R-beta: A new version of R-0.62.3 for Win3.1/Win32s, Win95 and WinNT

root root at sirio.stat.unipd.it
Thu Sep 17 23:12:58 CEST 1998

On Thu, Sep 17, 1998 at 08:54:06PM +0100, Douglas Steele wrote:
> Guido,
> Binary looks very promising.  Tried build of source.  Any idea what this
> error means ?
> gcc   -O2 -I../include  -o ../unix/devPS.o  -c ../unix/devPS.c
> gcc   -O2 -I../include  -o ../unix/devPicTeX.o  -c ../unix/devPicTeX.c
> -------- Building R.a --------
> ar cr R.a  .//*.o  ../graphics/*.o  ../nmath/*.o  ../appl/*.o
> ../main/*.o  ../unix/*.o
> ranlib R.a
> -------- Building R.exe from R.a --------
> echo EXPORTS > R.exp
> nm R.a | sed -n "/^........ [DT] _/s/^........ [DT] _/ /p" >> R.exp
> dlltool -k --as as --dllname R.exe  --output-exp R.e --def R.exp
> gcc   -s -mwindows  -o R.exe -Wl,--base-file,R.b R.e R.a -L. -lgraphapp
> -lregex -lmingw32 -lg2c
> ld: cannot open -lg2c: No such file or directory
> make.exe: *** [R.exe] Error 1
> [exited with 2]
> (/R-0.62.3/src/gnuwin32)$
> -- 
> Douglas Steele
This means that the linker (ld) cannot find the Fortran library.
You shoul have one somewhere, I suppose, since to arrive
to this point (final linking) you must have g77 (the Fortran
fron-end) installated.

Which version of the compiler have you installed? I suppose its egcs-1.02.
In egcs1.02 the Fortran lib is libf2c.a. But the g77 people has
recently changed the name to libg2c (I suppose since in this way 
people can have both g77 and f2c without problem with the library name).
If my hypothesis is correct, you shoul edit gnuwin32/MkRules
and substitute the line

Alternatively, you can try to use g77 as the linker (always in MkRules,
change LINKER=$(CC) to LINKER=g77 and delete the FLIBS row). In
this way you should get the Fortran lib automagically linked.

guido m.
(I will post this message to r-help hoping that can be useful for others)

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list