[R] Error in R-1.8.1 build

Prof Brian Ripley ripley at stats.ox.ac.uk
Sun Jan 11 09:22:40 CET 2004


You got a segmentation fault from the compiler.  It is the notorious gcc
`2.96' (see gcc.gnu.org -- there is no such version according to the gcc
developers).  It used to be a frequent source of grief (but then RedHat
7.2 is old now), and the advice is to use a released version of gcc.

At least some versions of `2.96' will compile R but the build will crash 
during the tests.

On Sun, 11 Jan 2004, Jim Lemon wrote:

> Hi,
> 
> Searchng the archives for the messages containing:
> 
> error & making & R-1.8
> error & make & R-1.8

Try searching for gcc & 2.96.

> I found two possibly relevant messages. Both concerned compression routines, 
> and the one reply that I could find indicated that the local libraries might 
> be out of date or corrupt.

Yes, but that was R not running correctly (later in the build), and your 
build is trying to compile libz from the sources since your libz *is* out 
of date (and a security risk).

> I think these are the relevant libraries:
> 
> /usr/lib/libz.so.1
> /usr/lib/libz.so.1.1.3
> /usr/lib/libzvt.so.2.2.10
> /usr/lib/libzvt.so.2
> /usr/lib/libz.a
> /usr/lib/libz.so
> 
> Building R-1.8.1 on:
> 
> RedHat v7.2
> gcc v2.96
> 
> produces the following error messages:
> 
> make[4]: Entering directory `/usr/local/R-1.8.1/src/extra/zlib'
> ...
> making infutil.d from infutil.c
> make[4]: *** [infutil.d] Segmentation fault
> make[4]: Leaving directory `/usr/local/R-1.8.1/src/extra/zlib'
> make[3]: *** [R] Error 2
> make[3]: Leaving directory `/usr/local/R-1.8.1/src/extra/zlib'
> make[2]: *** [R] Error 1
> make[2]: Leaving directory `/usr/local/R-1.8.1/src/extra'
> make[1]: *** [R] Error 1
> make[1]: Leaving directory `/usr/local/R-1.8.1/src'
> make: *** [R] Error 1
> 
> I suspect a type mismatch, but was unable to find it. The Makefile is so 
> cryptic that I couldn't work out the compilation command and thus couldn't 
> get more info. Thanks for any help.

Note: it is not compiling at this point but producing a dependency file.

-- 
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