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 445E63A1AFFB; Fri,  7 Jun 2024 11:44:06 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 445E63A1AFFB
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1717760646;
	bh=0sWUP3NEotBWEav06AuZdkQ34ZJz5/1i2z6845HBFF4=;
	h=From:To:Subject:Date:From;
	b=JmFQKmQGJmBGqerzHWmNvAJg2PZCIWgHy+EI+b6jY4KnBNEYPxHD8j0/Ar1jLTOiZ
	 0wT5C9Tomsiee8d9SwR3c9MDHAz2fT6YBM39uZhsklw+sWwIPvHBn861pijC357ICc
	 vDDn4lnAOlE1QzMFwTGWOa+3syO4/KYehbRhtoO4=
From: "rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/115385] New: Peeling for gaps can be
 optimized more or needs to peel more than one iteration
Date: Fri, 07 Jun 2024 11:44:05 +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: 15.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: rguenth 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: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status
 bug_severity priority component assigned_to reporter target_milestone
Message-ID: <bug-115385-4@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=3D115385

            Bug ID: 115385
           Summary: Peeling for gaps can be optimized more or needs to
                    peel more than one iteration
           Product: gcc
           Version: 15.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: rguenth at gcc dot gnu.org
  Target Milestone: ---

Consider

void __attribute__((noipa)) foo(unsigned char * __restrict x,
                                unsigned char *y, int n)
{
  for (int i =3D 0; i < n; ++i)
    {
      x[16*i+0] =3D y[3*i+0];
      x[16*i+1] =3D y[3*i+1];
      x[16*i+2] =3D y[3*i+2];
      x[16*i+3] =3D y[3*i+0];
      x[16*i+4] =3D y[3*i+1];
      x[16*i+5] =3D y[3*i+2];
      x[16*i+6] =3D y[3*i+0];
      x[16*i+7] =3D y[3*i+1];
      x[16*i+8] =3D y[3*i+2];
      x[16*i+9] =3D y[3*i+0];
      x[16*i+10] =3D y[3*i+1];
      x[16*i+11] =3D y[3*i+2];
      x[16*i+12] =3D y[3*i+0];
      x[16*i+13] =3D y[3*i+1];
      x[16*i+14] =3D y[3*i+2];
      x[16*i+15] =3D y[3*i+0];
    }
}

and

void __attribute__((noipa)) bar(unsigned char * __restrict x,
                                unsigned char *y, int n)
{
  for (int i =3D 0; i < n; ++i)
    {
      x[16*i+0] =3D y[5*i+0];
      x[16*i+1] =3D y[5*i+1];
      x[16*i+2] =3D y[5*i+2];
      x[16*i+3] =3D y[5*i+3];
      x[16*i+4] =3D y[5*i+4];
      x[16*i+5] =3D y[5*i+0];
      x[16*i+6] =3D y[5*i+1];
      x[16*i+7] =3D y[5*i+2];
      x[16*i+8] =3D y[5*i+3];
      x[16*i+9] =3D y[5*i+4];
      x[16*i+10] =3D y[5*i+0];
      x[16*i+11] =3D y[5*i+1];
      x[16*i+12] =3D y[5*i+2];
      x[16*i+13] =3D y[5*i+3];
      x[16*i+14] =3D y[5*i+4];
      x[16*i+15] =3D y[5*i+0];
    }
}

for both loops we currently cannot reduce the access for the load from 'y' =
to
not touch extra elements so we force peeling for gaps.  But in both cases
peeling a single scalar iteration is not sufficient and we get

t.c:5:21: note:   =3D=3D> examining statement: _3 =3D y[_1];
t.c:5:21: missed:   peeling for gaps insufficient for access
t.c:7:20: missed:   not vectorized: relevant stmt not supported: _3 =3D y[_=
1];

we can avoid this excessive peeling for gaps if we narrow the load from 'y'
to the next power-of-two size where then it's always sufficient to just
peel a single scalar iteration.  When the target cannot construct a vector
with those elements we'd have to peel more than one iteration.=