Compile, MBX, fails at make check

193 views
Skip to first unread message

Francisco Guzman

unread,
Aug 1, 2022, 5:34:09 PM8/1/22
to MBX-users
Hello,

I tried to compile MBX without MPI and I still run into JSON errors during the final "make check" step".  Could this be a compiler specific issue?  What versions are used to compile MBX at the UCSD systems? Attached is the config log.


./configure --disable-optimization
make -j16
make check





++ -DHAVE_CONFIG_H -I. -I../..    -L/nasa/pkgsrc/toss3/2021Q2/lib -I/nasa/pkgsrc/toss3/2021Q2/include -DHAVE_FFTWD=1  -fopenmp -I../../src/ -std=c++11 -O0   -MT unittest-buckingham.o -MD -MP -MF $depbase.Tpo -c -o unittest-buckingham.o unittest-buckingham.cpp &&\

mv -f $depbase.Tpo $depbase.Po

In file included from ../../src/potential/buckingham/buckingham.h:44,

                 from unittest-buckingham.cpp:37:

../../src/potential/buckingham/bucktools.h:113:26: error: 'nlohmann' does not name a type

  113 |                    const nlohmann::json& repdisp_j = {});

      |                          ^~~~~~~~

../../src/potential/buckingham/bucktools.h:113:40: error: expected unqualified-id before '&' token

  113 |                    const nlohmann::json& repdisp_j = {});

      |                                        ^

../../src/potential/buckingham/bucktools.h:113:40: error: expected ')' before '&' token

  113 |                    const nlohmann::json& repdisp_j = {});

      |                                        ^

      |                                        )

../../src/potential/buckingham/bucktools.h:111:19: note: to match this '('

  111 | bool GetBuckParams(std::string mon_id1, std::string mon_id2, size_t index1, size_t index2,

      |                   ^

../../src/potential/buckingham/bucktools.h:113:42: error: expected initializer before 'repdisp_j'

  113 |                    const nlohmann::json& repdisp_j = {});

      |                                          ^~~~~~~~~

../../src/potential/buckingham/bucktools.h:113:56: error: expected unqualified-id before ')' token

  113 |                    const nlohmann::json& repdisp_j = {});

      |                                                        ^

In file included from ../../src/bblock/sys_tools.h:48,

                 from ../../src/potential/buckingham/buckingham.h:46,

                 from unittest-buckingham.cpp:37:

../../src/potential/lj/ljtools.h:125:95: error: 'nlohmann' has not been declared

  125 |                  double& out_sigma, std::vector<std::pair<std::string, std::string> > use_lj, nlohmann::json lj_j = {});

      |                                                                                               ^~~~~~~~

../../src/potential/lj/ljtools.h:125:110: error: expected ',' or '...' before 'lj_j'

  125 |                  double& out_sigma, std::vector<std::pair<std::string, std::string> > use_lj, nlohmann::json lj_j = {});

      |                                                                                                              ^~~~

In file included from ../../src/potential/buckingham/buckingham.h:46,

                 from unittest-buckingham.cpp:37:

../../src/bblock/sys_tools.h:164:81: error: 'nlohmann' has not been declared

  164 |                      std::vector<size_t> &fi_at, std::vector<size_t> &fi_sites, nlohmann::json mon_j);

      |                                                                                 ^~~~~~~~

../../src/bblock/sys_tools.h:164:96: error: expected ',' or '...' before 'mon_j'

  164 |                      std::vector<size_t> &fi_at, std::vector<size_t> &fi_sites, nlohmann::json mon_j);

      |                                                                                                ^~~~~

../../src/bblock/sys_tools.h:290:35: error: 'nlohmann' has not been declared

  290 | void GetExcluded(std::string mon, nlohmann::json mon_j, excluded_set_type &exc12, excluded_set_type &exc13,

      |                                   ^~~~~~~~

../../src/bblock/sys_tools.h:290:50: error: expected ',' or '...' before 'mon_j'

  290 | void GetExcluded(std::string mon, nlohmann::json mon_j, excluded_set_type &exc12, excluded_set_type &exc13,

      |                                                  ^~~~~

../../src/bblock/sys_tools.h:459:63: error: 'nlohmann' has not been declared

  459 |                 size_t fst_ind, std::vector<double> &chg_der, nlohmann::json mon_j);

      |                                                               ^~~~~~~~

../../src/bblock/sys_tools.h:459:78: error: expected ',' or '...' before 'mon_j'

