Created attachment 523694 [details] build.log.xz In root-6.12.06/math/mathmore/src/cblas.cxx, cblas_dgemm is called directly. But the root-6.12.06/math/mathmore/CMakeLists.txt only specified to link against gsl. The error (in the attached build.log.xz) > CMakeFiles/MathMore.dir/src/cblas.cxx.o: In function ROOT::Math::Blas::AMultB(int, int, int, double const*, double const*, double*)': cblas.cxx:(.text+0x30): undefined reference to `cblas_dgemm' > CMakeFiles/MathMore.dir/src/cblas.cxx.o: In function `ROOT::Math::Blas::ATMultB(int, int, int, double const*, double const*, double*)': > cblas.cxx:(.text+0x70): undefined reference to `cblas_dgemm' > collect2: error: ld returned 1 exit status disappears upon adding '-lgslblas' to the linking command. # emerge -pv root These are the packages that would be merged, in order: Calculating dependencies ... done! [ebuild R ] sci-physics/root-6.12.06:6.12/06::gentoo USE="X asimage fortran gdml gsl math minuit python roofit tbb threads tiff tmva unuran xml -R (-aqua) -avahi -davix -emacs -examples -fftw -fits -graphviz -http -jemalloc -kerberos -ldap -libcxx -memstat -mysql -odbc -opengl* -oracle -postgres (-prefix) -pythia6 -pythia8 (-qt4) -root7 -shadow -sqlite -ssl -table {-test} -vc -xinetd -xrootd" PYTHON_SINGLE_TARGET="python3_6* -python2_7* -python3_4 -python3_5" PYTHON_TARGETS="python2_7 python3_6 -python3_4 -python3_5" 0 KiB
Hi Benda, how is your GSL configured? My GSL adds cblas to its flags: $ gsl-config Usage: gsl-config [OPTION] Known values for OPTION are: --prefix show GSL installation prefix --libs print library linking information, with cblas --libs-without-cblas print library linking information, without cblas --cflags print pre-processor and compiler flags --help display this help and exit --version output version information An external CBLAS library can be specified using the GSL_CBLAS_LIB environment variable. The GSL CBLAS library is used by default. epsft-53 master $ gsl-config --libs -L/usr/lib64 -lgsl -lgslcblas -lm Also, ROOT uses "gsl-config --libs" to set $GSL_LIBRARIES to link against.
(In reply to Guilherme Amadio from comment #1) > Hi Benda, how is your GSL configured? My GSL adds cblas to its flags: > > $ gsl-config > Usage: gsl-config [OPTION] > > Known values for OPTION are: > > --prefix show GSL installation prefix > --libs print library linking information, with cblas > --libs-without-cblas print library linking information, without cblas > --cflags print pre-processor and compiler flags > --help display this help and exit > --version output version information > > An external CBLAS library can be specified using the GSL_CBLAS_LIB > environment variable. The GSL CBLAS library is used by default. > > epsft-53 master $ gsl-config --libs > -L/usr/lib64 -lgsl -lgslcblas -lm > > Also, ROOT uses "gsl-config --libs" to set $GSL_LIBRARIES to link against. Interesting, my `gsl-config --libs` outputs, -L/usr/lib64 -lgsl -lm
Maybe you have USE=cblas-external for your gsl? $ equery u gsl [ Legend : U - final flag setting for installation] [ : I - package is installed with flag ] [ Colors : set, unset ] * Found these USE flags for sci-libs/gsl-2.4: U I - - cblas-external : Link gsl with external cblas provided by (virtual/cblas) instead of shipped internal version + + deprecated : Enable deprecated functions - - static-libs : Build static versions of dynamic libraries as well
(In reply to Guilherme Amadio from comment #3) > Maybe you have USE=cblas-external for your gsl? > > $ equery u gsl > [ Legend : U - final flag setting for installation] > [ : I - package is installed with flag ] > [ Colors : set, unset ] > * Found these USE flags for sci-libs/gsl-2.4: > U I > - - cblas-external : Link gsl with external cblas provided by > (virtual/cblas) instead of shipped internal version > + + deprecated : Enable deprecated functions > - - static-libs : Build static versions of dynamic libraries as well Looks like I don't, this is the output of emerge -pv gsl [ebuild R ] sci-libs/gsl-2.4:0/23::science USE="-cblas-external -static-libs" 0 KiB Oh, wait a minute, maybe it is the problem from gsl of science overlay.
Was this a local problem? Can we close?
Ok, let's close it for the moment.
