diff options
| author | Jan Hubicka <jh@suse.cz> | 2023-07-27 15:58:52 +0200 |
|---|---|---|
| committer | Jan Hubicka <jh@suse.cz> | 2023-07-27 15:58:52 +0200 |
| commit | 5b06b3b8771d0dc06426986be32b8599e97bb451 (patch) | |
| tree | f7590fbe8766ab3e6bf9dd7acc200f21e7fb5277 | |
| parent | 499b8079a6419bb8082de062ec30772296c6700c (diff) | |
| download | gcc-5b06b3b8771d0dc06426986be32b8599e97bb451.zip gcc-5b06b3b8771d0dc06426986be32b8599e97bb451.tar.gz gcc-5b06b3b8771d0dc06426986be32b8599e97bb451.tar.bz2 | |
Fix profile update in tree-ssa-loop-im.cc
This fixes two bugs in tree-ssa-loop-im.cc. First is that cap probability is not
reliable, but it is constructed with adjusted quality. Second is that sometimes
the conditional has wrong joiner BB count. This is visible on
testsuite/gcc.dg/pr102385.c however the testcase triggers another profile
update bug in pcom, so I will update it in followup patch.
gcc/ChangeLog:
* tree-ssa-loop-im.cc (execute_sm_if_changed): Turn cap probability
to guessed; fix count of new_bb.
| -rw-r--r-- | gcc/tree-ssa-loop-im.cc | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/gcc/tree-ssa-loop-im.cc b/gcc/tree-ssa-loop-im.cc index f5b01e9..268f466 100644 --- a/gcc/tree-ssa-loop-im.cc +++ b/gcc/tree-ssa-loop-im.cc @@ -2059,7 +2059,8 @@ execute_sm_if_changed (edge ex, tree mem, tree tmp_var, tree flag, nbbs++; } - profile_probability cap = profile_probability::always ().apply_scale (2, 3); + profile_probability cap + = profile_probability::guessed_always ().apply_scale (2, 3); if (flag_probability.initialized_p ()) ; @@ -2103,6 +2104,8 @@ execute_sm_if_changed (edge ex, tree mem, tree tmp_var, tree flag, old_dest = ex->dest; new_bb = split_edge (ex); + if (append_cond_position) + new_bb->count += last_cond_fallthru->count (); then_bb = create_empty_bb (new_bb); then_bb->count = new_bb->count.apply_probability (flag_probability); if (irr) |