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 66246385842A; Wed,  2 Feb 2022 19:30:28 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 66246385842A
From: "ckk at kvr dot at" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c/104353] New: ppc64le: Apparent reliance on undefined behavior
 of xvcvdpsxws
Date: Wed, 02 Feb 2022 19:30:28 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: new
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c
X-Bugzilla-Version: 11.2.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: ckk at kvr dot at
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_file_loc
 bug_status bug_severity priority component assigned_to reporter
 target_milestone attachments.created
Message-ID: <bug-104353-4@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, 02 Feb 2022 19:30:28 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D104353

            Bug ID: 104353
           Summary: ppc64le: Apparent reliance on undefined behavior of
                    xvcvdpsxws
           Product: gcc
           Version: 11.2.0
               URL: https://github.com/numpy/numpy/issues/20964#issuecomme
                    nt-1027865665
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ckk at kvr dot at
  Target Milestone: ---

Created attachment 52331
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=3D52331&action=3Dedit
Minimal test case for reproduction

I ran into a strange numpy error on ppc64le that only occurred inside a ppc=
64le
QEMU instance. In short, casting arrays of i doubles 1.0 to ints 1 worked as
expected on native hardware, but produced the following bogus results when
running inside a VM:

i =3D 1:   1
i =3D 2:   1 1
i =3D 3:   1 1 1
i =3D 4:   0 0 0 0
i =3D 5:   0 0 0 0 1
i =3D 6:   0 0 0 0 1 1
i =3D 7:   0 0 0 0 1 1 1
i =3D 8:   0 0 0 0 0 0 0 0
i =3D 9:   0 0 0 0 0 0 0 0 1
...


Guided by the numpy folks, a SIMD issue was suspected, and I managed to cre=
ate
a minimal test case (attached here) with which this could be reproduced. It
only occurs with -O3.

I then filed an issue with QEMU, where the issue was quickly rejected. This=
 led
to further analysis by the numpy folks. There, it was discovered that GCC is
apparently relying on undefined behavior of the xvcvdpsxws instruction, whi=
ch
happened to work on native hardware because it happen to exhibit that behav=
ior.

I'm only summarizing here; there's a great analysis in detail, and a much
better test case, on the GitHub issue, which I have linked in the URL as I'd
prefer not to reproduce the author's work here.=