MED, does not compile
-
hi, here the build log:
g++ "-DPACKAGE_NAME=\"Salome2 Project MED module\"" -DPACKAGE_TARNAME=\"salome\" -DPACKAGE_VERSION=\"4.1.4\" "-DPACKAGE_STRING=\"Salome2 Project MED module 4.1.4\"" -DPACKAGE_BUGREPORT=\"webmaster.salome@opencascade.com\" -DPACKAGE=\"salome\" -DVERSION=\"4.1.4\" -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 -DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -DHAVE_LIBDL=1 -DHAVE_LIBRT=1 -DHAVE_LIBM=1 -DHAVE_NAMESPACES= -DHAVE_PTHREAD=1 -DWITH_NUMPY= "-DF77_FUNC(name,NAME)=name ## _" "-DF77_FUNC_(name,NAME)=name ## _" -DSIZEOF_FORTRAN_INTEGER=4 -DSIZEOF_INT=4 -DYYTEXT_POINTER=1 -D__OSVERSION__=2 -DOMNIORB= -DCORBA_HAVE_POA= -DCORBA_ORB_INIT_HAVE_3_ARGS= -DCORBA_ORB_INIT_THIRD_ARG= -I. -DPCLINUX -DHAVE_SOCKET -march=i586 -mtune=i686 -fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -O2 -g -g -D_DEBUG_ -g -Wno-deprecated -Wparentheses -Wreturn-type -Wmissing-declarations -Wunused -pthread -MT libmed_V2_1_la-MEDdatasetNumEcrire.lo -MD -MP -MF .deps/libmed_V2_1_la-MEDdatasetNumEcrire.Tpo -c MEDdatasetNumEcrire.cxx -fPIC -DPIC -o .libs/libmed_V2_1_la-MEDdatasetNumEcrire.o
MEDdatasetNumEcrire.cxx: In function 'med_2_1::med_err med_2_1::_MEDdatasetNumEcrire(med_2_1::med_idt, char*, med_2_1::med_type_champ, med_2_1::med_mode_switch, med_2_1::med_size, med_2_1::med_size, med_2_1::med_size, med_2_1::med_ssize*, med_2_1::med_int, med_2_1::med_size*, unsigned char*, med_2_1::med_mode_acces)':
MEDdatasetNumEcrire.cxx:214: warning: comparison between signed and unsigned integer expressions
MEDdatasetNumEcrire.cxx:223: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
MEDdatasetNumEcrire.cxx:226: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
MEDdatasetNumEcrire.cxx:249: warning: comparison between signed and unsigned integer expressions
MEDdatasetNumEcrire.cxx:258: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
MEDdatasetNumEcrire.cxx:261: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
MEDdatasetNumEcrire.cxx:337: warning: comparison between signed and unsigned integer expressions
MEDdatasetNumEcrire.cxx:345: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
MEDdatasetNumEcrire.cxx:376: warning: comparison between signed and unsigned integer expressions
MEDdatasetNumEcrire.cxx:385: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
MEDdatasetNumEcrire.cxx:388: error: cannot convert 'const hsize_t**' to 'const hsize_t*' for argument '4' to 'herr_t H5Sselect_elements(hid_t, H5S_seloper_t, size_t, const hsize_t*)'
make[4]: *** [libmed_V2_1_la-MEDdatasetNumEcrire.lo] Error 1
make[4]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src/MEDWrapper/V2_1/Core'
make[3]: *** [all-recursive] Error 1
make[3]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src/MEDWrapper/V2_1'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src/MEDWrapper'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src'-
-
-
-
The HDF5 api change between 1.6.4 and 1.6.7 version.
MED is based on 1.6.4 version of hdf5.
A path is to replace all (const hsize_t **) by (const hsize_t *) for each call to H5Sselect_elements.
ei :
H5Sselect_elements(memspace,H5S_SELECT_SET, pcount[0], (const hsize_t **) pflmem )
become :
H5Sselect_elements(memspace,H5S_SELECT_SET, pcount[0], (const hsize_t *) pflmem )
and so on.-
Gautier Matthieu wrote:
The HDF5 api change between 1.6.4 and 1.6.7 version.
