From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
 id 3D0623857375; Fri, 22 Apr 2022 11:00:56 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 3D0623857375
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/105338] [12 Regression] Regression: jump or
 cmove generated for pattern (x ? CST : 0)
Date: Fri, 22 Apr 2022 11:00:56 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: rtl-optimization
X-Bugzilla-Version: 12.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: jakub at gcc dot gnu.org
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 12.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-105338-4-dkT1NvQhWo@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-105338-4@http.gcc.gnu.org/bugzilla/>
References: <bug-105338-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-BeenThere: gcc-bugs@gcc.gnu.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gcc-bugs mailing list <gcc-bugs.gcc.gnu.org>
List-Unsubscribe: <https://gcc.gnu.org/mailman/options/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=unsubscribe>
List-Archive: <https://gcc.gnu.org/pipermail/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-request@gcc.gnu.org?subject=help>
List-Subscribe: <https://gcc.gnu.org/mailman/listinfo/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Fri, 22 Apr 2022 11:00:56 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D105338

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
So, if I change passes.def
      NEXT_PASS (pass_sink_code, true /* unsplit edges */);
line
      NEXT_PASS (pass_sink_code, false /* unsplit edges */);
I get back the 11.x code.
Before ce1 pass, in that case say f looks like:
    8: flags:CCZ=3Dcmp(r83:SI,0)
    9: pc=3D{(flags:CCZ!=3D0)?L22:pc}
      REG_DEAD flags:CCZ
      REG_BR_PROB 375809644
   10: NOTE_INSN_BASIC_BLOCK 3
    5: r82:SI=3D0
      REG_DEAD r83:SI
      ; pc falls through to BB 5
   22: L22:
   21: NOTE_INSN_BASIC_BLOCK 4
    4: r82:SI=3D0x5
   15: L15:
   18: NOTE_INSN_BASIC_BLOCK 5
use (r82)
i.e. essentially
  if (tmp83 !=3D 0)
    tmp82 =3D 5;
  else
    tmp82 =3D 0;
But with vanilla trunk, it is instead:
    7: flags:CCZ=3Dcmp(r83:SI,0)
    8: pc=3D{(flags:CCZ=3D=3D0)?L10:pc}
      REG_DEAD flags:CCZ
      REG_BR_PROB 697932188
    9: NOTE_INSN_BASIC_BLOCK 4
    4: r83:SI=3D0x5
   10: L10:
   11: NOTE_INSN_BASIC_BLOCK 5
use (r83)
i.e.
  if (tmp83 !=3D 0)
    tmp83 =3D 5;

Before fwprop1, the code looks roughly the same except for the swapped
branches,
so in C
  if (tmp83 !=3D 0)
    tmp82 =3D 5;
  else
    tmp82 =3D tmp83;
vs.
  if (tmp83 =3D=3D 0)
    tmp82 =3D tmp83;
  else
    tmp82 =3D 5;
I think fwprop1 only works on extended basic blocks and therefore turns the
latter into
  if (tmp83 =3D=3D 0)
    tmp82 =3D 0;
  else
    tmp82 =3D 5;
and not the former.=