Am 08.12.2011 22:57, schrieb Jakub Jelinek:
Reading it again, isn't it overkill to keep the
vector? All you need is a bool and a way to restore its previous state.
Good catch. The vector was a leftover from the time when I was searching
up the call chain to check for any enclosing workshare.
Updated patch attached. I put a goto in there to simulate a double
"fall through", as you'll see in the patch. If anybody has strong
negative feelings about this, I will replace it with a switch within a
switch.
Regression-tested. OK for trunk (finally)?
Thomas
2011-12-11 Thomas Koenig <[email protected]>
PR fortran/50690
* frontend-passes.c (in_omp_workshare): New variable.
(cfe_expr_0): Don't eliminiate common function if it would put
the variable immediately into a WORKSHARE construct.
(optimize_namespace): Set in_omp_workshare.
(gfc_code_walker): Keep track of OMP PARALLEL and OMP WORKSHARE
constructs.
2011-12-11 Thomas Koenig <[email protected]>
PR fortran/50690
* gfortran.dg/gomp/workshare2.f90: New test.
* gfortran.dg/gomp/workshare3.f90: New test.
! { dg-do compile }
! { dg-options "-ffrontend-optimize -fdump-tree-original" }
! Test that common function elimination is done within the OMP parallel
! blocks even if there is a workshare around it.
program foo
implicit none
integer, parameter :: n = 10000000
real, parameter :: eps = 3e-7
integer :: i,j
real :: A(n), B(5), C(n)
real :: tmp
B(1) = 3.344
tmp = B(1)
do i=1,10
call random_number(a)
c = a
!$omp parallel workshare
!$omp parallel default(shared)
!$omp do
do j=1,n
A(j) = A(j)*cos(B(1))+A(j)*cos(B(1))
end do
!$omp end do
!$omp end parallel
!$omp end parallel workshare
end do
c = c*cos(b(1))+ c*cos(b(1))
do j=1,n
if (abs(a(j)-c(j)) > eps) then
print *,1,j,a(j), c(j)
call abort
end if
end do
end program foo
! { dg-final { scan-tree-dump-times "__builtin_cosf" 2 "original" } }
! { dg-final { cleanup-tree-dump "original" } }
! { dg-do compile }
! { dg-options "-ffrontend-optimize -fdump-tree-original" }
! PR 50690 - this used to ICE because workshare could not handle
! BLOCKs.
! To test for correct execution, run this program (but don't forget
! to unset the stack limit).
program foo
implicit none
integer, parameter :: n = 10000000
real, parameter :: eps = 3e-7
integer :: i,j
real :: A(n), B(5), C(n)
real :: tmp
B(1) = 3.344
tmp = B(1)
do i=1,10
call random_number(a)
c = a
!$omp parallel default(shared)
!$omp workshare
A(:) = A(:)*cos(B(1))+A(:)*cos(B(1))
!$omp end workshare nowait
!$omp end parallel ! sync is implied here
end do
c = c*tmp + c*tmp
do j=1,n
if (abs(a(j)-c(j)) > eps) then
print *,1,j,a(j), c(j)
call abort
end if
end do
do i=1,10
call random_number(a)
c = a
!$omp parallel workshare default(shared)
A(:) = A(:)*cos(B(1))+A(:)*cos(B(1))
!$omp end parallel workshare
end do
c = c*tmp + c*tmp
do j=1,n
if (abs(a(j)-c(j)) > eps) then
print *,2,j,a(j), c(j)
call abort
end if
end do
end program foo
! { dg-final { scan-tree-dump-times "__var" 0 "original" } }
! { dg-final { cleanup-tree-dump "original" } }
Index: frontend-passes.c
===================================================================
--- frontend-passes.c (Revision 181809)
+++ frontend-passes.c (Arbeitskopie)
@@ -66,6 +66,10 @@ static gfc_namespace *current_ns;
static int forall_level;
+/* Keep track of whether we are within an OMP workshare. */
+
+static bool in_omp_workshare;
+
/* Entry point - run all passes for a namespace. So far, only an
optimization pass is run. */
@@ -367,6 +371,14 @@ cfe_expr_0 (gfc_expr **e, int *walk_subtrees,
int i,j;
gfc_expr *newvar;
+ /* Don't do this optimization within OMP workshare. */
+
+ if (in_omp_workshare)
+ {
+ *walk_subtrees = 0;
+ return 0;
+ }
+
expr_count = 0;
gfc_expr_walker (e, cfe_register_funcs, NULL);
@@ -505,6 +517,7 @@ optimize_namespace (gfc_namespace *ns)
current_ns = ns;
forall_level = 0;
+ in_omp_workshare = false;
gfc_code_walker (&ns->code, convert_do_while, dummy_expr_callback, NULL);
gfc_code_walker (&ns->code, cfe_code, cfe_expr_0, NULL);
@@ -1150,11 +1163,13 @@ gfc_code_walker (gfc_code **c, walk_code_fn_t code
gfc_actual_arglist *a;
gfc_code *co;
gfc_association_list *alist;
+ bool saved_in_omp_workshare;
/* There might be statement insertions before the current code,
which must not affect the expression walker. */
co = *c;
+ saved_in_omp_workshare = in_omp_workshare;
switch (co->op)
{
@@ -1330,16 +1345,38 @@ gfc_code_walker (gfc_code **c, walk_code_fn_t code
WALK_SUBEXPR (co->ext.dt->extra_comma);
break;
- case EXEC_OMP_DO:
+ in_omp_workshare = true;
+
+ break;
+
case EXEC_OMP_PARALLEL:
case EXEC_OMP_PARALLEL_DO:
case EXEC_OMP_PARALLEL_SECTIONS:
+
+ in_omp_workshare = false;
+
+ /* This goto serves as a shortcut to avoid code
+ duplication or a larger if or switch statement. */
+ goto check_omp_clauses;
+
+ case EXEC_OMP_WORKSHARE:
case EXEC_OMP_PARALLEL_WORKSHARE:
+
+ in_omp_workshare = true;
+
+ /* Fall through */
+
+ case EXEC_OMP_DO:
case EXEC_OMP_SECTIONS:
case EXEC_OMP_SINGLE:
- case EXEC_OMP_WORKSHARE:
case EXEC_OMP_END_SINGLE:
case EXEC_OMP_TASK:
+
+ /* Come to this label only from the
+ EXEC_OMP_PARALLEL_* cases above. */
+
+ check_omp_clauses:
+
if (co->ext.omp_clauses)
{
WALK_SUBEXPR (co->ext.omp_clauses->if_expr);
@@ -1366,6 +1403,7 @@ gfc_code_walker (gfc_code **c, walk_code_fn_t code
if (co->op == EXEC_FORALL)
forall_level --;
+ in_omp_workshare = saved_in_omp_workshare;
}
}
return 0;
