Avoid recomputing the unique exit blocks and their insert points when doing

multiple scalar promotions on a single loop. This also has the effect of
preserving the order of stores sunk out of loops, which is aesthetically
pleasing, and it happens to fix the testcase in PR13542, though it doesn't
fix the underlying problem.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@161459 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Transforms/Scalar/LICM.cpp b/lib/Transforms/Scalar/LICM.cpp
index 582948ea14bbd8a1eb674ac06570dd4b991d53b7..0192e928fe8e3aaf1ae860c63b212c8e9bb6eb7a 100644 (file)
--- a/lib/Transforms/Scalar/LICM.cpp
+++ b/lib/Transforms/Scalar/LICM.cpp
@@ -175,7 +175,9 @@ namespace {
     bool canSinkOrHoistInst(Instruction &I);
     bool isNotUsedInLoop(Instruction &I);
 
-    void PromoteAliasSet(AliasSet &AS);
+    void PromoteAliasSet(AliasSet &AS,
+                         SmallVectorImpl<BasicBlock*> &ExitBlocks,
+                         SmallVectorImpl<Instruction*> &InsertPts);
   };
 }
 
@@ -256,10 +258,13 @@ bool LICM::runOnLoop(Loop *L, LPPassManager &LPM) {
   // Now that all loop invariants have been removed from the loop, promote any
   // memory references to scalars that we can.
   if (!DisablePromotion && Preheader && L->hasDedicatedExits()) {
+    SmallVector<BasicBlock *, 8> ExitBlocks;
+    SmallVector<Instruction *, 8> InsertPts;
+
     // Loop over all of the alias sets in the tracker object.
     for (AliasSetTracker::iterator I = CurAST->begin(), E = CurAST->end();
          I != E; ++I)
-      PromoteAliasSet(*I);
+      PromoteAliasSet(*I, ExitBlocks, InsertPts);
   }
 
   // Clear out loops state information for the next iteration
@@ -631,6 +636,7 @@ namespace {
     Value *SomePtr;  // Designated pointer to store to.
     SmallPtrSet<Value*, 4> &PointerMustAliases;
     SmallVectorImpl<BasicBlock*> &LoopExitBlocks;
+    SmallVectorImpl<Instruction*> &LoopInsertPts;
     AliasSetTracker &AST;
     DebugLoc DL;
     int Alignment;
@@ -638,11 +644,12 @@ namespace {
     LoopPromoter(Value *SP,
                  const SmallVectorImpl<Instruction*> &Insts, SSAUpdater &S,
                  SmallPtrSet<Value*, 4> &PMA,
-                 SmallVectorImpl<BasicBlock*> &LEB, AliasSetTracker &ast,
-                 DebugLoc dl, int alignment)
+                 SmallVectorImpl<BasicBlock*> &LEB,
+                 SmallVectorImpl<Instruction*> &LIP,
+                 AliasSetTracker &ast, DebugLoc dl, int alignment)
       : LoadAndStorePromoter(Insts, S), SomePtr(SP),
-        PointerMustAliases(PMA), LoopExitBlocks(LEB), AST(ast), DL(dl),
-        Alignment(alignment) {}
+        PointerMustAliases(PMA), LoopExitBlocks(LEB), LoopInsertPts(LIP),
+        AST(ast), DL(dl), Alignment(alignment) {}
 