  459 |                 size_t fst_ind, std::vector<double> &chg_der, nlohmann::json mon_j);

      |                                                                              ^~~~~

../../src/bblock/sys_tools.h:475:16: error: 'nlohmann' has not been declared

  475 |                nlohmann::json mon_j);

      |                ^~~~~~~~

../../src/bblock/sys_tools.h:475:31: error: expected ',' or '...' before 'mon_j'

  475 |                nlohmann::json mon_j);

      |                               ^~~~~

../../src/bblock/sys_tools.h:491:13: error: 'nlohmann' has not been declared

  491 |             nlohmann::json mon_j);

      |             ^~~~~~~~

../../src/bblock/sys_tools.h:491:28: error: expected ',' or '...' before 'mon_j'

  491 |             nlohmann::json mon_j);

      |                            ^~~~~

../../src/bblock/sys_tools.h:507:21: error: 'nlohmann' has not been declared

  507 |                     nlohmann::json mon_j);

      |                     ^~~~~~~~

../../src/bblock/sys_tools.h:507:36: error: expected ',' or '...' before 'mon_j'

  507 |                     nlohmann::json mon_j);

      |                                    ^~~~~

../../src/bblock/sys_tools.h:523:78: error: 'nlohmann' has not been declared

  523 |                     std::vector<std::pair<std::string, std::string>> use_lj, nlohmann::json repdisp_j);

      |                                                                              ^~~~~~~~

../../src/bblock/sys_tools.h:523:93: error: expected ',' or '...' before 'repdisp_j'

  523 |                     std::vector<std::pair<std::string, std::string>> use_lj, nlohmann::json repdisp_j);

      |                                                                                             ^~~~~~~~~

In file included from unittest-buckingham.cpp:37:

../../src/potential/buckingham/buckingham.h:106:37: error: 'nlohmann' has not been declared

  106 |     void SetJsonDispersionRepulsion(nlohmann::json repdisp_j);

      |                                     ^~~~~~~~

../../src/potential/buckingham/buckingham.h:106:52: error: expected ',' or '...' before 'repdisp_j'

  106 |     void SetJsonDispersionRepulsion(nlohmann::json repdisp_j);

      |                                                    ^~~~~~~~~

../../src/potential/buckingham/buckingham.h:107:26: error: 'nlohmann' has not been declared

  107 |     void SetJsonMonomers(nlohmann::json mon_j);

      |                          ^~~~~~~~

../../src/potential/buckingham/buckingham.h:107:41: error: expected ',' or '...' before 'mon_j'

  107 |     void SetJsonMonomers(nlohmann::json mon_j);

      |                                         ^~~~~

../../src/potential/buckingham/buckingham.h:197:5: error: 'nlohmann' does not name a type

  197 |     nlohmann::json GetJsonDispersionRepulsion();

      |     ^~~~~~~~

../../src/potential/buckingham/buckingham.h:201:5: error: 'nlohmann' does not name a type

  201 |     nlohmann::json GetJsonMonomers();

      |     ^~~~~~~~

../../src/potential/buckingham/buckingham.h:283:5: error: 'nlohmann' does not name a type

  283 |     nlohmann::json repdisp_j_;

      |     ^~~~~~~~

../../src/potential/buckingham/buckingham.h:286:5: error: 'nlohmann' does not name a type

  286 |     nlohmann::json mon_j_;

      |     ^~~~~~~~

unittest-buckingham.cpp: In function 'void ____C_A_T_C_H____T_E_S_T____4()':

unittest-buckingham.cpp:151:5: error: 'nlohmann' has not been declared

  151 |     nlohmann::json jsonDisp = {

      |     ^~~~~~~~

unittest-buckingham.cpp:166:34: error: 'jsonDisp' was not declared in this scope

  166 |     d.SetJsonDispersionRepulsion(jsonDisp);

      |                                  ^~~~~~~~

unittest-buckingham.cpp:167:5: error: 'nlohmann' has not been declared

  167 |     nlohmann::json getjsondisp = d.GetJsonDispersionRepulsion();

      |     ^~~~~~~~

In file included from ../../src/tools/testutils.h:35,

                 from unittest-buckingham.cpp:35:

unittest-buckingham.cpp:169:25: error: 'getjsondisp' was not declared in this scope

