[Bug target/51743] New: [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[Bug target/51743] New: [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

             Bug #: 51743
           Summary: [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with
                    -O2 -mbig-endian
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: ubizjak@gmail.com
            Target: ia64


Currently -mbig-endian -O2 fails almost all of gcc.dg/torture/vshuf*.c tests:

Native configuration is ia64-unknown-linux-gnu

                === gcc tests ===


Running target unix
FAIL: gcc.dg/torture/vshuf-v2si.c  -O2  (internal compiler error)
FAIL: gcc.dg/torture/vshuf-v2si.c  -O2  (test for excess errors)
UNRESOLVED: gcc.dg/torture/vshuf-v2si.c  -O2  compilation failed to produce
executable

                === gcc Summary for unix ===

# of expected passes            30
# of unexpected failures        2
# of unresolved testcases       1
# of unsupported tests          112

Running target unix/-mbig-endian
FAIL: gcc.dg/torture/vshuf-v16hi.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v16qi.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v2sf.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v2si.c  -O2  (internal compiler error)
FAIL: gcc.dg/torture/vshuf-v2si.c  -O2  (test for excess errors)
UNRESOLVED: gcc.dg/torture/vshuf-v2si.c  -O2  compilation failed to produce
executable
FAIL: gcc.dg/torture/vshuf-v32qi.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v4hi.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v4sf.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v4si.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v8hi.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v8qi.c  -O2  execution test
FAIL: gcc.dg/torture/vshuf-v8si.c  -O2  execution test

                === gcc Summary for unix/-mbig-endian ===

# of expected passes            20
# of unexpected failures        12
# of unresolved testcases       1
# of unsupported tests          112

                === gcc Summary ===

# of expected passes            50
# of unexpected failures        14
# of unresolved testcases       2
# of unsupported tests          224
/home/uros/gcc-build/gcc/xgcc  version 4.7.0 20120103 (experimental) [trunk
revision 182829] (GCC)

[Bug target/51743] [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

Uros Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rth at gcc dot gnu.org

--- Comment #1 from Uros Bizjak <ubizjak at gmail dot com> 2012-01-03 16:50:02 UTC ---
CC'd author.

[Bug target/51743] [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[rth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

--- Comment #2 from Richard Henderson <rth at gcc dot gnu.org> 2012-01-03 20:11:43 UTC ---
Uh... can you even force ia64-linux to run in big-endian mode?
Just because you said -mbig-endian doesn't mean it is.  I don't
see anything in the linux kernel that allows per-process endian
switching.

Do these same failures appear for ia64-hpux?

[Bug target/51743] [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

Uros Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sje at cup dot hp.com

--- Comment #3 from Uros Bizjak <ubizjak at gmail dot com> 2012-01-04 07:11:52 UTC ---
(In reply to comment #2)
> Uh... can you even force ia64-linux to run in big-endian mode?
> Just because you said -mbig-endian doesn't mean it is.  I don't
> see anything in the linux kernel that allows per-process endian
> switching.

These tests just shuffle bytes around, so I was under impression that the
functionality is isolated from OS. And Hello world executes correctly when
compiled with -mbig-endian.

> Do these same failures appear for ia64-hpux?

Hm, latest results are from 1st of December [1].

[1] http://gcc.gnu.org/ml/gcc-testresults/2011-12/msg00083.html

[Bug target/51743] [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

--- Comment #4 from Richard Guenther <rguenth at gcc dot gnu.org> 2012-01-04 10:34:43 UTC ---
(In reply to comment #2)
> Uh... can you even force ia64-linux to run in big-endian mode?
> Just because you said -mbig-endian doesn't mean it is.  I don't
> see anything in the linux kernel that allows per-process endian
> switching.

Hm, should we then reject this switch on linux?

> Do these same failures appear for ia64-hpux?

[Bug target/51743] [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[rth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

--- Comment #5 from Richard Henderson <rth at gcc dot gnu.org> 2012-01-05 00:31:35 UTC ---
(In reply to comment #3)
> These tests just shuffle bytes around, so I was under impression that the
> functionality is isolated from OS. And Hello world executes correctly when
> compiled with -mbig-endian.

Hello world only manipulates pointers and spends 99% of its time in libc.

Shuffling bytes around largely depends on how and what you do with it.

Try the more obvious

  int main() { 
    union { int i; char c[4]; } u;
    u.i = 0x01020304;
    printf("%d\n", u.c[0]);
    return 0;
  }

to convince yourself we aren't actually running in big-endian mode.

(In reply to comment #4)
> Hm, should we then reject this switch on linux?

We could probably remove it entirely and let it be controlled by the
OS config headers and get better code within gcc itself.

[Bug target/51743] [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51743

Uroš Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |INVALID

--- Comment #6 from Uroš Bizjak <ubizjak at gmail dot com> ---
Probably not a bug.
>From gcc-bugs-return-437432-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Thu Dec 12 18:17:08 2013
Return-Path: <gcc-bugs-return-437432-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 14954 invoked by alias); 12 Dec 2013 18:17:08 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 14925 invoked by uid 48); 12 Dec 2013 18:17:04 -0000
From: "ubizjak at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/51681] [4.7 Regression]: ICE in gcc.dg/torture/vshuf-v2si.c on ia64
Date: Thu, 12 Dec 2013 18:17:00 -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: 4.7.0
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: ubizjak at gmail dot com
X-Bugzilla-Status: RESOLVED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: ubizjak at gmail dot com
X-Bugzilla-Target-Milestone: 4.7.0
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: bug_status resolution
Message-ID: <bug-51681-4-RZBzqHIyC5@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-51681-4@http.gcc.gnu.org/bugzilla/>
References: <bug-51681-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-12/txt/msg01087.txt.bz2
Content-length: 492

http://gcc.gnu.org/bugzilla/show_bug.cgi?idQ681

Bug 51681 depends on bug 51743, which changed state.

Bug 51743 Summary: [ia64] Many gcc.dg/torture/vshuf*.c tests FAIL with -O2 -mbig-endian
http://gcc.gnu.org/bugzilla/show_bug.cgi?idQ743

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |INVALID