[LoopUnrollRuntime] Avoid high-cost trip count computation.

Summary:
Runtime unrolling of loops needs to emit an expression to compute the
loop's runtime trip-count.  Avoid runtime unrolling if this computation
will be expensive.

Depends on D8993.

Reviewers: atrick

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D8994

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@234846 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/Analysis/TargetTransformInfo.h b/include/llvm/Analysis/TargetTransformInfo.h
index 1a7f6f6cfc04fdb55a50f6929154e91a57346854..f4195fbb072c6592e707728dfa123a9429086aca 100644 (file)
--- a/include/llvm/Analysis/TargetTransformInfo.h
+++ b/include/llvm/Analysis/TargetTransformInfo.h
@@ -261,6 +261,9 @@ public:
     /// loop body even when the number of loop iterations is not known at
     /// compile time).
     bool Runtime;
+    /// Allow emitting expensive instructions (such as divisions) when computing
+    /// the trip count of a loop for runtime unrolling.
+    bool AllowExpensiveTripCount;
   };
 
   /// \brief Get target-customized preferences for the generic loop unrolling

diff --git a/include/llvm/Transforms/Utils/UnrollLoop.h b/include/llvm/Transforms/Utils/UnrollLoop.h
index 04d9ee14f9a6d3ee5c1b118bc5f7137e2ede017c..7f2cf8d7f59e598b0b8d4d95a3268b72a1a90a38 100644 (file)
--- a/include/llvm/Transforms/Utils/UnrollLoop.h
+++ b/include/llvm/Transforms/Utils/UnrollLoop.h
@@ -28,11 +28,13 @@ class MDNode;
 class Pass;
 
 bool UnrollLoop(Loop *L, unsigned Count, unsigned TripCount, bool AllowRuntime,
-                unsigned TripMultiple, LoopInfo *LI, Pass *PP,
-                LPPassManager *LPM, AssumptionCache *AC);
+                bool AllowExpensiveTripCount, unsigned TripMultiple,
+                LoopInfo *LI, Pass *PP, LPPassManager *LPM,
+                AssumptionCache *AC);
 
-bool UnrollRuntimeLoopProlog(Loop *L, unsigned Count, LoopInfo *LI,
-                             LPPassManager* LPM);
+bool UnrollRuntimeLoopProlog(Loop *L, unsigned Count,
+                             bool AllowExpensiveTripCount, LoopInfo *LI,
+                             LPPassManager *LPM);
 
 MDNode *GetUnrollMetadata(MDNode *LoopID, StringRef Name);
 }

diff --git a/lib/Transforms/Scalar/LoopUnrollPass.cpp b/lib/Transforms/Scalar/LoopUnrollPass.cpp
index 600cbde7b4219a06b3422e14e495c63584e65195..3de3b3d980b96461719b1c6fb15d9f676409dde7 100644 (file)
--- a/lib/Transforms/Scalar/LoopUnrollPass.cpp
+++ b/lib/Transforms/Scalar/LoopUnrollPass.cpp
@@ -165,6 +165,7 @@ namespace {
       UP.MaxCount = UINT_MAX;
       UP.Partial = CurrentAllowPartial;
       UP.Runtime = CurrentRuntime;
+      UP.AllowExpensiveTripCount = false;
       TTI.getUnrollingPreferences(L, UP);
     }
 
@@ -886,8 +887,8 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &LPM) {
   }
 
   // Unroll the loop.
-  if (!UnrollLoop(L, Count, TripCount, AllowRuntime, TripMultiple, LI, this,
-                  &LPM, &AC))
+  if (!UnrollLoop(L, Count, TripCount, AllowRuntime, UP.AllowExpensiveTripCount,
+                  TripMultiple, LI, this, &LPM, &AC))
     return false;
 
   return true;

