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 9337A3858CDA; Tue, 27 Sep 2022 09:20:31 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 9337A3858CDA
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1664270431;
	bh=ardkfBfPYQIAGclWX0W3EOqCUMzTJVcybtAxpqZwoJU=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=lCeoM0CuvrgHhoEednCWtxDR9d/4/Dkrhr+kskSp4AIgXU+Ofomw6QcrZAeqqRmM+
	 rCoxuGxaxLVJOjbrULXCPZCE6HoYjV1rGsseikR2qnWgZP5Zlwkhtwyv3BafXXfJCl
	 2qU2Vml7xF41FWv5tdnc0q1s/vFivG09Hrem4k3M=
From: "rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/107050] duplicate load of return value when
 facing multiple branches
Date: Tue, 27 Sep 2022 09:20:31 +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: 12.1.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: rguenth at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: everconfirmed cc cf_reconfirmed_on
 cf_known_to_fail cf_gcctarget keywords bug_status
Message-ID: <bug-107050-4-aNezBMgIHp@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107050-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107050-4@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
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D107050

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
                 CC|                            |segher at gcc dot gnu.org
   Last reconfirmed|                            |2022-09-27
      Known to fail|                            |13.0
             Target|                            |x86_64-*-*
           Keywords|                            |missed-optimization
             Status|UNCONFIRMED                 |NEW

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.  It's shrink-wrapping which is duplicating blocks and the load.=
=20
Before we have

    1: NOTE_INSN_DELETED
    5: NOTE_INSN_BASIC_BLOCK 2
    2: ax:SI=3Ddi:SI
    4: NOTE_INSN_FUNCTION_BEG
    7: flags:CCZ=3Dcmp(ax:SI,0)
    8: pc=3D{(flags:CCZ=3D=3D0)?L27:pc}
      REG_BR_PROB 536870916
    9: NOTE_INSN_BASIC_BLOCK 3
   10: ax:SI=3D[`f']
   11: flags:CCZ=3Dcmp(ax:SI,0)
   12: pc=3D{(flags:CCZ=3D=3D0)?L24:pc}
      REG_BR_PROB 708669604
   13: NOTE_INSN_BASIC_BLOCK 4
   14: flags:CCZ=3Dcmp(si:SI,0)
   15: pc=3D{(flags:CCZ=3D=3D0)?L19:pc}
      REG_BR_PROB 719407028
   16: NOTE_INSN_BASIC_BLOCK 5
   17: ax:QI=3D0
   18: call [`func_58'] argc:0
      REG_EH_REGION 0
   19: L19:
   20: NOTE_INSN_BASIC_BLOCK 6
   21: ax:SI=3D[`g_286']
   38: pc=3DL27
   39: barrier
   24: L24:
   25: NOTE_INSN_BASIC_BLOCK 7
   26: [`g_286']=3D0
   27: L27:
   28: NOTE_INSN_BASIC_BLOCK 8
   34: use ax:SI
   40: NOTE_INSN_DELETED

maybe shrink-wrapping should consider splitting blocks before doing the
transform?=