[MLton] bug report, vector of char problem

Matthew Fluet fluet@cs.cornell.edu
Wed, 6 Apr 2005 09:00:20 -0400 (EDT)


> Stephen Weeks wrote:
> > Thanks for the bug report Scott.  I have verified the bug and
> > determined the pass in MLton that is causing the problem.  I'm
> > thinking about a solution.  I hope to have something to report in a
> > day or two.
> 
> Just out of curiosity anyone have an informal list of subtle bugs that 
> didn't tickle a bug in the typed-IL? We were having a lunch discussion about 
> if typed-IL really prevented compiler bugs... I'm wondering what the MLton 
> experience has been.

I don't think we have an actual list, but you can always brows the
ChangeLog, which should track bugs fixed. Scanning back a year to
2003-03-21, I only saw 3 bugs that were in optimization passes and another
3 bugs that were in the codegens.  There are significantly more bug fixes
to the Basis Library implementation and to the front-end.

I think our experience has been very positive.  We throw in to the 
'type-checker' for our typed ILs a bit more than just types; for example, 
the SSA IL checks the def dominates use property as well.  The 
representation ILs also check that limit check insertions are good, etc.  
So, it's also a bit hard to pull out where 'types' helped, and where the 
more general notion of statically checked properties helped.