Hi Mattew, > I have recently gained access to v15.1 of Portland Fortran and > have been compiling up our software stack with it. Unfortunately > the Fortran wrapper for NetCDF fails to pass its own self-test. > Or rather it fails to build its own tests. > > The failure is an inability to link "tst03_f77_v2" to "Tests_". > This message > <https://www.unidata.ucar.edu/support/help/MailArchives/netcdf/msg12845.html> > appears to refer to the same problem. However the "don't compile > the v2 API" solution doesn't work for me as ESMF requires the > global error variable provided by it. > > The Makefile which builds the test states: "This fails with > pgifortran, but works with gfortran." which isn't particularly > helpful. Luckily 5 minutes work shows that if you make the > following change it works: > > am__f03tst_vars2_SOURCES_DIST = module_tests.F90 f03tst_vars2.F \ > f03lib_f_interfaces.f90 f03lib.c handle_err.F > <<< am_f03tst_vars2_OBJECTS = \ > >>> am_f03tst_vars2_OBJECTS = module_tests.$(OBJEXT) \ > f03tst_vars2.$(OBJEXT) \ > f03lib_f_interfaces.$(OBJEXT) \ > f03lib.$(OBJEXT) handle_err.$(OBJEXT) > f03tst_vars2_OBJECTS = $(am_f03tst_vars2_OBJECTS) > > I hope this is useful to you. Thanks, we'll fix that for the next release. Evidently, the file you are modifying is nf03_test/Makefile.in, but that file is generated automatically from by the automake tool from the Makefile.am file in the same directory, so that change won't work for us, but we'll figure out the analogous fix to Makefile.am. --Russ Russ Rew UCAR Unidata Program address@hidden http://www.unidata.ucar.edu Ticket Details =================== Ticket ID: PEG-669046 Department: Support netCDF Priority: Normal Status: Closed
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.