[petsc-dev] MPI type mismatches

2013-01-02 Thread Jed Brown
If you have Clang and MPICH-3.0, all MPI types should now be checked (including the types that PETSc creates). https://bitbucket.org/petsc/petsc-dev/commits/a7385c2b2cce59d13c376cee7ce8117a20c668ac On Sun, Dec 30, 2012 at 11:33 PM, Jed Brown wrote: > I've pushed fixes for the PetscBool and Pet

[petsc-dev] MPI type mismatches

2013-01-01 Thread Jed Brown
On Sun, Dec 30, 2012 at 11:33 PM, Jed Brown wrote: > I've pushed fixes for the PetscBool and PetscEnum instances. Please use > MPIU_BOOL and MPIU_ENUM in new code. The Complex cases still produce > warnings. Let me know if you object to my ripping out those special cases. > Done https://bitbuck

[petsc-dev] MPI type mismatches

2012-12-30 Thread Jed Brown
I've pushed fixes for the PetscBool and PetscEnum instances. Please use MPIU_BOOL and MPIU_ENUM in new code. The Complex cases still produce warnings. Let me know if you object to my ripping out those special cases. https://bitbucket.org/petsc/petsc-dev/commits/659ebb827654e17cd01d6fb5cf1a54155f4e

[petsc-dev] MPI type mismatches

2012-12-30 Thread Jed Brown
I just fixed a bunch of dangerous type mismatches. Presumably a number of things were broken with 64-bit integers, at least in cases with overflow. https://bitbucket.org/petsc/petsc-dev/commits/146d1df88e306d69e6ce797932b1c3352fd7851c These were warnings with clang-3.2. A year ago when people wer