[flang][OpenMP] Create unique reduction decls for different logical kinds #146558
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Some Fujitsu tests showed incorrect results because we were sharing reduction declarations for different kinds for logical variables.
llvmbot
added
flang
|
@llvm/pr-subscribers-flang-openmp @llvm/pr-subscribers-flang-fir-hlfir Author: Tom Eccles (tblah) ChangesSome Fujitsu tests showed incorrect results because we were sharing reduction declarations for different kinds for logical variables. Patch is 37.46 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/146558.diff 10 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/ReductionProcessor.cpp b/flang/lib/Lower/OpenMP/ReductionProcessor.cpp
index d7f520e86e532..330cef7b54c74 100644
--- a/flang/lib/Lower/OpenMP/ReductionProcessor.cpp
+++ b/flang/lib/Lower/OpenMP/ReductionProcessor.cpp
@@ -149,13 +149,17 @@ std::string ReductionProcessor::getReductionName(
reductionName = "multiply_reduction";
break;
case omp::clause::DefinedOperator::IntrinsicOperator::AND:
- return "and_reduction";
+ reductionName = "and_reduction";
+ break;
case omp::clause::DefinedOperator::IntrinsicOperator::EQV:
- return "eqv_reduction";
+ reductionName = "eqv_reduction";
+ break;
case omp::clause::DefinedOperator::IntrinsicOperator::OR:
- return "or_reduction";
+ reductionName = "or_reduction";
+ break;
case omp::clause::DefinedOperator::IntrinsicOperator::NEQV:
- return "neqv_reduction";
+ reductionName = "neqv_reduction";
+ break;
default:
reductionName = "other_reduction";
break;
@@ -673,9 +677,6 @@ void ReductionProcessor::processReductionArguments(
const auto &kindMap = firOpBuilder.getKindMap();
std::string reductionName;
ReductionIdentifier redId;
- mlir::Type redNameTy = redType;
- if (mlir::isa<fir::LogicalType>(redType.getEleTy()))
- redNameTy = builder.getI1Type();
if (const auto &redDefinedOp =
std::get_if<omp::clause::DefinedOperator>(&redOperator.u)) {
@@ -697,8 +698,7 @@ void ReductionProcessor::processReductionArguments(
break;
}
- reductionName =
- getReductionName(intrinsicOp, kindMap, redNameTy, isByRef);
+ reductionName = getReductionName(intrinsicOp, kindMap, redType, isByRef);
} else if (const auto *reductionIntrinsic =
std::get_if<omp::clause::ProcedureDesignator>(
&redOperator.u)) {
@@ -709,7 +709,7 @@ void ReductionProcessor::processReductionArguments(
redId = getReductionType(*reductionIntrinsic);
reductionName =
getReductionName(getRealName(*reductionIntrinsic).ToString(), kindMap,
- redNameTy, isByRef);
+ redType, isByRef);
} else {
TODO(currentLocation, "Unexpected reduction type");
}
diff --git a/flang/test/Lower/OpenMP/wsloop-reduction-logical-and-byref.f90 b/flang/test/Lower/OpenMP/wsloop-reduction-logical-and-byref.f90
index 18e252a787d78..cf5c1b127c34d 100644
--- a/flang/test/Lower/OpenMP/wsloop-reduction-logical-and-byref.f90
+++ b/flang/test/Lower/OpenMP/wsloop-reduction-logical-and-byref.f90
@@ -3,7 +3,7 @@
! NOTE: Assertions have been autogenerated by utils/generate-test-checks.py
-! CHECK-LABEL: omp.declare_reduction @and_reduction : !fir.ref<!fir.logical<4>>
+! CHECK-LABEL: omp.declare_reduction @and_reduction_byref_l32 : !fir.ref<!fir.logical<4>>
! CHECK-SAME: alloc {
! CHECK: %[[REF:.*]] = fir.alloca !fir.logical<4>
! CHECK: omp.yield(%[[REF]] : !fir.ref<!fir.logical<4>>)
@@ -42,7 +42,7 @@
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @and_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @and_reduction_byref_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reductionEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -90,7 +90,7 @@ end subroutine simple_reduction
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @and_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @and_reduction_byref_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reduction_switch_orderEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reduction_switch_orderEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -147,7 +147,7 @@ subroutine simple_reduction_switch_order(y)
! CHECK: %[[VAL_20:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_21:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_22:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(byref @and_reduction %[[VAL_7]]#0 -> %[[VAL_23:.*]], byref @and_reduction %[[VAL_9]]#0 -> %[[VAL_24:.*]], byref @and_reduction %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(byref @and_reduction_byref_l32 %[[VAL_7]]#0 -> %[[VAL_23:.*]], byref @and_reduction_byref_l32 %[[VAL_9]]#0 -> %[[VAL_24:.*]], byref @and_reduction_byref_l32 %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_26:.*]]) : i32 = (%[[VAL_20]]) to (%[[VAL_21]]) inclusive step (%[[VAL_22]]) {
