[PATCH] D40911: [OpenMP] Fix PR35542: Correct adjusting of private reduction variable

Jonas Hahnfeld via Phabricator via cfe-commits Wed, 06 Dec 2017 11:13:20 -0800

Hahnfeld updated this revision to Diff 125768.
Hahnfeld marked 3 inline comments as done.
Hahnfeld added a comment.


Get type from `Address`.


https://reviews.llvm.org/D40911

Files:
  lib/CodeGen/CGOpenMPRuntime.cpp
  test/OpenMP/for_reduction_codegen.cpp
  test/OpenMP/for_reduction_codegen_UDR.cpp

Index: test/OpenMP/for_reduction_codegen_UDR.cpp
===================================================================
--- test/OpenMP/for_reduction_codegen_UDR.cpp
+++ test/OpenMP/for_reduction_codegen_UDR.cpp
@@ -735,7 +735,8 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint ([[S_FLOAT_TY]]* getelementptr ([[S_FLOAT_TY]], [[S_FLOAT_TY]]* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VVAR2_PRIV:%.+]] = getelementptr [5 x [[S_FLOAT_TY]]], [5 x [[S_FLOAT_TY]]]* [[VVAR2_PRIV]], i64 [[OFFSET]]
+// CHECK: [[VVAR2_PRIV_PTR:%.+]] = bitcast [5 x [[S_FLOAT_TY]]]* [[VVAR2_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[VVAR2_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VVAR2_PRIV_PTR]], i64 [[OFFSET]]
 // CHECK: ret void
 
 // CHECK: define internal void [[MAIN_MICROTASK5]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}}, [4 x [[S_FLOAT_TY]]]* dereferenceable(48) %{{.+}})
@@ -761,8 +762,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint ([[S_FLOAT_TY]]* getelementptr ([[S_FLOAT_TY]], [[S_FLOAT_TY]]* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VAR3_PRIV:%.+]] = getelementptr [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]], i64 [[OFFSET]]
-// CHECK: [[VAR3_PRIV:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[PSEUDO_VAR3_PRIV]] to [4 x [[S_FLOAT_TY]]]*
+// CHECK: [[VAR3_PRIV_PTR:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[PSEUDO_VAR3_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VAR3_PRIV_PTR]], i64 [[OFFSET]]
+// CHECK: [[VAR3_PRIV:%.+]] = bitcast [[S_FLOAT_TY]]* [[PSEUDO_VAR3_PRIV]] to [4 x [[S_FLOAT_TY]]]*
 
 // CHECK: store [4 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]], [4 x [[S_FLOAT_TY]]]** %
 
@@ -985,8 +987,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_INT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint ([[S_INT_TY]]* getelementptr ([[S_INT_TY]], [[S_INT_TY]]* null, i32 1) to i64)
-// CHECK: [[PSEUDO_ARR_PRIV:%.+]] = getelementptr [40 x [[S_INT_TY]]], [40 x [[S_INT_TY]]]* [[ARR_PRIV]], i64 [[OFFSET]]
-// CHECK: [[ARR_PRIV:%.+]] = bitcast [40 x [[S_INT_TY]]]* [[PSEUDO_ARR_PRIV]] to [42 x [[S_INT_TY]]]*
+// CHECK: [[ARR_PRIV_PTR:%.+]] = bitcast [40 x [[S_INT_TY]]]* [[ARR_PRIV]] to [[S_INT_TY]]*
+// CHECK: [[PSEUDO_ARR_PRIV:%.+]] = getelementptr [[S_INT_TY]], [[S_INT_TY]]* [[ARR_PRIV_PTR]], i64 [[OFFSET]]
+// CHECK: [[ARR_PRIV:%.+]] = bitcast [[S_INT_TY]]* [[PSEUDO_ARR_PRIV]] to [42 x [[S_INT_TY]]]*
 
