CI builds for non-amd64 architectures #2401

mkuron · 2018-12-05T10:53:28Z

These tests will only run once a week and for tagged releases.

Should be tagged for Espresso 4.0.1 as it is part of the release checklist. We only need the .gitlab-ci.yml changes there though as I think everything else was caused by changes since the 4.0 release.

codecov · 2018-12-05T11:07:28Z

Codecov Report

Merging #2401 into python will not change coverage.
The diff coverage is n/a.

@@          Coverage Diff           @@
##           python   #2401   +/-   ##
======================================
  Coverage      73%     73%           
======================================
  Files         393     393           
  Lines       18477   18477           
======================================
  Hits        13533   13533           
  Misses       4944    4944

Impacted Files	Coverage Δ
src/core/unit_tests/field_coupling_fields_test.cpp	`100% <ø> (ø)`	⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b5e2b26...223c659. Read the comment docs.

mkuron · 2018-12-06T15:00:49Z

@RudolfWeeber, can you have a look at why the collision detection test fails on i386?

@fweik, can you have a look why I needed this terrible hack to make particle deletion work on i386?

I still need to fix up the s390x (big-endian) emulation problem.

mkuron · 2018-12-06T16:02:15Z

I think I finally reproduced the problem of hanging tests that @junghans reported for i386. There is an MPI deadlock which the npt testcase sometimes runs into. @fweik, any ideas?

rank 0:

#6  0xf5eb89a6 in PMPI_Test () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf5c07ab4 in boost::mpi::request::handle_serialized_irecv<ParticleList> (self=0xffd7ca48, action=boost::mpi::request::ra_test)
    at /usr/include/boost/mpi/communicator.hpp:1595
#8  0xf5e3f5a9 in boost::mpi::request::test() () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#9  0xf5c087e2 in boost::mpi::wait_all<boost::mpi::request*, boost::mpi::status*> (first=0xffd7ca34, last=0xffd7ca5c, out=0xffd7ca5c)
    at /usr/include/boost/mpi/nonblocking.hpp:196
#10 0xf5c0661d in Utils::Mpi::detail::sendrecv_impl<ParticleList> (rval=..., rtag=0, src=<optimized out>, sval=..., stag=0, dest=<optimized out>, comm=...)
    at /builds/espressomd/espresso/src/core/utils/mpi/sendrecv.hpp:60
#11 Utils::Mpi::sendrecv<ParticleList> (rval=..., rtag=0, src=<optimized out>, sval=..., stag=0, dest=<optimized out>, comm=...)
    at /builds/espressomd/espresso/src/core/utils/mpi/sendrecv.hpp:69
#12 (anonymous namespace)::exchange_neighbors (pl=pl@entry=0xffd7cbcc) at /builds/espressomd/espresso/src/core/domain_decomposition.cpp:811
#13 0xf5c06fbf in dd_exchange_and_sort_particles (global=0, pl=0xffd7cbcc) at /builds/espressomd/espresso/src/core/domain_decomposition.cpp:861
#14 0xf5bd8224 in cells_resort_particles (global_flag=0) at /builds/espressomd/espresso/src/core/cells.cpp:402
#15 0xf5bd8566 in cells_update_ghosts () at /builds/espressomd/espresso/src/core/cells.cpp:486
#16 0xf5c3d17e in integrate_vv (n_steps=2, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:370

rank 1:

#6  0xf5ea2348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf5c24208 in ghost_communicator (gc=0xf5ce4168 <cell_structure+8>, data_parts=64)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:557
#8  0xf5c24a5b in ghost_communicator (gc=0xf5ce4168 <cell_structure+8>)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:476
#9  0xf5bc91d4 in cells_resort_particles (global_flag=<optimized out>) at /builds/espressomd/espresso/src/core/cells.cpp:423
#10 0xf5bc9566 in cells_update_ghosts () at /builds/espressomd/espresso/src/core/cells.cpp:486
#11 0xf5c2e17e in integrate_vv (n_steps=2, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:370

fweik · 2018-12-10T09:01:14Z

#2410 contains a solution for the remove particles issue, there was actually a bug. I didn't have any luck with the deadlock on Friday, but looking into it again now.

mkuron · 2018-12-12T16:13:21Z

I have fixed the s390x issue and @fweik's fix for removing particles works.
That leaves two issues: the collision detection (@RudolfWeeber) and the deadlock (@fweik).