! CHECK: %[[VAL_19:.*]]:2 = hlfir.declare %[[VAL_18]] {uniq_name = "_QFmultiple_reductionsEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_27:.*]]:2 = hlfir.declare %[[VAL_23]] {uniq_name = "_QFmultiple_reductionsEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
diff --git a/flang/test/Lower/OpenMP/wsloop-reduction-logical-and.f90 b/flang/test/Lower/OpenMP/wsloop-reduction-logical-and.f90
index 7f3e3922ab29b..eff97aa5d0db0 100644
--- a/flang/test/Lower/OpenMP/wsloop-reduction-logical-and.f90
+++ b/flang/test/Lower/OpenMP/wsloop-reduction-logical-and.f90
@@ -3,7 +3,7 @@
! NOTE: Assertions have been autogenerated by utils/generate-test-checks.py
-! CHECK-LABEL: omp.declare_reduction @and_reduction : !fir.logical<4> init {
+! CHECK-LABEL: omp.declare_reduction @and_reduction_l32 : !fir.logical<4> init {
! CHECK: ^bb0(%[[VAL_0:.*]]: !fir.logical<4>):
! CHECK: %[[VAL_1:.*]] = arith.constant true
! CHECK: %[[VAL_2:.*]] = fir.convert %[[VAL_1]] : (i1) -> !fir.logical<4>
@@ -34,7 +34,7 @@
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@and_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@and_reduction_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reductionEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -82,7 +82,7 @@ end subroutine simple_reduction
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@and_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@and_reduction_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reduction_switch_orderEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reduction_switch_orderEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -139,7 +139,7 @@ subroutine simple_reduction_switch_order(y)
! CHECK: %[[VAL_20:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_21:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_22:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(@and_reduction %[[VAL_7]]#0 -> %[[VAL_23:.*]], @and_reduction %[[VAL_9]]#0 -> %[[VAL_24:.*]], @and_reduction %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(@and_reduction_l32 %[[VAL_7]]#0 -> %[[VAL_23:.*]], @and_reduction_l32 %[[VAL_9]]#0 -> %[[VAL_24:.*]], @and_reduction_l32 %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_26:.*]]) : i32 = (%[[VAL_20]]) to (%[[VAL_21]]) inclusive step (%[[VAL_22]]) {
! CHECK: %[[VAL_19:.*]]:2 = hlfir.declare %[[VAL_18]] {uniq_name = "_QFmultiple_reductionsEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_27:.*]]:2 = hlfir.declare %[[VAL_23]] {uniq_name = "_QFmultiple_reductionsEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
diff --git a/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv-byref.f90 b/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv-byref.f90
index e7654edffec8b..bbe4bf04b6827 100644
--- a/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv-byref.f90
+++ b/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv-byref.f90
@@ -3,7 +3,7 @@
! NOTE: Assertions have been autogenerated by utils/generate-test-checks.py
-! CHECK-LABEL: omp.declare_reduction @eqv_reduction : !fir.ref<!fir.logical<4>>
+! CHECK-LABEL: omp.declare_reduction @eqv_reduction_byref_l32 : !fir.ref<!fir.logical<4>>
! CHECK-SAME: alloc {
! CHECK: %[[REF:.*]] = fir.alloca !fir.logical<4>
! CHECK: omp.yield(%[[REF]] : !fir.ref<!fir.logical<4>>)
@@ -42,7 +42,7 @@
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @eqv_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @eqv_reduction_byref_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reductionEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -89,7 +89,7 @@ subroutine simple_reduction(y)
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @eqv_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(byref @eqv_reduction_byref_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reduction_switch_orderEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reduction_switch_orderEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -146,7 +146,7 @@ subroutine simple_reduction_switch_order(y)
! CHECK: %[[VAL_20:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_21:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_22:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(byref @eqv_reduction %[[VAL_7]]#0 -> %[[VAL_23:.*]], byref @eqv_reduction %[[VAL_9]]#0 -> %[[VAL_24:.*]], byref @eqv_reduction %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(byref @eqv_reduction_byref_l32 %[[VAL_7]]#0 -> %[[VAL_23:.*]], byref @eqv_reduction_byref_l32 %[[VAL_9]]#0 -> %[[VAL_24:.*]], byref @eqv_reduction_byref_l32 %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_26:.*]]) : i32 = (%[[VAL_20]]) to (%[[VAL_21]]) inclusive step (%[[VAL_22]]) {
