Bug in week-of-code: CreateSUBlingList.C:[135|137]

[jbordner]

Crashes or hangs at new[] or delete[], which indicates corruption of the dynamic memory heap.

Affects the following problems:

• SedovBlastAMR
• ImplosionAMR

#0  __lll_lock_wait ()
    at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:136
#1  0x00007f89673e0295 in _L_lock_949 () from /lib/libpthread.so.0
#2  0x00007f89673e00b8 in __pthread_mutex_lock (mutex=0x7f89684c0d00)
    at pthread_mutex_lock.c:61
#3  0x00007f896823836e in free () from /usr/lib/liblam.so.4
#4  0x00007f8968ea028e in H5MM_xfree () from /usr/lib/libhdf5-1.6.6.so.0
#5  0x00007f8968e9fb90 in H5I_destroy_group () from /usr/lib/libhdf5-1.6.6.so.0
#6  0x00007f8968e53acf in H5D_term_interface () from /usr/lib/libhdf5-1.6.6.so.0
#7  0x00007f8968e426a5 in H5_term_library () from /usr/lib/libhdf5-1.6.6.so.0
#8  0x00007f89670a0c12 in __run_exit_handlers (status=1) at exit.c:78
#9  *__GI_exit (status=1) at exit.c:100
#10 0x00007f89681e3ee9 in _cio_bail () from /usr/lib/liblam.so.4
#11 0x00007f8968233407 in ?? () from /usr/lib/liblam.so.4
#12 <signal handler called>
#13 0x00007f8968239bbb in _int_malloc () from /usr/lib/liblam.so.4
#14 0x00007f896823afcf in malloc () from /usr/lib/liblam.so.4
#15 0x00007f8967b5564d in operator new(unsigned long) ()
   from /usr/lib/libstdc++.so.6
#16 0x00007f8967b55769 in operator new[](unsigned long) ()
   from /usr/lib/libstdc++.so.6
#17 0x0000000000466adb in CreateSUBlingList (MetaData=0x7fff7186db40,
    LevelArray=0x7fff7186de50, level=1, SiblingList=0x514d880,
    SUBlingList=0x7fff7186cf28) at CreateSUBlingList.C:135
#18 0x000000000047e9cf in EvolveLevel (MetaData=0x7fff7186db40,
    LevelArray=0x7fff7186de50, level=1, dtLevelAbove=7.8791696844059342e-05,
    Exterior=0x7fff7186d2f0) at EvolveLevel.C:581
#19 0x000000000047e866 in EvolveLevel (MetaData=0x7fff7186db40,
    LevelArray=0x7fff7186de50, level=0, dtLevelAbove=7.8791696844059342e-05,
    Exterior=0x7fff7186d2f0) at EvolveLevel.C:545
#20 0x0000000000475201 in EvolveHierarchy (TopGrid=..., MetaData=...,
    Exterior=0x7fff7186d2f0, LevelArray=0x7fff7186de50, Initialdt=0)
    at EvolveHierarchy.C:455
#21 0x000000000040c196 in main (argc=3, argv=0x7fff7186e488) at enzo.C:631

[jwise]

I fixed this in r1323:504d771d0334, but this led to another bug when the boundary fluxes of the parents were being updated from SUBlings and child grids. This is commented out in v1.5 and the fluxfix-no option. In principle, we should correct for these refinements, but why is it commented out in other versions and not here?