     virtual bool isInstInList(Instruction *I,
                               const SmallVectorImpl<Instruction*> &) const {
@@ -662,7 +669,7 @@ namespace {
       for (unsigned i = 0, e = LoopExitBlocks.size(); i != e; ++i) {
         BasicBlock *ExitBlock = LoopExitBlocks[i];
         Value *LiveInValue = SSA.GetValueInMiddleOfBlock(ExitBlock);
-        Instruction *InsertPos = ExitBlock->getFirstInsertionPt();
+        Instruction *InsertPos = LoopInsertPts[i];
         StoreInst *NewSI = new StoreInst(LiveInValue, SomePtr, InsertPos);
         NewSI->setAlignment(Alignment);
         NewSI->setDebugLoc(DL);
@@ -684,7 +691,9 @@ namespace {
 /// looping over the stores in the loop, looking for stores to Must pointers
 /// which are loop invariant.
 ///
-void LICM::PromoteAliasSet(AliasSet &AS) {
+void LICM::PromoteAliasSet(AliasSet &AS,
+                           SmallVectorImpl<BasicBlock*> &ExitBlocks,
+                           SmallVectorImpl<Instruction*> &InsertPts) {
   // We can promote this alias set if it has a store, if it is a "Must" alias
   // set, if the pointer is loop invariant, and if we are not eliminating any
   // volatile loads or stores.
@@ -794,14 +803,20 @@ void LICM::PromoteAliasSet(AliasSet &AS) {
   // location is better than none.
   DebugLoc DL = LoopUses[0]->getDebugLoc();
 
-  SmallVector<BasicBlock*, 8> ExitBlocks;
-  CurLoop->getUniqueExitBlocks(ExitBlocks);
+  // Figure out the loop exits and their insertion points, if this is the
+  // first promotion.
+  if (ExitBlocks.empty()) {
+    CurLoop->getUniqueExitBlocks(ExitBlocks);
+    InsertPts.resize(ExitBlocks.size());
+    for (unsigned i = 0, e = ExitBlocks.size(); i != e; ++i)
+      InsertPts[i] = ExitBlocks[i]->getFirstInsertionPt();
+  }
 
   // We use the SSAUpdater interface to insert phi nodes as required.
   SmallVector<PHINode*, 16> NewPHIs;
   SSAUpdater SSA(&NewPHIs);
   LoopPromoter Promoter(SomePtr, LoopUses, SSA, PointerMustAliases, ExitBlocks,
-                        *CurAST, DL, Alignment);
+                        InsertPts, *CurAST, DL, Alignment);
 
   // Set up the preheader to have a definition of the value.  It is the live-out
   // value from the preheader that uses in the loop will use.

diff --git a/test/Transforms/LICM/promote-order.ll b/test/Transforms/LICM/promote-order.ll
new file mode 100644 (file)
index 0000000..b016265
--- /dev/null
+++ b/test/Transforms/LICM/promote-order.ll
@@ -0,0 +1,41 @@
+; RUN: opt -tbaa -basicaa -licm -S < %s | FileCheck %s
+
+; LICM should keep the stores in their original order when it sinks/promotes them.
+; rdar://12045203
+
+target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+@p = external global i8*
+
+define i32* @_Z4doiti(i32 %n, float* %tmp1, i32* %tmp3) nounwind {
+entry:
+  %cmp1 = icmp slt i32 0, %n
+  br i1 %cmp1, label %for.body.lr.ph, label %for.end
+
+for.body.lr.ph:                                   ; preds = %entry
+  br label %for.body
+
+for.body:                                         ; preds = %for.body, %for.body.lr.ph
+  %i.02 = phi i32 [ 0, %for.body.lr.ph ], [ %inc, %for.body ]
+  store float 1.000000e+00, float* %tmp1, align 4, !tbaa !1
+  store i32 1, i32* %tmp3, align 4, !tbaa !2
+  %inc = add nsw i32 %i.02, 1
+  %cmp = icmp slt i32 %inc, %n
+  br i1 %cmp, label %for.body, label %for.cond.for.end_crit_edge
+
+; CHECK: for.cond.for.end_crit_edge:
+; CHECK: store float 1.000000e+00, float* %tmp1
+; CHECK: store i32 1, i32* %tmp3
+for.cond.for.end_crit_edge:                       ; preds = %for.body
+  %split = phi i32* [ %tmp3, %for.body ]
+  br label %for.end
+
+for.end:                                          ; preds = %for.cond.for.end_crit_edge, %entry
+  %r.0.lcssa = phi i32* [ %split, %for.cond.for.end_crit_edge ], [ undef, %entry ]
+  ret i32* %r.0.lcssa
+}
+
+!0 = metadata !{metadata !"minimal TBAA"}
+!1 = metadata !{metadata !"float", metadata !0}
+!2 = metadata !{metadata !"int", metadata !0}