[R] choosing g/g++ compiler executable in R CMD SHLIB
Faheem Mitha
faheem at email.unc.edu
Sun Nov 18 00:57:35 CET 2001
On Sat, 17 Nov 2001, Prof Brian Ripley wrote:
> On Sat, 17 Nov 2001, Faheem Mitha wrote:
> > How can I get R CMD SHLIB to invoke gcc-3.0 instead of gcc and
> > g++-3.0 instead of g++?
[snip]
> It's nothing to do with "Writing R Extensions": it is an installation
> issue. If you build R with CC=gcc-3.0, CXX=g++-3.0 then these will be
> recorded in RHOME/etc/Makeconf and used by R CMD SHLIB (at least, that
> works for me). I suspect manually editing that file will do what you
> want.
Yes, this works. Thanks.
> However, a warning. FLIBS is also set at build time (configure time),
> and the correct specification for gcc 3.0.x is not the same as for gcc
> 2.95.x. I think it would be a lot safer to rebuild R under your favoured
> compilers.
>
> The FLIBS problem arises if compilers are updated so g77 refers to 3.0.x,
> BTW. It may not be relevant is you only want to compile C and C++.
If I understand you correctly, this will only be an issue if I was to use
R to do complilation of Fortran code. I have no such plans. I have changed
all the stuff that seems relevant, CC = gcc-3.0, CXX = g++-3.0, CXXCPP =
g++-3.0 -E, SHLIB_CXXLD = g++-3.0. I have left the Fortran stuff alone, so
FC = g77 still refers to 2.95.x. In any case, I don't even have g77
installed.
Just one more question. Is it possible to override ../etc/Makeconf
locally, in the directory in which I am doing the complilation? I can use
Makevars to set flags, but it doesn't seem to work with choosing the
compiler. Ie setting CC=.. in Makevars has no effect.
Thanks for the valuable help.
Sincerely, Faheem Mitha.
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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