Cbench - Scalable Cluster Benchmarking Code

for longstress_stress.in job tell cbench to use
8.5 hour walltime

updated mpigraph_mpigraph.in job with settings
found useful by Marcuse/John on sky with the
addition of 10x more iterations (so it will run
longer as a sustained test in the default Cbench
setup)

remove the (mostly) pointless usleep, and make the output formatting a little nicer

redoing makefile to be based on downloading the latest
IOR src tarball from sourceforge. get rid of outdated
lustre compiles. use latest IOR version which is IOR-2.10.2

updating ior make setup to be tarball based

Apply the fix of revision 673 to nodehwtest_output_parse.pl, too :).

Ignore bogus 0.000 min values being reported for tests in both statsmode and dplot.

fix a longstanding annoyance with cbench_output_parse.pl
when its output is piped and then teed to a file. there
would be hidden ansi escape codes that would break
normal grepping in weird ways unless you octaldumped
the teed file and remembered to filter the escape sequences.

the tweak is to not enable AUTORESET in ANSIColor until
detect_color_support() in cbench.pl decides to enable it.
associated with that is NOT setting AUTORESET after the
"use ANSIColor" in cbench_output_parse.pl

allow gen_jobs modules access to the %specialwalltimes hash

updated to better deal with edge cases that were
encountered

tune down amg memory requirements to fit 1.5GB/core better