MED is based on 1.6.4 version of hdf5.
A path is to replace all (const hsize_t **) by (const hsize_t *) for each call to H5Sselect_elements.
ei :
H5Sselect_elements(memspace,H5S_SELECT_SET, pcount[0], (const hsize_t **) pflmem )
become :
H5Sselect_elements(memspace,H5S_SELECT_SET, pcount[0], (const hsize_t *) pflmem )
and so on.
works fine!!-
hi, i still have some problems building that module... i notice it now only..so sorry...
configure does not like my libmed
i compiled MED-fichier 2.3.1 (the newst i found) with simple configure make make install
i run salome-med configure and i got it:
---------------------------------------------
testing MED2
---------------------------------------------
checking how to get verbose linking output from gfortran... -v
checking for Fortran 77 libraries of gfortran... -L/usr/lib/gcc/i586-suse-linux/4.3 -L/usr/lib/gcc/i586-suse-linux/4.3/../../../../i586-suse-linux/lib -L/usr/lib/gcc/i586-suse-linux/4.3/../../.. -lgfortranbegin -lgfortran -lm -lrt -ldl
checking for dummy main to link with Fortran 77 libraries... none
checking for Fortran 77 name-mangling scheme... lower case, underscore, no extra underscore
checking size of Fortran integer... 4
checking for int... yes
checking size of int... 4
configure: checking for MED2...
configure: WARNING: undefined MED2HOME variable which specify med2 installation directory
checking for mdump... /usr/bin/mdump
checking med.h usability... yes
checking med.h presence... yes
checking for med.h... yes
checking for MEDouvrir in -lmed... no
for med2: no
did i miss something on configure? i don't think so because i checked libmed.a file (the only one i can check, is a tar file
)
and MEDouvrir looks to exist:
anubis@Hp-Linux:~/Desktop/libmed> ls | grep MEDouvrir
MEDouvrir.o
so... what's wrong?-
Hello Andrea,
What is the result of :
nm libmed.so | grep MEDouvrir
You can download med-2.3.4 on the Code_Aster site http://www.code-aster.org/V2/spip.php?article102
I know it is really tricky to find it
-
thanks for link.. btw:
is that output useful?
anubis@Hp-Linux:~> nm /usr/lib/libmed.so | grep MEDouvrir
nm: /usr/lib/libmed.so: no symbols
anubis@Hp-Linux:~> nm /usr/lib/libmed.so
nm: /usr/lib/libmed.so: no symbols
anubis@Hp-Linux:~> locate libmed | grep /usr/lib
/usr/lib/gnome-settings-daemon-2.0/libmedia-keys.so
/usr/lib/libmed.a
/usr/lib/libmed.la
/usr/lib/libmed.so
/usr/lib/libmed.so.1
/usr/lib/libmed.so.1.1.1
/usr/lib/libmedC.a
/usr/lib/libmedC.la
/usr/lib/libmedC.so
/usr/lib/libmedC.so.1
/usr/lib/libmedC.so.1.1.1
anubis@Hp-Linux:~> cd /usr/lib/
anubis@Hp-Linux:/usr/lib> ls -lsa | grep libmed
1248 -rw-r--r-- 1 root root 1272632 10 gen 18:26 libmed.a
992 -rw-r--r-- 1 root root 1008940 10 gen 18:26 libmedC.a
4 -rwxr-xr-x 1 root root 816 10 gen 18:26 libmedC.la
0 lrwxrwxrwx 1 root root 16 11 gen 00:28 libmedC.so -> libmedC.so.1.1.1
0 lrwxrwxrwx 1 root root 16 11 gen 00:28 libmedC.so.1 -> libmedC.so.1.1.1
108 -rwxr-xr-x 1 root root 104076 10 gen 18:26 libmedC.so.1.1.1
4 -rwxr-xr-x 1 root root 994 10 gen 18:26 libmed.la
0 lrwxrwxrwx 1 root root 15 11 gen 00:29 libmed.so -> libmed.so.1.1.1
0 lrwxrwxrwx 1 root root 15 11 gen 00:27 libmed.so.1 -> libmed.so.1.1.1
156 -rwxr-xr-x 1 root root 151748 10 gen 18:26 libmed.so.1.1.1
anubis@Hp-Linux:~> nm /usr/lib/libmed.so.1.1.1
nm: /usr/lib/libmed.so.1.1.1: no symbols
anubis@Hp-Linux:~> nm /usr/lib/libmed.a | grep MEDouvrir
MEDouvrir.o:
00000000 T MEDouvrir
U MEDouvrir
-
your med is not compiled with debug informations so libmed.so doesn't contain symbol table. Output of nm is normal.
