public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "luoxhu at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/106126] [12 Regression] tree check fail in useless_type_conversion_p, at gimple-expr.cc:87 since r13-1184-g57424087e82db140
Date: Thu, 30 Jun 2022 03:34:51 +0000	[thread overview]
Message-ID: <bug-106126-4-b9eePvtFix@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-106126-4@http.gcc.gnu.org/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106126

--- Comment #13 from luoxhu at gcc dot gnu.org ---
Otherwise we need record first_bb when conditions_in_bbs->is_empty, then check
that in is_beneficial, ordered_remove the info entry if that bb is not the
first "if condition" with side_effect statement in it, the fix would be as
below, but I am not sure whether it is worth&correct way doing this to handle
both PR105740 and PR106126?


git diff
diff --git a/gcc/gimple-if-to-switch.cc b/gcc/gimple-if-to-switch.cc
index f7b0b02628b..44bb0228856 100644
--- a/gcc/gimple-if-to-switch.cc
+++ b/gcc/gimple-if-to-switch.cc
@@ -63,7 +63,7 @@ struct condition_info

   condition_info (gcond *cond): m_cond (cond), m_bb (gimple_bb (cond)),
     m_forwarder_bb (NULL), m_ranges (), m_true_edge (NULL), m_false_edge
(NULL),
-    m_true_edge_phi_mapping (), m_false_edge_phi_mapping ()
+    m_true_edge_phi_mapping (), m_false_edge_phi_mapping (), first_bb(false)
   {
     m_ranges.create (0);
   }
@@ -80,6 +80,7 @@ struct condition_info
   edge m_false_edge;
   mapping_vec m_true_edge_phi_mapping;
   mapping_vec m_false_edge_phi_mapping;
+  bool first_bb;
 };

 /* Recond PHI mapping for an original edge E and save these into vector VEC. 
*/
@@ -194,6 +195,16 @@ if_chain::is_beneficial ()
   auto_vec<cluster *> clusters;
   clusters.create (m_entries.length ());

+  for (unsigned i = 0; i < m_entries.length (); i++)
+    {
+      condition_info *info = m_entries[i];
+      if (info->first_bb && i != 0 && !no_side_effect_bb (info->m_bb))
+       {
+         m_entries.ordered_remove (i);
+         break;
+       }
+    }
+
   for (unsigned i = 0; i < m_entries.length (); i++)
     {
       condition_info *info = m_entries[i];
@@ -397,6 +408,8 @@ find_conditions (basic_block bb,
   tree_code code = gimple_cond_code (cond);

   condition_info *info = new condition_info (cond);
+  if (conditions_in_bbs->is_empty ())
+    info->first_bb = true;

  parent reply	other threads:[~2022-06-30  3:34 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-28 17:46 [Bug c/106126] New: tree check fail in useless_type_conversion_p, at gimple-expr.cc:87 dcb314 at hotmail dot com
2022-06-28 17:51 ` [Bug tree-optimization/106126] " dcb314 at hotmail dot com
2022-06-28 17:59 ` dcb314 at hotmail dot com
2022-06-28 20:06 ` dcb314 at hotmail dot com
2022-06-28 20:46 ` dcb314 at hotmail dot com
2022-06-28 21:33 ` dcb314 at hotmail dot com
2022-06-29  9:47 ` rguenth at gcc dot gnu.org
2022-06-29 11:45 ` dcb314 at hotmail dot com
2022-06-29 11:53 ` dcb314 at hotmail dot com
2022-06-29 11:59 ` [Bug tree-optimization/106126] [12 Regression] tree check fail in useless_type_conversion_p, at gimple-expr.cc:87 since r13-1184-g57424087e82db140 marxin at gcc dot gnu.org
2022-06-29 12:00 ` marxin at gcc dot gnu.org
2022-06-30  2:21 ` luoxhu at gcc dot gnu.org
2022-06-30  2:37 ` luoxhu at gcc dot gnu.org
2022-06-30  3:34 ` luoxhu at gcc dot gnu.org [this message]
2022-06-30 13:03 ` marxin at gcc dot gnu.org
2022-06-30 19:52 ` dcb314 at hotmail dot com
2022-07-01  6:48 ` cvs-commit at gcc dot gnu.org
2022-07-01  6:48 ` [Bug tree-optimization/106126] [13 " marxin at gcc dot gnu.org
2022-07-01  6:54 ` marxin at gcc dot gnu.org
2022-07-01  7:16 ` dcb314 at hotmail dot com
2022-07-01 17:48 ` dcb314 at hotmail dot com
2022-07-04 11:33 ` marxin at gcc dot gnu.org

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bug-106126-4-b9eePvtFix@http.gcc.gnu.org/bugzilla/ \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@gcc.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).