[R-pkg-devel] Errors in the pre-test after submitting the R package

[Dirk Eddelbuettel]

Hi Senren,

On 30 November 2017 at 16:49, Tan, Senren wrote:
| Thank you very much for your reply.

My pleasure. You did well here.
 
| I have been trying to look at the R package fftw developed by Olaf and Uwe over the last few days. And I have tried to make use of the configure.ac. file in their source package to generate the configure script. I tried R CMD Check on my Mac, as well as on some machines running Linux. It seems that I could pass all the R CMD Checks on these Unix / Linux machines.

Great!  [ Configure/autoconf are hated by many but it is really just a shell
scripting variant.  I don't mind it as much.  And great when it works. ]
 
| Also, I have come across the following website https://cran.r-project.org/bin/windows/contrib/ThirdPartySoftware.html, on which it seems to suggest that the C library fftw3 (with the make variable LIB_FFTW) is installed on the CRAN build machine for Windows.

Yes!
| 
| So, I tried the following Makevars.win.
| 
| PKG_CPPFLAGS = -I$(LIB_FFTW)/include  -I$(LIB_FFTW)
| CXX11 = g++
| PKG_CXXFLAGS = -std=c++11
| PKG_LIBS = -L$(LIB_FFTW)/lib${R_ARCH} -lfftw3

That looks correct to me (assuming LIB_FFTW is the chosen name). I do
something similar with my packages linking to the GNU GSL.
 
| However, when I tried to re-submit the package to CRAN, I still got the same error message on the pre-test as follows, suggesting that the machine had not detected the fftw3 library.
| 
|  **********************************************
|    WARNING: this package has a configure script
|          It probably needs manual configuration
|    **********************************************
| 
| ** libs
| 
| *** arch - i386
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c RcppExports.cpp -o RcppExports.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c cont_ks_dist.cpp -o cont_ks_dist.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c crossprob_new.cc -o crossprob_new.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c fftw_wrappers.cc -o fftw_wrappers.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c fftwconvolver.cc -o fftwconvolver.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c one_sided_noncrossing_probability.cc -o one_sided_noncrossing_probability.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c one_sided_noncrossing_probability_n2.cc -o one_sided_noncrossing_probability_n2.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c one_sided_noncrossing_probability_n2logn.cc -o one_sided_noncrossing_probability_n2logn.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c poisson_pmf.cc -o poisson_pmf.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c read_boundaries_file.cc -o read_boundaries_file.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c string_utils.cc -o string_utils.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -Id:/Rcompile/CRANpkg/extralibs215/local215/include -Id:/Rcompile/CRANpkg/extralibs215/local215 -I"d:/RCompile/CRANpkg/lib/3.5/Rcpp/include"   -I"d:/Compiler/gcc-4.9.3/local330/include"  -std=c++11   -O2 -Wall  -mtune=core2 -c two_sided_noncrossing_probability.cc -o two_sided_noncrossing_probability.o
| d:/Compiler/gcc-4.9.3/mingw_32/bin/g++ -shared -s -static-libgcc -o KSgeneral.dll tmp.def RcppExports.o cont_ks_dist.o crossprob_new.o fftw_wrappers.o fftwconvolver.o one_sided_noncrossing_probability.o one_sided_noncrossing_probability_n2.o one_sided_noncrossing_probability_n2logn.o poisson_pmf.o read_boundaries_file.o string_utils.o two_sided_noncrossing_probability.o -Ld:/Rcompile/CRANpkg/extralibs215/local215/lib/i386 -lfftw3 -Ld:/Compiler/gcc-4.9.3/local330/lib/i386 -Ld:/Compiler/gcc-4.9.3/local330/lib -LD:/RCompile/recent/R/bin/i386 -lR

That looks like it expanded right.

What I would do now is take one of the working packages, create a local
tarball _and put my name and email into Maintainer_ because that way the
win-builder test log will come back to me.  And then submit, and compare.

| fftw_wrappers.o:fftw_wrappers.cc:(.text+0x125): undefined reference to `_imp__fftw_plan_dft_r2c_1d'
| fftw_wrappers.o:fftw_wrappers.cc:(.text+0x14f): undefined reference to `_imp__fftw_destroy_plan'
| fftw_wrappers.o:fftw_wrappers.cc:(.text+0x29b): undefined reference to `_imp__fftw_execute'
| fftw_wrappers.o:fftw_wrappers.cc:(.text+0x315): undefined reference to `_imp__fftw_plan_dft_c2r_1d'
| fftw_wrappers.o:fftw_wrappers.cc:(.text+0x33f): undefined reference to `_imp__fftw_destroy_plan'
| fftw_wrappers.o:fftw_wrappers.cc:(.text+0x3bb): undefined reference to `_imp__fftw_execute'
| collect2.exe: error: ld returned 1 exit status
| no DLL was created

And I presume this works locally for you, ie these symbols are in fact in the
DLL you try to use?
 
| I was wondering whether I should include in the configure.ac file a few lines about downloading the library fftw3 when it is not available on the machine? Or have I made some mistakes in the Makevars.win file?

You could -- but if it works on *nix you can leave it alone.
 
| Thank you very much and I apologize for any inconvenience I may have caused you.

You are using the mailing list for the very purpose it was created for so no
apology needed or accepted!

Cheers, Dirk

-- 
http://dirk.eddelbuettel.com | @eddelbuettel | edd at debian.org