[Bug ada/22277] New: ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[Bug ada/22277] New: ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[laurent at guerby dot net]

On x86_64-linux as of LAST_UPDATED: Sat Jul  2 09:40:45 UTC 2005

+===========================GNAT BUG DETECTED==============================+
| 4.1.0 20050702 (experimental) (x86_64-unknown-linux-gnu) GCC error:      |
| in first_vi_for_offset, at tree-ssa-structalias.c:2566                   |
| Error detected at cc40001.adb:148:5                                      |

The test was PASS as of LAST_UPDATED: Tue Jun 28 20:57:27 UTC 2005.

Note: to bootstrap Ada on x86_64 you need the following patch to workaround PR22212.

Index: Makefile.in
===================================================================
RCS file: /cvs/gcc/gcc/gnattools/Makefile.in,v
retrieving revision 1.4
diff -u -r1.4 Makefile.in
--- Makefile.in 30 Mar 2005 08:56:55 -0000      1.4
+++ Makefile.in 2 Jul 2005 13:04:19 -0000
@@ -112,7 +112,7 @@
 TOOLS_FLAGS_TO_PASS_NATIVE= \
        "CC=../../xgcc -B../../" \
        "CFLAGS=$(CFLAGS)" \
-       "ADAFLAGS=$(ADAFLAGS)" \
+       "ADAFLAGS=$(ADAFLAGS) -O0" \
        "INCLUDES=$(INCLUDES_FOR_SUBDIR)" \
        "ADA_INCLUDES=-I../rts $(ADA_INCLUDES_FOR_SUBDIR)" \
        "exeext=$(exeext)" \

-- 
           Summary: ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-
                    structalias.c:2566
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P2
         Component: ada
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: laurent at guerby dot net
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|ada                         |tree-optimization


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug ada/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dberlin at gcc dot gnu dot
                   |                            |org, pinskia at gcc dot gnu
                   |                            |dot org
 GCC target triplet|                            |x86_64-unknown-linux-gnu
            Summary|ACATS ICE cc40001 in        |[4.1 Regression] ACATS ICE
                   |first_vi_for_offset, at     |cc40001 in
                   |tree-ssa-structalias.c:2566 |first_vi_for_offset, at
                   |                            |tree-ssa-structalias.c:2566
   Target Milestone|---                         |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-07-02 17:23 -------
Most likely related to PR 22279.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
  BugsThisDependsOn|                            |22279


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-07-04 10:10 -------
This fails the same way on powerpc-darwin which is 32bit too.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
 GCC target triplet|x86_64-unknown-linux-gnu    |x86_64-linux-gnu, powerpc-
                   |                            |darwin


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

-- 
Bug 22277 depends on bug 22279, which changed state.

Bug 22279 Summary: [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

           What    |Old Value                   |New Value
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pluto at agmk dot net]

------- Additional Comments From pluto at agmk dot net  2005-07-08 09:33 -------
(In reply to comment #1)
> Most likely related to PR 22279.

`first_vi_for_offset` bug still occurs and it's easy to reproduct:

[ qt4 build / gcc-20050707T1814UTC ]

(...)
g++ -c -pipe -I/usr/include/postgresql/server -I/usr/include/mysql -O2
-march=i686 -mtune=pentium4 -fvisibility=hidden -fvisibility-inlines-hidden
-D_REENTRANT -Wall -W -fPIC -DQT_CLEAN_NAMESPACE -DQT_COMPAT
-DQT_NO_CAST_TO_ASCII -DQT_NO_CAST_FROM_ASCII -DQT_NO_DEBUG -DQT_PLUGIN
-DQT_CORE_LIB -DQT_SQL_LIB -D_LARGEFILE64_SOURCE -D_LARGEFILE_SOURCE -DQT_SHARED
-I/home/users/pluto/rpm/BUILD/qt-x11-opensource-desktop-4.0.0/mkspecs/linux-g++
-I. -I../../../../include/QtSql -I../../../../include/QtCore
-I../../../../include -I.moc/release-shared -I. -o
.obj/release-shared/qsql_mysql.o ../../../sql/drivers/mysql/qsql_mysql.cpp
../../../sql/drivers/mysql/qsql_mysql.cpp: In member function 'virtual bool
QMYSQLResult::exec()':
../../../sql/drivers/mysql/qsql_mysql.cpp:685: internal compiler error: in
first_vi_for_offset, at tree-ssa-structalias.c:2585
(...)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[dberlin at dberlin dot org]