RudolfWeeber · 2018-12-12T17:07:23Z

My impression is that
folded_positoin()
does not substract a box-length in y direction
Traceback (most recent call last):
File "collision_detection.py", line 225, in test_bind_at_point_of_collision
self.run_test_bind_at_point_of_collision_for_pos(np.array((0, 0, 0)))
File "collision_detection.py", line 135, in run_test_bind_at_point_of_collisio
_pos
self.verify_state_after_bind_at_poc(expected_np)
File "collision_detection.py", line 176, in verify_state_after_bind_at_poc
self.verify_bind_at_poc_pair(base_p1, base_p2, p, p2)
File "collision_detection.py", line 220, in verify_bind_at_poc_pair
np.copy(p.pos_folded), expected_pos, atol=1E-4)
File "/usr/lib/python2.7/dist-packages/numpy/testing/utils.py", line 1395, in
t_allclose
verbose=verbose, header=header, equal_nan=equal_nan)
File "/usr/lib/python2.7/dist-packages/numpy/testing/utils.py", line 778, in a
_array_compare
raise AssertionError(msg)
AssertionError:
Not equal to tolerance rtol=1e-07, atol=0.0001

(mismatch 33.3333333333%)
x: array([ 6.000000e-02, 1.000000e+00, 8.881784e-18])
y: array([ 0.06, 0. , 0. ])

Ran 8 tests in 2.029s

FAILED (failures=1)

The box length is 1.
The code reads:
# Check placement
if rel_to == p1.id:
dist_centers = np.copy(p2.pos - p1.pos)
else:
dist_centers = p1.pos - p2.pos
expected_pos = self.s.part[rel_to].pos_folded +
self.s.collision_detection.vs_placement * dist_centers
np.testing.assert_allclose(
np.copy(p.pos_folded), expected_pos, atol=1E-4)

mkuron · 2018-12-18T09:21:50Z

Collision detection: fixed by #2415. MPI deadlock: fixed by #2421. I've updated this pull request and we'll see whether that makes all tests pass on all architectures.

fweik · 2018-12-18T10:47:39Z

Argh some of the collectives are also implemented in terms of isend and suffer from the same bug. I'll look into that, but this may be not feasible to fix.

