[Bug tree-optimization/65355] New: [4.8/4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[Bug tree-optimization/65355] New: [4.8/4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

            Bug ID: 65355
           Summary: [4.8/4.9 Regression] vectorizer increase alignment of
                    symbols already placed in anchors
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: hubicka at gcc dot gnu.org

Increasing alignment of symbol does not work once RTL containing anchor was
produced.

Index: symtab.c
===================================================================
--- symtab.c  (revision 221268)
+++ symtab.c  (working copy)
@@ -1924,6 +1924,13 @@ symtab_node::can_increase_alignment_p (v
   if (TREE_ASM_WRITTEN (target->decl))
     return false;

+  /* If target is already placed in an anchor, we can not touch its
+     alignment.  */
+  if (DECL_RTL_SET_P (target->decl)
+      && MEM_P (DECL_RTL (target->decl))
+      && SYMBOL_REF_HAS_BLOCK_INFO_P (XEXP (DECL_RTL (target->decl), 0)))
+    return false;
+
   /* Constant pool entries may be shared.  */
   if (DECL_IN_CONSTANT_POOL (target->decl))
     return false;

sadly introduce quite few testsuite regressions for ppc-linux (checking that
alignmnet was increased). This is because notice_global_symbol produces RTL
early and thus blocks this.

I suppose notice_global_symbol needs to be redone to work on tree only. For
next stage1 we also probably want to drop the increase_alignmnets IPA pass and
run all late tree passes prior RTL generation.

[Bug tree-optimization/65355] [4.8/4.9/5 Regression] vectorizer increase alignment of symbols already placed in anchors

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |wrong-code
   Target Milestone|---                         |4.8.5

[Bug tree-optimization/65355] [4.8/4.9/5 Regression] vectorizer increase alignment of symbols already placed in anchors

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

--- Comment #1 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
Author: hubicka
Date: Tue Mar 10 04:24:21 2015
New Revision: 221297

URL: https://gcc.gnu.org/viewcvs?rev=221297&root=gcc&view=rev
Log:

    PR tree-optimization/65355
    * varasm.c (notice_global_symbol): Do not produce RTL.
    * symtab.c (symtab_node::can_increase_alignment_p): Check for section
    anchor.
    * tree-vect-data-refs.c (vect_compute_data_ref_alignment): Do not
    check for section anchors.
    * gcc.dg/vect/section-anchors-vect-69.c: Update template.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/symtab.c
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gcc.dg/vect/section-anchors-vect-69.c
    trunk/gcc/tree-vect-data-refs.c
    trunk/gcc/varasm.c

[Bug tree-optimization/65355] [4.8/4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org
            Summary|[4.8/4.9/5 Regression]      |[4.8/4.9 Regression]
                   |vectorizer increase         |vectorizer increase
                   |alignment of symbols        |alignment of symbols
                   |already placed in anchors   |already placed in anchors

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed on the trunk then.
Not sure about the running all late gimple passes before all RTL generation,
I'd be afraid that would increase compiler memory consumption too much.

[Bug tree-optimization/65355] [4.8/4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[pthaugen at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

Pat Haugen <pthaugen at gcc dot gnu.org> changed:

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

--- Comment #3 from Pat Haugen <pthaugen at gcc dot gnu.org> ---
(In reply to Jan Hubicka from comment #1)
> Author: hubicka
> Date: Tue Mar 10 04:24:21 2015
> New Revision: 221297
> 
> URL: https://gcc.gnu.org/viewcvs?rev=221297&root=gcc&view=rev
> Log:
> 
> 	PR tree-optimization/65355
> 	* varasm.c (notice_global_symbol): Do not produce RTL.
> 	* symtab.c (symtab_node::can_increase_alignment_p): Check for section
> 	anchor.
> 	* tree-vect-data-refs.c (vect_compute_data_ref_alignment): Do not
> 	check for section anchors.
> 	* gcc.dg/vect/section-anchors-vect-69.c: Update template.
> 
> Modified:
>     trunk/gcc/ChangeLog
>     trunk/gcc/symtab.c
>     trunk/gcc/testsuite/ChangeLog
>     trunk/gcc/testsuite/gcc.dg/vect/section-anchors-vect-69.c
>     trunk/gcc/tree-vect-data-refs.c
>     trunk/gcc/varasm.c

This change introduced the following failures on powerpc64[le]:

FAIL: g++.dg/vect/pr36648.cc  -std=c++11  scan-tree-dump-times vect "vectorized
1 loops" 1
FAIL: g++.dg/vect/pr36648.cc  -std=c++11  scan-tree-dump-times vect
"vectorizing stmts using SLP" 1
FAIL: g++.dg/vect/pr36648.cc  -std=c++14  scan-tree-dump-times vect "vectorized
1 loops" 1
FAIL: g++.dg/vect/pr36648.cc  -std=c++14  scan-tree-dump-times vect
"vectorizing stmts using SLP" 1
FAIL: g++.dg/vect/pr36648.cc  -std=c++98  scan-tree-dump-times vect "vectorized
1 loops" 1
FAIL: g++.dg/vect/pr36648.cc  -std=c++98  scan-tree-dump-times vect
"vectorizing stmts using SLP" 1
>From gcc-bugs-return-480116-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Wed Mar 11 20:37:34 2015
Return-Path: <gcc-bugs-return-480116-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 109585 invoked by alias); 11 Mar 2015 20:37:34 -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 109520 invoked by uid 55); 11 Mar 2015 20:37:29 -0000
From: "mpolacek at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/65388] Wrong comparison in same_succ_def::equal() tree-ssa-tail-merge.c:590
Date: Wed, 11 Mar 2015 20:37:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: 4.9.1
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: mpolacek at gcc dot gnu.org
X-Bugzilla-Status: ASSIGNED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: mpolacek at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-65388-4-Kz4umqw9nh@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-65388-4@http.gcc.gnu.org/bugzilla/>
References: <bug-65388-4@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: 2015-03/txt/msg01260.txt.bz2
Content-length: 436

https://gcc.gnu.org/bugzilla/show_bug.cgi?ide388

--- Comment #6 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Author: mpolacek
Date: Wed Mar 11 20:36:56 2015
New Revision: 221359

URL: https://gcc.gnu.org/viewcvs?rev"1359&root=gcc&view=rev
Log:
    PR tree-optimization/65388
    * tree-ssa-tail-merge.c (same_succ_def::equal): Fix typo in comparison.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/tree-ssa-tail-merge.c

[Bug tree-optimization/65355] [4.8/4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

Jan Hubicka <hubicka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |hubicka at gcc dot gnu.org

--- Comment #4 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
I will take a look. looks like a testsuite issue.

Concering breaking up the optimization queue, I did some memory use
measurements on ltrans of firefox and they were not bad.  Tree optimizers are
quite effective to get code small back after inlining happens.  Will do more
next stage1.

[Bug tree-optimization/65355] [4.8/4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.8.5                       |4.9.3

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
The gcc-4_8-branch is being closed, re-targeting regressions to 4.9.3.

[Bug tree-optimization/65355] [4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 4.9.3 has been released.

[Bug tree-optimization/65355] [4.9 Regression] vectorizer increase alignment of symbols already placed in anchors

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65355

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.9.3                       |4.9.4