[PATCH] D132590: [SLP] Try to match reductions first in a vector build sequence.

[Alexey Bataev via Phabricator via llvm-commits]

ABataev added inline comments.


================
Comment at: llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp:11691-11693
+bool SLPVectorizerPass::vectorizeRootInstruction(
+    PHINode *P, Value *V, BasicBlock *BB, BoUpSLP &R, TargetTransformInfo *TTI,
+    SmallVectorImpl<WeakTrackingVH> &PostponedInsts) {
----------------
Can you put this and corresponding changes to the separate NFC patch?


================
Comment at: llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp:11963-11964
+
+    }
+    else if (isa<InsertElementInst>(I)) {
+      SmallVector<int> Mask;
----------------
Formatting


================
Comment at: llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp:11976-11980
+    // Try to vectorize postponed binops where reductions were not found.
+    for (Value *V : PostponedInsts)
+      if (auto *Inst = dyn_cast<Instruction>(V))
+        if (!R.isDeleted(Inst))
+          OpsChanged |= tryToVectorize(Inst, R);
----------------
Outline this code to the separate function? Same code is used in 2 places.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D132590/new/

https://reviews.llvm.org/D132590