  169 |     REQUIRE(jsonDisp == getjsondisp);

      |                         ^~~~~~~~~~~

In file included from ../../src/tools/testutils.h:35,

                 from unittest-buckingham.cpp:35:

unittest-buckingham.cpp:169:25: error: 'getjsondisp' was not declared in this scope

  169 |     REQUIRE(jsonDisp == getjsondisp);

      |                         ^~~~~~~~~~~

unittest-buckingham.cpp: In function 'void ____C_A_T_C_H____T_E_S_T____6()':

unittest-buckingham.cpp:177:5: error: 'nlohmann' has not been declared

  177 |     nlohmann::json jsonMon = {

      |     ^~~~~~~~

unittest-buckingham.cpp:190:23: error: 'jsonMon' was not declared in this scope

  190 |     d.SetJsonMonomers(jsonMon);

      |                       ^~~~~~~

unittest-buckingham.cpp:191:5: error: 'nlohmann' has not been declared

  191 |     nlohmann::json getjsonmon = d.GetJsonMonomers();

      |     ^~~~~~~~

In file included from ../../src/tools/testutils.h:35,

                 from unittest-buckingham.cpp:35:

unittest-buckingham.cpp:193:24: error: 'getjsonmon' was not declared in this scope

  193 |     REQUIRE(jsonMon == getjsonmon);

      |                        ^~~~~~~~~~

In file included from ../../src/tools/testutils.h:35,

                 from unittest-buckingham.cpp:35:

unittest-buckingham.cpp:193:24: error: 'getjsonmon' was not declared in this scope

  193 |     REQUIRE(jsonMon == getjsonmon);

      |                        ^~~~~~~~~~

make[3]: *** [unittest-buckingham.o] Error 1

make[3]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

make[2]: *** [check-am] Error 2

make[2]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

make[1]: *** [check-recursive] Error 1

make[1]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src'

make: *** [check-recursive] Error 1

config.log

Marc Riera

unread,
Aug 1, 2022, 5:49:28 PM8/1/22
to MBX-users
Hello!

It may be a compiler specific error. I am able to compile MBX with GCC 4.9, GCC 6 and GCC 7. I will add a test to do it with GCC 10, since I do not have it in my computer.
Can you try specifying another GCC compiler? And just in case, get a fresh copy of MBX before recompiling. And just to make sure, all the installation commands are run in MBX_HOME, right?

If you go to our github page, there should be the actions tab. In there, there is a test making sure that the unittests run. It seems that the tests are passing well so far.

However, the very first error is that the namespace of the json file is not found, which is pretty odd since it is actually there. Please run the following commands and let us know the output:

gcc --version
which gcc

I will dig up if I can make it work with gcc10 in the meantime.

Thanks!

Marc Riera

unread,
Aug 1, 2022, 5:56:20 PM8/1/22
to MBX-users
Also, another thing, try to do make install before make check. It should not matter and make check directly should work, but again, this is some odd error I have not seen yet...

On Monday, August 1, 2022 at 5:34:09 PM UTC-4 fg...@njit.edu wrote:

Francisco Guzman

unread,
Aug 2, 2022, 9:57:53 AM8/2/22
to MBX-users
Following up, I get the same errors (fresh install) with gcc 4.8.5, 7.5, and 10.4, all with --disable-optimization.  My guess is this could me something on our end.

If you can post the versions of gcc, make, and the GSL and FFTW libraries you can build  with (and anything else you think might affect it), I'll check those here and send it up our IT chain. Also, is there a particular JSON parser the code is using internally? 

