[flang][OpenMP] Enable delayed privatization by default omp.wsloop
#122471
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
llvmbot
added
mlir:llvm
mlir
flang
|
@llvm/pr-subscribers-flang-fir-hlfir @llvm/pr-subscribers-mlir-openmp Author: Kareem Ergawy (ergawy) ChangesThis enable delayed privatization by default for Patch is 251.21 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/122471.diff 73 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index cd4b25a17722c1..2ce753039880c1 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -2079,7 +2079,7 @@ static void genStandaloneDo(lower::AbstractConverter &converter,
DataSharingProcessor dsp(converter, semaCtx, item->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
- enableDelayedPrivatizationStaging, symTable);
+ enableDelayedPrivatization, symTable);
dsp.processStep1(&wsloopClauseOps);
mlir::omp::LoopNestOperands loopNestClauseOps;
diff --git a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90 b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
index 66fd120085c782..c98850b8000d36 100644
--- a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
+++ b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
@@ -1,6 +1,6 @@
-! RUN: %flang_fc1 -emit-hlfir -fopenmp -mmlir --openmp-enable-delayed-privatization-staging \
+! RUN: %flang_fc1 -emit-hlfir -fopenmp \
! RUN: -o - %s 2>&1 | FileCheck %s
-! RUN: bbc -emit-hlfir -fopenmp --openmp-enable-delayed-privatization-staging -o - %s 2>&1 \
+! RUN: bbc -emit-hlfir -fopenmp -o - %s 2>&1 \
! RUN: | FileCheck %s
subroutine wsloop_private
diff --git a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90 b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
index 77a1304f39a488..10879c53dc0c58 100644
--- a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
+++ b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
@@ -9,11 +9,10 @@
! The string "EXPECTED" denotes the expected FIR
! CHECK: omp.parallel private(@{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
-! CHECK: %[[TEMP:.*]] = fir.alloca i32 {bindc_name = "x", pinned, {{.*}}}
! CHECK: %[[const_1:.*]] = arith.constant 1 : i32
! CHECK: %[[const_2:.*]] = arith.constant 10 : i32
! CHECK: %[[const_3:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[TEMP:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[const_1]]) to (%[[const_2]]) inclusive step (%[[const_3]]) {
! CHECK: fir.store %[[ARG]] to %[[TEMP]] : !fir.ref<i32>
! EXPECTED: %[[temp_1:.*]] = fir.load %[[PRIVATE_Z]] : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/associate.f90 b/flang/test/Lower/OpenMP/associate.f90
index 4964890a6842c1..d497b4ade782e0 100644
--- a/flang/test/Lower/OpenMP/associate.f90
+++ b/flang/test/Lower/OpenMP/associate.f90
@@ -6,12 +6,12 @@
!CHECK: omp.parallel {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEa"}
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
!CHECK: omp.parallel {{.*}} {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
subroutine test_parallel_assoc()
diff --git a/flang/test/Lower/OpenMP/copyin.f90 b/flang/test/Lower/OpenMP/copyin.f90
index f3d147c10668f4..af41442e6870f6 100644
--- a/flang/test/Lower/OpenMP/copyin.f90
+++ b/flang/test/Lower/OpenMP/copyin.f90
@@ -154,14 +154,13 @@ subroutine copyin_derived_type()
! CHECK: omp.barrier
-! CHECK: %[[VAL_6:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_11:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_12:.*]] = fir.load %[[VAL_9]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_13:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_6:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_14:.*]]) : i32 = (%[[VAL_11]]) to (%[[VAL_12]]) inclusive step (%[[VAL_13]]) {
+! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_14]] to %[[VAL_7]]#1 : !fir.ref<i32>
! CHECK: fir.call @_QPsub4(%[[VAL_9]]#1) fastmath<contract> : (!fir.ref<i32>) -> ()
! CHECK: omp.yield
@@ -321,15 +320,12 @@ subroutine common_1()
! CHECK: %[[VAL_33:.*]] = fir.load %[[VAL_18]]#0 : !fir.ref<i32>
! CHECK: hlfir.assign %[[VAL_33]] to %[[VAL_31]]#0 : i32, !fir.ref<i32>
! CHECK: omp.barrier
-
-! CHECK: %[[VAL_19:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
! CHECK: %[[VAL_34:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_35:.*]] = fir.load %[[VAL_26]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_36:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_19:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_37:.*]]) : i32 = (%[[VAL_34]]) to (%[[VAL_35]]) inclusive step (%[[VAL_36]]) {
+! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_37]] to %[[VAL_20]]#1 : !fir.ref<i32>
! CHECK: %[[VAL_38:.*]] = fir.load %[[VAL_31]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_39:.*]] = fir.load %[[VAL_20]]#0 : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/critical.f90 b/flang/test/Lower/OpenMP/critical.f90
index 051d3782106469..99a4426ab04533 100644
--- a/flang/test/Lower/OpenMP/critical.f90
+++ b/flang/test/Lower/OpenMP/critical.f90
@@ -38,11 +38,10 @@ subroutine predetermined_privatization()
!CHECK: omp.parallel
!$omp parallel do
- !CHECK: %[[PRIV_I_ALLOC:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
- !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
do i = 2, 10
- !CHECK: omp.wsloop
+ !CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIV_I_ALLOC:.*]] : !fir.ref<i32>)
!CHECK: omp.loop_nest (%[[IV:[^[:space:]]+]])
+ !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
!CHECK: fir.store %[[IV]] to %[[PRIV_I_DECL]]#1
!CHECK: omp.critical
!$omp critical
diff --git a/flang/test/Lower/OpenMP/default-clause-byref.f90 b/flang/test/Lower/OpenMP/default-clause-byref.f90
index 654c13ada9e39f..10e62005f42ba0 100644
--- a/flang/test/Lower/OpenMP/default-clause-byref.f90
+++ b/flang/test/Lower/OpenMP/default-clause-byref.f90
@@ -346,7 +346,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/default-clause.f90 b/flang/test/Lower/OpenMP/default-clause.f90
index c004813a911f73..fcc8d033eea0fa 100644
--- a/flang/test/Lower/OpenMP/default-clause.f90
+++ b/flang/test/Lower/OpenMP/default-clause.f90
@@ -284,16 +284,13 @@ subroutine nested_default_clause_test4
!CHECK-LABEL: func @_QPnested_default_clause_test5
!CHECK: omp.parallel {
-!CHECK: %[[X_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "x", pinned, uniq_name = "_QFnested_default_clause_test5Ex"}
-!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
-
-!CHECK: %[[LOOP_VAR_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 50 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[X_ALLOCA:.*]], @{{.*}} %{{.*}} -> %[[LOOP_VAR_ALLOCA:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -321,13 +318,12 @@ subroutine nested_default_clause_test5
!CHECK: %[[Z_VAR_DECLARE:.*]]:2 = hlfir.declare %[[Z_VAR]] {{.*}}
-!CHECK: %[[LOOP_VAR:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 10 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[LOOP_VAR:.*]] : !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_VAR_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -386,7 +382,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/hlfir-wsloop.f90 b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
index f7b0ba681efebf..786ab916d000c3 100644
--- a/flang/test/Lower/OpenMP/hlfir-wsloop.f90
+++ b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
@@ -10,12 +10,11 @@ subroutine simple_loop
! CHECK-DAG: %[[WS_END:.*]] = arith.constant 9 : i32
! CHECK: omp.parallel
!$OMP PARALLEL
- ! CHECK-DAG: %[[ALLOCA_IV:.*]] = fir.alloca i32 {{{.*}}, pinned, {{.*}}}
- ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
- ! CHECK: omp.wsloop {
+ ! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[ALLOCA_IV:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%[[WS_ST]]) to (%[[WS_END]]) inclusive step (%[[WS_ST]]) {
!$OMP DO
do i=1, 9
+ ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
! CHECK: fir.store %[[I]] to %[[IV:.*]] : !fir.ref<i32>
! CHECK: %[[LOAD_IV:.*]] = fir.load %[[IV]] : !fir.ref<i32>
! CHECK: fir.call @_FortranAioOutputInteger32({{.*}}, %[[LOAD_IV]]) {{.*}}: (!fir.ref<i8>, i32) -> i1
diff --git a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90 b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
index 6b7d849fde93ca..fd8338393dd880 100644
--- a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
@@ -8,12 +8,11 @@
! CHECK: fir.store %[[VAL_2]] to %[[VAL_0]] : !fir.ref<!fir.box<!fir.heap<i32>>>
! CHECK: %[[VAL_3:.*]]:2 = hlfir.declare %[[VAL_0]] {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
! CHECK: omp.parallel {
-! create original copy of private variable
-! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
-! CHECK: %[[VAL_17:.*]] = fir.alloca i32 {bindc_name = "i", pinned, uniq_name = "_QFEi"}
-! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %{{.*}}, @{{.*}} %{{.*}} -> %[[VAL_17:.*]] : !fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<i32>) {
! CHECK: omp.loop_nest
+! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
+! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+
! [...]