Can you give me your config.log and config.status ?-
-
i still have to have some problems with that modules.
g++ -DPACKAGE_NAME=\"Salome2\ Project\ MED\ module\" -DPACKAGE_TARNAME=\"salome\" -DPACKAGE_VERSION=\"4.1.4\" -DPACKAGE_STRING=\"Salome2\ Project\ MED\ module\ 4.1.4\" -DPACKAGE_BUGREPORT=\"webmaster.salome@opencascade.com\" -DPACKAGE=\"salome\" -DVERSION=\"4.1.4\" -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 -DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -DHAVE_LIBDL=1 -DHAVE_LIBRT=1 -DHAVE_LIBM=1 -DHAVE_NAMESPACES= -DHAVE_PTHREAD=1 -DWITH_NUMPY= -DF77_FUNC\(name,NAME\)=name\ ##\ _ -DF77_FUNC_\(name,NAME\)=name\ ##\ _ -DSIZEOF_FORTRAN_INTEGER=4 -DSIZEOF_INT=4 -DYYTEXT_POINTER=1 -D__OSVERSION__=2 -DOMNIORB= -DCORBA_HAVE_POA= -DCORBA_ORB_INIT_HAVE_3_ARGS= -DCORBA_ORB_INIT_THIRD_ARG= -I. -DPCLINUX -I/usr/include -ftemplate-depth-42 -I./../MEDWrapper/Base -I./../MEDWrapper/V2_1/Core -I./../INTERPOLATION -I./../MEDMEM -I/opt/SALOME/include/salome -DHAVE_SOCKET -O2 -g -m32 -march=i586 -mtune=i686 -fmessage-length=0 -D_FORTIFY_SOURCE=2 -g -D_DEBUG_ -g -Wno-deprecated -Wparentheses -Wreturn-type -Wmissing-declarations -Wunused -pthread -MT test_grid.o -MD -MP -MF .deps/test_grid.Tpo -c -o test_grid.o test_grid.cxx
In file included from ./../MEDMEM/MEDMEM_Mesh.hxx:37,
from ./../MEDMEM/MEDMEM_Grid.hxx:35,
from test_grid.cxx:22:
./../MEDMEM/MEDMEM_Connectivity.hxx: In member function ‘virtual const int* MEDMEM::CONNECTIVITY::getReverseConnectivity(MED_EN::medConnectivity, MED_EN::medEntityMesh)’:
./../MEDMEM/MEDMEM_Connectivity.hxx:510: warning: suggest explicit braces to avoid ambiguous ‘else’
./../MEDMEM/MEDMEM_Connectivity.hxx: In member function ‘virtual const int* MEDMEM::CONNECTIVITY::getReverseConnectivityIndex(MED_EN::medConnectivity, MED_EN::medEntityMesh)’:
./../MEDMEM/MEDMEM_Connectivity.hxx:531: warning: suggest explicit braces to avoid ambiguous ‘else’
In file included from ./../MEDMEM/MEDMEM_Mesh.hxx:1042,
from ./../MEDMEM/MEDMEM_Grid.hxx:35,
from test_grid.cxx:22:
./../MEDMEM/MEDMEM_Support.hxx: In member function ‘void MEDMEM::SUPPORT::setNumberOfElements(const int*)’:
./../MEDMEM/MEDMEM_Support.hxx:462: warning: suggest explicit braces to avoid ambiguous ‘else’
test_grid.cxx: In function ‘int main(int, char**)’:
test_grid.cxx:67: error: ‘auto_ptr’ is not a member of ‘std’
test_grid.cxx:67: error: expected primary-expression before ‘>’ token
test_grid.cxx:67: error: ‘Mesh’ was not declared in this scope
test_grid.cxx:68: warning: unused variable ‘conn’
make[2]: *** [test_grid.o] Error 1
make[2]: Leaving directory `/home/anubis/src/rpm/BUILD/MED_SRC_4.1.4/src/MEDMEMBinTest'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/anubis/src/rpm/BUILD/MED_SRC_4.1.4/src'
make: *** [all-recursive] Error 1
i notice that:
test_grid.cxx:67: error: ‘Mesh’ was not declared in this scope
did i miss something?
