Changes between Version 24 and Version 25 of Introduction


Ignore:
Timestamp:
05/06/23 17:15:01 (3 years ago)
Author:
siegel
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • Introduction

    v24 v25  
    608608
    609609By default, CIVL attempts to simplify an OpenMP program by replacing parallel code with sequential code when it can determine that the two are equivalent.  In the best case, this can remove all of the OpenMP, resulting in a sequential program.   The option `-ompNoSimplify` can be used to disable such simplification.   Another option, `-ompLoopDecomp=X` can be used to specify the loop decomposition strategy, where `X` is one `ALL` (the default), `ROUND_ROBIN`, or `RANDOM`.
    610 
    611 A serious limitation in the current OpenMP transformation is that access to shared variables cannot take place through pointers.
    612 
    613 == Verifying CUDA-C Programs ==
    614 
    615 Link to main [[wiki:CUDA Documentation|CUDA-C Documentation.]]
    616 
    617 == Verifying C/Pthreads Programs ==
    618 
    619 Link to main [[wiki:Pthreads Documentation|Pthreads Documentation.]]
    620 
    621 == Verifying Fortran Programs ==
    622 
    623 Link to main [[wiki:Fortran Documentation|Fortran Documentation.]]
    624 
    625 (under development)
    626