[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [Gomp-discuss] Primal Subset
From: |
Scott Robert Ladd |
Subject: |
RE: [Gomp-discuss] Primal Subset |
Date: |
Mon, 10 Feb 2003 12:42:51 -0500 |
Lars Segerlund wrote:
> I have a question, is not
> #pragma omp parallel section
> 'primitive' in regard to
> #pragma omp parallel for
> ?
>
> Since if we did sections we didn't have to handle the loop variable,
> and a
> #pragma omp parallel for
> can be thought of as composed of sections with different loop indexes ?
Well, at the real bottom of the heap is "#pragma omp parallel", which simply
parallelizes a block of code. That would be the bottom of the complexity
tree, I believe -- no loop variables.
The parallel sections pragmas
#pragma omp sections
{
#pragma omp section
{ /* code block A*/}
#pragma omp section
{ /* code block B */}
}
involve assigning different code to different processors (block A is one
thread that runs parallle to another thread running block B), and so are an
entirely different kettle of fish.
..Scott