anubis@Hp-Linux:~> rpm -qa vtk-devel
vtk-devel-5.2.1-32.1
anubis@Hp-Linux:~> rpm -qa metis-devel
metis-devel-4.0-7.1
-
Andrea Florio wrote:
i notice that:
test_grid.cxx:67: error: ‘Mesh’ was not declared in this scope
did i miss something?
Yes :
test_grid.cxx:67: error: ‘auto_ptr’ is not a member of ‘std’
add "#include " to test_grid.cxx
This s a bug and the InstallWizard have no problem cause it doesn't compile tests.-
thanks matthieu, works fine indeed.. going on with compilation i must add other gcc4.3 fix, but now.. i really don't know how to fix it:
g++ "-DPACKAGE_NAME=\"Salome2 Project MED module\"" -DPACKAGE_TARNAME=\"salome\" -DPACKAGE_VERSION=\"4.1.4\" "-DPACKAGE_STRING=\"Salome2 Project MED module 4.1.4\"" -DPACKAGE_BUGREPORT=\"webmaster.salome@opencascade.com\" -DPACKAGE=\"salome\" -DVERSION=\"4.1.4\" -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 -DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -DHAVE_LIBDL=1 -DHAVE_LIBRT=1 -DHAVE_LIBM=1 -DHAVE_NAMESPACES= -DHAVE_PTHREAD=1 -DWITH_NUMPY= "-DF77_FUNC(name,NAME)=name ## _" "-DF77_FUNC_(name,NAME)=name ## _" -DSIZEOF_FORTRAN_INTEGER=4 -DSIZEOF_INT=4 -DYYTEXT_POINTER=1 -D__OSVERSION__=2 -DOMNIORB= -DCORBA_HAVE_POA= -DCORBA_ORB_INIT_HAVE_3_ARGS= -DCORBA_ORB_INIT_THIRD_ARG= -I. -DPCLINUX -I/usr/include -ftemplate-depth-42 -I/usr/include/libxml2 -I./../MEDMEM -I./../MEDWrapper/V2_1/Core -I/opt/SALOME/include/salome -DENABLE_METIS -I/usr/share/metis/Lib -DENABLE_SCOTCH -I/usr/share/scotch -DHAVE_SOCKET -O2 -g -m32 -march=i586 -mtune=i686 -fmessage-length=0 -D_FORTIFY_SOURCE=2 -g -D_DEBUG_ -g -Wno-deprecated -Wparentheses -Wreturn-type -Wmissing-declarations -Wunused -pthread -MT libmedsplitter_la-MEDSPLITTER_METISGraph.lo -MD -MP -MF .deps/libmedsplitter_la-MEDSPLITTER_METISGraph.Tpo -c MEDSPLITTER_METISGraph.cxx -fPIC -DPIC -o .libs/libmedsplitter_la-MEDSPLITTER_METISGraph.o
In file included from /usr/share/metis/Lib/metis.h:36,
from MEDSPLITTER_METISGraph.cxx:20:
/usr/share/metis/Lib/proto.h:462: error: declaration of C function ‘int __log2(int)’ conflicts with
/usr/include/bits/mathcalls.h:145: error: previous declaration ‘double __log2(double)’ here
make[3]: *** [libmedsplitter_la-MEDSPLITTER_METISGraph.lo] Error 1
make[3]: Leaving directory `/home/anubis/src/rpm/BUILD/MED_SRC_4.1.4/src/MEDSPLITTER'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/home/anubis/src/rpm/BUILD/MED_SRC_4.1.4/src/MEDSPLITTER'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/anubis/src/rpm/BUILD/MED_SRC_4.1.4/src'
make: *** [all-recursive] Error 1
proto.h is provided by metis 4.0 : http://pastebin.com/f407092b7
mathcalls.g instead, by glibc 2.8 : http://pastebin.com/m489c2ba9
if i have to patch one of the headers i can patch only metis..