! if this is the last iteration
! CHECK: fir.if %{{.*}} {
diff --git a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90 b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
index faa3d3e053f345..c059382bf634c5 100644
--- a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
@@ -11,12 +11,10 @@
!CHECK: %[[CB_C_Y_COOR:.*]] = fir.coordinate_of %[[CB_C_REF_CVT]], %{{.*}} : (!fir.ref<!fir.array<?xi8>>, index) -> !fir.ref<i8>
!CHECK: %[[CB_C_Y_ADDR:.*]] = fir.convert %[[CB_C_Y_COOR]] : (!fir.ref<i8>) -> !fir.ref<f32>
!CHECK: %[[Y_DECL:.*]]:2 = hlfir.declare %[[CB_C_Y_ADDR]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_X_REF:.*]] = fir.alloca f32 {bindc_name = "x", pinned, uniq_name = "_QFlastprivate_commonEx"}
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIVATE_X_REF:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y_REF:.*]], @{{.*}} %{{.*}} -> %{{.*}} : !{{.*}}, !{{.*}}, !{{.*}}) {
+!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[PRIVATE_X_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_X_REF]] {uniq_name = "_QFlastprivate_commonEx"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_Y_REF:.*]] = fir.alloca f32 {bindc_name = "y", pinned, uniq_name = "_QFlastprivate_commonEy"}
!CHECK: %[[PRIVATE_Y_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_Y_REF]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: omp.wsloop {
-!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[V:.*]] = arith.addi %[[I]], %{{.*}} : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
!CHECK: %[[NEG_STEP:.*]] = arith.cmpi slt, %{{.*}}, %[[C0]] : i32
diff --git a/flang/test/Lower/OpenMP/lastprivate-iv.f90 b/flang/test/Lower/OpenMP/lastprivate-iv.f90
index 63a81e818bc8ba..aacefd8b59c0f2 100644
--- a/flang/test/Lower/OpenMP/lastprivate-iv.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-iv.f90
@@ -6,14 +6,12 @@
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_incEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 4 : i32
!CHECK: %[[UB:.*]] = arith.constant 10 : i32
!CHECK: %[[STEP:.*]] = arith.constant 3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -42,15 +40,12 @@ subroutine lastprivate_iv_inc()
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_decEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 10 : i32
!CHECK: %[[UB:.*]] = arith.constant 1 : i32
!CHECK: %[[STEP:.*]] = arith.constant -3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -80,7 +75,7 @@ subroutine lastprivate_iv_dec()
subroutine lastprivate_iv_i1
integer*1 :: i1
i1=0
-!CHECK: omp.wsloop
+!CHECK: omp.wsloop private({{.*}})
!CHECK: omp.loop_nest
!CHECK: fir.if %{{.*}} {
!CHECK: %[[I8_VAL:.*]] = fir.convert %{{.*}} : (i32) -> i8
diff --git a/flang/test/Lower/OpenMP/location.f90 b/flang/test/Lower/OpenMP/location.f90
index 2dab22a1c1f90d..fc7dd434998638 100644
--- a/flang/test/Lower/OpenMP/location.f90
+++ b/flang/test/Lower/OpenMP/location.f90
@@ -28,7 +28,7 @@ subroutine sub_target()
!CHECK-LABEL: sub_loop
subroutine sub_loop()
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK-NEXT: omp.loop_nest {{.*}} {
!$omp do
do i=1,10
diff --git a/flang/test/Lower/OpenMP/order-clause.f90 b/flang/test/Lower/OpenMP/order-clause.f90
index 717d9740c56f80..75b3a27bbd7fe1 100644
--- a/flang/test/Lower/OpenMP/order-clause.f90
+++ b/flang/test/Lower/OpenMP/order-clause.f90
@@ -20,15 +20,15 @@ end subroutine simd_order
!CHECK-LABEL: func.func @_QPdo_order() {
subroutine do_order
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(reproducible:concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(unconstrained:concurrent) {
+ !CHECK: omp.wsloop order(unconstrained:concurrent) private({{.*}}) {
!$omp do order(unconstrained:concurrent)
do i = 1, 10
end do
diff --git a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90 b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
index 86309a24f91a03..531413c124f81b 100644
--- a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
+++ b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
@@ -10,12 +10,12 @@
!CHECK-DAG: %[[ARG1_DECL:.*]]:2 = hlfir.declare %[[ARG1_REF]] typeparams %[[FIVE]] dummy_scope %{{[0-9]+}} {uniq_name = "_QFlastprivate_characterEarg1"} : (!fir.ref<!fir.char<1,5>>, index, !fir.dscope) -> (!fir.ref<!fir.char<1,5>>, !fir.ref<!fir.char<1,5>>)
!CHECK: omp.parallel {
-!CHECK-DAG: %[[ARG1_PVT:.*]] = fir.alloca !fir.char<1,5> {bindc_name = "arg1", pin...
[truncated]
|
|
@llvm/pr-subscribers-mlir-llvm Author: Kareem Ergawy (ergawy) ChangesThis enable delayed privatization by default for Patch is 251.21 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/122471.diff 73 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index cd4b25a17722c1..2ce753039880c1 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -2079,7 +2079,7 @@ static void genStandaloneDo(lower::AbstractConverter &converter,
DataSharingProcessor dsp(converter, semaCtx, item->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
- enableDelayedPrivatizationStaging, symTable);
+ enableDelayedPrivatization, symTable);
dsp.processStep1(&wsloopClauseOps);
mlir::omp::LoopNestOperands loopNestClauseOps;
diff --git a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90 b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
index 66fd120085c782..c98850b8000d36 100644
--- a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
+++ b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
@@ -1,6 +1,6 @@
-! RUN: %flang_fc1 -emit-hlfir -fopenmp -mmlir --openmp-enable-delayed-privatization-staging \
+! RUN: %flang_fc1 -emit-hlfir -fopenmp \
! RUN: -o - %s 2>&1 | FileCheck %s
-! RUN: bbc -emit-hlfir -fopenmp --openmp-enable-delayed-privatization-staging -o - %s 2>&1 \
+! RUN: bbc -emit-hlfir -fopenmp -o - %s 2>&1 \
! RUN: | FileCheck %s
subroutine wsloop_private
diff --git a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90 b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
index 77a1304f39a488..10879c53dc0c58 100644
--- a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
+++ b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
@@ -9,11 +9,10 @@
! The string "EXPECTED" denotes the expected FIR
! CHECK: omp.parallel private(@{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
-! CHECK: %[[TEMP:.*]] = fir.alloca i32 {bindc_name = "x", pinned, {{.*}}}
! CHECK: %[[const_1:.*]] = arith.constant 1 : i32
! CHECK: %[[const_2:.*]] = arith.constant 10 : i32
! CHECK: %[[const_3:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[TEMP:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[const_1]]) to (%[[const_2]]) inclusive step (%[[const_3]]) {
