Hello Dear R Developers and Users: I'm trying to compile R with all the tools of the CRAN page. I compile the R-1.2.2 version without problem. But when I try to compile the development version 1.3.0 it doesn't work when I write the command: make BUILD=MINGW32 First it shows me something like this (I think without problems): As I make when I compile succesfuly the R-1.2.2 version: ---------------------------------------------------------------------------------------------------------------------------- mkdir -p ../../bin F:/GCC-29~1.2/BIN/MAKE.EXE -C ./fixed MAKE.EXE[1]: Entering directory `F:/R-devel/src/gnuwin32/fixed' zip -q bins bin/BATCH bin/INSTALL bin/REMOVE bin/Rd2dvi.sh bin/Rd2txt bin/SHLIB bin/helpPRINT.bat unzip -oaq bins -d ../../.. rm -f bins.zip echo done > fixbin MAKE.EXE[1]: Leaving directory `F:/R-devel/src/gnuwin32/fixed' cp -pr unzip ../.. rm -f ../../unzip/*.h F:/GCC-29~1.2/BIN/MAKE.EXE -C ../appl OPTFLAGS='-O2 -Wall -pedantic' -f Makefile.win MAKE.EXE[1]: Entering directory `F:/R-devel/src/appl' MAKE.EXE[1]: Leaving directory `F:/R-devel/src/appl' F:/GCC-29~1.2/BIN/MAKE.EXE -C ../main OPTFLAGS='-O2 -Wall -pedantic' -f Makefile.win MAKE.EXE[1]: Entering directory `F:/R-devel/src/main' MAKE.EXE[1]: Leaving directory `F:/R-devel/src/main' F:/GCC-29~1.2/BIN/MAKE.EXE -C ../nmath OPTFLAGS='-O2 -Wall -pedantic' -f Makefile.win MAKE.EXE[1]: Entering directory `F:/R-devel/src/nmath' MAKE.EXE[1]: Leaving directory `F:/R-devel/src/nmath' F:/GCC-29~1.2/BIN/MAKE.EXE -C ./graphapp MAKE.EXE[1]: Entering directory `F:/R-devel/src/gnuwin32/graphapp' MAKE.EXE[1]: `../ga.a' is up to date. MAKE.EXE[1]: Leaving directory `F:/R-devel/src/gnuwin32/graphapp' F:/GCC-29~1.2/BIN/MAKE.EXE -C ./getline MAKE.EXE[1]: Entering directory `F:/R-devel/src/gnuwin32/getline' MAKE.EXE[1]: `../gl.a' is up to date. MAKE.EXE[1]: Leaving directory `F:/R-devel/src/gnuwin32/getline' F:/GCC-29~1.2/BIN/MAKE.EXE -C ../extra/xdr MAKE.EXE[1]: Entering directory `F:/R-devel/src/extra/xdr' MAKE.EXE[1]: `../../gnuwin32/xdr.a' is up to date. MAKE.EXE[1]: Leaving directory `F:/R-devel/src/extra/xdr' F:/GCC-29~1.2/BIN/MAKE.EXE -C ./math MAKE.EXE[1]: Entering directory `F:/R-devel/src/gnuwin32/math' MAKE.EXE[1]: `../em.a' is up to date. MAKE.EXE[1]: Leaving directory `F:/R-devel/src/gnuwin32/math' ------- Building R.dll from R.a malloc.o libmain.a libappl.a libmath.a ga.a gl.a xdr.a em.a dllversion.o -------- echo LIBRARY R > R.def echo EXPORTS >> R.def nm R.a malloc.o libmain.a libappl.a libmath.a ga.a gl.a xdr.a em.a dllversion.o > Defs sed -n '/^........ [BCDRT] _/s/^........ [BCDRT] _/ /p' Defs >> R.def rm -f Defs gcc --shared -s -mwindows -o R.dll R.def R.a malloc.o libmain.a libappl.a libmath.a ga.a gl.a xdr.a em.a dllversion.o -L. -lg2c -lwsock32 -lcomctl32 -lversion ----------------------------------------------------------------------------------------------------------------------------------------- Then it stops with this error: ------------------------------------------------------------------------------------------------------------------------------------------ gcc: unrecognized option '-shared' ld:R.def: file format not recognized; treating as linker script ld:R.def:1: parse error F:\GCC-29~1.2\BIN\MAKE.EXE: *** [R.dll] Error 1 ------------------------------------------------------------------------------------------------------------------------------------------- What is my mistake? (Remember that the process work with the previous version 1.2.2) Thank you very much for you help Kenneth Cabrera -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Kenneth Cabrera <krcabrer@epm.net.co> writes:> gcc --shared -s -mwindows -o R.dll R.def R.a malloc.o libmain.a > libappl.a libmath.a ga.a gl.a xdr.a em.a dllversion.o -L. -lg2c > -lwsock32 -lcomctl32 -lversion > ----------------------------------------------------------------------------------------------------------------------------------------- > > Then it stops with this error: > > ------------------------------------------------------------------------------------------------------------------------------------------ > > gcc: unrecognized option '-shared'Hmm. Some stuff in MkRules has been changed: --- src/gnuwin32/MkRules 2001/02/03 20:11:37 1.26.2.1 +++ src/gnuwin32/MkRules 2001/02/08 07:25:59 1.29 @@ -3,8 +3,6 @@ # Alternatives CYGWIN (-mno-cygwin), MINGW32 (mingw32 gcc), CROSS (Linux) BUILD=MINGW32 -# Set this to YES if you use MINGW32 and gcc-2.95.2.1 or later, or cygwin-1.1. -HAVE_SHARED=YES # Where does 'HTML Help Workshop' live? (unused if compiled HTML help is # not requested. Spaces allowed.) @@ -88,16 +86,7 @@ $(NM) $^ > Defs $(SED) -n '/^........ [BCDRT] _/s/^........ [BCDRT] _/ /p' Defs >> $*.def $(RM) Defs -ifeq ($(strip $(HAVE_SHARED)),YES) $(DLL) --shared $(DLLFLAGS) $($*-DLLFLAGS) -o $@ $*.def $^ $($*-DLLLIBS) $(DLLLIBS) -else - $(DLL) -mdll -Wl,--base-file,$*.b $(DLLFLAGS) $($*-DLLFLAGS) -o $@ $^ $ ...> ld:R.def: file format not recognized; treating as linker script > ld:R.def:1: parse error > F:\GCC-29~1.2\BIN\MAKE.EXE: *** [R.dll] Error 1^^^^^^ Are you sure you are using the appropriate compiler? Looks like support for compilers before gcc-2.95.2.1 has gone away. (The flag looks a bit dodgy though. Isn't it normally "-shared" with a single hyphen?) Anyway, Brian will know...> What is my mistake? (Remember that the process work with the previous > version 1.2.2)-- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Mon, 26 Mar 2001, Kenneth Cabrera wrote:> Hello Dear R Developers and Users: > > I'm trying to compile R with all the tools of the CRAN page. > I compile the R-1.2.2 version without problem. > > But when I try to compile the development version 1.3.0 it > doesn't work when I write the command: > > make BUILD=MINGW32 > > First it shows me something like this (I think without problems): > As I make when I compile succesfuly the R-1.2.2 version:[...]> Then it stops with this error: > > ------------------------------------------------------------------------------------------------------------------------------------------ > > gcc: unrecognized option '-shared' > ld:R.def: file format not recognized; treating as linker script > ld:R.def:1: parse error > F:\GCC-29~1.2\BIN\MAKE.EXE: *** [R.dll] Error 1 > ------------------------------------------------------------------------------------------------------------------------------------------- > > What is my mistake? (Remember that the process work with the previous > version 1.2.2)Not reading the documentation! The file CHANGES says: rw1030 ===== No Windows-specific changes (nor are any planned). Support for earlier compilers has been dropped: you must use a compiler system that supports --shared and an in-memory dlltool. So, please get the compiler system as recommended in the INSTALL file. (Yours must be over a year old.) There's a good reason for this: I am not prepared to keep testing under obselete compilers like yours. If you want to play with development versions of R you do really need to read the documentation, and to be aware that it is an unstable version: there is no guarantee that it will compile anywhere, let alone under Windows (which you didn't mention: please do so). It is very likely that R-devel will be broken under Windows for several periods between now and the release of 1.3.0 as new features are added. And please do not post to both R-help and R-devel. -- Brian D. Ripley, ripley@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-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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._