thanks-
Hum.. Seems strange. Nothing in Salome compile with glibc.
How did you make the configure of MED ?
Where metis come from ?-
Gautier Matthieu wrote:
Hum.. Seems strange. Nothing in Salome compile with glibc.
How did you make the configure of MED ?
Where metis come from ?
metis comes from here:
i configure med like that:
export KERNEL_ROOT_DIR=/opt/SALOME
export GUI_ROOT_DIR=/opt/SALOME
export CASROOT=/opt/OpenCASCADE/
export METISDIR=/usr/share/metis
export SCOTCHDIR=/usr/share/scotch
export CFLAGS="$RPM_OPT_FLAGS"
export CXXFLAGS="$RPM_OPT_FLAGS"
./build_configure
./configure-
metis comes from here:
http://glaros.dtc.umn.edu/gkhome/metis/metis/overview
and i simply used "make" to compile it..-
I wrote nonsense on my last post on glibc and Salome. Of Course SALOME (and evrithing compiled with gcc) need glibc to compile.
Can you test this on your compile environment :
A file main.cpp:
#include "metis.h"
int main() {return 0;}
Compile this file with :
g++ main.cpp -I/usr/share/metis/Lib
Does it compile ?-
anubis@Hp-Linux:~/Desktop> touch main.cpp
anubis@Hp-Linux:~/Desktop> vi main.cpp
anubis@Hp-Linux:~/Desktop> cat main.cpp
#include "metis.h"
int main() {return 0;}
anubis@Hp-Linux:~/Desktop> g++ main.cpp -I/usr/share/metis/Lib
anubis@Hp-Linux:~/Desktop> ./a.out
looks to work, at least no output is show, a file a.out has been created but ./a.out show me nothing-
This is a known and very annoying issue of Metis 4.0... It redefines some internal standard C function, so that if you try to include "metis.h", it *may* fail to compile.
You will have to do some trick to bypass this issue.
-
David Monfort wrote:
This is a known and very annoying issue of Metis 4.0... It redefines some internal standard C function, so that if you try to include "metis.h", it *may* fail to compile.
You will have to do some trick to bypass this issue.
how can i do that?-
Difficult to say... it depends on the developer's choices.
I'll let the SALOME team answer, but I know that Code_Aster distributes a modified version of METIS and that Code_Saturne do not include the header (it redefines the METIS prototypes of the two fonctions it uses).