diff --git a/lib/Transforms/Utils/LoopUnroll.cpp b/lib/Transforms/Utils/LoopUnroll.cpp
index 6b3aa0218f0f0ca807a31866bd6187325fec2623..1dbce474683513ea2df00399948f84483634ac7c 100644 (file)
--- a/lib/Transforms/Utils/LoopUnroll.cpp
+++ b/lib/Transforms/Utils/LoopUnroll.cpp
@@ -146,6 +146,13 @@ FoldBlockIntoPredecessor(BasicBlock *BB, LoopInfo* LI, LPPassManager *LPM,
 /// Similarly, TripMultiple divides the number of times that the LatchBlock may
 /// execute without exiting the loop.
 ///
+/// If AllowRuntime is true then UnrollLoop will consider unrolling loops that
+/// have a runtime (i.e. not compile time constant) trip count.  Unrolling these
+/// loops require a unroll "prologue" that runs "RuntimeTripCount % Count"
+/// iterations before branching into the unrolled loop.  UnrollLoop will not
+/// runtime-unroll the loop if computing RuntimeTripCount will be expensive and
+/// AllowExpensiveTripCount is false.
+///
 /// The LoopInfo Analysis that is passed will be kept consistent.
 ///
 /// If a LoopPassManager is passed in, and the loop is fully removed, it will be
@@ -154,8 +161,9 @@ FoldBlockIntoPredecessor(BasicBlock *BB, LoopInfo* LI, LPPassManager *LPM,
 /// This utility preserves LoopInfo. If DominatorTree or ScalarEvolution are
 /// available from the Pass it must also preserve those analyses.
 bool llvm::UnrollLoop(Loop *L, unsigned Count, unsigned TripCount,
-                      bool AllowRuntime, unsigned TripMultiple, LoopInfo *LI,
-                      Pass *PP, LPPassManager *LPM, AssumptionCache *AC) {
+                      bool AllowRuntime, bool AllowExpensiveTripCount,
+                      unsigned TripMultiple, LoopInfo *LI, Pass *PP,
+                      LPPassManager *LPM, AssumptionCache *AC) {
   BasicBlock *Preheader = L->getLoopPreheader();
   if (!Preheader) {
     DEBUG(dbgs() << "  Can't unroll; loop preheader-insertion failed.\n");
@@ -218,7 +226,8 @@ bool llvm::UnrollLoop(Loop *L, unsigned Count, unsigned TripCount,
   // flag is specified.
   bool RuntimeTripCount = (TripCount == 0 && Count > 0 && AllowRuntime);
 
-  if (RuntimeTripCount && !UnrollRuntimeLoopProlog(L, Count, LI, LPM))
+  if (RuntimeTripCount &&
+      !UnrollRuntimeLoopProlog(L, Count, AllowExpensiveTripCount, LI, LPM))
     return false;
 
   // Notify ScalarEvolution that the loop will be substantially changed,

diff --git a/lib/Transforms/Utils/LoopUnrollRuntime.cpp b/lib/Transforms/Utils/LoopUnrollRuntime.cpp
index c8d47828ebafee9be6e56a0a99a37bfce4c54037..d1774dfa10d93031c90997522b6782d163d08d2c 100644 (file)
--- a/lib/Transforms/Utils/LoopUnrollRuntime.cpp
+++ b/lib/Transforms/Utils/LoopUnrollRuntime.cpp
@@ -278,7 +278,8 @@ static void CloneLoopBlocks(Loop *L, Value *NewIter, const bool UnrollProlog,
 /// ...
 /// End:
 ///
-bool llvm::UnrollRuntimeLoopProlog(Loop *L, unsigned Count, LoopInfo *LI,
+bool llvm::UnrollRuntimeLoopProlog(Loop *L, unsigned Count,
+                                   bool AllowExpensiveTripCount, LoopInfo *LI,
                                    LPPassManager *LPM) {
   // for now, only unroll loops that contain a single exit
   if (!L->getExitingBlock())
@@ -312,6 +313,12 @@ bool llvm::UnrollRuntimeLoopProlog(Loop *L, unsigned Count, LoopInfo *LI,
   if (isa<SCEVCouldNotCompute>(TripCountSC))
     return false;
 
+  BasicBlock *Header = L->getHeader();
+  const DataLayout &DL = Header->getModule()->getDataLayout();
+  SCEVExpander Expander(*SE, DL, "loop-unroll");
+  if (!AllowExpensiveTripCount && Expander.isHighCostExpansion(TripCountSC, L))
+    return false;
+
   // We only handle cases when the unroll factor is a power of 2.
   // Count is the loop unroll factor, the number of extra copies added + 1.
   if (!isPowerOf2_32(Count))
@@ -332,18 +339,15 @@ bool llvm::UnrollRuntimeLoopProlog(Loop *L, unsigned Count, LoopInfo *LI,
   auto *DT = DTWP ? &DTWP->getDomTree() : nullptr;
 