Francisco Guzman

unread,
Aug 2, 2022, 9:57:58 AM8/2/22
to MBX-users
Tried your suggestions with a fresh clone and I still see the JSON errors.  I changed gcc to gcc 7.5. Note that for our systems, the NAS team has several flavors of gcc which we just load as a module.  Below is the info for gcc -v, and which gcc. Commands are also executed from MBX_HOME (e.g, ~/MBX and NOT ~/MBX/src) else, it will not find the make file.

gcc -v
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/nasa/pkgsrc/toss3/2021Q2/gcc7/libexec/gcc/x86_64-redhat-linux/7.5.0/lto-wrapper
Target: x86_64-redhat-linux
Configured with: ../gcc-7.5.0/configure --disable-libstdcxx-pch --enable-nls --with-libiconv-prefix=/usr --disable-multilib --enable-__cxa_atexit --with-gxx-include-dir=/nasa/pkgsrc/toss3/2021Q2/gcc7/include/c++/ --enable-libssp --enable-languages='c obj-c++ objc fortran c++' --enable-shared --enable-long-long --with-local-prefix=/nasa/pkgsrc/toss3/2021Q2/gcc7 --enable-threads=posix --with-boot-ldflags='-static-libstdc++ -static-libgcc -Wl,-R/nasa/pkgsrc/toss3/2021Q2/lib  ' --with-arch=nocona --with-tune=nocona --with-fpmath=sse --prefix=/nasa/pkgsrc/toss3/2021Q2/gcc7 --build=x86_64-redhat-linux --host=x86_64-redhat-linux --infodir=/nasa/pkgsrc/toss3/2021Q2/gcc7/info --mandir=/nasa/pkgsrc/toss3/2021Q2/gcc7/man
Thread model: posix
gcc version 7.5.0 (GCC)

which gcc
/nasa/pkgsrc/toss3/2021Q2/gcc7/bin/gcc


Default version of gcc is 4.8.5, and modules available are gcc/10.2, gcc/10.3, gcc/6.5, gcc/7.5, gcc/8.4, gcc/9.3. 

I'll try a few more versions of gcc to see if anything works.

On Monday, August 1, 2022 at 5:56:20 PM UTC-4 marc.ri...@gmail.com wrote:

Marc Riera

unread,
Aug 2, 2022, 10:13:20 AM8/2/22
to MBX-users
Hello,

I have been able to install it with gcc10. The source of your error is a missing json syntax, which is located in src/json. The fact that your MBX compiles correctly means that those libraries are actually there. The file that is failing first, unittest-buckingham.cpp, has the include path properly: 
#include "json/json.h"

I am completely unable to reproduce this error, so most likely is a configuration problem on your end. Can you try passing the actual path to the compiler when doing the configure?

# Load modules
autoreconf -fi
./configure --disable-optimization CXX=PATH/TO/gcc10
make && make install
make check

I have compiled the code with several gcc compilers (tried 7, 9 and 10) in my ubuntu machine, using the latest FFTW3 (3.3.10), GSL 2.7. 

Let me know if this works. Thanks! And sorry for these problems, I really wish I would know the answer.

Francisco Guzman

unread,
Aug 5, 2022, 2:35:45 PM8/5/22
to MBX-users
Marc,

I was able to identify which of the tests fails my removing from the Makefile.in in MBX/unittests.    Any ideas? Perhaps it's something the removed tests have in common? 

All tests done with gcc 9.3

Changes to TESTS and check_PROGRAMS (all other tests are removed)