-
this patch to metis looks to be enought:
diff -uNr old-metis-4.0/Lib/proto.h metis-4.0/Lib/proto.h
--- old-metis-4.0/Lib/proto.h 1998-11-30 17:26:50.000000000 +0100
+++ metis-4.0/Lib/proto.h 2009-01-16 14:35:16.000000000 +0100
@@ -459,7 +459,7 @@
void srand48(long);
int ispow2(int);
void InitRandom(int);
-int log2(int);
+//int log2(int);
the gcc43 patch needed to compile MED module, instead is definetly longer:
http://pastebin.com/m226a671c
once i'll finish, i'll collect all patches into a tarball and i'll provide it to you!... in the meantime, mettiouh you can check patches in buildservice-
You can find some information on MED installation in the Code_Aster forum (see http://www.code-aster.org/) as well
-
opensuse 11.1 boost 1.36 (i think to be too new... I found that the same error is due to changes from 1.34 to 1.35 and next versions)
MED_Wrapper.cxx: In constructor 'MED::TLockProxy::TLockProxy(MED::TWrapper*)':
MED_Wrapper.cxx:43: error: 'boost::detail::thread' has not been declared
MED_Wrapper.cxx:43: error: expected primary-expression before '>' token
MED_Wrapper.cxx:43: error: '::lock' has not been declared
MED_Wrapper.cxx: In destructor 'MED::TLockProxy::~TLockProxy()':
MED_Wrapper.cxx:51: error: 'boost::detail::thread' has not been declared
MED_Wrapper.cxx:51: error: expected primary-expression before '>' token
MED_Wrapper.cxx:51: error: '::unlock' has not been declared
MED_SliceArray.hxx: In member function 'void MED::TCSlice::check_id(size_t) const [with TValueType = int]':
MED_SliceArray.hxx:82: instantiated from 'size_t MED::TCSlice::get_id(size_t) const [with TValueType = int]'
MED_SliceArray.hxx:123: instantiated from 'const TValueType& MED::TCSlice:
perator[](size_t) const [with TValueType = int]'
MED_Wrapper.cxx:183: instantiated from here
MED_SliceArray.hxx:65: warning: comparison between signed and unsigned integer expressions
MED_SliceArray.hxx: In member function 'void MED::TCSlice::check_id(size_t) const [with TValueType = double]':
MED_SliceArray.hxx:82: instantiated from 'size_t MED::TCSlice::get_id(size_t) const [with TValueType = double]'
MED_SliceArray.hxx:123: instantiated from 'const TValueType& MED::TCSlice:perator[](size_t) const [with TValueType = double]'
MED_Wrapper.cxx:483: instantiated from 'void MED:
rint(MED::SharedPtr) [with TimeStampValueType = MED::TTimeStampValue > > >]'
MED_Wrapper.cxx:510: instantiated from here
MED_SliceArray.hxx:65: warning: comparison between signed and unsigned integer expressions
make[3]: *** [libMEDWrapperBase_la-MED_Wrapper.lo] Error 1
make[3]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src/MEDWrapper/Base'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src/MEDWrapper'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/usr/src/packages/BUILD/MED_SRC_4.1.4/src'
make: *** [all-recursive] Error 1
error: Bad exit status from /var/tmp/rpm-tmp.4000 (%build)
RPM build errors:
cannot open Pubkeys index using db3 - No such file or directory (2)
Bad exit status from /var/tmp/rpm-tmp.4000 (%build)
System halted.
any idea on possible fix?-
In MED_Wrapper.cxx, please replace :
boost::detail::thread::lock_ops::lock(myWrapper->myMutex);
by
#if BOOST_VERSION >= 103500
myWrapper->myMutex.lock();
#else
boost::detail::thread::lock_ops::lock(myWrapper->myMutex);
#endif
and
boost::detail::thread::lock_ops::unlock(myWrapper->myMutex);
by
#if BOOST_VERSION >= 103500
myWrapper->myMutex.unlock();
#else
boost::detail::thread::lock_ops::unlock(myWrapper->myMutex);
#endif
Regards,
E.A.
-
works, thanks a lot...
btw also the module must be compiled with cppflag for hdf5 1.6 api, using that flag, gcc43 patch drammaticly increase.
in few days i'll send you all patches i collect, and also the exact url where reach the openSUSE 11.0/11.1 repo
thanks for everything.-
mmm.. probably the best idea is to port boost 1.34 to suse 11.1
output too long to spam in the forum, check here:
http://pastebin.com/f15102575
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Powered by
Ploneboard
