obtained gcc-4.6.4.tar.bz2 from upstream website;upstream

verified gcc-4.6.4.tar.bz2.sig;
imported gcc-4.6.4 source tree from verified upstream tarball.

downloading a git-generated archive based on the 'upstream' tag
should provide you with a source tree that is binary identical
to the one extracted from the above tarball.

if you have obtained the source via the command 'git clone',
however, do note that line-endings of files in your working
directory might differ from line-endings of the respective
files in the upstream repository.

1 files changed, 52 insertions, 0 deletions

diff --git a/gcc/testsuite/gfortran.dg/argument_checking_9.f90 b/gcc/testsuite/gfortran.dg/argument_checking_9.f90
new file mode 100644
index 000000000..fd7dde33d
--- /dev/null
+++ b/gcc/testsuite/gfortran.dg/argument_checking_9.f90
@@ -0,0 +1,52 @@
+! { dg-do compile }
+! { dg-options "-fmax-errors=40" }
+! PR33162 INTRINSIC functions as ACTUAL argument
+! Prepared by Jerry DeLisle <jvdelisle@gcc.gnu.org>
+program double_specs
+
+real(kind=4) :: rr, x, y
+real(kind=8) :: dr, dx, dy
+
+x = .5
+y = .7
+dx = .5d0
+dy = .5d0
+
+r = dabs(x) ! { dg-error "must be double precision" }
+r = dacos(x) ! { dg-error "must be double precision" }
+r = dacosh(x) ! { dg-error "must be double precision" }
+r = dasin(x) ! { dg-error "must be double precision" }
+r = dasinh(x) ! { dg-error "must be double precision" }
+r = datan(x) ! { dg-error "must be double precision" }
+r = datanh(x) ! { dg-error "must be double precision" }
+r = datan2(y, dx) ! { dg-error "must be double precision" }
+r = datan2(dy, x) ! { dg-error "must be double precision" }
+r = dbesj0(x) ! { dg-error "must be double precision" }
+r = dbesj1(x) ! { dg-error "must be double precision" }
+r = dbesy0(x) ! { dg-error "must be double precision" }
+r = dbesy1(x) ! { dg-error "must be double precision" }
+r = dcos(x) ! { dg-error "must be double precision" }
+r = dcosh(x) ! { dg-error "must be double precision" }
+r = ddim(x, dy) ! { dg-error "must be double precision" }
+r = ddim(dx, y) ! { dg-error "must be double precision" }
+r = derf(x) ! { dg-error "must be double precision" }
+r = derfc(x) ! { dg-error "must be double precision" }
+r = dexp(x) ! { dg-error "must be double precision" }
+r = dgamma(x) ! { dg-error "must be double precision" }
+r = dlgama(x) ! { dg-error "must be double precision" }
+r = dlog(x) ! { dg-error "must be double precision" }
+r = dlog10(x) ! { dg-error "must be double precision" }
+r = dmod(x, dy) ! { dg-error "must be double precision" }
+r = dmod(dx, y) ! { dg-error "must be double precision" }
+r = dsign(x, dy) ! { dg-error "must be double precision" }
+r = dsign(dx, y) ! { dg-error "must be double precision" }
+r = dsin(x) ! { dg-error "must be double precision" }
+r = dsinh(x) ! { dg-error "must be double precision" }
+r = dsqrt(x) ! { dg-error "must be double precision" }
+r = dtan(x) ! { dg-error "must be double precision" }
+r = dtanh(x) ! { dg-error "must be double precision" }
+dr = dprod(dx,y) ! { dg-error "must be default real" }
+dr = dprod(x,dy) ! { dg-error "must be default real" }
+dr = dprod(x,y)
+
+end program double_specs
\ No newline at end of file