TESTS = unittest-angle$(EXEEXT) unittest-bond$(EXEEXT) \

        unittest-connectivity$(EXEEXT) unittest-dihedral$(EXEEXT) \

        unittest-energy2b$(EXEEXT) unittest-energy3b$(EXEEXT) \

        unittest-energyff$(EXEEXT) unittest-ffcalculators$(EXEEXT) \

        unittest-gamma$(EXEEXT) unittest-inversion$(EXEEXT) \

        unittest-poly-1b$(EXEEXT) unittest-poly-2b$(EXEEXT) \

        unittest-poly-3b$(EXEEXT) unittest-poly-holder-1b$(EXEEXT) \

        unittest-poly-holder-2b$(EXEEXT) \

        unittest-poly-holder-3b$(EXEEXT) unittest-ps$(EXEEXT) \


check_PROGRAMS = unittest-angle$(EXEEXT) unittest-bond$(EXEEXT) \

        unittest-connectivity$(EXEEXT) unittest-dihedral$(EXEEXT) \

        unittest-energy2b$(EXEEXT) unittest-energy3b$(EXEEXT) \

        unittest-energyff$(EXEEXT) unittest-ffcalculators$(EXEEXT) \

        unittest-gamma$(EXEEXT) unittest-inversion$(EXEEXT) \

        unittest-poly-1b$(EXEEXT) unittest-poly-2b$(EXEEXT) \

        unittest-poly-3b$(EXEEXT) unittest-poly-holder-1b$(EXEEXT) \

        unittest-poly-holder-2b$(EXEEXT) \

        unittest-poly-holder-3b$(EXEEXT) unittest-ps$(EXEEXT) \


Tests that passsed

make  check-TESTS

make[3]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

make[4]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

PASS: unittest-angle

PASS: unittest-bond

PASS: unittest-connectivity

PASS: unittest-dihedral

PASS: unittest-energy2b

PASS: unittest-energy3b

PASS: unittest-energyff

PASS: unittest-ffcalculators

PASS: unittest-gamma

PASS: unittest-inversion

PASS: unittest-poly-1b

PASS: unittest-poly-2b

PASS: unittest-poly-3b

PASS: unittest-poly-holder-1b

PASS: unittest-poly-holder-2b

PASS: unittest-poly-holder-3b

PASS: unittest-ps

============================================================================

Testsuite summary for MBX 0.7

============================================================================

# TOTAL: 17

# PASS:  17

# SKIP:  0

# XFAIL: 0

# FAIL:  0

# XPASS: 0

# ERROR: 0

============================================================================

make[4]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

make[3]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

make[2]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'

make[1]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX/src'

make[1]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX'

make[1]: Leaving directory `/nobackupp19/fguzman/MB-MD/MBX'




On Monday, August 1, 2022 at 5:56:20 PM UTC-4 marc.ri...@gmail.com wrote:

Evan Dodge

unread,
Aug 5, 2022, 2:36:12 PM8/5/22
to MBX-users
Hello,

I'm helping Francisco with his MBX build on Pleiades. I was able to build MBX on my Ubuntu laptop as well. But when I try to build on the NAS systems I can confirm that I run into the same issue that Francisco runs into with the buckingham unit test. Is it possible that this has something to do with the operating system that we're using? We're running TOSS3, which is basically Redhat 7.

I have tried multiple gcc and Intel compilers to no avail.

Please let us know how to proceed.

Thanks,

Evan Dodge
NAS User Services

Marc Riera

unread,
Aug 5, 2022, 2:45:31 PM8/5/22
to MBX-users
Hello Francisco and Evan,

Evan, just to make sure, on Pleiades you are able to compile (make && make install) but fails with the tests (make check), right? If it fails the compilation, then it would be great to see where is it failing. 

Regarding the tests, it seems that any test that uses the JSON files is failing. I fear that maybe Redhat has some configuration issues that are not compatible with the autoconf building system.. Sincerely, I am not a big fan of having the tests not running well, but since MBX is actually compiling and properly running (we tested it with Francisco), I think he is good for now, but I would really like to try to fix it.

A possible test that we could try to see if it is a problem with autoconf or a problem with the repo itself is to compile one of the tests externally. You can copy the unittest-buckingham.cpp and unittest-main.cpp somewhere else, and make sure to include the Catch2 header file. Then you should be able to compile that source file as you would compile the examples/PEFs/src/C++ file. Francisco and I managed to compile that program, which uses the library MBX externally.  If this compiles, then it is a problem with autoconf, and not the repo itself. If this still does not compile, we can even add the json headers in the same folder and see if now they are detected.

Let me know how that goes, and if you think it would be better to have a quick chat, I am also open to it.

See you!

Francisco Guzman

unread,
Aug 9, 2022, 10:00:11 AM8/9/22
to MBX-users
Hello,

For some reason the issue was with the version of the pkgsrc module used. Although the code always compiled without issues, only  pkgsrc/2020Q4 build passed the make check.  It remains a mystery why this is the case, but the following should work for any future NAS users:


module load mpi-hpe/mpt  pkgsrc/2020Q4 gcc/9.3
./configure --disable-optimization
make -j14
make install
make check

for mpi,  change ./configure as follows:
./configure --enable-mpi CXX=mpicxx

Note that Intel MPI has not been tested. For MPI, there are a few checks that fail. Any suggestions here would be appreciated. 

