On 2015-04-01 11:51+0900 Hiroyasu Yasuda wrote: > Hello Alan and Arjen, > > Thank you very much for your kindness. Eventually, I’ve solved that problem. > > A way is to compile an code of plplot with ifort that three mod > files(plplot_str.mod, plplot_strutils.mod and plplot_types.mod) copy into > $PREFIX/lib/fortran/modules/plplot. Those mod files were generated in > bindings/f95 of executed place with cmake and make. Also it is important to > refuse conflict of build results between an older build result and a newer > build result when run make install if there is an built plplot library > already on an installed environment. > > Although there is an couple of plplot library in same environment as above > method, it is easy to can be switched those libraries with makefile: > > F95 = gfortran-mp-4.6 > F95 = ifort > > ifeq ($(F95),ifort) > PKG_CONFIG_ENV = PKG_CONFIG_PATH="/opt/intel/lib/pkgconfig" > else > PKG_CONFIG_ENV = PKG_CONFIG_PATH="/opt/local/lib/pkgconfig" > endif
Hi Hiro: It sounds from what you have said, that this problem could have been avoided by following these common-sense rules: (1) Always remove stale build results (i.e., start with an empty build tree) before a build. (2) Always remove stale installed results (i.e., completely remove the install tree for the designated install prefix) before redoing an install. (3) If you are trying to install two different PLplot builds (say one for ifort-compiled results and one for gfortran-compiled results), then use a different install prefix for each one so there can be no confusion between the two installations. Let's say those two install prefixes are /opt/ifort and /opt/gfortran. Then if you look in /opt/ifort/share/plplot5*/examples/f95/Makefile, you should see a reference to the ifort version of PKG_CONFIG_ENV which in turn will point to the ifort location for the module files, and if you look in /opt/gfortran/share/plplot5*/examples/f95/Makefile you should see a reference to the gfortran PKG_CONFIG_ENV which in turn will point to the gfortran version of the module files. In other words, each PLplot installation is internally consistent for a given installation prefix and does not refer to any other PLplot installation. Do you agree following these rules would have given you good results or is there still some underlying PLplot ifort bug (such as an internal inconsistency in install location) my summary is missing? Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net); the Time Ephemerides project (timeephem.sf.net); PLplot scientific plotting software package (plplot.sf.net); the libLASi project (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ ------------------------------------------------------------------------------ Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ Plplot-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/plplot-general
