From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugs-return-307576-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 30956 invoked by alias); 4 Feb 2010 23:22:44 -0000
Received: (qmail 30922 invoked by uid 48); 4 Feb 2010 23:22:28 -0000
Date: Thu, 04 Feb 2010 23:22:00 -0000
Subject: [Bug tree-optimization/42963]  New: [4,5 Regression] Redundant switch labels not cleaned up anymore
X-Bugzilla-Reason: CC
Message-ID: <bug-42963-280@http.gcc.gnu.org/bugzilla/>
Reply-To: gcc-bugzilla@gcc.gnu.org
To: gcc-bugs@gcc.gnu.org
From: "steven at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org>
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
X-SW-Source: 2010-02/txt/msg00328.txt.bz2

Test case is taken from bug 18046:
-----------------------------------------------------
extern void foo (void);
extern int i;
void
bar (void)
{
  switch (i)
    {
    case 0:
      foo ();
      break;
    case 1:
      break;
    }


  switch (i)
    {
    case 0:
      foo ();
      break;
    case 1:
      break;
    }
}
-----------------------------------------------------


the .012t.cfg dump with trunk (r156492) today looks like this:

bar ()
{
  int i.0;

<bb 2>:
  i.0 = i;
  switch (i.0) <default: <L2>, case 0: <L0>, case 1: <L2>>

<L0>:
  foo ();

<L2>:
  i.0 = i;
  switch (i.0) <default: <L6>, case 0: <L4>, case 1: <L6>>

<L4>:
  foo ();

<L6>:
  return;

}



The dump with "gcc (Debian 4.3.4-5) 4.3.4" looks better:

bar ()
{
  int i.0;

  # BLOCK 2
  # PRED: ENTRY (fallthru)
  i.0 = i;
  switch (i.0)
    {
      case 0: goto <L0>;
      default : goto <L1>;
    }
  # SUCC: 3 4

  # BLOCK 3
  # PRED: 2
<L0>:;
  foo ();
  # SUCC: 4 (fallthru)

  # BLOCK 4
  # PRED: 2 3 (fallthru)
<L1>:;
  i.0 = i;
  switch (i.0)
    {
      case 0: goto <L4>;
      default : goto <L5>;
    }
  # SUCC: 5 6

  # BLOCK 5
  # PRED: 4
<L4>:;
  foo ();
  # SUCC: 6 (fallthru)

  # BLOCK 6
  # PRED: 4 5 (fallthru)
<L5>:;
  return;
  # SUCC: EXIT

}


Note how GCC 4.5 fails to clean up  the "case 1" label.  The label is retained
all the way through the .137t.nrv dump (at -O2) and only disappears in the
.139t.optimized dump.

This confuses things in a couple of places and costs memory for no reason.


-- 
           Summary: [4,5 Regression] Redundant switch labels not cleaned up
                    anymore
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Keywords: missed-optimization
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: steven at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42963