gcc/libgomp/testsuite/libgomp.c++/scan-12.C
Jakub Jelinek 0ad7981cb4 re PR libgomp/91530 (Several libgomp.*/scan-* tests FAIL without avx_runtime)
PR libgomp/91530
	* testsuite/libgomp.c/scan-11.c: Add -msse2 option for sse2_runtime
	targets.
	* testsuite/libgomp.c/scan-12.c: Likewise.
	* testsuite/libgomp.c/scan-13.c: Likewise.
	* testsuite/libgomp.c/scan-14.c: Likewise.
	* testsuite/libgomp.c/scan-15.c: Likewise.
	* testsuite/libgomp.c/scan-16.c: Likewise.
	* testsuite/libgomp.c/scan-17.c: Likewise.
	* testsuite/libgomp.c/scan-18.c: Likewise.
	* testsuite/libgomp.c/scan-19.c: Likewise.
	* testsuite/libgomp.c/scan-20.c: Likewise.
	* testsuite/libgomp.c++/scan-9.C: Likewise.
	* testsuite/libgomp.c++/scan-10.C: Likewise.
	* testsuite/libgomp.c++/scan-11.C: Likewise.
	* testsuite/libgomp.c++/scan-12.C: Likewise.
	* testsuite/libgomp.c++/scan-14.C: Likewise.
	* testsuite/libgomp.c++/scan-15.C: Likewise.
	* testsuite/libgomp.c++/scan-13.C: Likewise.  Use sse2_runtime
	instead of i?86-*-* x86_64-*-* as target for scan-tree-dump-times.
	* testsuite/libgomp.c++/scan-16.C: Likewise.

From-SVN: r274947
2019-08-27 12:45:55 +02:00

155 lines
2.6 KiB
C

// { dg-require-effective-target size32plus }
// { dg-additional-options "-O2 -fopenmp -fdump-tree-vect-details" }
// { dg-additional-options "-msse2" { target sse2_runtime } }
// { dg-additional-options "-mavx" { target avx_runtime } }
// { dg-final { scan-tree-dump-times "vectorized \[2-6] loops" 2 "vect" { target sse2_runtime } } }
extern "C" void abort ();
struct S {
inline S ();
inline ~S ();
inline S (const S &);
inline S & operator= (const S &);
int s;
};
S::S () : s (0)
{
}
S::~S ()
{
}
S::S (const S &x)
{
s = x.s;
}
S &
S::operator= (const S &x)
{
s = x.s;
return *this;
}
static inline void
ini (S &x)
{
x.s = 0;
}
S r, a[1024], b[1024];
#pragma omp declare reduction (+: S: omp_out.s += omp_in.s)
#pragma omp declare reduction (plus: S: omp_out.s += omp_in.s) initializer (ini (omp_priv))
__attribute__((noipa)) void
foo (S *a, S *b, S &r)
{
#pragma omp for simd reduction (inscan, +:r)
for (int i = 0; i < 1024; i++)
{
r.s += a[i].s;
#pragma omp scan inclusive(r)
b[i] = r;
}
}
__attribute__((noipa)) S
bar ()
{
S s;
#pragma omp parallel
#pragma omp for simd reduction (inscan, plus:s)
for (int i = 0; i < 1024; i++)
{
s.s += 2 * a[i].s;
#pragma omp scan inclusive(s)
b[i] = s;
}
return s;
}
__attribute__((noipa)) void
baz (S *a, S *b, S &r)
{
#pragma omp parallel for simd if (simd: 0) reduction (inscan, +:r)
for (int i = 0; i < 1024; i++)
{
r.s += a[i].s;
#pragma omp scan inclusive(r)
b[i] = r;
}
}
__attribute__((noipa)) S
qux ()
{
S s;
#pragma omp parallel for simd reduction (inscan, plus:s) simdlen(1)
for (int i = 0; i < 1024; i++)
{
s.s += 2 * a[i].s;
#pragma omp scan inclusive(s)
b[i] = s;
}
return s;
}
int
main ()
{
S s;
for (int i = 0; i < 1024; ++i)
{
a[i].s = i;
b[i].s = -1;
asm ("" : "+g" (i));
}
#pragma omp parallel
foo (a, b, r);
if (r.s != 1024 * 1023 / 2)
abort ();
for (int i = 0; i < 1024; ++i)
{
s.s += i;
if (b[i].s != s.s)
abort ();
else
b[i].s = 25;
}
if (bar ().s != 1024 * 1023)
abort ();
s.s = 0;
for (int i = 0; i < 1024; ++i)
{
s.s += 2 * i;
if (b[i].s != s.s)
abort ();
}
r.s = 0;
baz (a, b, r);
if (r.s != 1024 * 1023 / 2)
abort ();
s.s = 0;
for (int i = 0; i < 1024; ++i)
{
s.s += i;
if (b[i].s != s.s)
abort ();
else
b[i].s = 25;
}
if (qux ().s != 1024 * 1023)
abort ();
s.s = 0;
for (int i = 0; i < 1024; ++i)
{
s.s += 2 * i;
if (b[i].s != s.s)
abort ();
}
}