! CHECK: fir.store %[[ARG]] to %[[TEMP]] : !fir.ref<i32>
! EXPECTED: %[[temp_1:.*]] = fir.load %[[PRIVATE_Z]] : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/associate.f90 b/flang/test/Lower/OpenMP/associate.f90
index 4964890a6842c1..d497b4ade782e0 100644
--- a/flang/test/Lower/OpenMP/associate.f90
+++ b/flang/test/Lower/OpenMP/associate.f90
@@ -6,12 +6,12 @@
!CHECK: omp.parallel {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEa"}
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
!CHECK: omp.parallel {{.*}} {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
subroutine test_parallel_assoc()
diff --git a/flang/test/Lower/OpenMP/copyin.f90 b/flang/test/Lower/OpenMP/copyin.f90
index f3d147c10668f4..af41442e6870f6 100644
--- a/flang/test/Lower/OpenMP/copyin.f90
+++ b/flang/test/Lower/OpenMP/copyin.f90
@@ -154,14 +154,13 @@ subroutine copyin_derived_type()
! CHECK: omp.barrier
-! CHECK: %[[VAL_6:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_11:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_12:.*]] = fir.load %[[VAL_9]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_13:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_6:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_14:.*]]) : i32 = (%[[VAL_11]]) to (%[[VAL_12]]) inclusive step (%[[VAL_13]]) {
+! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_14]] to %[[VAL_7]]#1 : !fir.ref<i32>
! CHECK: fir.call @_QPsub4(%[[VAL_9]]#1) fastmath<contract> : (!fir.ref<i32>) -> ()
! CHECK: omp.yield
@@ -321,15 +320,12 @@ subroutine common_1()
! CHECK: %[[VAL_33:.*]] = fir.load %[[VAL_18]]#0 : !fir.ref<i32>
! CHECK: hlfir.assign %[[VAL_33]] to %[[VAL_31]]#0 : i32, !fir.ref<i32>
! CHECK: omp.barrier
-
-! CHECK: %[[VAL_19:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
! CHECK: %[[VAL_34:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_35:.*]] = fir.load %[[VAL_26]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_36:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_19:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_37:.*]]) : i32 = (%[[VAL_34]]) to (%[[VAL_35]]) inclusive step (%[[VAL_36]]) {
+! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_37]] to %[[VAL_20]]#1 : !fir.ref<i32>
! CHECK: %[[VAL_38:.*]] = fir.load %[[VAL_31]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_39:.*]] = fir.load %[[VAL_20]]#0 : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/critical.f90 b/flang/test/Lower/OpenMP/critical.f90
index 051d3782106469..99a4426ab04533 100644
--- a/flang/test/Lower/OpenMP/critical.f90
+++ b/flang/test/Lower/OpenMP/critical.f90
@@ -38,11 +38,10 @@ subroutine predetermined_privatization()
!CHECK: omp.parallel
!$omp parallel do
- !CHECK: %[[PRIV_I_ALLOC:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
- !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
do i = 2, 10
- !CHECK: omp.wsloop
+ !CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIV_I_ALLOC:.*]] : !fir.ref<i32>)
!CHECK: omp.loop_nest (%[[IV:[^[:space:]]+]])
+ !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
!CHECK: fir.store %[[IV]] to %[[PRIV_I_DECL]]#1
!CHECK: omp.critical
!$omp critical
diff --git a/flang/test/Lower/OpenMP/default-clause-byref.f90 b/flang/test/Lower/OpenMP/default-clause-byref.f90
index 654c13ada9e39f..10e62005f42ba0 100644
--- a/flang/test/Lower/OpenMP/default-clause-byref.f90
+++ b/flang/test/Lower/OpenMP/default-clause-byref.f90
@@ -346,7 +346,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/default-clause.f90 b/flang/test/Lower/OpenMP/default-clause.f90
index c004813a911f73..fcc8d033eea0fa 100644
--- a/flang/test/Lower/OpenMP/default-clause.f90
+++ b/flang/test/Lower/OpenMP/default-clause.f90
@@ -284,16 +284,13 @@ subroutine nested_default_clause_test4
!CHECK-LABEL: func @_QPnested_default_clause_test5
!CHECK: omp.parallel {
-!CHECK: %[[X_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "x", pinned, uniq_name = "_QFnested_default_clause_test5Ex"}
-!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
-
-!CHECK: %[[LOOP_VAR_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 50 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[X_ALLOCA:.*]], @{{.*}} %{{.*}} -> %[[LOOP_VAR_ALLOCA:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -321,13 +318,12 @@ subroutine nested_default_clause_test5
!CHECK: %[[Z_VAR_DECLARE:.*]]:2 = hlfir.declare %[[Z_VAR]] {{.*}}
-!CHECK: %[[LOOP_VAR:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 10 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[LOOP_VAR:.*]] : !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_VAR_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -386,7 +382,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/hlfir-wsloop.f90 b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
index f7b0ba681efebf..786ab916d000c3 100644
--- a/flang/test/Lower/OpenMP/hlfir-wsloop.f90
+++ b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
@@ -10,12 +10,11 @@ subroutine simple_loop
! CHECK-DAG: %[[WS_END:.*]] = arith.constant 9 : i32
! CHECK: omp.parallel
!$OMP PARALLEL
- ! CHECK-DAG: %[[ALLOCA_IV:.*]] = fir.alloca i32 {{{.*}}, pinned, {{.*}}}
- ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
- ! CHECK: omp.wsloop {
+ ! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[ALLOCA_IV:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%[[WS_ST]]) to (%[[WS_END]]) inclusive step (%[[WS_ST]]) {
!$OMP DO
do i=1, 9
+ ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
! CHECK: fir.store %[[I]] to %[[IV:.*]] : !fir.ref<i32>
! CHECK: %[[LOAD_IV:.*]] = fir.load %[[IV]] : !fir.ref<i32>
! CHECK: fir.call @_FortranAioOutputInteger32({{.*}}, %[[LOAD_IV]]) {{.*}}: (!fir.ref<i8>, i32) -> i1
diff --git a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90 b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
index 6b7d849fde93ca..fd8338393dd880 100644
--- a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
@@ -8,12 +8,11 @@
! CHECK: fir.store %[[VAL_2]] to %[[VAL_0]] : !fir.ref<!fir.box<!fir.heap<i32>>>
! CHECK: %[[VAL_3:.*]]:2 = hlfir.declare %[[VAL_0]] {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
! CHECK: omp.parallel {
-! create original copy of private variable
-! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
-! CHECK: %[[VAL_17:.*]] = fir.alloca i32 {bindc_name = "i", pinned, uniq_name = "_QFEi"}
-! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %{{.*}}, @{{.*}} %{{.*}} -> %[[VAL_17:.*]] : !fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<i32>) {
! CHECK: omp.loop_nest
+! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
+! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+
! [...]
