STORM-105 : improve decompression perf gathering, allow perf name to be passed on the command line, fix crash in analysis phase

Change Linux fasttimer implementation back to RDTSC - using a reliable syscall was REALLY chewing CPU time.  Sigh.  I didn't realize how incredibly often this gets called.  So, back to the assembly.
But be more careful with CPU clock count on linux, so the fasttimer values are much more accurate than they were the last time we were with RDTSC, in absolute terms - back in the right order of magnitude anyway.

Also change many instances of Mhz to MHz.
Also some minor comment fixes.

Checker: UNINIT_CTOR
Function: NamedTimerFactory::NamedTimerFactory()
File: /indra/llcommon/llfasttimer_class.cpp

more reliable fix based on feedback from Richard.  dicked with the Darwin results too since those seemed wrong based on the same feedback (also covered in test plan).