sequence of operations with CTH data causes client *and* server to lock
This issue was created automatically from an original Mantis Issue. Further discussion may take place here.
Note: This happens with a CVS trunk checkout of source from around Sept 17, and running client on Windows XP, server on linux cluster at Sandia.
Note 2: We've repeatedly seen this error also on Linux client (with linux cluster)
These were the steps that caused problems on windows client run:
-- Opened Golevka CTH data. -- Checked the two Mass variables on. -- Threshold on Mass 1, resulting in about 28K cells. -- Clip on half of data. -- Probe on location (used default location).
on Windows, the probe froze, did not return control to app within a few minutes, server spun.
on Linux, only got to probe once, rotating before probe caused app to freeze, and server to die.
On windows run we attached to the cluster/server root node, attached gdb, and saw the stack spinning on an open MPI call.