[R] fortran compile problem with R-1.1.0

[Prof Brian D Ripley]

On Mon, 19 Jun 2000, Jim Robison-Cox wrote:

> Hello List,
> 
>    Knowing I would eventually need to ask for help on the list, I've held
> off my thanks and congratulations on the 1.+ releases of R.  I am amazed
> at the dedication of the R-core team, and do add my applause and thanks.
> 
>   I've had a compile problem with the latest version which has not
> occurred before.  I'm compiling under:
> 
> platform sparc-sun-solaris2.6
> arch     sparc               
> os       solaris2.6          
> system   sparc, solaris2.6   

That is one of the test systems (although not 2.6 for much longer) so this
has been pretty thoroughly tested.

>  and have used f77 without problems in the past.  Now --with-f77 gives a
> compile error, --with-g77 gives a different error, but --with-f2c does
> work.
> 
> Specifying  --with-f77 I got the following in config.log:
> *****  config.log file:  *****
> configure:2187: checking whether the Fortran 77 compiler (f77  ) works
> configure:2200: f77 -o conftest   conftest.f -L/usr/local/lib 1>&5
> conftest.f:
>  MAIN conftest:
> Undefined                       first referenced
>  symbol                             in file
> __f77_exit                          conftest.o
> f77_init                            conftest.o
> __f77_init                          conftest.o
> ld: fatal: Symbol referencing errors. No output written to conftest
> configure: failed program was:
> 
>       program conftest
>       end
> 
> *** EOF config.log  ****

Worked for me with both SunPro SC4.0 and SunPro SC5.0 Fortran compilers.
Looks to me as if you have a problematic Fortran installation.  One
possible problem is that `-L/usr/local/lib': you could try
configuring with `LIBS=' to remove it.

What version is your f77?  (f77 -V should tell you.)

> And if I try --with-g77 I get:
>  ***  config.log contains: ***
> configure:2187: checking whether the Fortran 77 compiler (g77  ) works
> configure:2200: g77 -o conftest   conftest.f -L/usr/local/lib 1>&5
> gcc: language f77 not recognized
> ld: elf error: file conftest.f: unknown type, unable to process using
> elf(3E) libraries
> ld: fatal: File processing errors. No output written to conftest
> configure: failed program was:
> 
>       program conftest
>       end
> 
>   *** EOF config.log
> 
> 
>   I've seen no reports of compile problem under Solaris, so I assume I'm
> the only one with this problem.  Does anyone have suggestions about where
> to look for a solution?  

Well, the second one is easy to diagnose: you seem to have a broken gcc
installation, in that g77 exists but the internal support does not.  And
that will be why you can't just specify nothing, as configure will try g77.
I have gcc-2.95.2 on my one remaining Solaris2.6 system, and that works out
of the box.

My suggestion would be to try LIBS= ./configure --with-f77 and if that does
not work, to get gcc-2.95.2 installed with Fortran support.  If you don't
like building compilers and have access to /opt or /usr/local you can add
gcc (and much else) as Solaris packages from sunfreeware.com.  Perhaps more
usefully, it is often easier to persuade Solaris system staff to do that
for you than to compile for you.

-- 
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 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._