[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |5.0

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

David Binderman <dcb314 at hotmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dcb314 at hotmail dot com

--- Comment #2 from David Binderman <dcb314 at hotmail dot com> ---
(In reply to Joost VandeVondele from comment #1)
> a similar error is seen in PR63926 (slightly different line numbers),
> without lto, but pgo.

I am seeing something similar, without lto and without pgo.

FTjackSupport.cpp:695:1: internal compiler error: in estimate_edge_growth, at
ipa-inline.h:300

Test case available on request.

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[Joost.VandeVondele at mat dot ethz.ch]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #3 from Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> ---
(In reply to David Binderman from comment #2)
> Test case available on request.

since non of the PRs have a testcase yet, would be good, if only to be added to
the testsuite.

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #4 from David Binderman <dcb314 at hotmail dot com> ---
Created attachment 34015
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34015&action=edit
gzipped C++ source code

FTjackSupport.cpp:695:1: internal compiler error: in estimate_edge_growth, at
ipa-inline.h:300
0xa70ba9 estimate_edge_growth(cgraph_edge*)
    ../../src/trunk/gcc/ipa-inline.h:299
0xa70abe estimate_size_after_inlining(cgraph_node*, cgraph_edge*)
    ../../src/trunk/gcc/ipa-inline-analysis.c:3834
0x17e677c caller_growth_limits(cgraph_edge*)
    ../../src/trunk/gcc/ipa-inline.c:201
0x17e677c can_inline_edge_p(cgraph_edge*, bool, bool)
    ../../src/trunk/gcc/ipa-inline.c:371

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[dcb314 at hotmail dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="UTF-8", Size: 3190 bytes --]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #5 from David Binderman <dcb314 at hotmail dot com> ---
(In reply to David Binderman from comment #4)
> Created attachment 34015 [details]
> gzipped C++ source code
> 
> FTjackSupport.cpp:695:1: internal compiler error: in estimate_edge_growth,
> at ipa-inline.h:300
> 0xa70ba9 estimate_edge_growth(cgraph_edge*)
> 	../../src/trunk/gcc/ipa-inline.h:299
> 0xa70abe estimate_size_after_inlining(cgraph_node*, cgraph_edge*)
> 	../../src/trunk/gcc/ipa-inline-analysis.c:3834
> 0x17e677c caller_growth_limits(cgraph_edge*)
> 	../../src/trunk/gcc/ipa-inline.c:201
> 0x17e677c can_inline_edge_p(cgraph_edge*, bool, bool)
> 	../../src/trunk/gcc/ipa-inline.c:371

Flags -O2 -finline-functions required.
>From gcc-bugs-return-467127-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Tue Nov 18 09:02:11 2014
Return-Path: <gcc-bugs-return-467127-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 15300 invoked by alias); 18 Nov 2014 09:02:11 -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 15228 invoked by uid 48); 18 Nov 2014 09:02:08 -0000
From: "ebotcazou at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/54779] split FRAME variables back into pieces
Date: Tue, 18 Nov 2014 09:02:00 -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: 4.9.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: ebotcazou at gcc dot gnu.org
X-Bugzilla-Status: NEW
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: attachments.isobsolete attachments.created
Message-ID: <bug-54779-4-iQLvceu3vk@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-54779-4@http.gcc.gnu.org/bugzilla/>
References: <bug-54779-4@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: 2014-11/txt/msg01599.txt.bz2
Content-length: 557

https://gcc.gnu.org/bugzilla/show_bug.cgi?idT779

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #32249|0                           |1
        is obsolete|                            |

--- Comment #8 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
Created attachment 34016
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id4016&actioníit
Updated implementation

For a 4.9-based compiler.

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[trippels at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

Markus Trippelsdorf <trippels at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-11-18
                 CC|                            |trippels at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #6 from Markus Trippelsdorf <trippels at gcc dot gnu.org> ---
markus@x4 test % cat test.ii
class A
{
public:
  virtual bool m_fn1 ();
  virtual const char **m_fn2 (int);
  virtual int m_fn3 ();
};
class FTjackSupport : A
{
  ~FTjackSupport ();
  bool m_fn1 ();
  bool m_fn4 ();
  const char **
  m_fn2 (int)
  {
  }
  int _inited;
  int *_jackClient;
  int _activePathCount;
}

* a;
void fn1 (...);
void fn2 (void *);
int fn3 (int *);
FTjackSupport::~FTjackSupport () { m_fn4 (); }

bool
FTjackSupport::m_fn1 ()
{
  if (!_jackClient)
    return 0;
  for (int i=0; _activePathCount; ++i)
    if (m_fn2 (i))
      fn2 (a);
  if (m_fn3 ())
    fn2 (a);
  if (fn3 (_jackClient))
    fn1 (0);
}

bool
FTjackSupport::m_fn4 ()
{
  if (_inited && _jackClient)
    {
      m_fn1 ();
      return 0;
    }
}

