[Bug testsuite/108810] gcc.target/powerpc/fold-vec-extract-double.p9.c fails on power 9 BE

["linkw at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108810

Kewen Lin <linkw at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2023-02-17
             Status|UNCONFIRMED                 |NEW
          Component|target                      |testsuite
                 CC|                            |linkw at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #1 from Kewen Lin <linkw at gcc dot gnu.org> ---
This is an test case issue:

double
testd_cst (vector double vd2)
{
  return vec_extract (vd2, 1);
}

got xxlor on LE but xxpermdi on BE.

The scan insn xxlor is for the high order element, the index is 0 on BE and 1
on LE.

The below diff can fix it:

diff --git a/gcc/testsuite/gcc.target/powerpc/fold-vec-extract-double.p9.c
b/gcc/testsuite/gcc.target/powerpc/fold-vec-extract-double.p9.c
index 6c515035d1a..100f680fd02 100644
--- a/gcc/testsuite/gcc.target/powerpc/fold-vec-extract-double.p9.c
+++ b/gcc/testsuite/gcc.target/powerpc/fold-vec-extract-double.p9.c
@@ -18,9 +18,15 @@ testd_var (vector double vd2, signed int si)
   return vec_extract (vd2, si);
 }

+#ifdef __BIG_ENDIAN__
+#define HIGH_ORDER_ELEMENT_INDEX 0
+#else
+#define HIGH_ORDER_ELEMENT_INDEX 1
+#endif
+
 double
 testd_cst (vector double vd2)
 {
-  return vec_extract (vd2, 1);
+  return vec_extract (vd2, HIGH_ORDER_ELEMENT_INDEX);
 }