Re: Structure of large link libraries in f95

Charles Russell wrote:
....

Seems I heard of some potential problem with ALLOCATABLEs, but I can't remember what. Was it a problem with bounds check?

Using allocatables as defined in F90/F95 is quite reliable. There are sometimes problems with using the TR-15581 allocatable features, which is still not supported on all compilers. This allows complex things like allocatables inside of allocatables, and allocating an allocatable dummy-argument in a subroutine.

Joe
.

Relevant Pages

  • Re: Is Fortran faster than C?
    ... So I think F2003 has the benefits of a language with garbage ... So you can use allocatables to set up ... compiler from implementing GC for pointer variables, ... subroutine add_node ...
    (comp.lang.fortran)
  • Re: ALLOCATABLE arrays
    ... > temporary work arrays: ... > subroutine sub1 ... > On the contrast, frequent allocation/deallocation of allocatables ... I often use use allocatable arrays instead of automatic arrays as a ...
    (comp.lang.fortran)
  • Re: Is Fortran faster than C?
    ... type, pointer:: next ... subroutine add_node ... The idea being to use pointers to implement aliasing, and allocatables ... extended types and finalizers have nothing to do with the issue at ...
    (comp.lang.fortran)
  • Re: Is Fortran faster than C?
    ... type, pointer:: next ... subroutine add_node ... The idea being to use pointers to implement aliasing, and allocatables ... extended types and finalizers have nothing to do with the issue at ...
    (comp.lang.fortran)
  • Re: ALLOCATABLE arrays
    ... I have a subroutine that is called from ... > | of the work array into a separate module? ... > On the contrast, frequent allocation/deallocation of allocatables ...
    (comp.lang.fortran)