diff options
| author | Richard Biener <rguenther@suse.de> | 2021-02-09 10:11:35 +0100 |
|---|---|---|
| committer | Richard Biener <rguenther@suse.de> | 2021-02-09 10:11:35 +0100 |
| commit | 283653f45595f53db486cbc2f1f10091cb45ea3b (patch) | |
| tree | efe55e0054e2b64740d67c450d9b61c47f07f701 /gcc | |
| parent | 57d705da0b98f5d398c4b8f9bd76fe8ad98e13bc (diff) | |
| download | gcc-283653f45595f53db486cbc2f1f10091cb45ea3b.zip gcc-283653f45595f53db486cbc2f1f10091cb45ea3b.tar.gz gcc-283653f45595f53db486cbc2f1f10091cb45ea3b.tar.bz2 | |
tree-optimization/99017 - be more forgiving in BB vect costing
This works around a SLP graph partitioning or cost collecting issue
by being more forgiving in vect_bb_vectorization_profitable_p.
2021-02-09 Richard Biener <rguenther@suse.de>
PR tree-optimization/99017
* tree-vect-slp.c (vect_bb_vectorization_profitable_p): Allow
zero vector cost entries.
Diffstat (limited to 'gcc')
| -rw-r--r-- | gcc/tree-vect-slp.c | 5 |
1 files changed, 2 insertions, 3 deletions
diff --git a/gcc/tree-vect-slp.c b/gcc/tree-vect-slp.c index b9f12c3..ea8a97b 100644 --- a/gcc/tree-vect-slp.c +++ b/gcc/tree-vect-slp.c @@ -4427,7 +4427,8 @@ vect_bb_vectorization_profitable_p (bb_vec_info bb_vinfo, /* Now cost the portions individually. */ unsigned vi = 0; unsigned si = 0; - do + while (si < li_scalar_costs.length () + && vi < li_vector_costs.length ()) { unsigned sl = li_scalar_costs[si].first; unsigned vl = li_vector_costs[vi].first; @@ -4497,8 +4498,6 @@ vect_bb_vectorization_profitable_p (bb_vec_info bb_vinfo, return false; } } - while (si < li_scalar_costs.length () - && vi < li_vector_costs.length ()); if (vi < li_vector_costs.length ()) { if (dump_enabled_p ()) |