[Rd] PGI 6.1 compile error
Paul Gilbert
pgilbert at bank-banque-canada.ca
Thu May 11 22:32:29 CEST 2006
I am trying to compile R-2.3.0 (both first release and patched
2006-05-10) on Red Hat Enterprise Linux AS release 3 (Taroon Update 7)
using the Portland Group compiler version 6.1 and the notes from
Jennifer Lai on p33 of "R Installation and Administration" version 2.3.0
(2006-04-24). I have not used this compiler before, so it is possible
things are messed up more than usual, even for me.
Following the R Installation and Administration notes I set
export PGI=/usr/pgi/linux86-64/6.1
export PATH=$PGI/bin:$PATH
export MANPATH=$MANPATH=$PGI/man:/
export CC=pgcc
export CFLAGS="-g -O2 -Kieee"
export CPPFLAGS="-I$PGI/include -I$PGI/include/CC"
export F77=pgf77
export FFLAGS="-g -O2 -Kieee"
export CXX="pgCC"
export CXXFLAGS="-g -O2 -Kieee"
export F95=pgf95
export FCFLAGS="-g -O2 -Kieee"
export SHLIB_CXXLDFLAGS=-shared
export SHLIB_LDFLAGS=-shared
export LDFLAGS="-L$PGI/libso -L/usr/lib64"
However, the last line causes
:../../src/R-2.3.0-patched/configure
checking build system type... x86_64-unknown-linux-gnu
checking host system type... x86_64-unknown-linux-gnu
loading site script '../../src/R-2.3.0-patched/config.site'
....
checking for gcc... pgcc
checking for C compiler default output file name... a.out
checking whether the C compiler works... configure: error: cannot run C
compiled programs.
If you meant to cross compile, use `--host'.
See `config.log' for more details.
Either unsetting LDFLAGS, or setting it to "-L/usr/lib64"
.../../src/R-2.3.0-patched/configure --x-libraries=/usr/X11R6/lib64
configures but I get
checking whether C99 double complex is supported...
checking complex.h usability... no
checking complex.h presence... yes
configure: WARNING: complex.h: present but cannot be compiled
configure: WARNING: complex.h: check for missing prerequisite headers?
configure: WARNING: complex.h: see the Autoconf documentation
configure: WARNING: complex.h: section "Present But Cannot Be Compiled"
configure: WARNING: complex.h: proceeding with the preprocessor's result
configure: WARNING: complex.h: in the future, the compiler will take
precedence
configure: WARNING: ## ----------------------------------- ##
configure: WARNING: ## Report this to r-bugs at R-project.org ##
configure: WARNING: ## ----------------------------------- ##
checking for complex.h... yes
checking for double complex... no
no
checking for cblas_cdotu_sub in vecLib framework... no
....
R is now configured for x86_64-unknown-linux-gnu
Source directory: ../../src/R-2.3.0-patched
Installation directory: /usr/local
C compiler: pgcc -g -O2 -Kieee
Fortran 77 compiler: pgf77 -g -O2 -Kieee
C++ compiler: pgCC -g -O2 -Kieee
Fortran 90/95 compiler: pgf95 -g -O2 -Kieee
Interfaces supported: X11, tcltk
External libraries: readline
Additional capabilities: PNG, JPEG, iconv, MBCS, NLS
Options enabled: R profiling
Recommended packages: yes
configure: WARNING: you cannot build info or html versions of the R manuals
However, when I make I get
....
pgcc -shared -L/usr/lib64 -o tools.so text.o init.o Rmd5.o md5.o
mkdir -p -- ../../../../library/tools/libs
make[5]: Leaving directory
`/home/mfa/gilp/toolchain/R/mfa04559/R-2.3.0PGI/src/library/tools/src'
make[4]: Leaving directory
`/home/mfa/gilp/toolchain/R/mfa04559/R-2.3.0PGI/src/library/tools/src'
**ERROR: in routine alloca() there is a
stack overflow: thread 0, max 10228KB, used 0KB, request 16B
make[3]: *** [all] Error 1
make[3]: Leaving directory
`/home/mfa/gilp/toolchain/R/mfa04559/R-2.3.0PGI/src/library/tools'
make[2]: *** [R] Error 1
make[2]: Leaving directory
`/home/mfa/gilp/toolchain/R/mfa04559/R-2.3.0PGI/src/library'
make[1]: *** [R] Error 1
make[1]: Leaving directory
`/home/mfa/gilp/toolchain/R/mfa04559/R-2.3.0PGI/src'
make: *** [R] Error 1
Any help would be appreciated.
Paul Gilbert
====================================================================================
La version française suit le texte anglais.
------------------------------------------------------------------------------------
This email may contain privileged and/or confidential inform...{{dropped}}
More information about the R-devel
mailing list