[PATCH] D49491: [RFC][VPlan, SLP] Add simple SLP analysis on top of VPlan.

[Florian Hahn via Phabricator via llvm-commits]

fhahn added inline comments.


================
Comment at: lib/Transforms/Vectorize/VPlanSLP.cpp:118
+
+  // For loads, check that there are no stores between them.
+  // FIXME: we only have to forbid stores that could interfere with any of the
----------------
ABataev wrote:
> Seems to me the comment does not match the functionality
I've updated the comment and removed the code to check if all instructions are in the same BB. That's check earlier on. Does it make sense now?


================
Comment at: lib/Transforms/Vectorize/VPlanSLP.cpp:300
+  for (auto &Operands : MultiNodeOps) {
+    FinalOrder.push_back({Operands.first, {Operands.second[0]}});
+    if (cast<VPInstruction>(Operands.second[0])->getOpcode() ==
----------------
ABataev wrote:
> Try `FinalOrder.emplace_back(Operands.first, Operands.second[0]);`
I tried, but neither that nor `emplace_back(Operands.first, {Operands.second[0])};` gets accepted unfortunately.


https://reviews.llvm.org/D49491