[llvm] r320407 - [InstCombine] Fix PR35618: Instcombine hangs on single minmax load bitcast.

[Alexey Bataev via llvm-commits]

Author: abataev
Date: Mon Dec 11 11:11:16 2017
New Revision: 320407

URL: http://llvm.org/viewvc/llvm-project?rev=320407&view=rev
Log:
[InstCombine] Fix PR35618: Instcombine hangs on single minmax load bitcast.

Summary:
If we have pattern `store (load(bitcast(select (cmp(V1, V2), &V1,
&V2)))), bitcast)`, but the load is used in other instructions, it leads
to looping in InstCombiner. Patch adds additional check that all users
of the load instructions are stores and then replaces all uses of load
instruction by the new one with new type.

Reviewers: RKSimon, spatel, majnemer

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D41072

Added:
    llvm/trunk/test/Transforms/InstCombine/multiple-uses-load-bitcast-select.ll
Modified:
    llvm/trunk/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp

Modified: llvm/trunk/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp?rev=320407&r1=320406&r2=320407&view=diff
==============================================================================
--- llvm/trunk/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp (original)
+++ llvm/trunk/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp Mon Dec 11 11:11:16 2017
@@ -1344,9 +1344,24 @@ bool removeBitcastsFromLoadStoreOnMinMax
   if (!isMinMaxWithLoads(LoadAddr))
     return false;
 
+  if (!all_of(LI->users(), [LI](User *U) {
+        auto *SI = dyn_cast<StoreInst>(U);
+        return SI && SI->getPointerOperand() != LI &&
+               !SI->getPointerOperand()->isSwiftError();
+      }))
+    return false;
+
+  IC.Builder.SetInsertPoint(LI);
   LoadInst *NewLI = combineLoadToNewType(
       IC, *LI, LoadAddr->getType()->getPointerElementType());
-  combineStoreToNewValue(IC, SI, NewLI);
+  // Replace all the stores with stores of the newly loaded value.
+  for (auto *UI : LI->users()) {
+    auto *SI = cast<StoreInst>(UI);
+    IC.Builder.SetInsertPoint(SI);
+    combineStoreToNewValue(IC, *SI, NewLI);
+    IC.eraseInstFromFunction(*SI);
+  }
+  IC.eraseInstFromFunction(*LI);
   return true;
 }
 
@@ -1375,7 +1390,7 @@ Instruction *InstCombiner::visitStoreIns
     return eraseInstFromFunction(SI);
 
   if (removeBitcastsFromLoadStoreOnMinMax(*this, SI))
-    return eraseInstFromFunction(SI);
+    return nullptr;
 
   // Replace GEP indices if possible.
   if (Instruction *NewGEPI = replaceGEPIdxWithZero(*this, Ptr, SI)) {

Added: llvm/trunk/test/Transforms/InstCombine/multiple-uses-load-bitcast-select.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/InstCombine/multiple-uses-load-bitcast-select.ll?rev=320407&view=auto
==============================================================================
--- llvm/trunk/test/Transforms/InstCombine/multiple-uses-load-bitcast-select.ll (added)
+++ llvm/trunk/test/Transforms/InstCombine/multiple-uses-load-bitcast-select.ll Mon Dec 11 11:11:16 2017
@@ -0,0 +1,30 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -instcombine -S -data-layout="E-m:e-i1:8:16-i8:8:16-i64:64-f128:64-v128:64-a:8:16-n32:64" | FileCheck %s
+
+define void @PR35618(i64* %st1, double* %st2) {
+; CHECK-LABEL: @PR35618(
+; CHECK-NEXT:    [[Y1:%.*]] = alloca double, align 8
+; CHECK-NEXT:    [[Z1:%.*]] = alloca double, align 8
+; CHECK-NEXT:    [[LD1:%.*]] = load double, double* [[Y1]], align 8
+; CHECK-NEXT:    [[LD2:%.*]] = load double, double* [[Z1]], align 8
+; CHECK-NEXT:    [[TMP10:%.*]] = fcmp olt double [[LD1]], [[LD2]]
+; CHECK-NEXT:    [[TMP121:%.*]] = select i1 [[TMP10]], double [[LD1]], double [[LD2]]
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast i64* [[ST1:%.*]] to double*
+; CHECK-NEXT:    store double [[TMP121]], double* [[TMP1]], align 8
+; CHECK-NEXT:    store double [[TMP121]], double* [[ST2:%.*]], align 8
+; CHECK-NEXT:    ret void
+;
+  %y1 = alloca double
+  %z1 = alloca double
+  %ld1 = load double, double* %y1
+  %ld2 = load double, double* %z1
+  %tmp10 = fcmp olt double %ld1, %ld2
+  %sel = select i1 %tmp10, double* %y1, double* %z1
+  %tmp11 = bitcast double* %sel to i64*
+  %tmp12 = load i64, i64* %tmp11
+  store i64 %tmp12, i64* %st1
+  %bc = bitcast double* %st2 to i64*
+  store i64 %tmp12, i64* %bc
+  ret void
+}
+