[SLP] fix verification check for valid IR

[spatel]

Summary

This is a fix for PR45965 - https://bugs.llvm.org/show_bug.cgi?id=45965 - which was left out of D80106 because of a test failure.
SLP does its own mini-CSE after potentially creating redundant instructions, so we need to wait for that to complete before running the verifier. Otherwise, we will see a test failure for test/Transforms/SLPVectorizer/X86/crash_vectorizeTree.ll (not changed here) because a phi temporarily has identical but different incoming values for the same incoming block.

The test that is altered here could be fixed independently because the incoming input metadata is invalid. AFAICT, that fix does not change the motivation for that test. The test was escaping verification in SLP without this change because we were not running verifyFunction() unless SLP actually changed the IR.

Note: "-disable-verify" on the RUN line doesn't seem to suppress the verifyFunction; not sure if that's another bug.

[ABataev]

LG

[ebrevnov]

This change caused verification to catch "invalid" IR produced by instcombine. Filed a bug for that https://bugs.llvm.org/show_bug.cgi?id=46362

llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp

for (const auto &Pair : DeletedInstructions) {
     }
     Pair.getFirst()->dropAllReferences();
   }
   for (const auto &Pair : DeletedInstructions) {
     assert(Pair.getFirst()->use_empty() &&
            "trying to erase instruction with users.");
     Pair.getFirst()->eraseFromParent();
   }
+  assert(!verifyFunction(*F, &dbgs()));
 }
 
 void BoUpSLP::eraseInstructions(ArrayRef<Value *> AV) {
   for (auto *V : AV) {
     if (auto *I = dyn_cast<Instruction>(V))
       eraseInstruction(I, /*ReplaceWithUndef=*/true);
   };
 }
if (!GEPs.empty()) {
                         << " underlying objects.\n");
       Changed |= vectorizeGEPIndices(BB, R);
     }
   }
 
   if (Changed) {
     R.optimizeGatherSequence();
     LLVM_DEBUG(dbgs() << "SLP: vectorized \"" << F.getName() << "\"\n");
-    LLVM_DEBUG(verifyFunction(F));
   }
   return Changed;
 }
 
 bool SLPVectorizerPass::vectorizeStoreChain(ArrayRef<Value *> Chain, BoUpSLP &R,
                                             unsigned Idx) {
   LLVM_DEBUG(dbgs() << "SLP: Analyzing a store chain of length " << Chain.size()
                     << "\n");

llvm/test/Transforms/SLPVectorizer/X86/crash_scheduling.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
-; RUN: opt < %s -basicaa -disable-verify -slp-vectorizer -S -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7 | FileCheck %s
+; RUN: opt < %s -basicaa -slp-vectorizer -S -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7 | FileCheck %s
 
 target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-apple-darwin13.3.0"
 
 define void @_foo(double %p1, double %p2, double %p3) #0 {
 ; CHECK-LABEL: @_foo(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[TAB1:%.*]] = alloca [256 x i32], align 16
for.body:
   br i1 %exitcond, label %return, label %for.body
 
 return:
   ret void
 }
 
 declare i32 @_xfn(<2 x double>) #4
 
-!3 = !{!"int", !4, i64 0}
+!3 = !{!"int", !5, i64 0}
 !4 = !{!3, !3, i64 0}
+!5 = !{!"omnipotent char", !6, i64 0}
+!6 = !{!"Simple C/C++ TBAA"}