! if this is the last iteration
! CHECK: fir.if %{{.*}} {
diff --git a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90 b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
index faa3d3e053f345..c059382bf634c5 100644
--- a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
@@ -11,12 +11,10 @@
!CHECK: %[[CB_C_Y_COOR:.*]] = fir.coordinate_of %[[CB_C_REF_CVT]], %{{.*}} : (!fir.ref<!fir.array<?xi8>>, index) -> !fir.ref<i8>
!CHECK: %[[CB_C_Y_ADDR:.*]] = fir.convert %[[CB_C_Y_COOR]] : (!fir.ref<i8>) -> !fir.ref<f32>
!CHECK: %[[Y_DECL:.*]]:2 = hlfir.declare %[[CB_C_Y_ADDR]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_X_REF:.*]] = fir.alloca f32 {bindc_name = "x", pinned, uniq_name = "_QFlastprivate_commonEx"}
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIVATE_X_REF:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y_REF:.*]], @{{.*}} %{{.*}} -> %{{.*}} : !{{.*}}, !{{.*}}, !{{.*}}) {
+!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[PRIVATE_X_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_X_REF]] {uniq_name = "_QFlastprivate_commonEx"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_Y_REF:.*]] = fir.alloca f32 {bindc_name = "y", pinned, uniq_name = "_QFlastprivate_commonEy"}
!CHECK: %[[PRIVATE_Y_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_Y_REF]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: omp.wsloop {
-!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[V:.*]] = arith.addi %[[I]], %{{.*}} : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
!CHECK: %[[NEG_STEP:.*]] = arith.cmpi slt, %{{.*}}, %[[C0]] : i32
diff --git a/flang/test/Lower/OpenMP/lastprivate-iv.f90 b/flang/test/Lower/OpenMP/lastprivate-iv.f90
index 63a81e818bc8ba..aacefd8b59c0f2 100644
--- a/flang/test/Lower/OpenMP/lastprivate-iv.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-iv.f90
@@ -6,14 +6,12 @@
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_incEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 4 : i32
!CHECK: %[[UB:.*]] = arith.constant 10 : i32
!CHECK: %[[STEP:.*]] = arith.constant 3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -42,15 +40,12 @@ subroutine lastprivate_iv_inc()
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_decEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 10 : i32
!CHECK: %[[UB:.*]] = arith.constant 1 : i32
!CHECK: %[[STEP:.*]] = arith.constant -3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -80,7 +75,7 @@ subroutine lastprivate_iv_dec()
subroutine lastprivate_iv_i1
integer*1 :: i1
i1=0
-!CHECK: omp.wsloop
+!CHECK: omp.wsloop private({{.*}})
!CHECK: omp.loop_nest
!CHECK: fir.if %{{.*}} {
!CHECK: %[[I8_VAL:.*]] = fir.convert %{{.*}} : (i32) -> i8
diff --git a/flang/test/Lower/OpenMP/location.f90 b/flang/test/Lower/OpenMP/location.f90
index 2dab22a1c1f90d..fc7dd434998638 100644
--- a/flang/test/Lower/OpenMP/location.f90
+++ b/flang/test/Lower/OpenMP/location.f90
@@ -28,7 +28,7 @@ subroutine sub_target()
!CHECK-LABEL: sub_loop
subroutine sub_loop()
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK-NEXT: omp.loop_nest {{.*}} {
!$omp do
do i=1,10
diff --git a/flang/test/Lower/OpenMP/order-clause.f90 b/flang/test/Lower/OpenMP/order-clause.f90
index 717d9740c56f80..75b3a27bbd7fe1 100644
--- a/flang/test/Lower/OpenMP/order-clause.f90
+++ b/flang/test/Lower/OpenMP/order-clause.f90
@@ -20,15 +20,15 @@ end subroutine simd_order
!CHECK-LABEL: func.func @_QPdo_order() {
subroutine do_order
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(reproducible:concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(unconstrained:concurrent) {
+ !CHECK: omp.wsloop order(unconstrained:concurrent) private({{.*}}) {
!$omp do order(unconstrained:concurrent)
do i = 1, 10
end do
diff --git a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90 b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
index 86309a24f91a03..531413c124f81b 100644
--- a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
+++ b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
@@ -10,12 +10,12 @@
!CHECK-DAG: %[[ARG1_DECL:.*]]:2 = hlfir.declare %[[ARG1_REF]] typeparams %[[FIVE]] dummy_scope %{{[0-9]+}} {uniq_name = "_QFlastprivate_characterEarg1"} : (!fir.ref<!fir.char<1,5>>, index, !fir.dscope) -> (!fir.ref<!fir.char<1,5>>, !fir.ref<!fir.char<1,5>>)
!CHECK: omp.parallel {
-!CHECK-DAG: %[[ARG1_PVT:.*]] = fir.alloca !fir.char<1,5> {bindc_name = "arg1", pin...
[truncated]
|
|
@llvm/pr-subscribers-flang-openmp Author: Kareem Ergawy (ergawy) ChangesThis enable delayed privatization by default for Patch is 251.21 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/122471.diff 73 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index cd4b25a17722c1..2ce753039880c1 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -2079,7 +2079,7 @@ static void genStandaloneDo(lower::AbstractConverter &converter,
DataSharingProcessor dsp(converter, semaCtx, item->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
- enableDelayedPrivatizationStaging, symTable);
+ enableDelayedPrivatization, symTable);
dsp.processStep1(&wsloopClauseOps);
mlir::omp::LoopNestOperands loopNestClauseOps;
diff --git a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90 b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
index 66fd120085c782..c98850b8000d36 100644
--- a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
+++ b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
@@ -1,6 +1,6 @@
-! RUN: %flang_fc1 -emit-hlfir -fopenmp -mmlir --openmp-enable-delayed-privatization-staging \
+! RUN: %flang_fc1 -emit-hlfir -fopenmp \
! RUN: -o - %s 2>&1 | FileCheck %s
-! RUN: bbc -emit-hlfir -fopenmp --openmp-enable-delayed-privatization-staging -o - %s 2>&1 \
+! RUN: bbc -emit-hlfir -fopenmp -o - %s 2>&1 \
! RUN: | FileCheck %s
subroutine wsloop_private
diff --git a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90 b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
index 77a1304f39a488..10879c53dc0c58 100644
--- a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
+++ b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
@@ -9,11 +9,10 @@
! The string "EXPECTED" denotes the expected FIR
! CHECK: omp.parallel private(@{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
-! CHECK: %[[TEMP:.*]] = fir.alloca i32 {bindc_name = "x", pinned, {{.*}}}
! CHECK: %[[const_1:.*]] = arith.constant 1 : i32
! CHECK: %[[const_2:.*]] = arith.constant 10 : i32
! CHECK: %[[const_3:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[TEMP:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[const_1]]) to (%[[const_2]]) inclusive step (%[[const_3]]) {
! CHECK: fir.store %[[ARG]] to %[[TEMP]] : !fir.ref<i32>
! EXPECTED: %[[temp_1:.*]] = fir.load %[[PRIVATE_Z]] : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/associate.f90 b/flang/test/Lower/OpenMP/associate.f90
index 4964890a6842c1..d497b4ade782e0 100644
--- a/flang/test/Lower/OpenMP/associate.f90
+++ b/flang/test/Lower/OpenMP/associate.f90
@@ -6,12 +6,12 @@
!CHECK: omp.parallel {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEa"}
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
!CHECK: omp.parallel {{.*}} {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
subroutine test_parallel_assoc()
diff --git a/flang/test/Lower/OpenMP/copyin.f90 b/flang/test/Lower/OpenMP/copyin.f90
index f3d147c10668f4..af41442e6870f6 100644
--- a/flang/test/Lower/OpenMP/copyin.f90
+++ b/flang/test/Lower/OpenMP/copyin.f90
@@ -154,14 +154,13 @@ subroutine copyin_derived_type()
! CHECK: omp.barrier
-! CHECK: %[[VAL_6:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_11:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_12:.*]] = fir.load %[[VAL_9]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_13:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_6:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_14:.*]]) : i32 = (%[[VAL_11]]) to (%[[VAL_12]]) inclusive step (%[[VAL_13]]) {
+! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_14]] to %[[VAL_7]]#1 : !fir.ref<i32>
! CHECK: fir.call @_QPsub4(%[[VAL_9]]#1) fastmath<contract> : (!fir.ref<i32>) -> ()
! CHECK: omp.yield
@@ -321,15 +320,12 @@ subroutine common_1()
! CHECK: %[[VAL_33:.*]] = fir.load %[[VAL_18]]#0 : !fir.ref<i32>
! CHECK: hlfir.assign %[[VAL_33]] to %[[VAL_31]]#0 : i32, !fir.ref<i32>
! CHECK: omp.barrier
-
-! CHECK: %[[VAL_19:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
! CHECK: %[[VAL_34:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_35:.*]] = fir.load %[[VAL_26]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_36:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_19:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_37:.*]]) : i32 = (%[[VAL_34]]) to (%[[VAL_35]]) inclusive step (%[[VAL_36]]) {
+! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_37]] to %[[VAL_20]]#1 : !fir.ref<i32>
