[PATCH] D87502: [DAGCombiner] Use known bits to fold extract_vector_elt with const index

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:17887
+    APInt DemandedElts = APInt::getOneBitSet(NumElts, IndexC->getZExtValue());
+    KnownBits Known = DAG.computeKnownBits(VecOp, DemandedElts);
+    if (Known.isConstant()) {
----------------
RKSimon wrote:
> Would this be better handled by SimplifyDemandedBits?
Good idea. D88569, D88570.


================
Comment at: llvm/test/CodeGen/X86/vector-trunc-math.ll:2713
 ; SSE-NEXT:    packuswb %xmm1, %xmm0
+; SSE-NEXT:    pand {{.*}}(%rip), %xmm0
 ; SSE-NEXT:    retq
----------------
foad wrote:
> foad wrote:
> > Regression. Perhaps we need better known bits analysis.
> D87912 is my attempt at fixing this.
This regression got fixed by D88225.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D87502/new/

https://reviews.llvm.org/D87502