This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] do not lower a/b to a*(1/b)


> Dan, could PRE strength reduction catch this "optimization"
> that we'd miss with Paolo's patch?

In general, PRE strength reduction can look for any computation that is
cheaper to replace with a few computations somewhere else, and do that
transform.

All PRE strength reduction does is say, for example "Oh, I can keep that
multiply up to date till here with a simple addition in a few places,
let's do that".
Think of it as advanced rematerialization during redundancy elimination.



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]