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 D857339DA014; Wed, 21 Apr 2021 05:48:46 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org D857339DA014
From: "crazylht at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/100173] New: telecom/viterb00data_1 has
 16.92% regression compared O2 -ftree-vectorize -fvect-cost-model=very-cheap
 to O2  on CLX/ICX, 9% regression on znver3
Date: Wed, 21 Apr 2021 05:48:46 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: new
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: 11.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: crazylht at gmail dot com
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: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status
 bug_severity priority component assigned_to reporter cc target_milestone
 cf_gcchost cf_gcctarget attachments.created
Message-ID: <bug-100173-4@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: Wed, 21 Apr 2021 05:48:47 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D100173

            Bug ID: 100173
           Summary: telecom/viterb00data_1 has 16.92% regression compared
                    O2 -ftree-vectorize -fvect-cost-model=3Dvery-cheap to O2
                     on CLX/ICX, 9% regression on znver3
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: crazylht at gmail dot com
                CC: hjl.tools at gmail dot com
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu
            Target: x86_64-*-* i?86-*-*

Created attachment 50647
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=3D50647&action=3Dedit
ACS.cpp

cat testcase

void
__attribute__ ((noipa))
ACS(e_s16 *pBranchMetric)
{
  n_int i;
  e_s16 esMetricIn, esMetric1, esMetric2;

  StatePathMetricData *pIn1 =3D BufPtr[BufSelector];
  StatePathMetricData *pIn2 =3D pIn1 + (1<<5)/2;
  StatePathMetricData *pOut =3D BufPtr[1 - BufSelector];

  BufSelector ^=3D 1;

  for (i =3D 0; i < (1<<5)/2; i++) {

    esMetricIn =3D *pBranchMetric++;

    esMetric1 =3D pIn1->m_esPathMetric - esMetricIn;
    esMetric2 =3D pIn2->m_esPathMetric + esMetricIn;

    if (esMetric1 >=3D esMetric2) {
      pOut->m_esPathMetric =3D esMetric1;
      pOut->m_esState =3D (pIn1->m_esState << 1);
    }
    else {
      pOut->m_esPathMetric =3D esMetric2;
      pOut->m_esState =3D (pIn2->m_esState << 1);
    }
    pOut++;

    esMetric1 =3D pIn1->m_esPathMetric + esMetricIn;
    esMetric2 =3D pIn2->m_esPathMetric - esMetricIn;

    if (esMetric1 >=3DesMetric2) {
      pOut->m_esPathMetric =3DesMetric1;
      pOut->m_esState =3D (pIn1->m_esState << 1) | 1;
    }
    else {
      pOut->m_esPathMetric =3DesMetric2;
      pOut->m_esState =3D (pIn2->m_esState << 1) | 1;
    }
    pOut++;

    pIn1++;
    pIn2++;
  }
}

It is if conditional store replacement plays here, it sinks 2 stores from I=
F_BB
and ELSE_BB to JOIN_BB since they have same address. But failed to vectorize
them with -fvect-cost-model=3Dvery-cheap, and it causes worse IPC for conse=
cutive
stores in JOIN_BB on both ICX and znver3. With -fvect-cost-model=3Dcheap, t=
he
loop can be vectorized and 2.6x faster than O2.

So I think we should either vectorize this loop or not sink conditional sto=
res
when cost model is very-cheap.

and the codes related are here:=20

  /* If either vectorization or if-conversion is disabled then do
     not sink any stores.  */
  if (param_max_stores_to_sink =3D=3D 0
      || (!flag_tree_loop_vectorize && !flag_tree_slp_vectorize)
      || !flag_tree_loop_if_convert)
    return false;=