On Thu, 7 Nov 2013, Satish Balay wrote: > On Thu, 7 Nov 2013, Barry Smith wrote: > > > > > Rather than making a massive change to PETSc, how about the following > > modest proposal > > > > Add a configure option that causes PETSc to compile Fortran code with > > the free-form option. When built with this option, of course, possibly many > > of the current examples just won’t be built in that configuration. > > It already works this way - doesn't it? > > All petsc sources and all .F examples fixed/free-form complaint. > > So if one were to use FFLAGS=-fixed-form or FFLAGS=-free-form things > should work? > > Perhaps some examples got broke - and they need fixing?
BTW: if none of the options are specified - compilers usually default to 'fixed form' for .f/.F and free form for .f90/.F90. So perhaps 'FFLAGS=-fixed-form' will conflict with our .F90 examples - so some extra check [to not run .F90 examples] might be needed somehwere.. Satish
