search

PHAREHUB / PHARE

💫 Parallel Hybrid Particle In Cell code with Adaptive mesh REfinement
https://phare.readthedocs.io
GNU General Public License v3.0
71 stars 24 forks source link

2d MPI SAMRAI assertion failure #217

Closed PhilipDeegan closed 4 years ago

PhilipDeegan commented 4 years ago

``/mkn/r/llnl/samrai/master/samrai/source/SAMRAI/tbox/MessageStream.C'' at line 43 P=0000001:ERROR MESSAGE: P=0000001:Failed assertion: num_bytes >= 1

Stacktrace suggests inside: https://github.com/PHAREHUB/PHARE/blob/master/src/amr/messengers/hybrid_hybrid_messenger_strategy.h#L236

PhilipDeegan commented 4 years ago

image

in https://github.com/PHAREHUB/PHARE/blob/master/src/amr/data/particles/particles_data.h#L254 is returning 0

PhilipDeegan commented 4 years ago

image The boxes which return 0 particles, consistently have matching Y indices for lower/upper e.g. box.lower = (40,46) box.upper = (51,46)

"domainParticles" is empty so there are no "isInBox" checks

PhilipDeegan commented 4 years ago

stacktrace

/path/to/master/samrai/source/SAMRAI/tbox/MessageStream.C:30:26
/path/to/master/samrai/source/SAMRAI/tbox/MessageStream.C:67:4
/path/to/master/samrai/source/SAMRAI/tbox/Schedule.C:377:7
/path/to/master/samrai/source/SAMRAI/tbox/Schedule.C:214:4
/path/to/master/samrai/source/SAMRAI/tbox/Schedule.C:189:25
/path/to/master/samrai/source/SAMRAI/xfer/RefineSchedule.C:2167:32
/path/to/master/samrai/source/SAMRAI/xfer/RefineSchedule.C:2093:31
/path/to/phare/master/src/amr/messengers/communicators.h:185:21
/path/to/phare/master/src/amr/messengers/hybrid_hybrid_messenger_strategy.h:236:41
/path/to/phare/master/src/amr/messengers/hybrid_messenger.h:143:9
/path/to/phare/master/src/solver/level_initializer/hybrid_level_initializer.h:66:13
/path/to/phare/master/src/solver/multiphysics_integrator.h:294:13
/path/to/master/samrai/source/SAMRAI/mesh/StandardTagAndInitialize.C:164:42
/path/to/master/samrai/source/SAMRAI/mesh/GriddingAlgorithm.C:1040:50
/path/to/master/samrai/source/SAMRAI/algs/TimeRefinementIntegrator.C:430:11
/path/to/master/samrai/source/SAMRAI/algs/TimeRefinementIntegrator.C:244:48
/path/to/phare/master/src/amr/wrappers/integrator.h:99:26
/path/to/phare/master/src/simulator/simulator.h:120:5
/path/to/phare/master/src/amr/wrappers/integrator.h:99:26
/path/to/phare/master/tests/simulator/per_test.h:82:25
/path/to/phare/master/src/simulator/simulator.h:120:5
/path/to/phare/master/tests/diagnostic/test-diagnostics_2d.cpp:11:15
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:2613:28
/path/to/phare/master/tests/simulator/per_test.h:82:25gtest.cc:2649:77
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:2695:30
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:2868:3
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:3020:31
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:2649:77
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:5712:37
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:2615:1
/path/to/phare/master/subprojects/googletest/googletest/src/gtest.cc:5297:64
/path/to/phare/master/subprojects/googletest/googletest/include/gtest/gtest.h:2486:1
/path/to/phare/master/tests/diagnostic/test-diagnostics_2d.cpp:36:29
PhilipDeegan commented 4 years ago

I've tested this with the 2d splitting PR, and it seems to be resolved. so not really an issue.