This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Peter, Sorry to have sent you off on a wild goose chase, I hope trying the Intel instructions to rebuild from source didn't take too much extra time. In the meantime, I've looked at your original compiler flags and noticed you are specifying -r8 and -i8 as Fortran compiler flags to the ifort compiler, but not specifying any corresponding flags to the icc C compiler. So Fortran thinks integers and reals are 8 bytes, but the C code has 4byte ints and floats. That could easily be the source of the problem. The Fortran-77 interface is implemented in C, and the Fortran-90 interface uses the Fortran-77 interface in netCDF. If a Fortran function declares a netCDF ID to be an integer which is stored in 8 bytes and it passes this to a C function which is expecting a 4 byte integer, things won't work. You must specify compatible flags for both the Fortran and C compiler, so that corresponding types take the same amount of storage. If you use -r8 and -i8 flags to the Fortran compiler, you must use equivalent flags for C compilations. --Russ Russ Rew UCAR Unidata Program address@hidden http://www.unidata.ucar.edu Ticket Details =================== Ticket ID: ZTO-391659 Department: Support netCDF Priority: Normal Status: Closed