#Results (without MPI)

make  check-TESTS
make[3]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'
make[4]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'
PASS: unittest-angle
PASS: unittest-bond
PASS: unittest-buckingham
PASS: unittest-bucktools
PASS: unittest-connectivity
PASS: unittest-dihedral
PASS: unittest-dispersion
PASS: unittest-disptools
PASS: unittest-electrostatics
PASS: unittest-energy1b

PASS: unittest-energy2b
PASS: unittest-energy3b
PASS: unittest-energyff
PASS: unittest-ffcalculators
PASS: unittest-gamma
PASS: unittest-inversion
PASS: unittest-lj
PASS: unittest-ljtools
PASS: unittest-pme-nopolarization
PASS: unittest-pme-nopolarization-findif
PASS: unittest-pme-withpolarization

PASS: unittest-pme-withpolarization-findif

PASS: unittest-poly-1b
PASS: unittest-poly-2b
PASS: unittest-poly-3b
PASS: unittest-poly-holder-1b
PASS: unittest-poly-holder-2b
PASS: unittest-poly-holder-3b
PASS: unittest-ps
PASS: unittest-system
PASS: unittest-systools
PASS: unittest-tools
PASS: unittest-topology

============================================================================
Testsuite summary for MBX 0.7
============================================================================
# TOTAL: 33
# PASS:  33

# SKIP:  0
# XFAIL: 0
# FAIL:  0
# XPASS: 0
# ERROR: 0
============================================================================


# Results with MPI
make  check-TESTS
make[3]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'
make[4]: Entering directory `/nobackupp19/fguzman/MB-MD/MBX/src/unittests'
PASS: unittest-angle
PASS: unittest-bond
PASS: unittest-buckingham
PASS: unittest-bucktools
PASS: unittest-connectivity
PASS: unittest-dihedral
PASS: unittest-dispersion
PASS: unittest-disptools
../../build-aux/test-driver: line 109: 56699 Floating point exception(core dumped) "$@" > $log_file 2>&1
FAIL: unittest-electrostatics
PASS: unittest-energy1b

PASS: unittest-energy2b
PASS: unittest-energy3b
PASS: unittest-energyff
PASS: unittest-ffcalculators
PASS: unittest-gamma
PASS: unittest-inversion
PASS: unittest-lj
PASS: unittest-ljtools
../../build-aux/test-driver: line 109: 57813 Floating point exception(core dumped) "$@" > $log_file 2>&1
FAIL: unittest-pme-nopolarization
../../build-aux/test-driver: line 109: 57850 Floating point exception(core dumped) "$@" > $log_file 2>&1
FAIL: unittest-pme-nopolarization-findif
../../build-aux/test-driver: line 109: 57887 Floating point exception(core dumped) "$@" > $log_file 2>&1
FAIL: unittest-pme-withpolarization
../../build-aux/test-driver: line 109: 57924 Floating point exception(core dumped) "$@" > $log_file 2>&1
FAIL: unittest-pme-withpolarization-findif

PASS: unittest-poly-1b
PASS: unittest-poly-2b
PASS: unittest-poly-3b
PASS: unittest-poly-holder-1b
PASS: unittest-poly-holder-2b
PASS: unittest-poly-holder-3b
PASS: unittest-ps
../../build-aux/test-driver: line 109: 58008 Floating point exception(core dumped) "$@" > $log_file 2>&1
FAIL: unittest-system
PASS: unittest-systools
PASS: unittest-tools
PASS: unittest-topology

============================================================================
Testsuite summary for MBX 0.7
============================================================================
# TOTAL: 33
# PASS:  27

# SKIP:  0
# XFAIL: 0
# FAIL:  6

# XPASS: 0
# ERROR: 0
============================================================================
See src/unittests/test-suite.log
Please report to mbx-...@googlegroups.com
============================================================================

Marc Riera

unread,
Aug 9, 2022, 10:03:37 AM8/9/22
to MBX-users
Hello!

It is indeed a mistery. If the serial tests are passing, then everything is fine. The unittests are not supposed to be executed with the MPI build. All the unittests must be run in the serial build. I will make a note for myself to make sure to clarify that either when running make check or properly documented in the README.

Thanks a lot for helping with this issue. And pleas elet us know if there is anything else that shows odd behavior.

All the best,
Marc

Reply all
Reply to author
Forward
0 new messages