aeaafce464
The intention of this work is to give MLIR->LLVMIR conversion freedom to control how the private variable is allocated so that it can be allocated on the stack in ordinary cases or as part of a structure used to give closure context for tasks which might outlive the current stack frame. See RFC: https://discourse.llvm.org/t/rfc-openmp-supporting-delayed-task-execution-with-firstprivate-variables/83084 For example, a privatizer for an integer used to look like ```mlir omp.private {type = private} @x.privatizer : !fir.ref<i32> alloc { ^bb0(%arg0: !fir.ref<i32>): %0 = ... allocate proper memory for the private clone ... omp.yield(%0 : !fir.ref<i32>) } ``` After this change, allocation become implicit in the operation: ```mlir omp.private {type = private} @x.privatizer : i32 ``` For more complex types that require initialization after allocation, an init region can be used: ``` mlir omp.private {type = private} @x.privatizer : !some.type init { ^bb0(%arg0: !some.pointer<!some.type>, %arg1: !some.pointer<!some.type>): // initialize %arg1, using %arg0 as a mold for allocations omp.yield(%arg1 : !some.pointer<!some.type>) } dealloc { ^bb0(%arg0: !some.pointer<!some.type>): ... deallocate memory allocated by the init region ... omp.yield } ``` This patch lays the groundwork for delayed task execution but is not enough on its own. After this patch all gfortran tests which previously passed still pass. There are the following changes to the Fujitsu test suite: - 0380_0009 and 0435_0009 are fixed - 0688_0041 now fails at runtime. This patch is testing firstprivate variables with tasks. Previously we got lucky with the undefined behavior and won the race. After these changes we no longer get lucky. This patch lays the groundwork for a proper fix for this issue. In flang the lowering re-uses the existing lowering used for reduction init and dealloc regions. In flang, before this patch we hit a TODO with the same wording when generating the copy region for firstprivate polymorphic variables. After this patch the box-like fir.class is passed by reference into the copy region, leading to a different path that didn't hit that old TODO but the generated code still didn't work so I added a new TODO in DataSharingProcessor.
182 lines
5.3 KiB
Fortran
182 lines
5.3 KiB
Fortran
! This test checks lowering of OpenMP loop Directive.
|
|
|
|
! RUN: bbc -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
|
|
! RUN: %flang_fc1 -emit-hlfir -fopenmp -fopenmp-version=50 -o - %s 2>&1 | FileCheck %s
|
|
|
|
! CHECK: omp.declare_reduction @[[RED:add_reduction_i32]] : i32
|
|
! CHECK: omp.private {type = private} @[[DUMMY_PRIV:.*test_privateEdummy_private.*]] : i32
|
|
! CHECK: omp.private {type = private} @[[I_PRIV:.*test_no_clausesEi.*]] : i32
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_no_clauses
|
|
subroutine test_no_clauses()
|
|
integer :: i, j, dummy = 1
|
|
|
|
! CHECK: omp.simd private(@[[I_PRIV]] %{{.*}}#0 -> %[[ARG:.*]] : !fir.ref<i32>) {
|
|
! CHECK-NEXT: omp.loop_nest (%[[IV:.*]]) : i32 = (%{{.*}}) to (%{{.*}}) {{.*}} {
|
|
! CHECK: %[[ARG_DECL:.*]]:2 = hlfir.declare %[[ARG]]
|
|
! CHECK: fir.store %[[IV]] to %[[ARG_DECL]]#1 : !fir.ref<i32>
|
|
! CHECK: }
|
|
! CHECK: }
|
|
!$omp loop
|
|
do i=1,10
|
|
dummy = dummy + 1
|
|
end do
|
|
!$omp end loop
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_collapse
|
|
subroutine test_collapse()
|
|
integer :: i, j, dummy = 1
|
|
! CHECK: omp.simd private(@{{.*}} %{{.*}}#0 -> %{{.*}}, @{{.*}} %{{.*}}#0 -> %{{.*}} : {{.*}}) {
|
|
! CHECK-NEXT: omp.loop_nest (%{{.*}}, %{{.*}}) : i32 {{.*}} {
|
|
! CHECK: }
|
|
! CHECK: }
|
|
!$omp loop collapse(2)
|
|
do i=1,10
|
|
do j=2,20
|
|
dummy = dummy + 1
|
|
end do
|
|
end do
|
|
!$omp end loop
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_private
|
|
subroutine test_private()
|
|
integer :: i, dummy = 1
|
|
! CHECK: omp.simd private(@[[DUMMY_PRIV]] %{{.*}}#0 -> %[[DUMMY_ARG:.*]], @{{.*}} %{{.*}}#0 -> %{{.*}} : {{.*}}) {
