fortcom terminated by kill signal - multi-file optimization

[wimberjc]

I'm trying to compile a utility called MM5toGrADS, which converts output data from the MM5 atmospheric model into a format suitable for input into the GrADS display suite. My machine (Celeron D with 256MB RAM) runs Debian Linux (2.4.27 kernel). MM5toGrADS requires the use of either the Portland Group or Intel Fortran compiler; I chose to use Intel. I tried to install the compiler from the RPM, but it failed, so I converted the package to a .deb and installed it that way. I had to do some hacking to get it to work, but it seems to work correctly (I successfully compiled the example program included in the doc/samples directory).



I seem to get through the compilation of the individual .f source files successfully. (I initially had problems with grads.f, but another post in this forum suggested that the problem could be solved if I added the -ipo flag to ifort. I did this and it stopped giving that error.) However, when it combines the individual files to make the final executable, I get an error. In fact, this error ("Fatal error in /opt/intel/fc/9.0/bin/fortcom, terminated by kill signal") is the same as the one that grads.f gave before I added the -ipo flag.



Is there anybody who has had a similar problem or has a suggestion on solving this one? I will paste the output from 'make' at the bottom of this message.



I really appreciate any assistance here. I am a Fortran newbie so forgive me if this is a dumb question.



Regards,

Jake





Output from make:



[...]

Compiling for Linux

make[1]: Entering directory `/home/jcw/gz/MM5toGrADS/src'

/bin/rm -f calcclfr.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calcclfr.F > calcclfr.

f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calcclfr.f

/bin/rm -f calcclfr.f

/bin/rm -f calcdbz.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calcdbz.F > calcdbz.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calcdbz.f

/bin/rm -f calcdbz.f

/bin/rm -f calcpt.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calcpt.F > calcpt.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calcpt.f

/bin/rm -f calcpt.f

/bin/rm -f calcrh.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calcrh.F > calcrh.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calcrh.f

/bin/rm -f calcrh.f

/bin/rm -f calctd.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calctd.F > calctd.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calctd.f

/bin/rm -f calctd.f

/bin/rm -f calcthe.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calcthe.F > calcthe.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calcthe.f

/bin/rm -f calcthe.f

/bin/rm -f calcvordiv.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux calcvordiv.F > calcvor

div.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian calcvordiv.f

/bin/rm -f calcvordiv.f

/bin/rm -f grads.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux grads.F > grads.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian grads.f

/bin/rm -f grads.f

/bin/rm -f interp.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux interp.F > interp.f

ifort -c -ipo -I. -FR -pc 32 -conve rt big_endian interp.f

/bin/rm -f interp.f

/bin/rm -f lninterp.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux lninterp.F > lninterp.

f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian lninterp.f

/bin/rm -f lninterp.f

/bin/rm -f nhgeosig.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux nhgeosig.F > nhgeosig.

f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian nhgeosig.f

/bin/rm -f nhgeosig.f

/bin/rm -f tdlninterp.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux tdlninterp.F > tdlnint

erp.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian tdlninterp.f

/bin/rm -f tdlninterp.f

/bin/rm -f vect.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux vect.F > vect.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian vect.f

/bin/rm -f vect.f

/bin/rm -f seaprsnh.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux seaprsnh.F > seaprsnh.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian seaprsnh.f

/bin/rm -f seaprsnh.f

/bin/rm -f dircomp.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux dircomp.F > dircomp.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian dircomp.f

/bin/rm -f dircomp.f

/bin/rm -f pvp.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux pvp.F > pvp.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian pvp.f

/bin/rm -f pvp.f

/bin/rm -f integrat.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux integrat.F > integrat.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian integrat.f

/bin/rm -f integrat.f

/bin/rm -f pvs.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux pvs.F > pvs.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian pvs.f

/bin/rm -f pvs.f

/bin/rm -f advec.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux advec.F > advec.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian advec.f

/bin/rm -f advec.f

/bin/rm -f fillit.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux fillit.F > fillit.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian fillit.f

/bin/rm -f fillit.f

/bin/rm -f writeout.o

/lib/cpp -I. -C -traditional -DRECLENBYTE -Drecl -Dlinux writeout.F > writeout.f

ifort -c -ipo -I. -FR -pc 32 -convert big_endian writeout.f

/bin/rm -f writeout.f

ifort -o grads.exe calcclfr.o calcdbz.o calcpt.o calcrh.o calctd.o calcthe.o calcvordiv.o grads.o interp.o lninterp.o nhgeosig.o tdlninterp.o vect.o seaprsnh.o dircomp.o pvp.o integrat.o pvs.o advec.o fillit.o writeout.o

ifort: error: Fatal error in /opt/intel/fc/9.0/bin/fortcom, terminated by kill signal

ifort: error: problem during multi-file optimization compilation (code 1)

make[1]: [grads.exe] Error 1 (ignored)

make[1]: Leaving directory `/home/jcw/gz/MM5toGrADS/src'