sPuReMD: fixes to SAI preconditioner to correctly apply threshold to negative entries and to handle EE/ACKS2 cases when a threshold value of 1.0 is chosen.

sPuReMD: fix bug with diagonal scaling during FG_ICHOL/FG_ILUT producing incorrect factorizations when diagonal entries were negative.

sPuReMD: complete fine-grained Cholesky implementation. TODO: fix numeric breakdowns with water 6540 atom system.

sPuReMD: workarounds for memory issues with ILUT factors. Implement ILU(0) and call when selecting ILUT with a droptol of 0.

sPuReMD: fix issue with hbonds list indices not being reset correctly. Small algebraic rearrangements of valence angle calculations to match Fortran ReaxFF code (no numeric changes). Fix issue with rvec_isZero being incorrect. Fix issue with debugging arrays for TEST_FORCES not having indicies initialized. Other clean-up work.

sPuReMD: revert force tabulation changes. Temporarily work around out-of-bound memory accesses in cubic spline interpolation routes (for force tabulation) by expanding array sizes. Revisit at a later date.

sPuReMD: fix several more warning from Valgrind (out-of-bounds accesses in force tabulation lookup tables, sparse matrices with EE/ACKS2 with tabulation). Remove unused variables. Fixes to string handling. Other code clean-up.

sPuReMD: fix several issues with strings not being null-terminated. Fix use of uninitialized values reported by Valgrind (heap memory). Other general clean-ups.

sPuReMD: fix issue with list for debugging forces not having indices initialized (dDelta, dBO). Fix interaction counting in OpenMP code (greater than 1 thread). Other general clean-up.

sPuReMD: fix small differences in valence angle energy term with Fortran ReaxFF code. Other minor code updates.

sPuReMD: change log file column headers to be contigious strings of non-whitespace characters to facilitate post-processing.