…

On Tue, Dec 18, 2018, 11:38 Michael Kuron ***@***.*** wrote: i386 still has a deadlock during *interactions_non-bonded*: rank 0: #8 0xf70a2f61 in PMPI_Gather () from target:/usr/lib/i386-linux-gnu/libmpi.so.20 #9 0xf6e89d2c in mpi_who_has () at /builds/espressomd/espresso/src/core/particle_data.cpp:144 #10 0xf6e89fc4 in build_particle_node () at /builds/espressomd/espresso/src/core/particle_data.cpp:171 #11 0xf6e8a047 in get_particle_node (id=<optimized out>) at /builds/espressomd/espresso/src/core/particle_data.cpp:181 #12 0xf6e8b03c in get_particle_data (part=<optimized out>) at /builds/espressomd/espresso/src/core/particle_data.cpp:375 #13 0xe8179841 in __pyx_f_10espressomd_13particle_data_14ParticleHandle_update_particle_data (__pyx_v_self=0xd93273f8) at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:3550 #14 0xe8180946 in __pyx_pf_10espressomd_13particle_data_14ParticleHandle_3pos_2__get__ (__pyx_v_self=0xd93273f8) at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:4571 #15 __pyx_pw_10espressomd_13particle_data_14ParticleHandle_3pos_3__get__ (__pyx_v_self=0xd93273f8) at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:4550 #16 __pyx_getprop_10espressomd_13particle_data_14ParticleHandle_pos (o=0xd93273f8, x=0x0) at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:33957 rank 1: #6 0xf7080348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20 #7 0xf7a800de in boost::mpi::detail::packed_archive_recv(ompi_communicator_t*, int, int, boost::mpi::packed_iarchive&, ompi_status_public_t&) () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1 #8 0xf7a73226 in void boost::mpi::broadcast<boost::mpi::packed_iarchive>(boost::mpi::communicator const&, boost::mpi::packed_iarchive&, int) () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1 #9 0xf6dfec40 in boost::mpi::detail::broadcast_impl<IA_parameters> (values=0x57aab3d0, root=0, n=1, comm=...) at /usr/include/boost/mpi/collectives/broadcast.hpp:118 #10 0xf6dfee52 in boost::mpi::broadcast<IA_parameters> (root=0, value=..., comm=...) at /usr/include/boost/mpi/collectives/broadcast.hpp:128 #11 mpi_bcast_ia_params_slave (i=0, j=0) at /builds/espressomd/espresso/src/core/communication.cpp:1204 #12 0xf6e540eb in std::function<void (int, int)>::operator()(int, int) const (__args#1=<optimized out>, __args#0=<optimized out>, this=<optimized out>) at /usr/include/c++/7/bits/std_function.h:706 #13 Communication::MpiCallbacks::loop (this=0x57a74820) at /builds/espressomd/espresso/src/core/MpiCallbacks.cpp:91 #14 0xf6dfc8b9 in mpi_loop () at /builds/espressomd/espresso/src/core/communication.cpp:2312 #15 0xf7ef4447 in init_init () at /builds/espressomd/espresso/build/src/python/espressomd/_init.cpp:1149 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2401 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA2i3xku4Hq3ZH8ra3bLOPZ5C0N7vCw1ks5u6MWmgaJpZM4ZCenN> .

mkuron · 2018-12-18T11:05:41Z

https://gitlab.icp.uni-stuttgart.de/espressomd/espresso/pipelines/4521

@RudolfWeeber, the collision detection problem still exists.

@fweik, i386 still has four deadlocks, two of which involve Boost:

during interactions_non-bonded:

rank 0:

#8  0xf70a2f61 in PMPI_Gather () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#9  0xf6e89d2c in mpi_who_has () at /builds/espressomd/espresso/src/core/particle_data.cpp:144
#10 0xf6e89fc4 in build_particle_node () at /builds/espressomd/espresso/src/core/particle_data.cpp:171
#11 0xf6e8a047 in get_particle_node (id=<optimized out>) at /builds/espressomd/espresso/src/core/particle_data.cpp:181
#12 0xf6e8b03c in get_particle_data (part=<optimized out>) at /builds/espressomd/espresso/src/core/particle_data.cpp:375
#13 0xe8179841 in __pyx_f_10espressomd_13particle_data_14ParticleHandle_update_particle_data (__pyx_v_self=0xd93273f8)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:3550
#14 0xe8180946 in __pyx_pf_10espressomd_13particle_data_14ParticleHandle_3pos_2__get__ (__pyx_v_self=0xd93273f8)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:4571
#15 __pyx_pw_10espressomd_13particle_data_14ParticleHandle_3pos_3__get__ (__pyx_v_self=0xd93273f8)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:4550
#16 __pyx_getprop_10espressomd_13particle_data_14ParticleHandle_pos (o=0xd93273f8, x=0x0)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:33957

rank 1:

#6  0xf7080348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf7a800de in boost::mpi::detail::packed_archive_recv(ompi_communicator_t*, int, int, boost::mpi::packed_iarchive&, ompi_status_public_t&) () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#8  0xf7a73226 in void boost::mpi::broadcast<boost::mpi::packed_iarchive>(boost::mpi::communicator const&, boost::mpi::packed_iarchive&, int) () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#9  0xf6dfec40 in boost::mpi::detail::broadcast_impl<IA_parameters> (values=0x57aab3d0, root=0, n=1, comm=...)
    at /usr/include/boost/mpi/collectives/broadcast.hpp:118
#10 0xf6dfee52 in boost::mpi::broadcast<IA_parameters> (root=0, value=..., comm=...)
    at /usr/include/boost/mpi/collectives/broadcast.hpp:128
#11 mpi_bcast_ia_params_slave (i=0, j=0) at /builds/espressomd/espresso/src/core/communication.cpp:1204

during rigid_bond:

rank 0:

#6  0xf7189348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf6f05a58 in ghost_communicator (gc=0xf6fc5180 <cell_structure+32>, data_parts=14)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:557
#8  0xf6f062ab in ghost_communicator (gc=0xf6fc5180 <cell_structure+32>)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:476
#9  0xf6f3adef in correct_pos_shake () at /builds/espressomd/espresso/src/core/rattle.cpp:179
#10 0xf6f0f837 in integrate_vv (n_steps=5000, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:348
#11 0xf6ebca6f in mpi_integrate (n_steps=5000, reuse_forces=0)
    at /builds/espressomd/espresso/src/core/communication.cpp:1155

rank 1:

#6  0xf70dd348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf6e59a58 in ghost_communicator (gc=0xf6f19180 <cell_structure+32>, data_parts=14)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:557
#8  0xf6e5a2ab in ghost_communicator (gc=0xf6f19180 <cell_structure+32>)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:476
#9  0xf6e8edef in correct_pos_shake () at /builds/espressomd/espresso/src/core/rattle.cpp:179
#10 0xf6e63837 in integrate_vv (n_steps=5000, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:348
#11 0xf6e0e0cc in mpi_integrate_slave (n_steps=5000, reuse_forces=0)
    at /builds/espressomd/espresso/src/core/communication.cpp:1162

during virtual_sites_tracers:

rank 0:

#6  0xf7170d6d in PMPI_Sendrecv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf662d3a6 in halo_push_communication (lbfluid=...)
    at /builds/espressomd/espresso/src/core/grid_based_algorithms/lb.cpp:1652
#8  0xf6631040 in lb_collide_stream () at /builds/espressomd/espresso/src/core/grid_based_algorithms/lb.cpp:2529
#9  0xf662aedf in lattice_boltzmann_update () at /builds/espressomd/espresso/src/core/grid_based_algorithms/lb.cpp:2562
#10 0xf6ef189d in integrate_vv (n_steps=500, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:410
#11 0xf6e9ea6f in mpi_integrate (n_steps=500, reuse_forces=0) at /builds/espressomd/espresso/src/core/communication.cpp:1155

rank 1:

#6  0xf70b8d8b in PMPI_Sendrecv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf6575171 in halo_push_communication (lbfluid=...)
    at /builds/espressomd/espresso/src/core/grid_based_algorithms/lb.cpp:1611
#8  0xf6579040 in lb_collide_stream () at /builds/espressomd/espresso/src/core/grid_based_algorithms/lb.cpp:2529
#9  0xf6572edf in lattice_boltzmann_update () at /builds/espressomd/espresso/src/core/grid_based_algorithms/lb.cpp:2562
#10 0xf6e3989d in integrate_vv (n_steps=500, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:410
#11 0xf6de40cc in mpi_integrate_slave (n_steps=500, reuse_forces=0)

during observable_cylindrical:

rank 0:

#6  0xf5e64348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf5df20de in boost::mpi::detail::packed_archive_recv(ompi_communicator_t*, int, int, boost::mpi::packed_iarchive&, ompi_status_public_t&) () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#8  0xf5dea589 in boost::mpi::status boost::mpi::communicator::recv<boost::mpi::packed_iarchive>(int, int, boost::mpi::packed_iarchive&) const () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#9  0xf5ba08c2 in boost::mpi::communicator::recv_impl<Particle> (this=0xf5ca6cfc <comm_cart>, tag=42, value=..., source=1)
    at /usr/include/boost/mpi/communicator.hpp:1265
#10 boost::mpi::communicator::recv<Particle> (value=..., tag=42, source=1, this=0xf5ca6cfc <comm_cart>)
    at /usr/include/boost/mpi/communicator.hpp:1278
#11 mpi_recv_part (pnode=1, part=76) at /builds/espressomd/espresso/src/core/communication.cpp:1110
#12 0xf5bfc0b8 in get_particle_data (part=<optimized out>) at /builds/espressomd/espresso/src/core/particle_data.cpp:391
#13 0xe87cd841 in __pyx_f_10espressomd_13particle_data_14ParticleHandle_update_particle_data (__pyx_v_self=0xd96353c8)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:3550
#14 0xe87d4946 in __pyx_pf_10espressomd_13particle_data_14ParticleHandle_3pos_2__get__ (__pyx_v_self=0xd96353c8)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:4571
#15 __pyx_pw_10espressomd_13particle_data_14ParticleHandle_3pos_3__get__ (__pyx_v_self=0xd96353c8)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:4550
#16 __pyx_getprop_10espressomd_13particle_data_14ParticleHandle_pos (o=0xd96353c8, x=0x0)
    at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:33957

rank 1:

#9  0xf5ee7cdc in PMPI_Bcast () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#10 0xf5c9106d in boost::mpi::detail::broadcast_impl<int> (n=3, root=0, values=0xfff4b70c, comm=...)
    at /usr/include/boost/mpi/collectives/broadcast.hpp:96
#11 boost::mpi::broadcast<int> (root=0, n=3, values=<optimized out>, comm=...)
    at /usr/include/boost/mpi/collectives/broadcast.hpp:134
#12 Communication::MpiCallbacks::loop (this=0x587b33e0) at /builds/espressomd/espresso/src/core/MpiCallbacks.cpp:85
#13 0xf5c398b9 in mpi_loop () at /builds/espressomd/espresso/src/core/communication.cpp:2312

mkuron · 2019-01-09T12:58:46Z

I've merged the current master and #2431. We'll see how it goes.

junghans · 2019-01-09T14:08:19Z

I can re-package for the next release and see what still fails.

mkuron · 2019-01-09T14:21:08Z

@fweik, the above MPI deadlock on i386 during virtual_sites_tracers still happens. So that one wasn't Boost's fault and still needs to be fixed. And there seems to be a new one in dpd.

Rank 0:

#6  0xf799a9a6 in PMPI_Test () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf59009f0 in boost::mpi::request::handle_serialized_array_irecv<Particle> (self=0x57255780, 
    action=boost::mpi::request::ra_test) at /usr/include/boost/mpi/communicator.hpp:1640
#8  0xf7c2f5a9 in boost::mpi::request::test() () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#9  0xf589645e in boost::mpi::wait_all<__gnu_cxx::__normal_iterator<boost::mpi::request*, std::vector<boost::mpi::request, std::allocator<boost::mpi::request> > > > (first=..., last=...) at /usr/include/boost/mpi/nonblocking.hpp:268
#10 0xf5900dea in Utils::Mpi::detail::gatherv_impl<Particle> (comm=..., in_values=<optimized out>, in_size=<optimized out>, 
    out_values=<optimized out>, sizes=0x571bca30, displs=0x571b7a00, root=<optimized out>)
    at /builds/espressomd/espresso/src/core/utils/mpi/gatherv.hpp:70
#11 0xf5900ff0 in Utils::Mpi::gatherv<Particle> (comm=..., in_values=<optimized out>, in_size=<optimized out>, 
    out_values=<optimized out>, sizes=<optimized out>, root=<optimized out>)
    at /builds/espressomd/espresso/src/core/utils/mpi/gatherv.hpp:97
#12 0xf58fcdd1 in mpi_get_particles (ids=...) at /builds/espressomd/espresso/src/core/particle_data.cpp:448
#13 0xf58fd126 in prefetch_particle_data (ids=...) at /builds/espressomd/espresso/src/core/particle_data.cpp:477
#14 0xe882b017 in __pyx_gb_10espressomd_13particle_data_18_ParticleSliceImpl_6generator (__pyx_generator=<optimized out>, 
    __pyx_sent_value=<optimized out>) at /builds/espressomd/espresso/build/src/python/espressomd/particle_data.cpp:18535

rank 1:

#6  0xf7902348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf7b9e0de in boost::mpi::detail::packed_archive_recv(ompi_communicator_t*, int, int, boost::mpi::packed_iarchive&, ompi_status_public_t&) () from target:/usr/lib/i386-linux-gnu/libboost_mpi.so.1.65.1
#8  0xf586a914 in boost::mpi::detail::scatter_impl<std::vector<int, std::allocator<int> > > (comm=..., 
    out_values=out_values@entry=0xffd99ce4, root=root@entry=0, n=1) at /usr/include/boost/mpi/collectives/scatter.hpp:92
#9  0xf586b87e in boost::mpi::scatter<std::vector<int, std::allocator<int> > > (root=0, out_value=..., comm=...)
    at /usr/include/boost/mpi/collectives/scatter.hpp:125
#10 mpi_get_particles_slave () at /builds/espressomd/espresso/src/core/particle_data.cpp:397

And a new one in collision_detection.

Rank 0:

#6  0xf79ce348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf6b4e8c8 in ghost_communicator (gc=0xf6c0e18c <cell_structure+44>, data_parts=16)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:557
#8  0xf6b4f11b in ghost_communicator (gc=0xf6c0e18c <cell_structure+44>)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:476
#9  0xf6b3a50b in force_calc () at /builds/espressomd/espresso/src/core/forces.cpp:183
#10 0xf6b58572 in integrate_vv (n_steps=5000, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:379
#11 0xf6b05b4f in mpi_integrate (n_steps=5000, reuse_forces=0)
    at /builds/espressomd/espresso/src/core/communication.cpp:1155

Rank 1:

#6  0xf7927348 in PMPI_Recv () from target:/usr/lib/i386-linux-gnu/libmpi.so.20
#7  0xf6aa78c8 in ghost_communicator (gc=0xf6b6718c <cell_structure+44>, data_parts=16)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:557
#8  0xf6aa811b in ghost_communicator (gc=0xf6b6718c <cell_structure+44>)
    at /builds/espressomd/espresso/src/core/ghosts.cpp:476
#9  0xf6a93536 in force_calc () at /builds/espressomd/espresso/src/core/forces.cpp:190
#10 0xf6ab1572 in integrate_vv (n_steps=5000, reuse_forces=0) at /builds/espressomd/espresso/src/core/integrate.cpp:379
#11 0xf6a5c1ac in mpi_integrate_slave (n_steps=5000, reuse_forces=0)
    at /builds/espressomd/espresso/src/core/communication.cpp:1162

Also, the one reported for npt above has come back.

fweik · 2019-01-10T08:23:54Z

I rechecked with a newer boost (and ubuntu), with cosmic and boost 1.67 on i386 only the collision detection test fails with [0300cee22312:14302] Read -1, expected 17088, errno = 1 which was fixed somewhere else iirc. My best guess is that there were more bugs in boost than the one I patched.
(I checked 3a66795 with the ubuntu:i386 docker image, where I changed the base image to cosmic)

mkuron · 2019-01-10T09:27:45Z

Read -1, expected 17088, errno = 1 sounds similar to #2271, but that isn't actually related. We could of course switch these containers to cosmic, but I'm not convinced this is (exclusively) Boost's fault. The dpd and npt deadlocks looks like it might require a similar fix to Boost, but _ virtual_sites_tracers_ and collision_detection look like genuine bugs

fweik · 2019-01-10T09:34:26Z

I think the virtual_sites_tracers is genuine, that deadlocks in the lb halo comm. Unfortunately this code is not easy to read (for me). The other one looks like the issues we had' before. (One of the failure modes of the boost problem is that one communication is skipped on the sending side, which leads to a deadlock somewhere later).