markus@x4 test % g++ -c -O2 -finline-functions test.ii
test.ii:50:1: internal compiler error: in estimate_edge_growth, at
ipa-inline.h:300
 }
 ^
0xb6e70d estimate_edge_growth
        ../../gcc/gcc/ipa-inline.h:299
0xb6e70d estimate_size_after_inlining(cgraph_node*, cgraph_edge*)
        ../../gcc/gcc/ipa-inline-analysis.c:3835
0x1354db4 caller_growth_limits
        ../../gcc/gcc/ipa-inline.c:201
0x1354db4 can_inline_edge_p
        ../../gcc/gcc/ipa-inline.c:371
0x1356fa2 update_callee_keys
        ../../gcc/gcc/ipa-inline.c:1252
0x1358bfd inline_small_functions
        ../../gcc/gcc/ipa-inline.c:1834
0x1358bfd ipa_inline
        ../../gcc/gcc/ipa-inline.c:2198
0x1358bfd execute
        ../../gcc/gcc/ipa-inline.c:2568
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.

[Bug ipa/63470] [5 Regression] lto1: internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[Joost.VandeVondele at mat dot ethz.ch]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> changed:

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

--- Comment #7 from Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> ---
Jan, as mentioned in PR63967, could you have a look ?

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #9 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
The ICE still reproduces for me even though at different place now.  I am
looking ito it.

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[trippels at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #10 from Markus Trippelsdorf <trippels at gcc dot gnu.org> ---
Author: trippels
Date: Mon Jan 12 07:48:50 2015
New Revision: 219445

URL: https://gcc.gnu.org/viewcvs?rev=219445&root=gcc&view=rev
Log:
ipa/63470 add testcase

2015-01-11  Markus Trippelsdorf  <markus@trippelsdorf.de>

    PR ipa/63470
    * ipa/pr63470.C: New testcase.

Added:
    trunk/gcc/testsuite/g++.dg/ipa/pr63470.C
Modified:
    trunk/gcc/testsuite/ChangeLog

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[trippels at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #11 from Markus Trippelsdorf <trippels at gcc dot gnu.org> ---
Oh, well. I will revert the addition of the testcase.

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #12 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
The other bug is double ajdusting of call_stmt_size when the call is turned
direct. Fixed thus:

Index: ipa-prop.c
===================================================================
--- ipa-prop.c  (revision 219430)
+++ ipa-prop.c  (working copy)
@@ -2736,6 +2736,7 @@
                       "converting indirect call in %s to direct call to %s\n",
                       ie->caller->name (), callee->name ());
     }
+  struct cgraph_edge *orig = ie;
   if (!speculative)
     ie = ie->make_direct (callee);
   else
@@ -2750,11 +2751,16 @@
       ie = ie->make_speculative
             (callee, ie->count * 8 / 10, ie->frequency * 8 / 10);
     }
-  es = inline_edge_summary (ie);
-  es->call_stmt_size -= (eni_size_weights.indirect_call_cost
-                        - eni_size_weights.call_cost);
-  es->call_stmt_time -= (eni_time_weights.indirect_call_cost
-                        - eni_time_weights.call_cost);
+  /* If we resolved speuclative edge the cost is already up to date
+     for direct call (adjusted by inline_edge_duplication_hook).  */
+  if (ie == orig)
+    {
+      es = inline_edge_summary (ie);
+      es->call_stmt_size -= (eni_size_weights.indirect_call_cost
+                            - eni_size_weights.call_cost);
+      es->call_stmt_time -= (eni_time_weights.indirect_call_cost
+                            - eni_time_weights.call_cost);
+    }

   return ie;
 }
Index: ipa-inline-analysis.c
===================================================================
--- ipa-inline-analysis.c       (revision 219430)
+++ ipa-inline-analysis.c       (working copy)
@@ -1312,6 +1312,14 @@
   info->predicate = NULL;
   edge_set_predicate (dst, srcinfo->predicate);
   info->param = srcinfo->param.copy ();
+  if (!dst->indirect_unknown_callee && src->indirect_unknown_callee)
+    {
+      gcc_checking_assert (src->caller == dst->caller);
+      info->call_stmt_size -= (eni_size_weights.indirect_call_cost
+                              - eni_size_weights.call_cost);
+      info->call_stmt_time -= (eni_time_weights.indirect_call_cost
+                              - eni_time_weights.call_cost);
+    }
 }

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

--- Comment #13 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
Author: hubicka
Date: Mon Jan 12 09:24:18 2015
New Revision: 219451

URL: https://gcc.gnu.org/viewcvs?rev=219451&root=gcc&view=rev
Log:


    PR ipa/63470
    * ipa-inline-analysis.c (inline_edge_duplication_hook): Adjust
    cost when edge becomes direct.
    * ipa-prop.c (make_edge_direct): Do not adjust when speculation
    is resolved or when introducing new speculation.
    * testsuite/g++.dg/ipa/pr63470.C: New testcase.

Added:
    trunk/gcc/testsuite/g++.dg/ipa/pr63470.C
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/ipa-inline-analysis.c
    trunk/gcc/ipa-prop.c
    trunk/gcc/testsuite/ChangeLog

[Bug ipa/63470] [5 Regression] internal compiler error: in estimate_edge_growth, at ipa-inline.h:308

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63470

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #14 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
Fixed.