diff options
| author | Jan Hubicka <jh@suse.cz> | 2023-07-16 23:55:14 +0200 |
|---|---|---|
| committer | Jan Hubicka <jh@suse.cz> | 2023-07-16 23:55:14 +0200 |
| commit | c62791fa413a49fc6476ce186b324250f8ae6d40 (patch) | |
| tree | 81610bd4d94290e37ad0952945883d67a1c18383 | |
| parent | 1d203d4c90adb064edfa9680768d1f83a41f17e0 (diff) | |
| download | gcc-c62791fa413a49fc6476ce186b324250f8ae6d40.zip gcc-c62791fa413a49fc6476ce186b324250f8ae6d40.tar.gz gcc-c62791fa413a49fc6476ce186b324250f8ae6d40.tar.bz2 | |
Fix optimize_mask_stores profile update
While looking into sphinx3 regression I noticed that vectorizer produces
BBs with overall probability count 120%. This patch fixes it.
Richi, I don't know how to create a testcase, but having one would
be nice.
Bootstrapped/regtested x86_64-linux, will commit it shortly.
gcc/ChangeLog:
PR tree-optimization/110649
* tree-vect-loop.cc (optimize_mask_stores): Set correctly
probability of the if-then-else construct.
| -rw-r--r-- | gcc/tree-vect-loop.cc | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/gcc/tree-vect-loop.cc b/gcc/tree-vect-loop.cc index 7d917bf..61343ea 100644 --- a/gcc/tree-vect-loop.cc +++ b/gcc/tree-vect-loop.cc @@ -11680,6 +11680,7 @@ optimize_mask_stores (class loop *loop) efalse = make_edge (bb, store_bb, EDGE_FALSE_VALUE); /* Put STORE_BB to likely part. */ efalse->probability = profile_probability::unlikely (); + e->probability = efalse->probability.invert (); store_bb->count = efalse->count (); make_single_succ_edge (store_bb, join_bb, EDGE_FALLTHRU); if (dom_info_available_p (CDI_DOMINATORS)) |