[polly] r249010 - [FIX] Parallel codegen for invariant loads

[Johannes Doerfert via llvm-commits]

Author: jdoerfert
Date: Thu Oct  1 08:40:36 2015
New Revision: 249010

URL: http://llvm.org/viewvc/llvm-project?rev=249010&view=rev
Log:
[FIX] Parallel codegen for invariant loads

  Hand down all preloaded values to the parallel subfunction.

Added:
    polly/trunk/test/Isl/CodeGen/OpenMP/invariant_base_pointer_preloaded.ll
Modified:
    polly/trunk/lib/CodeGen/IslNodeBuilder.cpp

Modified: polly/trunk/lib/CodeGen/IslNodeBuilder.cpp
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/lib/CodeGen/IslNodeBuilder.cpp?rev=249010&r1=249009&r2=249010&view=diff
==============================================================================
--- polly/trunk/lib/CodeGen/IslNodeBuilder.cpp (original)
+++ polly/trunk/lib/CodeGen/IslNodeBuilder.cpp Thu Oct  1 08:40:36 2015
@@ -590,6 +590,11 @@ void IslNodeBuilder::createForParallel(_
     SubtreeValues.insert(V);
   }
 
+  // Values preloaded prior to the SCoP need to be available in the subfunction.
+  const auto &InvariantAccesses = S.getInvariantAccesses();
+  for (const InvariantAccessTy &IA : InvariantAccesses)
+    SubtreeValues.insert(ValueMap[IA.first->getAccessInstruction()]);
+
   ParallelLoopGenerator::ValueToValueMapTy NewValues;
   ParallelLoopGenerator ParallelLoopGen(Builder, P, LI, DT, DL);
 

Added: polly/trunk/test/Isl/CodeGen/OpenMP/invariant_base_pointer_preloaded.ll
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/test/Isl/CodeGen/OpenMP/invariant_base_pointer_preloaded.ll?rev=249010&view=auto
==============================================================================
--- polly/trunk/test/Isl/CodeGen/OpenMP/invariant_base_pointer_preloaded.ll (added)
+++ polly/trunk/test/Isl/CodeGen/OpenMP/invariant_base_pointer_preloaded.ll Thu Oct  1 08:40:36 2015
@@ -0,0 +1,33 @@
+; RUN: opt %loadPolly -polly-codegen -polly-parallel \
+; RUN: -polly-parallel-force -polly-detect-unprofitable -S < %s | FileCheck %s
+;
+; Test to verify that we hand down the preloaded A[0] to the OpenMP subfunction.
+;
+;    void f(float *A) {
+;      for (int i = 1; i < 1000; i++)
+;        A[i] += A[0];
+;    }
+;
+; CHECK:  %polly.subfn.storeaddr.polly.access.A.load = getelementptr inbounds { float*, float }, { float*, float }* %polly.par.userContext, i32 0, i32 1
+; CHECK:  store float %polly.access.A.load, float* %polly.subfn.storeaddr.polly.access.A.load
+;
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+
+define void @f(float* nocapture %A) {
+entry:
+  br label %for.body
+
+for.cond.cleanup:                                 ; preds = %for.body
+  ret void
+
+for.body:                                         ; preds = %for.body, %entry
+  %indvars.iv = phi i64 [ 1, %entry ], [ %indvars.iv.next, %for.body ]
+  %tmp = load float, float* %A, align 4
+  %arrayidx1 = getelementptr inbounds float, float* %A, i64 %indvars.iv
+  %tmp1 = load float, float* %arrayidx1, align 4
+  %add = fadd float %tmp, %tmp1
+  store float %add, float* %arrayidx1, align 4
+  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
+  %exitcond = icmp eq i64 %indvars.iv.next, 1000
+  br i1 %exitcond, label %for.cond.cleanup, label %for.body
+}