!fixup address latest comments, thanks!

Author: fhahn

llvm/lib/Transforms/Vectorize/LoopVectorize.cpp

@@ -2449,12 +2449,9 @@ void InnerLoopVectorizer::emitIterationCountCheck(BasicBlock *Bypass) {
     } else if (!SE.isKnownPredicate(CmpInst::getInversePredicate(P),
                                     TripCountSCEV, SE.getSCEV(Step))) {
       // Generate the minimum iteration check only if we cannot prove the
-      // check is known to be true, or known to be false
+      // check is known to be true, or known to be false.
       CheckMinIters = Builder.CreateICmp(P, Count, Step, "min.iters.check");
-
-      // else step is known to be smaller than trip count, use CheckMinIters
-      // preset to false.
-    }
+    } // else step known to be < trip count, use CheckMinIters preset to false.
   } else if (VF.isScalable() &&
              !isIndvarOverflowCheckKnownFalse(Cost, VF, UF) &&
              Style != TailFoldingStyle::DataAndControlFlowWithoutRuntimeCheck) {
@@ -2470,11 +2467,12 @@ void InnerLoopVectorizer::emitIterationCountCheck(BasicBlock *Bypass) {
     Value *Step = CreateStep();
 #ifndef NDEBUG
     ScalarEvolution &SE = *PSE.getSE();
-    const SCEV *TripCountSCEV = SE.applyLoopGuards(SE.getSCEV(LHS), OrigLoop);
+    const SCEV *TC2OverflowSCEV = SE.applyLoopGuards(SE.getSCEV(LHS), OrigLoop);
     assert(
+        !isIndvarOverflowCheckKnownFalse(Cost, VF * UF) &&
         !SE.isKnownPredicate(CmpInst::getInversePredicate(ICmpInst::ICMP_ULT),
-                             TripCountSCEV, SE.getSCEV(Step)) &&
-        "SCEV unexpectedly proved overflow check to be known");
+                             TC2OverflowSCEV, SE.getSCEV(Step)) &&
+        "unexpectedly proved overflow check to be known");
 #endif
     // Don't execute the vector loop if (UMax - n) < (VF * UF).
     CheckMinIters = Builder.CreateICmp(ICmpInst::ICMP_ULT, LHS, Step);

llvm/test/Transforms/LoopVectorize/if-reduction.ll

@@ -1659,6 +1659,7 @@ for.end:                                          ; preds = %for.body, %entry
   ret i64 %1
 }
 
+; FIXME: %indvars.iv.next is poison on first iteration due to sub nuw 0, 1.
 define i32 @fcmp_0_sub_select1(ptr noalias %x, i32 %N) nounwind readonly {
 ; CHECK-LABEL: define i32 @fcmp_0_sub_select1(
 ; CHECK-SAME: ptr noalias [[X:%.*]], i32 [[N:%.*]]) #[[ATTR0]] {

llvm/test/Transforms/LoopVectorize/version-stride-with-integer-casts.ll

@@ -415,6 +415,7 @@ exit:
 
 ; Test case to make sure that uses of versioned strides of type i1 are properly
 ; extended. From https://github.com/llvm/llvm-project/issues/91369.
+; TODO: Better check (udiv i64 15, %g.64) after checking if %g == 1.
 define void @zext_of_i1_stride(i1 %g, ptr %dst) mustprogress {
 ; CHECK-LABEL: define void @zext_of_i1_stride(
 ; CHECK-SAME: i1 [[G:%.*]], ptr [[DST:%.*]]) #[[ATTR0:[0-9]+]] {