[PATCH] D24511: [LV] Process pointer IVs with PHINodes in collectLoopUniforms

[Michael Kuperstein via llvm-commits]

mkuper added inline comments.

================
Comment at: lib/Transforms/Vectorize/LoopVectorize.cpp:5397
@@ +5396,3 @@
+      // True if all users of the pointer operand are memory accesses.
+      auto UsersAreMemAccesses = all_of(Ptr->users(), [&](User *U) -> bool {
+        return isa<LoadInst>(U) || isa<StoreInst>(U);
----------------
Do we care about bitcasts / GEP chains? (I'm ok with being conservative with this, just making sure it's by choice.)

================
Comment at: lib/Transforms/Vectorize/LoopVectorize.cpp:5398
@@ +5397,3 @@
+      auto UsersAreMemAccesses = all_of(Ptr->users(), [&](User *U) -> bool {
+        return isa<LoadInst>(U) || isa<StoreInst>(U);
+      });
----------------
Do we need to check that if U is a store, then Ptr is actually the pointer operand of the store, not the value?

I'm thinking about:
store i32* %p, i32** %q,
Where both %p and %q are consecutive.

================
Comment at: lib/Transforms/Vectorize/LoopVectorize.cpp:5466
@@ +5465,3 @@
+      return I == IndUpdate || !TheLoop->contains(I) || Worklist.count(I) ||
+             isVectorizedMemAccessUse(I, Ind);
+    });
----------------
The only thing that changed here, functionally, is the addition of isVectorizedMemAccessUse right?
The rest is cleanup?


https://reviews.llvm.org/D24511