[MLton] cvs commit: critical sections during thread switch

[Matthew Fluet]

> Here's a revised gc.c.  It seems to pass the regressions (with the slow
> GC_switchToThread), but I don't want to check it in without someone else
> taking a look at it.

I looked at the final set of regressions this morning; everything passed
with -debug true  except mutex.sml, once.sml, and prodcons.sml.  (Of
course, 95% of the regressions are single-threaded, so we never encounter
a stack in the heap (that isn't the current thread).)