On Wed, 15 Mar 2023, Alexandre Oliva wrote: > > This test is similar to pr103116-1.c, but instead of writing to > 4*COUNT elements of x, it writes to 8*COUNT elements, but the > definition of x seems to have been adjusted along with the loop. Fix > the array size so that it doesn't scribble over unrelated > statically-allocated objects. > > Regstrapped on ppc64-linux-gnu. Also tested with gcc-11 on vxworks7r2 > (x86- and x86_64-), where the scribbling caused visible runtime effects. > Ok to install? I'm tempted to put this in as obvious.
OK (it's really obvious). Richard. > > for gcc/testsuite/ChangeLog > > * gcc.dg/vect/pr103116-2.c (x): Fix array size. > --- > gcc/testsuite/gcc.dg/vect/pr103116-2.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/gcc/testsuite/gcc.dg/vect/pr103116-2.c > b/gcc/testsuite/gcc.dg/vect/pr103116-2.c > index 2f4ed0f404c76..aa9797a94074c 100644 > --- a/gcc/testsuite/gcc.dg/vect/pr103116-2.c > +++ b/gcc/testsuite/gcc.dg/vect/pr103116-2.c > @@ -31,7 +31,7 @@ loop (TYPE *restrict x, TYPE *restrict y) > } > } > > -TYPE x[COUNT * 4]; > +TYPE x[COUNT * 8]; > > int > main (void) > > -- Richard Biener <rguent...@suse.de> SUSE Software Solutions Germany GmbH, Frankenstrasse 146, 90461 Nuernberg, Germany; GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman; HRB 36809 (AG Nuernberg)