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 8C9F23857C6B; Thu, 13 Jan 2022 10:22:51 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 8C9F23857C6B
From: "crazylht at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/103771] [12 Regression] Missed vectorization under
 -mavx512f -mavx512vl after r12-5489
Date: Thu, 13 Jan 2022 10:22:51 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 12.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: crazylht at gmail dot com
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: 12.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-103771-4-gk8kNZUZtQ@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-103771-4@http.gcc.gnu.org/bugzilla/>
References: <bug-103771-4@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: Thu, 13 Jan 2022 10:22:51 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D103771

--- Comment #10 from Hongtao.liu <crazylht at gmail dot com> ---
with
@@ -12120,7 +12120,8 @@ supportable_narrowing_operation (enum tree_code cod=
e,
       c1 =3D VEC_PACK_TRUNC_EXPR;
       if (VECTOR_BOOLEAN_TYPE_P (narrow_vectype)
          && VECTOR_BOOLEAN_TYPE_P (vectype)
-         && TYPE_MODE (narrow_vectype) =3D=3D TYPE_MODE (vectype)
+         && (TYPE_MODE (narrow_vectype) =3D=3D TYPE_MODE (vectype)
+             || known_lt (TYPE_VECTOR_SUBPARTS (vectype), BITS_PER_UNIT))
          && SCALAR_INT_MODE_P (TYPE_MODE (vectype)))
        optab1 =3D vec_pack_sbool_trunc_optab;
       else
@@ -12213,6 +12214,7 @@ supportable_narrowing_operation (enum tree_code cod=
e,
       if (VECTOR_BOOLEAN_TYPE_P (intermediate_type)
          && VECTOR_BOOLEAN_TYPE_P (prev_type)
          && intermediate_mode =3D=3D prev_mode
+         && known_lt (TYPE_VECTOR_SUBPARTS (intermediate_type), BITS_PER_U=
NIT)
          && SCALAR_INT_MODE_P (prev_mode))
        interm_optab =3D vec_pack_sbool_trunc_optab;
       else

-march=3Dicelake-server -O3 -mprefer-vector-width=3D128 now can get vectori=
zed
loop.


        vmovdqu8        (%rsi,%rax), %xmm0
        vpmovzxbw       %xmm0, %xmm2
        vpmovzxwd       %xmm2, %xmm1
        vpsrldq $8, %xmm0, %xmm0
        vpsrldq $8, %xmm2, %xmm2
        vpmovzxbw       %xmm0, %xmm0
        vpmovzxwd       %xmm2, %xmm2
        vpmulld %xmm9, %xmm1, %xmm1
        vpmulld %xmm9, %xmm2, %xmm2
        vpmovzxwd       %xmm0, %xmm4
        vpsrldq $8, %xmm0, %xmm0
        vpmovzxwd       %xmm0, %xmm0
        vpmulld %xmm9, %xmm4, %xmm4
        vpmulld %xmm9, %xmm0, %xmm0
        vpcmpud $6, %xmm6, %xmm1, %k0
        vpsubd  %xmm1, %xmm7, %xmm3
        vpcmpud $6, %xmm6, %xmm2, %k1
        vpsubd  %xmm2, %xmm7, %xmm5
        vpsrad  $31, %xmm5, %xmm5
        vpsrad  $31, %xmm3, %xmm3
        vpermt2w        %xmm5, %xmm8, %xmm3
        vpsubd  %xmm0, %xmm7, %xmm10
        vpsubd  %xmm4, %xmm7, %xmm5
        kshiftlb        $4, %k1, %k1
        vpcmpud $6, %xmm6, %xmm0, %k2
        vpsrad  $31, %xmm5, %xmm5
        vpsrad  $31, %xmm10, %xmm10
        kandb   %k3, %k0, %k0
        korb    %k1, %k0, %k0
        vpcmpud $6, %xmm6, %xmm4, %k1
        vpermt2w        %xmm10, %xmm8, %xmm5
        vpermt2w        %xmm2, %xmm8, %xmm1
        vpermt2w        %xmm0, %xmm8, %xmm4
        vpermt2b        %xmm5, %xmm11, %xmm3
        vpermt2b        %xmm4, %xmm11, %xmm1
        kandb   %k3, %k1, %k1
        kshiftlb        $4, %k2, %k2
        korb    %k2, %k1, %k1
        kunpckbw        %k0, %k1, %k1
        vmovdqu8        %xmm3, %xmm1{%k1}
        vmovdqu8        %xmm1, (%rdi,%rax)
        addq    $16, %rax
        cmpq    %rax, %r8
        jne     .L4=