[PATCH] D27518: Moving isComplex decision to TTI

[Elena Demikhovsky via Phabricator via llvm-commits]

delena added inline comments.


================
Comment at: include/llvm/Analysis/TargetTransformInfo.h:606
+  /// of address-computation cost 
+  struct AddressAccessInfo {
+    bool  isStrided;        /// True in case the access is strided (AddRec).
----------------
mkuper wrote:
> delena wrote:
> > Why do you need a struct? "Value *" will contain everything.
> > 
> > Value *Stride;
> > 
> > if (Stride == nullptr) - no stride,
> > dyn_cast<ConstantInt>(Stride) - answers the question isConstant()
> >  
> IIUC, we don't necessarily have a Value* in hand, this originates from a SCEV, right?
yes. We can use const SCEV*. 


https://reviews.llvm.org/D27518