This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: regression with openmp
- From: Daniel Franke <franke dot daniel at gmail dot com>
- To: fortran at gcc dot gnu dot org
- Date: Fri, 29 Jun 2007 10:10:32 +0200
- Subject: Re: regression with openmp
- Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:from:to:subject:date:user-agent:references:in-reply-to:mime-version:content-type:message-id; b=IhlcGoh0avtI+3aS0xn022mjoiWyoII8STXV+rifUN6dPdhrq2Ho096WBg9uTHBY+mx7cO7N+OmOAP1lkQxA1gTouQQAP/0L+Z/v7TwBM/CWobUH4X0kD7FMuxPC9E2haOG9eur1hN7Yfi0gLarsGFydGvGwI+lNOJPV5u4Kx6Y=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:from:to:subject:date:user-agent:references:in-reply-to:mime-version:content-type:message-id; b=H/VV2GEHfNiCTsQZ9X1w94cGLrU9EOTvN4E95HiRmtOV3N/BwQCEW1uXumyTKrR7z5dxGVDsQWkLSS22f3zXLu6tzhpmXNy8J8MV5VsTalanYgxOOI6MaCPYlkyzJ/9coEDTCjXjEC8EHo1eoFdBjGin9AQ+z0ppwKb0b9JmkCM=
- References: <20070628231033.GA96915@troutmask.apl.washington.edu> <46844C8D.8050900@verizon.net>
On Friday 29 June 2007 02:04:29 Jerry DeLisle wrote:
> Steve Kargl wrote:
> > Daniel,
> >
> > It appears your recent patch [...]
> > has broken testsuite/gfortran.dg/gomp/reduction1.f90.
>
> On closer look it appeared that some of the error messages changed, so
> maybe the dg-error directives just need to be modified.
Jerry is right. I somehow managed to change the error message after regtesting
and forgot to double-check. Will forget my head next.
Attached changes fix it.
Daniel
Index: fortran/openmp.c
===================================================================
--- fortran/openmp.c (revision 126079)
+++ fortran/openmp.c (working copy)
@@ -850,7 +850,7 @@
case OMP_LIST_MULT:
case OMP_LIST_SUB:
if (!gfc_numeric_ts (&n->sym->ts))
- gfc_error ("%c REDUCTION variable '%s' at %L must be of intrinsic type, got %s",
+ gfc_error ("%c REDUCTION variable '%s' at %L must be of numeric type, got %s",
list == OMP_LIST_PLUS ? '+'
: list == OMP_LIST_MULT ? '*' : '-',
n->sym->name, &code->loc,
Index: testsuite/gfortran.dg/gomp/allocatable_components_1.f90
===================================================================
--- testsuite/gfortran.dg/gomp/allocatable_components_1.f90 (revision 126079)
+++ testsuite/gfortran.dg/gomp/allocatable_components_1.f90 (working copy)
@@ -49,7 +49,7 @@
TYPE(t) :: a(10)
INTEGER :: i
- !$omp parallel do reduction(+: a) ! { dg-error "must be of intrinsic type" }
+ !$omp parallel do reduction(+: a) ! { dg-error "must be of numeric type" }
DO i = 1, SIZE(a)
END DO
!$omp end parallel do
Index: testsuite/gfortran.dg/gomp/reduction1.f90
===================================================================
--- testsuite/gfortran.dg/gomp/reduction1.f90 (revision 126079)
+++ testsuite/gfortran.dg/gomp/reduction1.f90 (working copy)
@@ -60,15 +60,15 @@
!$omp end parallel
!$omp parallel reduction (*:ia1) ! { dg-error "Assumed size" }
!$omp end parallel
-!$omp parallel reduction (+:l1) ! { dg-error "is LOGICAL" }
+!$omp parallel reduction (+:l1) ! { dg-error "must be of numeric type, got LOGICAL" }
!$omp end parallel
-!$omp parallel reduction (*:la1) ! { dg-error "is LOGICAL" }
+!$omp parallel reduction (*:la1) ! { dg-error "must be of numeric type, got LOGICAL" }
!$omp end parallel
-!$omp parallel reduction (-:a1) ! { dg-error "is CHARACTER" }
+!$omp parallel reduction (-:a1) ! { dg-error "must be of numeric type, got CHARACTER" }
!$omp end parallel
-!$omp parallel reduction (+:t1) ! { dg-error "is TYPE" }
+!$omp parallel reduction (+:t1) ! { dg-error "must be of numeric type, got TYPE" }
!$omp end parallel
-!$omp parallel reduction (*:ta1) ! { dg-error "is TYPE" }
+!$omp parallel reduction (*:ta1) ! { dg-error "must be of numeric type, got TYPE" }
!$omp end parallel
!$omp parallel reduction (.and.:i3) ! { dg-error "must be LOGICAL" }
!$omp end parallel