This adds more guards to the VEC_PERM_EXPR scan, namely that
we also could end up with load-lanes and of course no vectorization
at all. Need dependent scans (scan-if-scan-X PASSed ...).
Pushed.
2021-01-21 Richard Biener <rguent...@suse.de>
PR testsuite/97299
* gcc.dg/vect/slp-reduc-3.c: Amend target selectors.
---
gcc/testsuite/gcc.dg/vect/slp-reduc-3.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/gcc/testsuite/gcc.dg/vect/slp-reduc-3.c
b/gcc/testsuite/gcc.dg/vect/slp-reduc-3.c
index fc875865208..9e29757717f 100644
--- a/gcc/testsuite/gcc.dg/vect/slp-reduc-3.c
+++ b/gcc/testsuite/gcc.dg/vect/slp-reduc-3.c
@@ -46,6 +46,7 @@ int main (void)
for (i=0; i<N; i++) {
X[i] = i;
Y[i] = 64-i;
+ asm volatile ("" ::: "memory");
}
foo1 (N/2, &dot1, &dot2);
@@ -58,7 +59,7 @@ int main (void)
/* The initialization loop in main also gets vectorized. */
/* { dg-final { scan-tree-dump-times "vect_recog_dot_prod_pattern: detected" 1
"vect" { xfail *-*-* } } } */
-/* { dg-final { scan-tree-dump-times "vectorized 1 loops" 2 "vect" { target {
vect_short_mult && { vect_widen_sum_hi_to_si && vect_unpack } } } } } */
-/* { dg-final { scan-tree-dump-times "vectorizing stmts using SLP" 1 "vect" {
xfail { vect_widen_sum_hi_to_si_pattern || { ! vect_unpack } } } } } */
+/* { dg-final { scan-tree-dump-times "vectorized 1 loops" 1 "vect" { target {
vect_short_mult && { vect_widen_sum_hi_to_si && vect_unpack } } } } } */
+/* { dg-final { scan-tree-dump-times "vectorizing stmts using SLP" 1 "vect" {
xfail { vect_widen_sum_hi_to_si_pattern || { ! { vect_short_mult && {
vect_widen_sum_hi_to_si && vect_unpack } } } } } } } */
/* Check we can elide permutes if SLP vectorizing the reduction. */
-/* { dg-final { scan-tree-dump-times "VEC_PERM_EXPR" 0 "vect" { xfail {
vect_widen_sum_hi_to_si_pattern || { ! vect_unpack } } } } } */
+/* { dg-final { scan-tree-dump-times "VEC_PERM_EXPR" 0 "vect" { xfail { { {
vect_widen_sum_hi_to_si_pattern || { ! vect_unpack } } && { ! vect_load_lanes }
} && { vect_short_mult && { vect_widen_sum_hi_to_si && vect_unpack } } } } } }
*/
--
2.26.2
