[gcc r17-3416] improve vect_optimize_slp_pass::containing_loop

https://gcc.gnu.org/g:157cf7898dbdd43d39b2f2a9cf4390f0338d2ad3

commit r17-3416-g157cf7898dbdd43d39b2f2a9cf4390f0338d2ad3
Author: Richard Biener <[email protected]>
Date:   Tue Aug 18 14:21:39 2026 +0200

    improve vect_optimize_slp_pass::containing_loop
    
    This aovids falling back to the loop root node when handling
    a VEC_PERM node.
    
            * tree-vect-slp.cc (vect_optimize_slp_pass::containing_loop):
            For nodes w/o representative use the region entry block
            to indicate the containing loop.

Diff:
---
 gcc/tree-vect-slp.cc | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/gcc/tree-vect-slp.cc b/gcc/tree-vect-slp.cc
index 1cff56f3099a..ab29e8996d20 100644
--- a/gcc/tree-vect-slp.cc
+++ b/gcc/tree-vect-slp.cc
@@ -6823,7 +6823,7 @@ vect_optimize_slp_pass::containing_loop (slp_tree node)
 {
   stmt_vec_info rep = SLP_TREE_REPRESENTATIVE (node);
   if (!rep)
-    return ENTRY_BLOCK_PTR_FOR_FN (cfun)->loop_father;
+    return m_vinfo->bbs[0]->loop_father;
   return gimple_bb (vect_orig_stmt (rep)->stmt)->loop_father;
 }