! CHECK: %[[VAL_19:.*]]:2 = hlfir.declare %[[VAL_18]] {uniq_name = "_QFmultiple_reductionsEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_27:.*]]:2 = hlfir.declare %[[VAL_23]] {uniq_name = "_QFmultiple_reductionsEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
diff --git a/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv.f90 b/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv.f90
index dffd6545b95bf..304c7eefd58b0 100644
--- a/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv.f90
+++ b/flang/test/Lower/OpenMP/wsloop-reduction-logical-eqv.f90
@@ -3,7 +3,7 @@
! NOTE: Assertions have been autogenerated by utils/generate-test-checks.py
-! CHECK-LABEL: omp.declare_reduction @eqv_reduction : !fir.logical<4> init {
+! CHECK-LABEL: omp.declare_reduction @eqv_reduction_l32 : !fir.logical<4> init {
! CHECK: ^bb0(%[[VAL_0:.*]]: !fir.logical<4>):
! CHECK: %[[VAL_1:.*]] = arith.constant true
! CHECK: %[[VAL_2:.*]] = fir.convert %[[VAL_1]] : (i1) -> !fir.logical<4>
@@ -34,7 +34,7 @@
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@eqv_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@eqv_reduction_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reductionEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reductionEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -81,7 +81,7 @@ subroutine simple_reduction(y)
! CHECK: %[[VAL_12:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_13:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_14:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@eqv_reduction %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_10:.*]] : !fir.ref<i32>) reduction(@eqv_reduction_l32 %[[VAL_4]]#0 -> %[[VAL_15:.*]] : !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_16:.*]]) : i32 = (%[[VAL_12]]) to (%[[VAL_13]]) inclusive step (%[[VAL_14]]) {
! CHECK: %[[VAL_11:.*]]:2 = hlfir.declare %[[VAL_10]] {uniq_name = "_QFsimple_reduction_switch_orderEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_17:.*]]:2 = hlfir.declare %[[VAL_15]] {uniq_name = "_QFsimple_reduction_switch_orderEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
@@ -138,7 +138,7 @@ subroutine simple_reduction_switch_order(y)
! CHECK: %[[VAL_20:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_21:.*]] = arith.constant 100 : i32
! CHECK: %[[VAL_22:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(@eqv_reduction %[[VAL_7]]#0 -> %[[VAL_23:.*]], @eqv_reduction %[[VAL_9]]#0 -> %[[VAL_24:.*]], @eqv_reduction %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}}#0 -> %[[VAL_18:.*]] : !fir.ref<i32>) reduction(@eqv_reduction_l32 %[[VAL_7]]#0 -> %[[VAL_23:.*]], @eqv_reduction_l32 %[[VAL_9]]#0 -> %[[VAL_24:.*]], @eqv_reduction_l32 %[[VAL_11]]#0 -> %[[VAL_25:.*]] : !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_26:.*]]) : i32 = (%[[VAL_20]]) to (%[[VAL_21]]) inclusive step (%[[VAL_22]]) {
! CHECK: %[[VAL_19:.*]]:2 = hlfir.declare %[[VAL_18]] {uniq_name = "_QFmultiple_reductionsEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_27:.*]]:2 = hlfir.declare %[[VAL_23]] {uniq_name = "_QFmultiple_reductionsEx"} : (!fir.ref<!fir.logical<4>>) -> (!fir.ref<!fir.logical<4>>, !fir.ref<!fir.logical<4>>)
diff --git a/flang/test/Lower/OpenMP/wsloop-reduction-logical-kinds.f90 b/flang/test/Lower/OpenMP/wsloop-reduction-logical-kinds.f90
new file mode 100644
index 0000000000000..f7fd2a52060e6
--- /dev/null
+++ b/flang/test/Lower/OpenMP/wsloop-reduction-logical-kinds.f90
@@ -0,0 +1,24 @@
+! RUN: %flang_fc1 -emit-hlfir -fopenmp -o - %s | FileCheck %s
+
+! Check that logical reductions of different kinds do not end up using the same
+! reduction declaration
+
+! CHECK-LABEL: omp.declare_reduction @or_reduction_l64 : !fir.logical<8> init {
+! CHECK-LABEL: omp.declare_reduction @or_reduction_l32 : !fir.logical<4> init {
+
+subroutine test(a4, a8, sz)
+ integer :: sz
+ logical(4)...
[truncated]
|
abhinavgaba
reviewed
Jul 8, 2025
There was a problem hiding this comment.
In case you aren't aware of this, the test seems to be failing in flang Windows buildbots off and on. See https://lab.llvm.org/buildbot/#/builders/207/builds/3523, https://lab.llvm.org/buildbot/#/builders/207/builds/3524, https://lab.llvm.org/buildbot/#/builders/207/builds/3530 for reference.
Please take a look.
There was a problem hiding this comment.
Thank you for the bug report. I don't think flang/test/Parser/OpenMP/declare-reduction-operator.f90 could have been effected by this PR because that is testing a much earlier stage of the compiler and will not even execute the changed lines. The stack trace has very little information but both look like they could be the same issue.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Some Fujitsu tests showed incorrect results because we were sharing reduction declarations for different kinds for logical variables.