mkuron · 2019-01-11T16:44:03Z

After upgrading to Boost 1.67, all segfaults are actually gone, so maybe we can still get this pull request into Espresso 4.0.1. However, the coordinate folding issue in collision detection is back, @RudolfWeeber:

68/113 Test  #68: collision_detection .....................***Failed    4.66 sec
.F.....{'bond_centers': HarmonicBond(0): {'r_0': 0.1, 'k': 5000.0, 'r_cut': 0.0}, 'distance': 0.11, 'mode': 'bind_centers'}
.
======================================================================
FAIL: test_bind_at_point_of_collision (__main__.CollisionDetection)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/builds/espressomd/espresso/build/testsuite/python/collision_detection.py", line 225, in test_bind_at_point_of_collision
    self.run_test_bind_at_point_of_collision_for_pos(np.array((0, 0, 0)))
  File "/builds/espressomd/espresso/build/testsuite/python/collision_detection.py", line 135, in run_test_bind_at_point_of_collision_for_pos
    self.verify_state_after_bind_at_poc(expected_np)
  File "/builds/espressomd/espresso/build/testsuite/python/collision_detection.py", line 176, in verify_state_after_bind_at_poc
    self.verify_bind_at_poc_pair(base_p1, base_p2, p, p2)
  File "/builds/espressomd/espresso/build/testsuite/python/collision_detection.py", line 220, in verify_bind_at_poc_pair
    np.copy(p.pos_folded), expected_pos, atol=1E-4)
  File "/usr/lib/python2.7/dist-packages/numpy/testing/nose_tools/utils.py", line 1396, in assert_allclose
    verbose=verbose, header=header, equal_nan=equal_nan)
  File "/usr/lib/python2.7/dist-packages/numpy/testing/nose_tools/utils.py", line 779, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Not equal to tolerance rtol=1e-07, atol=0.0001

(mismatch 33.3333333333%)
 x: array([6.000000e-02, 1.000000e+00, 8.881784e-18])
 y: array([0.06, 0.  , 0.  ])

…ctures

RudolfWeeber · 2019-01-14T10:29:19Z

Please merge pr2422 into this one.

mkuron · 2019-01-14T11:42:55Z

@RudolfWeeber, it still fails after merging the current master

fweik · 2019-01-14T11:57:52Z

I don't think it's in the master, you have to merge the mentioned pr. Should be a change in grid.hpp

mkuron · 2019-01-14T12:17:36Z

No, it's there. Also, I meant python, not master.

junghans · 2019-01-18T18:12:40Z

@mkuron I set this up for another project, but might be nice here to as you would get testing on x86_64, i386 and ppc64le and a couple of different distros. You can use https://src.fedoraproject.org/rpms/espresso/blob/master/f/espresso.spec and a starting point.

Otherwise the collision detection test fails due to some unfolded coordinates

mkuron · 2019-01-21T13:25:02Z

ready to merge

CI builds for non-amd64 architectures

9c4b715

mkuron added 2 commits December 6, 2018 11:49

DPD test can't be better than machine precision

e62e96b

Fix tests broken by optimization on i386

213101a

mkuron mentioned this pull request Dec 7, 2018

Fix s390x QEMU container espressomd/docker#49

Merged

Merge branch 'python' of github.com:espressomd/espresso into patch-8

d159edd

Merge remote-tracking branch 'upstream/python' into patch-8

7fcaea4

Merge branch 'python' of github.com:espressomd/espresso into patch-8

edffac0

mkuron mentioned this pull request Jan 11, 2019

Upgrade non-amd64 images to Ubuntu 18.10 for Boost 1.67 espressomd/docker#54

Merged

Require Boost 1.67 on 32-bit architectures

d79c491

mkuron mentioned this pull request Jan 13, 2019

Revert ppc64le to bionic espressomd/docker#55

Merged

mkuron added 2 commits January 13, 2019 15:26

Fix arm64 CI build

59fb4d8

Reduce some tests' precision requirements so they pass on ARM archite…

02e0fea

…ctures

Merge branch 'python' of github.com:espressomd/espresso into patch-8

9229a7a

mkuron changed the title ~~WIP: CI builds for non-amd64 architectures~~ CI builds for non-amd64 architectures Jan 14, 2019

