yade-dev team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

One of your bodies has state.velocity==NaN. That is not problem in the
collider.

-- 
InsertionSortCollider segfault after O.load(..)
https://bugs.launchpad.net/bugs/600574
You received this bug notification because you are a member of Yade
developers, which is subscribed to Yade.

Status in Yet Another Dynamic Engine: New

Bug description:
bzr2310 has some problems with InsertionSortCollider, after loading from a saved file, it crashes:

Running script test2.py
2370  FATAL yade.VelocityBins pkg/common/DataClass/VelocityBins.cpp:77 setBins: Body #2, velSq=nan not put in any bin!?
SIGSEGV/SIGABRT handler called; gdb batch file is `/tmp/yade-15sOSi/tmp-0'

#7  0x00007f08f192f630 in VelocityBins::setBins (this=0x7f08e1439870, rootBody=0x2f9a000, currMaxVelSq=<value optimized out>, refSweepLength=<value optimized out>) at pkg/common/DataClass/VelocityBins.cpp:77
#8  0x00007f08efcce372 in InsertionSortCollider::action (this=0x7f08dc001c20) at pkg/common/Engine/GlobalEngine/InsertionSortCollider.cpp:199
#9  0x00007f08fb3b5153 in Scene::moveToNextTimeStep (this=0x2f9a000) at /home/janek/20-Programowanie/10-cpp/50-Yade/Code/HEAD/trunk/core/Scene.cpp:87


A am attaching a short script that will reproduce this bug, it happens about 50% of the time. Please run it several times to reproduce this. Sometimes this bug manifests in a different way: after loading the simulation explodes.