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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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.
>From: Patrick Dills <address@hidden> >Organization: UCAR/COMET >Keywords: 200407020423.i624NgaW002282 McIDAS-X build Hi Patrick, >Well, after a abort yesterday of my 'make mcx' on the 'gizmo' machine due >to an unexpected reboot...another attempt today and a few errors seen to >be causing 'make mcx' to come to an unhappy termination. > >Seems to be some trouble in the hdf and possibly even (higher up) netcdf >portions of the 'make'. I've attached the 'makelog' from this afternoon. > >Below are the last dozen lines or so before the abort. Looks like a >problem building libraries...troubles appear to start well before that >point. Any insights from your end ? It looks like it thinks it should have the library in the HDF subdirectory, but didn't. Since the build was interrupted in the middle, it may be a good idea to simply start the build process from scratch -- if, of course, that isn't what you already tried. When you say you made another attempt at the build, do you mean that you started from scratch, or did you pick up where things left off? If you started from scratch: cd mcidas2003/src make clobber make mcx then the problem probably lies in the configuration files for the HDF build. What I have in mind is that the config.guess (or some similar file) in the HDF tree does not have an entry that will match your new Sun F95 compiler. If this is the case, I will need to see exactly where the build starts to fail so I can figure out what new entry is needed in the configuration file(s). For reference, I had to make a modification in two of the HDF configuration files for our HP-UX 11.0 system. >Env vars, paths, directories look OK as far as I can tell. I may try to >go back and give the gcc/g77 approach a spin just to see if that still >works. OK, can't hurt. You have to start from scratch to do that. >Looking like a long holiday weekend indeed. You should be able to do the build while being logged in remotely. Then the process is just setting the environment variables, running 'make clobber' followed by 'make mcx'. I still have a login going on gizmo in my office. When I get in later today (trying to catch-up on work), I will take a quick look. Cheers, and happy 4th of July! Tom >------------------------------------------------------------------------ >ld: fatal: file /pub/mcidas/mcidas2003/hdf/hdf/src/libdf.a: open failed: >No such file or directory >ld: fatal: File processing errors. No output written to testhdf >make[7]: *** [testhdf] Error 1 >make[7]: Leaving directory `/pub/mcidas/mcidas2003/hdf/hdf/test' >make[6]: *** [rall] Error 2 >make[6]: Leaving directory `/pub/mcidas/mcidas2003/hdf/hdf/test' >make[5]: *** [all] Error 2 >make[5]: Leaving directory `/pub/mcidas/mcidas2003/hdf/hdf/test' >make[4]: *** [subd] Error 2 >make[4]: Leaving directory `/pub/mcidas/mcidas2003/hdf/hdf' >make[3]: *** [all] Error 2 >make[3]: Leaving directory `/pub/mcidas/mcidas2003/hdf/hdf' >make[2]: *** [subd] Error 2 >make[2]: Leaving directory `/pub/mcidas/mcidas2003/hdf' >make[1]: *** [all] Error 2 >make[1]: Leaving directory `/pub/mcidas/mcidas2003/hdf' >make: *** [../hdf/hdf/src/libdf.a] Error 2 -- NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publically 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.