[petsc-dev] Merge request error

Satish Balay balay at mcs.anl.gov
Sat Dec 5 11:16:34 CST 2020


Its best to send the patch file as attachment - so that the format is preserved.

Satish

On Sat, 5 Dec 2020, Ali Reza Khaz'ali wrote:

> Dear Matt,
> 
>  
> 
> Thanks a lot for your kind answer. It is just a simple fix for Pragma directives, which are a bit different for windows icl.
> 
>  
> 
>  
> 
> diff --git a/include/petscsys.h b/include/petscsys.h
> 
> index 1b414e7dce..381ad1c0cf 100644
> 
> --- a/include/petscsys.h
> 
> +++ b/include/petscsys.h
> 
> @@ -388,8 +388,10 @@ M*/
> 
> #  define PetscPragmaSIMD _Pragma("vector")
> 
> #elif defined(__GNUC__) && __GNUC__ >= 5 && !defined(__PGI)
> 
> #  define PetscPragmaSIMD _Pragma("GCC ivdep")
> 
> -#elif defined(_OPENMP) && _OPENMP >= 201307
> 
> +#elif defined(_OPENMP) && _OPENMP >= 201307 && !defined(_WIN32)
> 
> #  define PetscPragmaSIMD _Pragma("omp simd")
> 
> +#elif defined(_OPENMP) && _OPENMP >= 201307 && defined(_WIN32)
> 
> +#  define PetscPragmaSIMD __pragma(omp simd)
> 
> #elif defined(PETSC_HAVE_CRAY_VECTOR)
> 
> #  define PetscPragmaSIMD _Pragma("_CRI ivdep")
> 
> #else
> 
>  
> 
>  
> 
> Best wishes,
> 
> Ali
> 
>  
> 
>  
> 
> From: Matthew Knepley <knepley at gmail.com> 
> Sent: Saturday, December 05, 2020 7:10 PM
> To: Ali Reza Khaz'ali <arkhazali at iut.ac.ir>
> Cc: petsc-dev <petsc-dev at mcs.anl.gov>
> Subject: Re: [petsc-dev] Merge request error
> 
>  
> 
> On Sat, Dec 5, 2020 at 4:12 AM Ali Reza Khaz'ali <arkhazali at iut.ac.ir <mailto:arkhazali at iut.ac.ir> > wrote:
> 
> Hi,
> 
>  
> 
> I want to fix an issue in petscsys.h that prevents the code from being compiled by icl on Windows. However, I cannot create a merge request, and GitLab web interface link shows a 404 error. I would be grateful if you could provide guidance on the matter.
> 
>  
> 
> Hi,
> 
>  
> 
> This would be a grat contribution. I see two ways to solve this:
> 
>  
> 
> 1) Are you trying to create a Merge Request from your fork of PETSc? Just making sure, because you might not have permission for an MR from the PETSc repo itself.
> 
>  
> 
> 2) If Gitlab is just not working, you can send your diff to me, and I will check it in under your name.
> 
>  
> 
>    Thanks,
> 
>  
> 
>       Matt
> 
>  
> 
>  
> 
> Best wishes,
> 
> Dr. Ali Reza Khaz’ali
> 
> Assistant Professor of Petroleum Engineering & Director of Student Affairs,
> 
> Department of Chemical Engineering
> 
> Isfahan University of Technology
> 
> Isfahan, Iran
> 
>  
> 
> 
> 
> 
>  
> 
> 


More information about the petsc-dev mailing list