mkuron changed the title ~~CI builds for non-amd64 architectures~~ WIP: CI builds for non-amd64 architectures Jan 14, 2019

mkuron mentioned this pull request Jan 14, 2019

backport support for non-amd64 architectures to 4.0.1 #2438

Merged

prevent 80-bit arithmetic on old Intel processors

91a79ae

Otherwise the collision detection test fails due to some unfolded coordinates

mkuron mentioned this pull request Jan 21, 2019

prevent 80-bit arithmetic on old Intel processors #2454

Merged

CI: don't build exotic architectures for pull requests

223c659

mkuron changed the title ~~WIP: CI builds for non-amd64 architectures~~ CI builds for non-amd64 architectures Jan 21, 2019

fweik approved these changes Jan 23, 2019

View reviewed changes

fweik merged commit 413491f into espressomd:python Jan 23, 2019

mkuron deleted the patch-8 branch April 24, 2019 09:53

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

CI builds for non-amd64 architectures #2401

CI builds for non-amd64 architectures #2401

mkuron commented Dec 5, 2018 •

edited

Loading

codecov bot commented Dec 5, 2018 •

edited

Loading

mkuron commented Dec 6, 2018

mkuron commented Dec 6, 2018 •

edited

Loading

fweik commented Dec 10, 2018

mkuron commented Dec 12, 2018

RudolfWeeber commented Dec 12, 2018

mkuron commented Dec 18, 2018

fweik commented Dec 18, 2018 via email

mkuron commented Dec 18, 2018 •

edited

Loading

mkuron commented Jan 9, 2019

junghans commented Jan 9, 2019

mkuron commented Jan 9, 2019 •

edited

Loading

fweik commented Jan 10, 2019

mkuron commented Jan 10, 2019

fweik commented Jan 10, 2019

mkuron commented Jan 11, 2019

RudolfWeeber commented Jan 14, 2019

mkuron commented Jan 14, 2019

fweik commented Jan 14, 2019

mkuron commented Jan 14, 2019

junghans commented Jan 18, 2019

mkuron commented Jan 21, 2019

CI builds for non-amd64 architectures #2401

CI builds for non-amd64 architectures #2401

Conversation

mkuron commented Dec 5, 2018 • edited Loading

codecov bot commented Dec 5, 2018 • edited Loading

Codecov Report

mkuron commented Dec 6, 2018

mkuron commented Dec 6, 2018 • edited Loading

fweik commented Dec 10, 2018

mkuron commented Dec 12, 2018

RudolfWeeber commented Dec 12, 2018

mkuron commented Dec 18, 2018

fweik commented Dec 18, 2018 via email

mkuron commented Dec 18, 2018 • edited Loading

mkuron commented Jan 9, 2019

junghans commented Jan 9, 2019

mkuron commented Jan 9, 2019 • edited Loading

fweik commented Jan 10, 2019

mkuron commented Jan 10, 2019

fweik commented Jan 10, 2019

mkuron commented Jan 11, 2019

RudolfWeeber commented Jan 14, 2019

mkuron commented Jan 14, 2019

fweik commented Jan 14, 2019

mkuron commented Jan 14, 2019

junghans commented Jan 18, 2019

mkuron commented Jan 21, 2019

mkuron commented Dec 5, 2018 •

edited

Loading

codecov bot commented Dec 5, 2018 •

edited

Loading

mkuron commented Dec 6, 2018 •

edited

Loading

mkuron commented Dec 18, 2018 •

edited

Loading

mkuron commented Jan 9, 2019 •

edited

Loading