! CHECK: %[[VAL_38:.*]] = fir.load %[[VAL_31]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_39:.*]] = fir.load %[[VAL_20]]#0 : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/critical.f90 b/flang/test/Lower/OpenMP/critical.f90
index 051d3782106469..99a4426ab04533 100644
--- a/flang/test/Lower/OpenMP/critical.f90
+++ b/flang/test/Lower/OpenMP/critical.f90
@@ -38,11 +38,10 @@ subroutine predetermined_privatization()
!CHECK: omp.parallel
!$omp parallel do
- !CHECK: %[[PRIV_I_ALLOC:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
- !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
do i = 2, 10
- !CHECK: omp.wsloop
+ !CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIV_I_ALLOC:.*]] : !fir.ref<i32>)
!CHECK: omp.loop_nest (%[[IV:[^[:space:]]+]])
+ !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
!CHECK: fir.store %[[IV]] to %[[PRIV_I_DECL]]#1
!CHECK: omp.critical
!$omp critical
diff --git a/flang/test/Lower/OpenMP/default-clause-byref.f90 b/flang/test/Lower/OpenMP/default-clause-byref.f90
index 654c13ada9e39f..10e62005f42ba0 100644
--- a/flang/test/Lower/OpenMP/default-clause-byref.f90
+++ b/flang/test/Lower/OpenMP/default-clause-byref.f90
@@ -346,7 +346,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/default-clause.f90 b/flang/test/Lower/OpenMP/default-clause.f90
index c004813a911f73..fcc8d033eea0fa 100644
--- a/flang/test/Lower/OpenMP/default-clause.f90
+++ b/flang/test/Lower/OpenMP/default-clause.f90
@@ -284,16 +284,13 @@ subroutine nested_default_clause_test4
!CHECK-LABEL: func @_QPnested_default_clause_test5
!CHECK: omp.parallel {
-!CHECK: %[[X_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "x", pinned, uniq_name = "_QFnested_default_clause_test5Ex"}
-!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
-
-!CHECK: %[[LOOP_VAR_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 50 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[X_ALLOCA:.*]], @{{.*}} %{{.*}} -> %[[LOOP_VAR_ALLOCA:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -321,13 +318,12 @@ subroutine nested_default_clause_test5
!CHECK: %[[Z_VAR_DECLARE:.*]]:2 = hlfir.declare %[[Z_VAR]] {{.*}}
-!CHECK: %[[LOOP_VAR:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 10 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[LOOP_VAR:.*]] : !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_VAR_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -386,7 +382,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/hlfir-wsloop.f90 b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
index f7b0ba681efebf..786ab916d000c3 100644
--- a/flang/test/Lower/OpenMP/hlfir-wsloop.f90
+++ b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
@@ -10,12 +10,11 @@ subroutine simple_loop
! CHECK-DAG: %[[WS_END:.*]] = arith.constant 9 : i32
! CHECK: omp.parallel
!$OMP PARALLEL
- ! CHECK-DAG: %[[ALLOCA_IV:.*]] = fir.alloca i32 {{{.*}}, pinned, {{.*}}}
- ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
- ! CHECK: omp.wsloop {
+ ! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[ALLOCA_IV:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%[[WS_ST]]) to (%[[WS_END]]) inclusive step (%[[WS_ST]]) {
!$OMP DO
do i=1, 9
+ ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
! CHECK: fir.store %[[I]] to %[[IV:.*]] : !fir.ref<i32>
! CHECK: %[[LOAD_IV:.*]] = fir.load %[[IV]] : !fir.ref<i32>
! CHECK: fir.call @_FortranAioOutputInteger32({{.*}}, %[[LOAD_IV]]) {{.*}}: (!fir.ref<i8>, i32) -> i1
diff --git a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90 b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
index 6b7d849fde93ca..fd8338393dd880 100644
--- a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
@@ -8,12 +8,11 @@
! CHECK: fir.store %[[VAL_2]] to %[[VAL_0]] : !fir.ref<!fir.box<!fir.heap<i32>>>
! CHECK: %[[VAL_3:.*]]:2 = hlfir.declare %[[VAL_0]] {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
! CHECK: omp.parallel {
-! create original copy of private variable
-! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
-! CHECK: %[[VAL_17:.*]] = fir.alloca i32 {bindc_name = "i", pinned, uniq_name = "_QFEi"}
-! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %{{.*}}, @{{.*}} %{{.*}} -> %[[VAL_17:.*]] : !fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<i32>) {
! CHECK: omp.loop_nest
+! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
+! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+
! [...]
! if this is the last iteration
! CHECK: fir.if %{{.*}} {
diff --git a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90 b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
index faa3d3e053f345..c059382bf634c5 100644
--- a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
@@ -11,12 +11,10 @@
!CHECK: %[[CB_C_Y_COOR:.*]] = fir.coordinate_of %[[CB_C_REF_CVT]], %{{.*}} : (!fir.ref<!fir.array<?xi8>>, index) -> !fir.ref<i8>
!CHECK: %[[CB_C_Y_ADDR:.*]] = fir.convert %[[CB_C_Y_COOR]] : (!fir.ref<i8>) -> !fir.ref<f32>
!CHECK: %[[Y_DECL:.*]]:2 = hlfir.declare %[[CB_C_Y_ADDR]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_X_REF:.*]] = fir.alloca f32 {bindc_name = "x", pinned, uniq_name = "_QFlastprivate_commonEx"}
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIVATE_X_REF:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y_REF:.*]], @{{.*}} %{{.*}} -> %{{.*}} : !{{.*}}, !{{.*}}, !{{.*}}) {
+!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[PRIVATE_X_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_X_REF]] {uniq_name = "_QFlastprivate_commonEx"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_Y_REF:.*]] = fir.alloca f32 {bindc_name = "y", pinned, uniq_name = "_QFlastprivate_commonEy"}
!CHECK: %[[PRIVATE_Y_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_Y_REF]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: omp.wsloop {
-!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[V:.*]] = arith.addi %[[I]], %{{.*}} : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
!CHECK: %[[NEG_STEP:.*]] = arith.cmpi slt, %{{.*}}, %[[C0]] : i32
diff --git a/flang/test/Lower/OpenMP/lastprivate-iv.f90 b/flang/test/Lower/OpenMP/lastprivate-iv.f90
index 63a81e818bc8ba..aacefd8b59c0f2 100644
--- a/flang/test/Lower/OpenMP/lastprivate-iv.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-iv.f90
@@ -6,14 +6,12 @@
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_incEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 4 : i32
!CHECK: %[[UB:.*]] = arith.constant 10 : i32
!CHECK: %[[STEP:.*]] = arith.constant 3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -42,15 +40,12 @@ subroutine lastprivate_iv_inc()
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_decEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 10 : i32
!CHECK: %[[UB:.*]] = arith.constant 1 : i32
!CHECK: %[[STEP:.*]] = arith.constant -3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -80,7 +75,7 @@ subroutine lastprivate_iv_dec()
subroutine lastprivate_iv_i1
integer*1 :: i1
i1=0
-!CHECK: omp.wsloop
+!CHECK: omp.wsloop private({{.*}})
!CHECK: omp.loop_nest
!CHECK: fir.if %{{.*}} {
!CHECK: %[[I8_VAL:.*]] = fir.convert %{{.*}} : (i32) -> i8
diff --git a/flang/test/Lower/OpenMP/location.f90 b/flang/test/Lower/OpenMP/location.f90
index 2dab22a1c1f90d..fc7dd434998638 100644
--- a/flang/test/Lower/OpenMP/location.f90
+++ b/flang/test/Lower/OpenMP/location.f90
@@ -28,7 +28,7 @@ subroutine sub_target()
!CHECK-LABEL: sub_loop
subroutine sub_loop()
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK-NEXT: omp.loop_nest {{.*}} {
!$omp do
do i=1,10
diff --git a/flang/test/Lower/OpenMP/order-clause.f90 b/flang/test/Lower/OpenMP/order-clause.f90
index 717d9740c56f80..75b3a27bbd7fe1 100644
--- a/flang/test/Lower/OpenMP/order-clause.f90
+++ b/flang/test/Lower/OpenMP/order-clause.f90
@@ -20,15 +20,15 @@ end subroutine simd_order
!CHECK-LABEL: func.func @_QPdo_order() {
subroutine do_order
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(reproducible:concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(unconstrained:concurrent) {
+ !CHECK: omp.wsloop order(unconstrained:concurrent) private({{.*}}) {
!$omp do order(unconstrained:concurrent)
do i = 1, 10
end do
diff --git a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90 b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
index 86309a24f91a03..531413c124f81b 100644
--- a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
+++ b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
@@ -10,12 +10,12 @@
!CHECK-DAG: %[[ARG1_DECL:.*]]:2 = hlfir.declare %[[ARG1_REF]] typeparams %[[FIVE]] dummy_scope %{{[0-9]+}} {uniq_name = "_QFlastprivate_characterEarg1"} : (!fir.ref<!fir.char<1,5>>, index, !fir.dscope) -> (!fir.ref<!fir.char<1,5>>, !fir.ref<!fir.char<1,5>>)
!CHECK: omp.parallel {
-!CHECK-DAG: %[[ARG1_PVT:.*]] = fir.alloca !fir.char<1,5> {bindc_name = "arg1", pin...
