Re: Problems compiling R65.0 (PR#262)

About this list Date view Thread view Subject view Author view Other groups

Subject: Re: Problems compiling R65.0 (PR#262)
From: Peter Dalgaard BSA (p.dalgaard@biostat.ku.dk)
Date: Wed 01 Sep 1999 - 20:18:15 EST


Message-ID: <x21zcjdjfc.fsf@blueberry.kubism.ku.dk>

plindsey@alpha.luc.ac.be writes:

> Hi,
> I appear to have a slight problem compiling R0.65.0 under Linux
> (slackware) using g77. By simply typing configure then make, I obtain
> a binary which when "ldd" does not contain the f2c library. I
> therefore cannot load any libraries containing fortran code (I obtain
> unresolved symbol messages from R.X11 and an unable to load shared
> library from dyn.load(file)).

Yes, this is a bit of a pain. The issue is that libf2c is not a
dynamic library on all systems, and hence does not show on ldd, and
only the routines that are actually used by R get loaded into the
binary.

What should happen in that case is that R SHLIB/INSTALL would generate
so files with the relevant Fortran linked in, e.g. if that doesn't
happen for you, there's a bug in the configure script. With egsc/g77 I
get the following:

# R INSTALL multiv
Installing package `multiv' ...
 libs
g77 -fPIC -g -O2 -c bea.f -o bea.o
g77 -fPIC -g -O2 -c ca.f -o ca.o
g77 -fPIC -g -O2 -c hc.f -o hc.o
g77 -fPIC -g -O2 -c hcmovie.f -o hcmovie.o
g77 -fPIC -g -O2 -c members.f -o members.o
g77 -fPIC -g -O2 -c partition.f -o partition.o
g77 -fPIC -g -O2 -c pca.f -o pca.o
g77 -fPIC -g -O2 -c sammon.f -o sammon.o
gcc -shared -o /usr/local/lib/R/library/multiv/libs/multiv.so bea.o ca.o hc.o hcmovie.o members.o partition.o pca.o sammon.o -L/usr/lib/gcc-lib/i386-redhat-linux/egcs-2.91.66 -L/usr/i386-redhat-linux/lib -lg2c -lm

# R SHLIB ca.f
gcc -shared -o ca.so ca.o -L/usr/lib/gcc-lib/i386-redhat-linux/egcs-2.91.66 -L/usr/i386-redhat-linux/lib -lg2c -lm

> If I change the "-lf2c" flag on the FLIBS line of the Makeconf to
> "-L/usr/lib/libf2c", I then obtain the following error when R.X11 is
> being created:

That couldn't possibly work. '-L' specifies a directory to search, so
you get the same effect as if you had just removed -lf2c.

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel 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-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._


About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b25 : Tue 04 Jan 2000 - 14:16:08 EST