◢ <實作一> | <回課程大綱> ▲ | <實作三> ◣

實作二

多核心程式設計 : OpenMP
Multi-Cores Programming : OpenMP

編譯環境檢驗 Compiler Version Check

• GCC 4.2 以上版本就有支援 OpenMP 了....用 "gcc -fopenmp" 就可以編譯
• 確認目前 gcc 版本，須大於 4.2 以後的版本

  ~$ gcc -v
  Using built-in specs.
  Target: x86_64-linux-gnu
  Configured with: ../src/configure -v --with-pkgversion='Debian 4.4.5-8' --with-bugurl=file:///usr/share/doc/gcc-4.4/README.Bugs --enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr --program-suffix=-4.4 --enable-shared --enable-multiarch --enable-linker-build-id --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.4 --libdir=/usr/lib --enable-nls --enable-clocale=gnu --enable-libstdcxx-debug --enable-objc-gc --with-arch-32=i586 --with-tune=generic --enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu --target=x86_64-linux-gnu
  Thread model: posix
  gcc version 4.4.5 (Debian 4.4.5-8) 
  

Example 1 : HelloWorld

• 首先從 ​OpenMP Exercise 下載 omp_hello.c

  ~$ wget https://computing.llnl.gov/tutorials/openMP/samples/C/omp_hello.c
  

• 使用 GCC 編譯 omp_hello.c

  ~$ gcc -o omp_hello -fopenmp omp_hello.c
  

• 執行 omp_hello 執行檔

  ~$ ./omp_hello 
  Hello World from thread = 2
  Hello World from thread = 3
  Hello World from thread = 0
  Number of threads = 4
  Hello World from thread = 1
  

Example 2 : Loop work-sharing

• 首先從 ​OpenMP Exercise 下載 omp_workshare1.c

  ~$ wget https://computing.llnl.gov/tutorials/openMP/samples/C/omp_workshare1.c
  

• 使用 GCC 編譯 omp_workshare1.c

  ~$ gcc -o omp_workshare1 -fopenmp omp_workshare1.c
  

• 執行 omp_workshare1 執行檔

  ~$ ./omp_workshare1
  ~$ ./omp_workshare1
  Thread 3 starting...
  Thread 1 starting...
  Number of threads = 4
  Thread 0 starting...
  Thread 0: c[20]= 40.000000
  Thread 0: c[21]= 42.000000
  Thread 0: c[22]= 44.000000
  Thread 0: c[23]= 46.000000
  Thread 0: c[24]= 48.000000
  Thread 0: c[25]= 50.000000
  
   ... skipped ... 
  
  Thread 2 starting...
  Thread 2: c[90]= 180.000000
  Thread 2: c[91]= 182.000000
  Thread 2: c[92]= 184.000000
  Thread 2: c[93]= 186.000000
  Thread 2: c[94]= 188.000000
  
   ... skipped ... 
  
  Thread 3: c[0]= 0.000000
  Thread 3: c[1]= 2.000000
  Thread 3: c[2]= 4.000000
  Thread 3: c[3]= 6.000000
  Thread 3: c[4]= 8.000000
  Thread 3: c[5]= 10.000000
   
   ... skipped ... 
  
  Thread 1: c[10]= 20.000000
  Thread 1: c[11]= 22.000000
  Thread 1: c[12]= 24.000000
  Thread 1: c[13]= 26.000000
  Thread 1: c[14]= 28.000000
  
  

OpenMP 參考資料

• 多核心平行化程式設計 Multi-Cores Programming
• <官網> ​OpenMP Tutorial
• <官網> ​OpenMP Exercise
• ​OpenMP Tutorial at National Energy Research Scientific Computing Center (NERSC)
• ​簡易的程式平行化方法－OpenMP（一）簡介
• ​簡易的程式平行化－OpenMP（二）語法說明
• ​簡易的程式平行化－OpenMP（三）範例 parallel、section
• ​簡易的程式平行化－OpenMP（四）範例 for
• ​簡易的程式平行化－OpenMP（五） 變數的平行化

Perl v.s. multicore

• ​Parallel::ForkManager - A simple parallel processing fork manager
  • ​How can I make my Perl script use multiple cores for child processes?
  • ​Harnessing the power of multicore

Python v.s. multicore

• ​Does python support multiprocessor/multicore programming?
• Parallel Python - ​Examples
• ​python-multiprocessing - Python 2.5/2.4 back port of the multiprocessing package