|Code optimization questions email@example.com (1992-11-04)|
|Re: Code optimization questions firstname.lastname@example.org (1992-11-11)|
|Re: Code optimization questions Jonathan.Bowen@prg.oxford.ac.uk (1992-11-13)|
|Re: Code Optimization questions email@example.com (1992-11-14)|
|Re: Code optimization questions firstname.lastname@example.org (1992-11-16)|
|Re: Code optimization questions email@example.com (1992-11-18)|
|From:||firstname.lastname@example.org (Preston Briggs)|
|Organization:||Rice University, Houston|
|Date:||Wed, 11 Nov 1992 16:53:57 GMT|
email@example.com (What Is) writes:
> Is there any info on optimizers that take advantage of runtime
> assertions? This can help with anti dependencies, but it may have
> other advantages; for instance, if an integer is guaranteed to be
> within a certain range, then some operations on it may be cheaper,
A big assertion that optimizers take advantage of are the type
declarations. After all, choosing an adequate number representation was a
tough problem in the old days for Lisp compilers. They're always glad to
be able to use a short integer rather than some bignum representation.
There's also the various schemes people use to tell their compiler to
vectorize or parallelize a particular loop, regardless of how unsafe it
appears to the compiler.
On the other hand, I don't know of specific papers in the area.
BTW, the plural of "dependence" is "dependences".
> If statement A has a def-use chain to statement B, does it follow
> that statement B has a use-def chain to statement A?
Return to the
Search the comp.compilers archives again.