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 A4DA7382EF1F; Mon, 14 Nov 2022 20:08:57 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org A4DA7382EF1F
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1668456537;
	bh=2UcqUX0oh5ml+kNAIc7fd6BrE6BOjPuHKCiY398DhMg=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=jfr6nI1y5wNQJEYN5iOPPsCr7jp3Ax9YecITUoAsq6nX9y5yDHxCwtdptUi5Y7Co/
	 mfM+vhjmhFPdFNEKReN4XRv4jrC3zCYdaBb8Eab4AglKkidttXY9SGZ/bJ/MgjkFbU
	 FQpxWrst4efsfGVbvODx9TUiuckV9yf6kgZqeRJs=
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/107690] [12/13 Regression] vectorization fails
 for std::ranges::transform due to IR changes
Date: Mon, 14 Nov 2022 20:08:57 +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: 13.0
X-Bugzilla-Keywords: missed-optimization, needs-bisection
X-Bugzilla-Severity: normal
X-Bugzilla-Who: pinskia 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: 12.3
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: target_milestone short_desc cf_reconfirmed_on
 blocked everconfirmed bug_status keywords
Message-ID: <bug-107690-4-1r92xNirIP@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107690-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107690-4@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=3D107690

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.3
            Summary|Regression in               |[12/13 Regression]
                   |ranges::transform           |vectorization fails for
                   |vectorization               |std::ranges::transform due
                   |                            |to IR changes
   Last reconfirmed|                            |2022-11-14
             Blocks|                            |53947
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |missed-optimization,
                   |                            |needs-bisection

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
In GCC 12 before the vectorizer we have:
  <bb 2> [local count: 114863530]:
  _4 =3D v_2(D) + 64;
  _5 =3D &v_2(D)->_M_elems;
  if (_4 !=3D _5)
    goto <bb 5>; [89.30%]
  else
    goto <bb 4>; [10.70%]

  <bb 5> [local count: 102576004]:

  <bb 3> [local count: 958878296]:
  # __first1_24 =3D PHI <_11(6), &u._M_elems(5)>
  # __first2_25 =3D PHI <_12(6), _5(5)>
  _7 =3D MEM[(const int &)__first1_24];
  _9 =3D *__first2_25;
  _10 =3D _7 + _9;
  *__first1_24 =3D _10;
  _11 =3D __first1_24 + 4;
  _12 =3D __first2_25 + 4;
  _15 =3D _4 !=3D _12;
  _18 =3D &MEM <struct array> [(void *)&u + 64B] !=3D _11;
  _16 =3D _15 & _18;
  if (_16 !=3D 0)
    goto <bb 6>; [89.30%]
  else
    goto <bb 4>; [10.70%]

  <bb 6> [local count: 856302294]:
  goto <bb 3>; [100.00%]


But with GCC 11 we had:

  <bb 2> [local count: 114863530]:
  _2 =3D &MEM <const int[16]> [(void *)v_3(D) + 64B];
  _5 =3D &v_3(D)->_M_elems;
  goto <bb 5>; [100.00%]

  <bb 4> [local count: 114863532]:
  <retval> =3D u;
  return <retval>;

  <bb 6> [local count: 899822495]:

  <bb 5> [local count: 1014686026]:
  # __first1_22 =3D PHI <_11(6), &u._M_elems(2)>
  # __first2_23 =3D PHI <_12(6), _5(2)>
  # ivtmp_24 =3D PHI <ivtmp_13(6), 16(2)>
  _7 =3D MEM[(const int &)__first1_22];
  _9 =3D *__first2_23;
  _10 =3D _7 + _9;
  *__first1_22 =3D _10;
  _11 =3D __first1_22 + 4;
  _12 =3D __first2_23 + 4;
  ivtmp_13 =3D ivtmp_24 - 1;
  if (ivtmp_13 !=3D 0)
    goto <bb 6>; [93.84%]
  else
    goto <bb 4>; [6.16%]

There is a missing optimization before the vectorizer which is causing the
vectorizer not to know how many iterations the loop is for.

I am tries tracking down which passes the IR changes to make things worse b=
ut I
didn't do a good at doing that.


Referenced Bugs:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D53947
[Bug 53947] [meta-bug] vectorizer missed-optimizations=