[PATCH] [SCEV] make SCEV smarter about proving no-wrap.
sanjoy at playingwithpointers.com
Tue Mar 3 14:44:14 PST 2015
In http://reviews.llvm.org/D7980#133867, @atrick wrote:
> The logic is great. But we really need to determine whether computing SCEV for each recurrence 5 times is worthwhile. Please do some compile time experiments where SCEV is taking more time than usual to find out how much worse this makes it.
Sure, I'll run some experiments and measure the compile time.
> Why did you do this rather than checking SCEV's for existing phis?
Do you mean just checking the folding set for an _existing_ SCEVAddRec instead of `SE.getAddRecExpr(PreStart, Step, AR->getLoop(), SCEV::FlagAnyWrap)`? That sounds like a very good idea, I'll try that.
More information about the llvm-commits