== Sections Directive ====
@@ Linea 243: / Linea 263: @@
 There is an implied barrier at the end of a SECTIONS directive
-[[http://www.fis.unipr.it/home/roberto.alfieri/didattica/matdid/prog/omp/esempi/ | Examples]]: sections.c sections2.c
+ Examples: sections.c
 ==== Critical  Directive ====
@@ Linea 268: / Linea 288: @@
 }
 </code>
+Examples: ex2.c
@@ Linea 292: / Linea 312: @@
-[[http://www.fis.unipr.it/home/roberto.alfieri/didattica/matdid/prog/omp/esempi/ | Examples]]: reduction.c
+ Examples: reduction.c
 == Combining FOR and reduce ==
@@ Linea 319: / Linea 339: @@
-[[http://www.fis.unipr.it/home/roberto.alfieri/didattica/matdid/prog/omp/esempi/ | Examples]]: for2.c
+ Examples: for-schedule.c
@@ Linea 327: / Linea 347: @@
-==== Esempio LIFE ====
-Parallelizzazione con openMP di
-[[http://www.fis.unipr.it/dokuwiki/doku.php?id=roberto.alfieri:user:prog_seq#simulazioni_di_sistemi_dinamici_discreti
-| life-ser.c]]
-== Metodo 1 ==
-Parallelizzazione dei cicli for per l'aggiornamento delle funzioni do_step() e grid_copy()
-<code>
-void do_step(..) {
-#pragma omp parallel for private (i,j)
-for (i=1;i<=nrows;i++)
-  for (j=1;j<=nrows;j++)
-      update(i,j);
-}
-void grid_copy(..) {
-#pragma omp parallel for private (i)
-for (i=1;i<=nrows;i++)
-  memcpy(..); // copia riga
-}
-</code>
-== Metodo 2 ==
-Matrice suddivisa in N righe. Ogni riga e' gestita da un thread.
-{{  roberto.alfieri:pub:life-omp.png?100|}}
-<code>
-init(lattice);
-omp_set_num_threads(NUM_THREADS);
-#pragma omp parallel
-for(step=0; step<max; step++)
- {
-  do_step();
-  copy_row();
-  #pragma omp barrier
-  #pragma omp single
-   {
-    cp_border();
-   }
- }
-save("lattice.dat");
-</code>