[PATCH] D94056: [CodeGen] Update transformations to use poison for shufflevector/insertelem's initial vector elem

[Juneyoung Lee via Phabricator via llvm-commits]

aqjune added inline comments.


================
Comment at: llvm/lib/CodeGen/InterleavedAccessPass.cpp:362
       return false;
 
     if (cast<Instruction>(Shuffle->getOperand(0))->getOperand(0) == LI)
----------------
spatel wrote:
> Could we also assert here (and/or in the loop above this one) that:
>   VecTy->getNumElements() < Shuffle->getShuffleMask().size() 
> ?
> 
> Ie, the shuffle is always reducing the number of elements from the input vector operands?
Yes, I'll add one.


================
Comment at: llvm/lib/CodeGen/InterleavedAccessPass.cpp:401
+               cast<FixedVectorType>(BI->getOperand(0)->getType())
+                   ->getNumElements());
 
----------------
nikic wrote:
> Why is this checking only the back() element? Better check all? `assert(llvm::all_of(Mask, [&](int Idx) { return Idx < ... });`
It was simply because of its cost. Actually I was wondering whether it is acceptable to have such formula in the assertion. 


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D94056/new/

https://reviews.llvm.org/D94056