[truncated]
|
|
@llvm/pr-subscribers-mlir Author: Kareem Ergawy (ergawy) ChangesThis enable delayed privatization by default for Patch is 251.21 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/122471.diff 73 Files Affected:
diff --git a/flang/lib/Lower/OpenMP/OpenMP.cpp b/flang/lib/Lower/OpenMP/OpenMP.cpp
index cd4b25a17722c1..2ce753039880c1 100644
--- a/flang/lib/Lower/OpenMP/OpenMP.cpp
+++ b/flang/lib/Lower/OpenMP/OpenMP.cpp
@@ -2079,7 +2079,7 @@ static void genStandaloneDo(lower::AbstractConverter &converter,
DataSharingProcessor dsp(converter, semaCtx, item->clauses, eval,
/*shouldCollectPreDeterminedSymbols=*/true,
- enableDelayedPrivatizationStaging, symTable);
+ enableDelayedPrivatization, symTable);
dsp.processStep1(&wsloopClauseOps);
mlir::omp::LoopNestOperands loopNestClauseOps;
diff --git a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90 b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
index 66fd120085c782..c98850b8000d36 100644
--- a/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
+++ b/flang/test/Lower/OpenMP/DelayedPrivatization/wsloop.f90
@@ -1,6 +1,6 @@
-! RUN: %flang_fc1 -emit-hlfir -fopenmp -mmlir --openmp-enable-delayed-privatization-staging \
+! RUN: %flang_fc1 -emit-hlfir -fopenmp \
! RUN: -o - %s 2>&1 | FileCheck %s
-! RUN: bbc -emit-hlfir -fopenmp --openmp-enable-delayed-privatization-staging -o - %s 2>&1 \
+! RUN: bbc -emit-hlfir -fopenmp -o - %s 2>&1 \
! RUN: | FileCheck %s
subroutine wsloop_private
diff --git a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90 b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
index 77a1304f39a488..10879c53dc0c58 100644
--- a/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
+++ b/flang/test/Lower/OpenMP/Todo/omp-default-clause-inner-loop.f90
@@ -9,11 +9,10 @@
! The string "EXPECTED" denotes the expected FIR
! CHECK: omp.parallel private(@{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
-! CHECK: %[[TEMP:.*]] = fir.alloca i32 {bindc_name = "x", pinned, {{.*}}}
! CHECK: %[[const_1:.*]] = arith.constant 1 : i32
! CHECK: %[[const_2:.*]] = arith.constant 10 : i32
! CHECK: %[[const_3:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[TEMP:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[const_1]]) to (%[[const_2]]) inclusive step (%[[const_3]]) {
! CHECK: fir.store %[[ARG]] to %[[TEMP]] : !fir.ref<i32>
! EXPECTED: %[[temp_1:.*]] = fir.load %[[PRIVATE_Z]] : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/associate.f90 b/flang/test/Lower/OpenMP/associate.f90
index 4964890a6842c1..d497b4ade782e0 100644
--- a/flang/test/Lower/OpenMP/associate.f90
+++ b/flang/test/Lower/OpenMP/associate.f90
@@ -6,12 +6,12 @@
!CHECK: omp.parallel {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEa"}
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
!CHECK: omp.parallel {{.*}} {
!CHECK-NOT: hlfir.declare {{.*}} {uniq_name = "_QFtest_parallel_assocEb"}
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK: }
!CHECK: }
subroutine test_parallel_assoc()
diff --git a/flang/test/Lower/OpenMP/copyin.f90 b/flang/test/Lower/OpenMP/copyin.f90
index f3d147c10668f4..af41442e6870f6 100644
--- a/flang/test/Lower/OpenMP/copyin.f90
+++ b/flang/test/Lower/OpenMP/copyin.f90
@@ -154,14 +154,13 @@ subroutine copyin_derived_type()
! CHECK: omp.barrier
-! CHECK: %[[VAL_6:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: %[[VAL_11:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_12:.*]] = fir.load %[[VAL_9]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_13:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_6:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_14:.*]]) : i32 = (%[[VAL_11]]) to (%[[VAL_12]]) inclusive step (%[[VAL_13]]) {
+! CHECK: %[[VAL_7:.*]]:2 = hlfir.declare %[[VAL_6]] {uniq_name = "_QFcombined_parallel_worksharing_loopEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_14]] to %[[VAL_7]]#1 : !fir.ref<i32>
! CHECK: fir.call @_QPsub4(%[[VAL_9]]#1) fastmath<contract> : (!fir.ref<i32>) -> ()
! CHECK: omp.yield
@@ -321,15 +320,12 @@ subroutine common_1()
! CHECK: %[[VAL_33:.*]] = fir.load %[[VAL_18]]#0 : !fir.ref<i32>
! CHECK: hlfir.assign %[[VAL_33]] to %[[VAL_31]]#0 : i32, !fir.ref<i32>
! CHECK: omp.barrier
-
-! CHECK: %[[VAL_19:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
! CHECK: %[[VAL_34:.*]] = arith.constant 1 : i32
! CHECK: %[[VAL_35:.*]] = fir.load %[[VAL_26]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_36:.*]] = arith.constant 1 : i32
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[VAL_19:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[VAL_37:.*]]) : i32 = (%[[VAL_34]]) to (%[[VAL_35]]) inclusive step (%[[VAL_36]]) {
+! CHECK: %[[VAL_20:.*]]:2 = hlfir.declare %[[VAL_19]] {uniq_name = "_QFcommon_2Ei"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
! CHECK: fir.store %[[VAL_37]] to %[[VAL_20]]#1 : !fir.ref<i32>
! CHECK: %[[VAL_38:.*]] = fir.load %[[VAL_31]]#0 : !fir.ref<i32>
! CHECK: %[[VAL_39:.*]] = fir.load %[[VAL_20]]#0 : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenMP/critical.f90 b/flang/test/Lower/OpenMP/critical.f90
index 051d3782106469..99a4426ab04533 100644
--- a/flang/test/Lower/OpenMP/critical.f90
+++ b/flang/test/Lower/OpenMP/critical.f90
@@ -38,11 +38,10 @@ subroutine predetermined_privatization()
!CHECK: omp.parallel
!$omp parallel do
- !CHECK: %[[PRIV_I_ALLOC:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
- !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
do i = 2, 10
- !CHECK: omp.wsloop
+ !CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIV_I_ALLOC:.*]] : !fir.ref<i32>)
!CHECK: omp.loop_nest (%[[IV:[^[:space:]]+]])
+ !CHECK: %[[PRIV_I_DECL:.*]]:2 = hlfir.declare %[[PRIV_I_ALLOC]]
!CHECK: fir.store %[[IV]] to %[[PRIV_I_DECL]]#1
!CHECK: omp.critical
!$omp critical
diff --git a/flang/test/Lower/OpenMP/default-clause-byref.f90 b/flang/test/Lower/OpenMP/default-clause-byref.f90
index 654c13ada9e39f..10e62005f42ba0 100644
--- a/flang/test/Lower/OpenMP/default-clause-byref.f90
+++ b/flang/test/Lower/OpenMP/default-clause-byref.f90
@@ -346,7 +346,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(byref @min_byref_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/default-clause.f90 b/flang/test/Lower/OpenMP/default-clause.f90
index c004813a911f73..fcc8d033eea0fa 100644
--- a/flang/test/Lower/OpenMP/default-clause.f90
+++ b/flang/test/Lower/OpenMP/default-clause.f90
@@ -284,16 +284,13 @@ subroutine nested_default_clause_test4
!CHECK-LABEL: func @_QPnested_default_clause_test5
!CHECK: omp.parallel {
-!CHECK: %[[X_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "x", pinned, uniq_name = "_QFnested_default_clause_test5Ex"}
-!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
-
-!CHECK: %[[LOOP_VAR_ALLOCA:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 50 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[X_ALLOCA:.*]], @{{.*}} %{{.*}} -> %[[LOOP_VAR_ALLOCA:.*]] : !fir.ref<i32>, !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[X_DECLARE:.*]]:2 = hlfir.declare %[[X_ALLOCA]] {{.*}}
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR_ALLOCA]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -321,13 +318,12 @@ subroutine nested_default_clause_test5
!CHECK: %[[Z_VAR_DECLARE:.*]]:2 = hlfir.declare %[[Z_VAR]] {{.*}}
-!CHECK: %[[LOOP_VAR:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
-
!CHECK: %[[CONST_LB:.*]] = arith.constant 1 : i32
!CHECK: %[[CONST_UB:.*]] = arith.constant 10 : i32
!CHECK: %[[CONST_STEP:.*]] = arith.constant 1 : i32
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[LOOP_VAR:.*]] : !fir.ref<i32>) {
!CHECK: omp.loop_nest (%[[ARG:.*]]) : i32 = (%[[CONST_LB]]) to (%[[CONST_UB]]) inclusive step (%[[CONST_STEP]]) {
+!CHECK: %[[LOOP_VAR_DECLARE:.*]]:2 = hlfir.declare %[[LOOP_VAR]] {{.*}}
!CHECK: fir.store %[[ARG]] to %[[LOOP_VAR_DECLARE]]#1 : !fir.ref<i32>
!CHECK: %[[LOADED_X:.*]] = fir.load %[[X_VAR_DECLARE]]#0 : !fir.ref<i32>
!CHECK: %[[CONST:.*]] = arith.constant 1 : i32
@@ -386,7 +382,7 @@ subroutine skipped_default_clause_checks()
type(it)::iii
!CHECK: omp.parallel {{.*}} {