 // CHECK: ret void
 
Index: test/OpenMP/for_reduction_codegen.cpp
===================================================================
--- test/OpenMP/for_reduction_codegen.cpp
+++ test/OpenMP/for_reduction_codegen.cpp
@@ -944,8 +944,8 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint i32* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (i32* getelementptr (i32, i32* null, i32 1) to i64)
-// CHECK: [[PSEUDO_ARR_PRIV:%.+]] = getelementptr [1 x [2 x i32]], [1 x [2 x i32]]* [[ARR_PRIV]], i64 [[OFFSET]]
-// CHECK: [[ARR_PRIV:%.+]] = bitcast [1 x [2 x i32]]* [[PSEUDO_ARR_PRIV]] to i32*
+// CHECK: [[ARR_PRIV_PTR:%.+]] = bitcast [1 x [2 x i32]]* [[ARR_PRIV]] to i32*
+// CHECK: [[ARR_PRIV:%.+]] = getelementptr i32, i32* [[ARR_PRIV_PTR]], i64 [[OFFSET]]
 
 // CHECK: ret void
 
@@ -1000,9 +1000,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (float* getelementptr (float, float* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VAR2_PRIV:%.+]] = getelementptr [1 x [6 x [[S_FLOAT_TY]]]], [1 x [6 x [[S_FLOAT_TY]]]]* [[VAR2_PRIV]], i64 [[OFFSET]]
+// CHECK: [[VAR2_PRIV_PTR:%.+]] = bitcast [1 x [6 x [[S_FLOAT_TY]]]]* [[VAR2_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[VAR2_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VAR2_PRIV_PTR]], i64 [[OFFSET]]
 // CHECK: store [[S_FLOAT_TY]]** [[REF:.+]], [[S_FLOAT_TY]]*** %
-// CHECK: [[VAR2_PRIV:%.+]] = bitcast [1 x [6 x [[S_FLOAT_TY]]]]* [[PSEUDO_VAR2_PRIV]] to [[S_FLOAT_TY]]*
 // CHECK: store [[S_FLOAT_TY]]* [[VAR2_PRIV]], [[S_FLOAT_TY]]** [[REF]]
 // CHECK: ret void
 
@@ -1029,9 +1029,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (float* getelementptr (float, float* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VAR2_PRIV:%.+]] = getelementptr [1 x [6 x [[S_FLOAT_TY]]]], [1 x [6 x [[S_FLOAT_TY]]]]* [[VAR2_PRIV]], i64 [[OFFSET]]
+// CHECK: [[VAR2_PRIV_PTR:%.+]] = bitcast [1 x [6 x [[S_FLOAT_TY]]]]* [[VAR2_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[VAR2_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VAR2_PRIV_PTR]], i64 [[OFFSET]]
 // CHECK: store [[S_FLOAT_TY]]** [[REF:.+]], [[S_FLOAT_TY]]*** %
-// CHECK: [[VAR2_PRIV:%.+]] = bitcast [1 x [6 x [[S_FLOAT_TY]]]]* [[PSEUDO_VAR2_PRIV]] to [[S_FLOAT_TY]]*
 // CHECK: store [[S_FLOAT_TY]]* [[VAR2_PRIV]], [[S_FLOAT_TY]]** [[REF]]
 // CHECK: ret void
 
@@ -1080,7 +1080,8 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (float* getelementptr (float, float* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VVAR2_PRIV:%.+]] = getelementptr [5 x [[S_FLOAT_TY]]], [5 x [[S_FLOAT_TY]]]* [[VVAR2_PRIV]], i64 [[OFFSET]]
+// CHECK: [[VVAR2_PRIV_PTR:%.+]] = bitcast [5 x [[S_FLOAT_TY]]]* [[VVAR2_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[VVAR2_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VVAR2_PRIV_PTR]], i64 [[OFFSET]]
 // CHECK: ret void
 
 // CHECK: define internal void [[MAIN_MICROTASK9]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}}, [4 x [[S_FLOAT_TY]]]* dereferenceable(16) %{{.+}})
@@ -1106,8 +1107,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (float* getelementptr (float, float* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VAR3_PRIV:%.+]] = getelementptr [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]], i64 [[OFFSET]]
-// CHECK: [[VAR3_PRIV:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[PSEUDO_VAR3_PRIV]] to [4 x [[S_FLOAT_TY]]]*
+// CHECK: [[VAR3_PRIV_PTR:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[PSEUDO_VAR3_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VAR3_PRIV_PTR]], i64 [[OFFSET]]
+// CHECK: [[VAR3_PRIV:%.+]] = bitcast [[S_FLOAT_TY]]* [[PSEUDO_VAR3_PRIV]] to [4 x [[S_FLOAT_TY]]]*
 
