major linker errors | gcc (gfortran)

Evan Cooch cooch17@verizon.net
Mon Dec 17 01:31:00 GMT 2007


Solved -turns out I needed to run a clean first, and then recompile 
everything. Thought I'd tried that, but apparently not.

Evan Cooch wrote:
> Greetings -
>
> Apologies if this should go to the fortran list, but as its not really 
> a gfortran issue, I'm hoping/guessing this is the most appropriate 
> spot to pursue this.
>
> I have a very large application that I have compiled/linked 
> successfully on a Linux box running Fedora 5. Recent upgrade to Fedora 
> 7. Runs gcc 4.1.2-27. The compilation of individual functions/routines 
> in the makefile works perfectly (as it did before). But, what doesn't 
> work anymore is the linking. Here is the basic structure of the makefile:
>
> COMPILER = gfortran
> LINKER = gfortran
> COPTIONS = -c -fimplicit-none -fbounds-check -march=nocona 
> -funroll-loops -O3
> OBJECTS = mark.o btest.o cvtcas.o headng.o \
> error.o fillsg.o forset.o getcmd.o oindex.o \
> ibclr.o ibset.o ivalue.o linchk.o mindex. etc. etc.
> mark.exe: mark.f $(OBJECTS) MODELC STATUS
>        $(LINKER) $(OBJECTS) -o mark.exe -L/home/egc/Desktop/mark 
> -lLinpack
> hyperdist.mod: hyperdist.f
>        $(COMPILER) $(COPTIONS) hyperdist.f
> mark.o: mark.f HEADST MODELC STATUS
>        $(COMPILER) $(COPTIONS) mark.f
> multistrata.mod: multistrata.f
>        $(COMPILER) $(COPTIONS) multistrata.f
> <etc...etc...etc>
>
>
> As noted, the compilation of the individual routines goes fine, but 
> thinks go to heck at the link step. I get lots (hundreds) of the 
> following sorts of errors:
>
> hyperread.f:(.text+0x3091): undefined reference to 
> `__hyperdist__nhypdesignpars'
> hyperread.f:(.text+0x3179): undefined reference to `__hyperdist__nvcsize'
> hyperread.f:(.text+0x31a6): undefined reference to `__hyperdist__nvcsize'
> hyperread.f:(.text+0x31d3): undefined reference to `__hyperdist__nvcsize'
> etc...etc...etc
>
> In other words, hundreds of 'undefined reference' errors, which I've 
> never seen before (perhaps lucky me, but...).
>
> Suggestions? Pointers to the obvious?
>
> Much thanks in advance...
>



More information about the Gcc-help mailing list