-!CHECK: omp.wsloop reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
+!CHECK: omp.wsloop private({{.*}}) reduction(@min_i32 %[[VAL_Z_DECLARE]]#0 -> %[[PRV:.+]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[ARG:.*]]) {{.*}} {
!CHECK: omp.yield
!CHECK: }
diff --git a/flang/test/Lower/OpenMP/hlfir-wsloop.f90 b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
index f7b0ba681efebf..786ab916d000c3 100644
--- a/flang/test/Lower/OpenMP/hlfir-wsloop.f90
+++ b/flang/test/Lower/OpenMP/hlfir-wsloop.f90
@@ -10,12 +10,11 @@ subroutine simple_loop
! CHECK-DAG: %[[WS_END:.*]] = arith.constant 9 : i32
! CHECK: omp.parallel
!$OMP PARALLEL
- ! CHECK-DAG: %[[ALLOCA_IV:.*]] = fir.alloca i32 {{{.*}}, pinned, {{.*}}}
- ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
- ! CHECK: omp.wsloop {
+ ! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[ALLOCA_IV:.*]] : !fir.ref<i32>) {
! CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%[[WS_ST]]) to (%[[WS_END]]) inclusive step (%[[WS_ST]]) {
!$OMP DO
do i=1, 9
+ ! CHECK: %[[IV:.*]] = fir.declare %[[ALLOCA_IV]] {uniq_name = "_QFsimple_loopEi"} : (!fir.ref<i32>) -> !fir.ref<i32>
! CHECK: fir.store %[[I]] to %[[IV:.*]] : !fir.ref<i32>
! CHECK: %[[LOAD_IV:.*]] = fir.load %[[IV]] : !fir.ref<i32>
! CHECK: fir.call @_FortranAioOutputInteger32({{.*}}, %[[LOAD_IV]]) {{.*}}: (!fir.ref<i8>, i32) -> i1
diff --git a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90 b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
index 6b7d849fde93ca..fd8338393dd880 100644
--- a/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-allocatable.f90
@@ -8,12 +8,11 @@
! CHECK: fir.store %[[VAL_2]] to %[[VAL_0]] : !fir.ref<!fir.box<!fir.heap<i32>>>
! CHECK: %[[VAL_3:.*]]:2 = hlfir.declare %[[VAL_0]] {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
! CHECK: omp.parallel {
-! create original copy of private variable
-! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
-! CHECK: %[[VAL_17:.*]] = fir.alloca i32 {bindc_name = "i", pinned, uniq_name = "_QFEi"}
-! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-! CHECK: omp.wsloop {
+! CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %{{.*}}, @{{.*}} %{{.*}} -> %[[VAL_17:.*]] : !fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<i32>) {
! CHECK: omp.loop_nest
+! CHECK: %[[VAL_16:.*]]:2 = hlfir.declare %{{.*}} {fortran_attrs = {{.*}}<allocatable>, uniq_name = "_QFEa"} : (!fir.ref<!fir.box<!fir.heap<i32>>>) -> (!fir.ref<!fir.box<!fir.heap<i32>>>, !fir.ref<!fir.box<!fir.heap<i32>>>)
+! CHECK: %[[VAL_18:.*]]:2 = hlfir.declare %[[VAL_17]] {uniq_name = "_QFEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
+
! [...]
! if this is the last iteration
! CHECK: fir.if %{{.*}} {
diff --git a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90 b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
index faa3d3e053f345..c059382bf634c5 100644
--- a/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-commonblock.f90
@@ -11,12 +11,10 @@
!CHECK: %[[CB_C_Y_COOR:.*]] = fir.coordinate_of %[[CB_C_REF_CVT]], %{{.*}} : (!fir.ref<!fir.array<?xi8>>, index) -> !fir.ref<i8>
!CHECK: %[[CB_C_Y_ADDR:.*]] = fir.convert %[[CB_C_Y_COOR]] : (!fir.ref<i8>) -> !fir.ref<f32>
!CHECK: %[[Y_DECL:.*]]:2 = hlfir.declare %[[CB_C_Y_ADDR]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_X_REF:.*]] = fir.alloca f32 {bindc_name = "x", pinned, uniq_name = "_QFlastprivate_commonEx"}
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[PRIVATE_X_REF:.*]], @{{.*}} %{{.*}} -> %[[PRIVATE_Y_REF:.*]], @{{.*}} %{{.*}} -> %{{.*}} : !{{.*}}, !{{.*}}, !{{.*}}) {
+!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[PRIVATE_X_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_X_REF]] {uniq_name = "_QFlastprivate_commonEx"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: %[[PRIVATE_Y_REF:.*]] = fir.alloca f32 {bindc_name = "y", pinned, uniq_name = "_QFlastprivate_commonEy"}
!CHECK: %[[PRIVATE_Y_DECL:.*]]:2 = hlfir.declare %[[PRIVATE_Y_REF]] {uniq_name = "_QFlastprivate_commonEy"} : (!fir.ref<f32>) -> (!fir.ref<f32>, !fir.ref<f32>)
-!CHECK: omp.wsloop {
-!CHECK-NEXT: omp.loop_nest (%[[I:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) inclusive step (%{{.*}}) {
!CHECK: %[[V:.*]] = arith.addi %[[I]], %{{.*}} : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
!CHECK: %[[NEG_STEP:.*]] = arith.cmpi slt, %{{.*}}, %[[C0]] : i32
diff --git a/flang/test/Lower/OpenMP/lastprivate-iv.f90 b/flang/test/Lower/OpenMP/lastprivate-iv.f90
index 63a81e818bc8ba..aacefd8b59c0f2 100644
--- a/flang/test/Lower/OpenMP/lastprivate-iv.f90
+++ b/flang/test/Lower/OpenMP/lastprivate-iv.f90
@@ -6,14 +6,12 @@
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_incEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 4 : i32
!CHECK: %[[UB:.*]] = arith.constant 10 : i32
!CHECK: %[[STEP:.*]] = arith.constant 3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_incEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -42,15 +40,12 @@ subroutine lastprivate_iv_inc()
!CHECK: %[[I2_MEM:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFlastprivate_iv_decEi"}
!CHECK: %[[I2:.*]]:2 = hlfir.declare %[[I2_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
-!CHECK: %[[I_MEM:.*]] = fir.alloca i32 {bindc_name = "i", pinned, {{.*}}}
-!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-
!CHECK: %[[LB:.*]] = arith.constant 10 : i32
!CHECK: %[[UB:.*]] = arith.constant 1 : i32
!CHECK: %[[STEP:.*]] = arith.constant -3 : i32
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private(@{{.*}} %{{.*}} -> %[[I_MEM:.*]] : !fir.ref<i32>) {
!CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%[[LB]]) to (%[[UB]]) inclusive step (%[[STEP]]) {
+!CHECK: %[[I:.*]]:2 = hlfir.declare %[[I_MEM]] {uniq_name = "_QFlastprivate_iv_decEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
!CHECK: fir.store %[[IV]] to %[[I]]#1 : !fir.ref<i32>
!CHECK: %[[V:.*]] = arith.addi %[[IV]], %[[STEP]] : i32
!CHECK: %[[C0:.*]] = arith.constant 0 : i32
@@ -80,7 +75,7 @@ subroutine lastprivate_iv_dec()
subroutine lastprivate_iv_i1
integer*1 :: i1
i1=0
-!CHECK: omp.wsloop
+!CHECK: omp.wsloop private({{.*}})
!CHECK: omp.loop_nest
!CHECK: fir.if %{{.*}} {
!CHECK: %[[I8_VAL:.*]] = fir.convert %{{.*}} : (i32) -> i8
diff --git a/flang/test/Lower/OpenMP/location.f90 b/flang/test/Lower/OpenMP/location.f90
index 2dab22a1c1f90d..fc7dd434998638 100644
--- a/flang/test/Lower/OpenMP/location.f90
+++ b/flang/test/Lower/OpenMP/location.f90
@@ -28,7 +28,7 @@ subroutine sub_target()
!CHECK-LABEL: sub_loop
subroutine sub_loop()
-!CHECK: omp.wsloop {
+!CHECK: omp.wsloop private({{.*}}) {
!CHECK-NEXT: omp.loop_nest {{.*}} {
!$omp do
do i=1,10
diff --git a/flang/test/Lower/OpenMP/order-clause.f90 b/flang/test/Lower/OpenMP/order-clause.f90
index 717d9740c56f80..75b3a27bbd7fe1 100644
--- a/flang/test/Lower/OpenMP/order-clause.f90
+++ b/flang/test/Lower/OpenMP/order-clause.f90
@@ -20,15 +20,15 @@ end subroutine simd_order
!CHECK-LABEL: func.func @_QPdo_order() {
subroutine do_order
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(reproducible:concurrent) {
+ !CHECK: omp.wsloop order(reproducible:concurrent) private({{.*}}) {
!$omp do order(reproducible:concurrent)
do i = 1, 10
end do
- !CHECK: omp.wsloop order(unconstrained:concurrent) {
+ !CHECK: omp.wsloop order(unconstrained:concurrent) private({{.*}}) {
!$omp do order(unconstrained:concurrent)
do i = 1, 10
end do
diff --git a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90 b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
index 86309a24f91a03..531413c124f81b 100644
--- a/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
+++ b/flang/test/Lower/OpenMP/parallel-lastprivate-clause-scalar.f90
@@ -10,12 +10,12 @@
!CHECK-DAG: %[[ARG1_DECL:.*]]:2 = hlfir.declare %[[ARG1_REF]] typeparams %[[FIVE]] dummy_scope %{{[0-9]+}} {uniq_name = "_QFlastprivate_characterEarg1"} : (!fir.ref<!fir.char<1,5>>, index, !fir.dscope) -> (!fir.ref<!fir.char<1,5>>, !fir.ref<!fir.char<1,5>>)
!CHECK: omp.parallel {
-!CHECK-DAG: %[[ARG1_PVT:.*]] = fir.alloca !fir.char<1,5> {bindc_name = "arg1", pin...