|
|
! CHECK-NEXT: omp.loop_nest (%{{.*}}) : i32 = (%{{.*}}) to (%{{.*}}) {{.*}} {
|
|
! CHECK: %[[DUMMY_DECL:.*]]:2 = hlfir.declare %[[DUMMY_ARG]] {uniq_name = "_QFtest_privateEdummy"}
|
|
! CHECK: %{{.*}} = fir.load %[[DUMMY_DECL]]#0
|
|
! CHECK: hlfir.assign %{{.*}} to %[[DUMMY_DECL]]#0
|
|
! CHECK: }
|
|
! CHECK: }
|
|
!$omp loop private(dummy)
|
|
do i=1,10
|
|
dummy = dummy + 1
|
|
end do
|
|
!$omp end loop
|
|
end subroutine
|
|
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_order
|
|
subroutine test_order()
|
|
integer :: i, dummy = 1
|
|
! CHECK: omp.loop order(reproducible:concurrent) private(@{{.*}} %{{.*}}#0 -> %{{.*}} : {{.*}}) {
|
|
! CHECK: }
|
|
!$omp loop order(concurrent)
|
|
do i=1,10
|
|
dummy = dummy + 1
|
|
end do
|
|
!$omp end loop
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_reduction
|
|
subroutine test_reduction()
|
|
integer :: i, dummy = 1
|
|
|
|
! CHECK: omp.loop private(@{{.*}} %{{.*}}#0 -> %{{.*}} : !{{.*}}) reduction
|
|
! CHECK-SAME: (@[[RED]] %{{.*}}#0 -> %[[DUMMY_ARG:.*]] : !{{.*}}) {
|
|
! CHECK-NEXT: omp.loop_nest (%{{.*}}) : i32 = (%{{.*}}) to (%{{.*}}) {{.*}} {
|
|
! CHECK: %[[DUMMY_DECL:.*]]:2 = hlfir.declare %[[DUMMY_ARG]] {uniq_name = "_QFtest_reductionEdummy"}
|
|
! CHECK: %{{.*}} = fir.load %[[DUMMY_DECL]]#0
|
|
! CHECK: hlfir.assign %{{.*}} to %[[DUMMY_DECL]]#0
|
|
! CHECK: }
|
|
! CHECK: }
|
|
!$omp loop reduction(+:dummy)
|
|
do i=1,10
|
|
dummy = dummy + 1
|
|
end do
|
|
!$omp end loop
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_bind
|
|
subroutine test_bind()
|
|
integer :: i, dummy = 1
|
|
! CHECK: omp.simd private(@{{.*}} %{{.*}}#0 -> %{{.*}} : {{.*}}) {
|
|
! CHECK: }
|
|
!$omp loop bind(thread)
|
|
do i=1,10
|
|
dummy = dummy + 1
|
|
end do
|
|
!$omp end loop
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_nested_directives
|
|
subroutine test_nested_directives
|
|
implicit none
|
|
integer, parameter :: N = 100000
|
|
integer a(N), b(N), c(N)
|
|
integer j,i, num, flag;
|
|
num = N
|
|
|
|
! CHECK: omp.teams {
|
|
|
|
! Verify the first `loop` directive was combined with `target teams` into
|
|
! `target teams distribute parallel do`.
|
|
! CHECK: omp.parallel {{.*}} {
|
|
! CHECK: omp.distribute {
|
|
! CHECK: omp.wsloop {
|
|
! CHECK: omp.loop_nest {{.*}} {
|
|
|
|
! Very the second `loop` directive was rewritten to `simd`.
|
|
! CHECK: omp.simd {{.*}} {
|
|
! CHECK: omp.loop_nest {{.*}} {
|
|
! CHECK: }
|
|
! CHECK: }
|
|
|
|
! CHECK: }
|
|
! CHECK: } {omp.composite}
|
|
! CHECK: } {omp.composite}
|
|
! CHECK: } {omp.composite}
|
|
! CHECK: }
|
|
!$omp target teams map(to: a,b) map(from: c)
|
|
!$omp loop
|
|
do j=1,1000
|
|
!$omp loop
|
|
do i=1,N
|
|
c(i) = a(i) * b(i)
|
|
end do
|
|
end do
|
|
!$omp end target teams
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_standalone_bind_teams
|
|
subroutine test_standalone_bind_teams
|
|
implicit none
|
|
integer, parameter :: N = 100000
|
|
integer a(N), b(N), c(N)
|
|
integer j,i, num, flag;
|
|
num = N
|
|
|
|
! CHECK: omp.distribute
|
|
! CHECK-SAME: private(@{{.*}}Ea_private_box_100000xi32 {{[^,]*}},
|
|
! CHECK-SAME: @{{.*}}Ei_private_i32 {{.*}} : {{.*}}) {
|
|
! CHECK: omp.loop_nest {{.*}} {
|
|
! CHECK: }
|
|
! CHECK: }
|
|
!$omp loop bind(teams) private(a)
|
|
do i=1,N
|
|
c(i) = a(i) * b(i)
|
|
end do
|
|
end subroutine
|
|
|
|
! CHECK-LABEL: func.func @_QPtest_standalone_bind_parallel
|
|
subroutine test_standalone_bind_parallel
|
|
implicit none
|
|
integer, parameter :: N = 100000
|
|
integer a(N), b(N), c(N)
|
|
integer j,i, num, flag;
|
|
num = N
|
|
|
|
! CHECK: omp.wsloop
|
|
! CHECK-SAME: private(@{{.*}}Ea_private_box_100000xi32 {{[^,]*}},
|
|
! CHECK-SAME: @{{.*}}Ei_private_i32 {{.*}} : {{.*}}) {
|
|
! CHECK: omp.loop_nest {{.*}} {
|
|
! CHECK: }
|
|
! CHECK: }
|
|
!$omp loop bind(parallel) private(a)
|
|
do i=1,N
|
|
c(i) = a(i) * b(i)
|
|
end do
|
|
end subroutine
|