[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[sje at cup dot hp.com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

Steve Ellcey <sje at cup dot hp.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2010.09.24 21:06:43
               date|                            |
                 CC|                            |sje at cup dot hp.com
     Ever Confirmed|0                           |1

-- 
Configure bugmail: http://gcc.gnu.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[sje at cup dot hp.com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

--- Comment #5 from Steve Ellcey <sje at cup dot hp.com> 2010-09-24 22:21:38 UTC ---
I have verified that the bug shows up in r163443.  Looking at the assembly
language differences between 163442 and 163443, both versions have
_GLOBAL__I_65535_0__ZN2c12f6Ev,
a global routine to call the static initializer
(_Z41__static_initialization_and_destruction_0ii), but the new version also has
_GLOBAL__I__ZN2c12f6Ev to call the static initializer and this routine is not
global.  If I change the assembly language code to change the
GLOBAL__I__ZN2c12f6Ev function from static (.PARAM) to global (.EXPORT) then
the code will compile.

-- 
Configure bugmail: http://gcc.gnu.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[sje at cup dot hp.com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

--- Comment #6 from Steve Ellcey <sje at cup dot hp.com> 2010-09-24 23:45:28 UTC ---
I have a patch I am testing.  It worked on the test case but I haven't fully
bootstrapped it.

Index: ipa.c
===================================================================
--- ipa.c       (revision 164578)
+++ ipa.c       (working copy)
@@ -1480,6 +1480,7 @@ build_cdtor (bool ctor_p, VEC (tree, hea
            DECL_STATIC_CONSTRUCTOR (fn) = 0;
          else
            DECL_STATIC_DESTRUCTOR (fn) = 0;
+         TREE_PUBLIC (fn) = 1;
          /* We do not want to optimize away pure/const calls here.
             When optimizing, these should be already removed, when not
             optimizing, we want user to be able to breakpoint in them.  */

-- 
Configure bugmail: http://gcc.gnu.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[dave at hiauly1 dot hia.nrc.ca]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

--- Comment #7 from dave at hiauly1 dot hia.nrc.ca 2010-09-26 23:21:44 UTC ---
Hi Steve,

> I have a patch I am testing.  It worked on the test case but I haven't fully
> bootstrapped it.

Keep plugging away on this.

I have the sense that the breakage is intentional.  I fought the same
battle in 4.5, so Honza should be aware what's happened.  I believe that
he doesn't want the constructors to be global to improve interprocedurial
optimisations.  I'm sure this could be fixed with a target hook, but
Honza doesn't really work on anything but x86.

Regards,
Dave

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED

--- Comment #8 from Jan Hubicka <hubicka at gcc dot gnu.org> 2010-12-12 19:54:28 UTC ---
Mine, patch sent http://gcc.gnu.org/ml/gcc-patches/2010-12/msg00961.html

Sorry about the delay. 

The problem is not that I would not want constructors to be global. The problem
was that since the original constructor merging was introduced (in 2004), C++
FE produced collect2 recognizable names for no reason. This problem was masked
by fact that the code set always inline flag on all functions. This is wrong,
since not all functions are inlinable so it can lead to ICE when non-inlinable
static constructor is introduced.

The patch should avoid those recognizable names until final merging pass and
thus avoid the problem.

Honza

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

--- Comment #9 from Jan Hubicka <hubicka at gcc dot gnu.org> 2010-12-13 17:29:17 UTC ---
Author: hubicka
Date: Mon Dec 13 17:29:14 2010
New Revision: 167753

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=167753
Log:

    PR middle-end/45388
    * decl2.c (start_objects): Do not generate collect2 recognicable name
    for static ctor.
    * ipa.c (cgraph_build_static_cdtor_1): Break out from ... ; add FINAL
parameter.
    (cgraph_build_static_cdtor): ... here.
    (build_cdtor): Use cgraph_build_static_cdtor_1.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/cp/ChangeLog

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

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

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

--- Comment #10 from Jan Hubicka <hubicka at gcc dot gnu.org> 2010-12-13 17:31:47 UTC ---
Fixed by my patch.

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[amonakov at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

Alexander Monakov <amonakov at gcc dot gnu.org> changed:

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

--- Comment #11 from Alexander Monakov <amonakov at gcc dot gnu.org> 2010-12-13 18:05:21 UTC ---
Only Changelog changed have been checked in.

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

--- Comment #12 from Jan Hubicka <hubicka at gcc dot gnu.org> 2010-12-14 01:26:51 UTC ---
Author: hubicka
Date: Tue Dec 14 01:26:47 2010
New Revision: 167781

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=167781
Log:
This time really commit
    PR middle-end/45388
    * decl2.c (start_objects): Do not generate collect2 recognicable name
    for static ctor.
    * ipa.c (cgraph_build_static_cdtor_1): Break out from ... ; add FINAL
parameter.
    (cgraph_build_static_cdtor): ... here.
    (build_cdtor): Use cgraph_build_static_cdtor_1.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/cp/decl2.c
    trunk/gcc/ipa.c

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
                 CC|                            |jakub at gcc dot gnu.org
         Resolution|                            |FIXED

--- Comment #13 from Jakub Jelinek <jakub at gcc dot gnu.org> 2010-12-15 11:55:07 UTC ---
Assuming this is fixed now.

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #3 from dave at hiauly1 dot hia dot nrc dot ca  2010-08-24 15:06 -------
Subject: Re:  [4.6 Regression] Global constructor not found

> Is there another consturctor calling this one? (or can you attach complette .s
> file?) Perhaps it is result of my constructor merging patch.

Unfortunately, the .s is gone for the moment.  What I showed contained
all the references to the constructor.  There is no `.EXPORT' statement,
so the constructor is local.

I believe it was caused by your constructor merging patch, but I haven't
explicitly verified this.

Collect2 builds a table of constructors to call at startup.  This startup
file is then linked into the final executable.  This only works if the
constructors are global symbols.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[hubicka at gcc dot gnu dot org]

------- Comment #2 from hubicka at gcc dot gnu dot org  2010-08-24 12:12 -------
Is there another consturctor calling this one? (or can you attach complette .s
file?) Perhaps it is result of my constructor merging patch.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.6.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388

[Bug middle-end/45388] [4.6 Regression] Global constructor not found

[danglin at gcc dot gnu dot org]

------- Comment #1 from danglin at gcc dot gnu dot org  2010-08-23 20:58 -------
Constructor is no longer global:

        .PARAM _GLOBAL__I__ZN2c12f6Ev,PRIV_LEV=3
L$FB0035:
_GLOBAL__I__ZN2c12f6Ev:

As a result, collect2 does not find the constructor.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45388