[truncated]
|
ergawy
force-pushed
the
enable_dealyed_priv_ws.loop
branch
2 times, most recently
from
98eaa4b to
705865a
Compare
ergawy
requested review from
jsjodin,
tblah,
kiranchandramohan,
luporl and
skatrak
|
Ping! Can you give this PR a look 👀? Specially the "temporary" workaround until @tblah finishes updating the |
skatrak
approved these changes
Jan 13, 2025
mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp
Outdated
Show resolved
Hide resolved
ergawy
force-pushed
the
enable_dealyed_priv_ws.loop
branch
2 times, most recently
from
b79bcf5 to
a50e46f
Compare
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
ergawy
force-pushed
the
enable_dealyed_priv_ws.loop
branch
2 times, most recently
from
5754fe0 to
423fdaf
Compare
|
I think the workaround looks okay. I found a bug (I think) unrelated to the workaround: This only reproduces when lastprivate is used. |
Seems to be working fine with me: |
luporl
reviewed
Jan 14, 2025
ergawy
changed the title
omp.wsloopomp.wsloop
Sorry I didn't explain: the generated code is wrong. On my system I hit the |
Ah, I see. I will take a look in a bit then ... |
It is indeed a bug and a weird one 🤔. When so it stores from the private copy to the original host one and not the other way around like it should. The weird thing is, the last private update is emitted properly in the body of the op itself. Might be a problem with how values are managed by the symbol table. But not sure yet, looking .... |
|
Yup, I had fun finding that out :). I have a fix and will open a pr soon. |
This enable delayed privatization by default for `omp.wsloop` ops, with one caveat! I had to workaround the "impure" alloc region issue that being resolved at the moment. The workaround detects whether the alloc region's argument is used in the region and at the same time defined in block that does not dominate the chosen alloca insertion point. If so, we move the alloca insertion point below the defining instruction of the alloc region argument. This basically reverts to the non-delayed-privatizaiton behavior.
ergawy
force-pushed
the
enable_dealyed_priv_ws.loop
branch
from
423fdaf to
dc2b91f
Compare
|
Created a follow up PR with a possible fix: #123162. |
ergawy
added a commit
that referenced
this pull request
Jan 16, 2025
…r` (#123162) This fixes a bug when the same variable is used in `firstprivate` and `lastprivate` clauses on the same construct. The issue boils down to the fact that `copyHostAssociateVar` was deciding the direction of the copy assignment (i.e. the `lhs` and `rhs`) based on whether the `copyAssignIP` parameter is set. This is not the best way to do it since it is not related to whether we doing a copy from host to localized copy or the other way around. When we set the insertion for `firstprivate` in delayed privatization, this resulted in switching the direction of the copy assignment. Instead, this PR adds a new paramter to explicitely tell the function the direction of the assignment. This is a follow up PR for #122471, only the latest commit is relevant.
github-actions bot
pushed a commit
to arm/arm-toolchain
that referenced
this pull request
Jan 16, 2025
…AssociateVar` (#123162) This fixes a bug when the same variable is used in `firstprivate` and `lastprivate` clauses on the same construct. The issue boils down to the fact that `copyHostAssociateVar` was deciding the direction of the copy assignment (i.e. the `lhs` and `rhs`) based on whether the `copyAssignIP` parameter is set. This is not the best way to do it since it is not related to whether we doing a copy from host to localized copy or the other way around. When we set the insertion for `firstprivate` in delayed privatization, this resulted in switching the direction of the copy assignment. Instead, this PR adds a new paramter to explicitely tell the function the direction of the assignment. This is a follow up PR for llvm/llvm-project#122471, only the latest commit is relevant.
ergawy
added a commit
to ergawy/llvm-project
that referenced
this pull request
Jan 17, 2025
…wsloop` (llvm#122471)" This seems to have caused some regressions in Fujitsu's test-suite: https://linaro.atlassian.net/browse/LLVM-1521 This reverts commit 6f82408.
|
Reverted while investigating the Fujitsu regressions: #123324 |
ergawy
added a commit
to ergawy/llvm-project
that referenced
this pull request
Jan 20, 2025
…wsloop` (llvm#122471)" This seems to have caused some regressions in Fujitsu's test-suite: https://linaro.atlassian.net/browse/LLVM-1521 This reverts commit 6f82408.
ergawy
added a commit
to ergawy/llvm-project
that referenced
this pull request
Jan 22, 2025
…wsloop` (llvm#122471)" This seems to have caused some regressions in Fujitsu's test-suite: https://linaro.atlassian.net/browse/LLVM-1521 This reverts commit 6f82408.
ergawy
added a commit
that referenced
this pull request
Jan 22, 2025
…wsloop` (#122471)" (#123324) This seems to have caused some regressions in Fujitsu's test-suite: https://linaro.atlassian.net/browse/LLVM-1521 This reverts commit 6f82408.
ergawy
added a commit
that referenced
this pull request
Feb 6, 2025
…125732) Reapplies #122471 This is based on #125699, only the latest commit is relevant. With changes in this PR and the parent one, the previously reported failures in the Fujitsu(*) test suite should hopefully be resolved (I verified all the 14 reported failures and they pass now). (*) https://linaro.atlassian.net/browse/LLVM-1521
Icohedron
pushed a commit
to Icohedron/llvm-project
that referenced
this pull request
Feb 11, 2025
…lvm#125732) Reapplies llvm#122471 This is based on llvm#125699, only the latest commit is relevant. With changes in this PR and the parent one, the previously reported failures in the Fujitsu(*) test suite should hopefully be resolved (I verified all the 14 reported failures and they pass now). (*) https://linaro.atlassian.net/browse/LLVM-1521
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.
This enable delayed privatization by default for
omp.wsloopops, with one caveat! I had to workaround the "impure" alloc region issue that being resolved at the moment. The workaround detects whether the alloc region's argument is used in the region and at the same time defined in block that does not dominate the chosen alloca insertion point. If so, we move the alloca insertion point below the defining instruction of the alloc region argument. This basically reverts to the non-delayed-privatizaiton behavior.