 // CHECK: store [4 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]], [4 x [[S_FLOAT_TY]]]** %
 
@@ -1136,8 +1138,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_FLOAT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (float* getelementptr (float, float* null, i32 1) to i64)
-// CHECK: [[PSEUDO_VAR3_PRIV:%.+]] = getelementptr [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]], i64 [[OFFSET]]
-// CHECK: [[VAR3_PRIV:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[PSEUDO_VAR3_PRIV]] to [4 x [[S_FLOAT_TY]]]*
+// CHECK: [[VAR3_PRIV_PTR:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[PSEUDO_VAR3_PRIV:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[VAR3_PRIV_PTR]], i64 [[OFFSET]]
+// CHECK: [[VAR3_PRIV:%.+]] = bitcast [[S_FLOAT_TY]]* [[PSEUDO_VAR3_PRIV]] to [4 x [[S_FLOAT_TY]]]*
 
 // CHECK: store [4 x [[S_FLOAT_TY]]]* [[VAR3_PRIV]], [4 x [[S_FLOAT_TY]]]** %
 
@@ -1453,8 +1456,9 @@
 // CHECK: [[LOW_BOUND:%.+]] = ptrtoint [[S_INT_TY]]* [[LOW]] to i64
 // CHECK: [[OFFSET_BYTES:%.+]] = sub i64 [[START]], [[LOW_BOUND]]
 // CHECK: [[OFFSET:%.+]] = sdiv exact i64 [[OFFSET_BYTES]], ptrtoint (i32* getelementptr (i32, i32* null, i32 1) to i64)
-// CHECK: [[PSEUDO_ARR_PRIV:%.+]] = getelementptr [40 x [[S_INT_TY]]], [40 x [[S_INT_TY]]]* [[ARR_PRIV]], i64 [[OFFSET]]
-// CHECK: [[ARR_PRIV:%.+]] = bitcast [40 x [[S_INT_TY]]]* [[PSEUDO_ARR_PRIV]] to [42 x [[S_INT_TY]]]*
+// CHECK: [[ARR_PRIV_PTR:%.+]] = bitcast [40 x [[S_INT_TY]]]* [[ARR_PRIV]] to [[S_INT_TY]]*
+// CHECK: [[PSEUDO_ARR_PRIV:%.+]] = getelementptr [[S_INT_TY]], [[S_INT_TY]]* [[ARR_PRIV_PTR]], i64 [[OFFSET]]
+// CHECK: [[ARR_PRIV:%.+]] = bitcast [[S_INT_TY]]* [[PSEUDO_ARR_PRIV]] to [42 x [[S_INT_TY]]]*
 
 // CHECK: ret void
 
Index: lib/CodeGen/CGOpenMPRuntime.cpp
===================================================================
--- lib/CodeGen/CGOpenMPRuntime.cpp
+++ lib/CodeGen/CGOpenMPRuntime.cpp
@@ -1104,11 +1104,14 @@
                     OriginalBaseLValue);
     llvm::Value *Adjustment = CGF.Builder.CreatePtrDiff(
         BaseLValue.getPointer(), SharedAddresses[N].first.getPointer());
-    llvm::Value *Ptr =
-        CGF.Builder.CreateGEP(PrivateAddr.getPointer(), Adjustment);
+    llvm::Value *PrivatePointer =
+        CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
+            PrivateAddr.getPointer(),
+            SharedAddresses[N].first.getAddress().getType());
+    llvm::Value *Ptr = CGF.Builder.CreateGEP(PrivatePointer, Adjustment);
     return castToBase(CGF, OrigVD->getType(),
                       SharedAddresses[N].first.getType(),
-                      OriginalBaseLValue.getPointer()->getType(),
+                      OriginalBaseLValue.getAddress().getType(),
                       OriginalBaseLValue.getAlignment(), Ptr);
   }
   BaseDecls.emplace_back(

_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D40911: [OpenMP] Fix PR35542: Correct adjusting of private reduction variable

Reply via email to