sun, lei, I've carefully read your question and here are some of the suggestions:
The first of all C pseudo code is *NOT* the either pthread or simd intrinsic functions.
Just for your interest: the C pseudo code is just a plain text explanation of how a certain code in C/C++ works.
The C pseudo code for the specific OpenMP pragmas listed below against the code in C/C++:
#pragma omp parallel for:
#pragma omp parallel for private(i)
for (int i = 0; i < N; i++)
{
// do some work in parallel
}
initialize i to zero (i = 0);
while i less than N do:
spawn i-th thread;
do some work in parallel;
end spawn;
increment i by 1;
end while;
#pragma omp simd:
#pragma omp simd
{
// An associated loop
for (int i =0 ; i < N; i++)
{
// do some work in parallel
}
}
initialize i to zero (i = 0);
spawn each of N-threads in parallel;
while i less than N / 4 do:
do the loop vectorization using SIMD instructions:
do some work in parallel #1 (i);
do some work in parallel #2 (i + 1);
do some work in parallel #3 (i + 2);
do some work in parallel #4 (i + 3);
end do;
increment i by 4;
end while;
Also, it's recommended to use the following OpenMP construct instead of using #pragma parallel for simd:
#pragma omp parallel
{
#pragma omp for simd
for (int i = 0; i < N; i++)
{
// do some work in parallel
}
}
That's all. Have a good day ahead!