   BasicBlock *PH = L->getLoopPreheader();
-  BasicBlock *Header = L->getHeader();
   BasicBlock *Latch = L->getLoopLatch();
   // It helps to splits the original preheader twice, one for the end of the
   // prolog code and one for a new loop preheader
   BasicBlock *PEnd = SplitEdge(PH, Header, DT, LI);
   BasicBlock *NewPH = SplitBlock(PEnd, PEnd->getTerminator(), DT, LI);
   BranchInst *PreHeaderBR = cast<BranchInst>(PH->getTerminator());
-  const DataLayout &DL = Header->getModule()->getDataLayout();
 
   // Compute the number of extra iterations required, which is:
   //  extra iterations = run-time trip count % (loop unroll factor + 1)
-  SCEVExpander Expander(*SE, DL, "loop-unroll");
   Value *TripCount = Expander.expandCodeFor(TripCountSC, TripCountSC->getType(),
                                             PreHeaderBR);
   Value *BECount = Expander.expandCodeFor(BECountSC, BECountSC->getType(),

diff --git a/test/Transforms/LoopUnroll/high-cost-trip-count-computation.ll b/test/Transforms/LoopUnroll/high-cost-trip-count-computation.ll
new file mode 100644 (file)
index 0000000..6b1943f
--- /dev/null
+++ b/test/Transforms/LoopUnroll/high-cost-trip-count-computation.ll
@@ -0,0 +1,27 @@
+; RUN: opt -S -unroll-runtime -loop-unroll < %s | FileCheck %s
+
+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"
+
+;; Check that we don't emit expensive instructions to compute trip
+;; counts when unrolling loops.
+
+define i32 @test(i64 %v12, i8* %array, i64* %loc) {
+; CHECK-LABEL: @test(
+; CHECK-NOT: udiv
+entry:
+  %step = load i64, i64* %loc, !range !0
+  br label %loop
+
+loop:                                           ; preds = %entry, %loop
+  %k.015 = phi i64 [ %v15, %loop ], [ %v12, %entry ]
+  %v14 = getelementptr inbounds i8, i8* %array, i64 %k.015
+  store i8 0, i8* %v14
+  %v15 = add nuw nsw i64 %k.015, %step
+  %v16 = icmp slt i64 %v15, 8193
+  br i1 %v16, label %loop, label %loopexit
+
+loopexit:                             ; preds = %loop
+  ret i32 0
+}
+
+!0 = !{i64 1, i64 100}

diff --git a/test/Transforms/LoopUnroll/runtime-loop.ll b/test/Transforms/LoopUnroll/runtime-loop.ll
index 9e78edfb25e3a5fcaed00feaf53041fc59546f8d..fea15b65d5a55cfffd6a861c3529379ccd694e80 100644 (file)
--- a/test/Transforms/LoopUnroll/runtime-loop.ll
+++ b/test/Transforms/LoopUnroll/runtime-loop.ll
@@ -1,5 +1,7 @@
 ; RUN: opt < %s -S -loop-unroll -unroll-runtime=true | FileCheck %s
 
+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"
+
 ; Tests for unrolling loops with run-time trip counts
 
 ; CHECK: %xtraiter = and i32 %n

diff --git a/test/Transforms/LoopUnroll/runtime-loop4.ll b/test/Transforms/LoopUnroll/runtime-loop4.ll
index 9be0ffd23d328a3e30622d9b56b226a16addf437..5014c739e4c962b67c9d270841be76c3c537a4ad 100644 (file)
--- a/test/Transforms/LoopUnroll/runtime-loop4.ll
+++ b/test/Transforms/LoopUnroll/runtime-loop4.ll
@@ -20,7 +20,8 @@ loop1:
   br label %loop2.header
 
 loop2.header:
-  br label %loop2
+  %e = icmp uge i32 %iter, 1
+  br i1 %e, label %loop2, label %exit2
 
 loop2:
   %iv2 = phi i32 [ 0, %loop2.header ], [ %inc2, %loop2 ]