[petsc-dev] more issues

Looks like we need a volunteer to finish the job of finding every floating point literal that needs to be cast to PetscScalar. I'm not going to get to it in the next few days. -- next part -- An HTML attachment was scrubbed... URL:

[petsc-dev] more issues

On Fri, Jun 3, 2011 at 13:20, Lisandro Dalcin wrote: > Can any of you try to reproduce (using single precision + complex > scalars) ? I'm using system GCC 4.5.5 from Fedora 13. > This issue is present with both gcc-4.6 and clang-2.9. As an example, consider PetscIsInfOrNanScalar(PetscScalar a)

[petsc-dev] more issues

I agree, even though the castes are ugly, I think we have to live with them. Barry On Jun 3, 2011, at 7:37 AM, Jed Brown wrote: > On Fri, Jun 3, 2011 at 13:20, Lisandro Dalcin wrote: > Can any of you try to reproduce (using single precision + complex > scalars) ? I'm using system GCC 4.5.

[petsc-dev] more issues

On 2 June 2011 13:51, Barry Smith wrote: > > ? I guess. But I think I would argue that you have a buggy compiler/templated > C++ class; Can any of you try to reproduce (using single precision + complex scalars) ? I'm using system GCC 4.5.5 from Fedora 13. > > it should be able to do those compa

[petsc-dev] more issues

I guess. But I think I would argue that you have a buggy compiler/templated C++ class; it should be able to do those comparisons. Barry On Jun 2, 2011, at 8:14 AM, Lisandro Dalcin wrote: > Barry, I'm trying to build like below: > > $ ./configure PETSC_ARCH=arch-test --with-shared-lib

[petsc-dev] more issues

Barry, I'm trying to build like below: $ ./configure PETSC_ARCH=arch-test --with-shared-libraries=1 --with-64-bit-indices=1 --with-clanguage=C++ --with-precision=single --with-scalar-type=complex and of course it fails. I'm working on a patch, fixed a few things, but I got to this: [ 39%] Buildi