[R] Source installation error: "gfortran and gcc disagree on int and double ...

Prof Brian Ripley ripley at stats.ox.ac.uk
Tue Aug 8 11:08:42 CEST 2006


First, you replied to the list and not to me, which was discourteous.

Your error does indeed appear to be an incorrectly installed compiler.

> conftestf.o: In function `cftest_':
> /home/mike/Desktop/R-2.3.1/conftestf.f:9: undefined reference to
> `_gfortran_pow_r8_i4'

this is in -lgfortran, so that is not being found or is broken.
One possibility is a missing symlink that is in a -devel RPM.

This is not an R issue.

On Mon, 7 Aug 2006, Mike wrote:

> Prof Brian Ripley wrote:
> 
> > First, you do not need -fPIC: it is the same as -fpic which R selects on
> > your platform.
> > 
> 
> Right, I commented the flags and configure goes just as far
> 
> 
> > Second, please look at config.log to find the exact problem: it well be
> > that your compilers are not properly installed (as was the case in the
> > reference you quote below).
> 
> from config.log it seem that there are a few missing inclides and syntax
> errors in confdefs.h. (I can quote the specifics, if necessary). The last
> lines before the error message are:
> 
> configure:27770: checking whether gfortran and gcc agree on int and double
> conftest.c: In function 'main':
> conftest.c:28: warning: implicit declaration of function 'printf'
> conftest.c:28: warning: incompatible implicit declaration of built-in
> function 'printf'
> conftest.c:29: warning: implicit declaration of function 'exit'
> conftest.c:29: warning: incompatible implicit declaration of built-in
> function 'exit'
> conftestf.o: In function `cftest_':
> /home/mike/Desktop/R-2.3.1/conftestf.f:9: undefined reference to
> `_gfortran_pow_r8_i4'
> collect2: ld returned 1 exit status
> configure:27848: WARNING: gfortran and gcc disagree on int and double
> configure:27850: error: Maybe change CFLAGS or FFLAGS?
> 
> > 
> > Finally, your compilers are pretty obselete (there have been 4.0.2, 4.0.3,
> > 4.1.0 and 4.1.1), so you should be updating them.
> > 
> Changing compiler is not an option for me. Could you advise me how to
> configure the flags?


-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595



More information about the R-help mailing list