------- Additional Comments From dberlin at gcc dot gnu dot org  2005-07-08 12:54 -------
Subject: Re:  [4.1 Regression] ACATS ICE
	cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

On Fri, 2005-07-08 at 09:33 +0000, pluto at agmk dot net wrote:
> ------- Additional Comments From pluto at agmk dot net  2005-07-08 09:33 -------
> (In reply to comment #1)
> > Most likely related to PR 22279.
> 
> `first_vi_for_offset` bug still occurs and it's easy to reproduct

There's a few bugs involved, it's just that the first_vi_for_offset is
the common place for the assert to trigger.

> :
> 
> [ qt4 build / gcc-20050707T1814UTC ]
> 
> (...)
> g++ -c -pipe -I/usr/include/postgresql/server -I/usr/include/mysql -O2
> -march=i686 -mtune=pentium4 -fvisibility=hidden -fvisibility-inlines-hidden
> -D_REENTRANT -Wall -W -fPIC -DQT_CLEAN_NAMESPACE -DQT_COMPAT
> -DQT_NO_CAST_TO_ASCII -DQT_NO_CAST_FROM_ASCII -DQT_NO_DEBUG -DQT_PLUGIN
> -DQT_CORE_LIB -DQT_SQL_LIB -D_LARGEFILE64_SOURCE -D_LARGEFILE_SOURCE -DQT_SHARED
> -I/home/users/pluto/rpm/BUILD/qt-x11-opensource-desktop-4.0.0/mkspecs/linux-g++
> -I. -I../../../../include/QtSql -I../../../../include/QtCore
> -I../../../../include -I.moc/release-shared -I. -o
> .obj/release-shared/qsql_mysql.o ../../../sql/drivers/mysql/qsql_mysql.cpp
> ../../../sql/drivers/mysql/qsql_mysql.cpp: In member function 'virtual bool
> QMYSQLResult::exec()':
> ../../../sql/drivers/mysql/qsql_mysql.cpp:685: internal compiler error: in
> first_vi_for_offset, at tree-ssa-structalias.c:2585
> (...)
> 
> 

Can you -save-temps that and attach the preprocessed source?
If so, i'll fix the bug :)




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pluto at agmk dot net]

------- Additional Comments From pluto at agmk dot net  2005-07-09 07:12 -------
Created an attachment (id=9234)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9234&action=view)
testcase


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pluto at agmk dot net]

------- Additional Comments From pluto at agmk dot net  2005-07-09 07:14 -------
(In reply to comment #4) 
 
> Can you -save-temps that and attach the preprocessed source? 
> If so, i'll fix the bug :) 
 
$ g++ qsql_mysql.ii -c -O1 
../../../sql/drivers/mysql/qsql_mysql.cpp: In member function 'virtual bool 
QMYSQLResult::exec()': 
../../../sql/drivers/mysql/qsql_mysql.cpp:685: internal compiler error: 
in first_vi_for_offset, at tree-ssa-structalias.c:2585 
 
with -O0 gcc works. 
 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[dberlin at dberlin dot org]

------- Additional Comments From dberlin at gcc dot gnu dot org  2005-07-10 01:59 -------
Subject: Re:  [4.1 Regression] ACATS ICE
	cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

On Fri, 2005-07-08 at 12:54 +0000, dberlin at dberlin dot org wrote:
> ------- Additional Comments From dberlin at gcc dot gnu dot org  2005-07-08 12:54 -------
> Subject: Re:  [4.1 Regression] ACATS ICE
> 	cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566
> 
> On Fri, 2005-07-08 at 09:33 +0000, pluto at agmk dot net wrote:
> > ------- Additional Comments From pluto at agmk dot net  2005-07-08 09:33 -------
> > (In reply to comment #1)
> > > Most likely related to PR 22279.
> > 
> > `first_vi_for_offset` bug still occurs and it's easy to reproduct
> 
> There's a few bugs involved, it's just that the first_vi_for_offset is
> the common place for the assert to trigger.


Here is the reduced testcase for the C++ file you sent, i'm working on a
fix.


------- Additional Comments From dberlin at gcc dot gnu dot org  2005-07-10 01:59 -------
Created an attachment (id=9236)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9236&action=view)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2005-07-11 08:48 -------
More compact C++ testcase:

======================================
struct A
{
  char c;
  int i;
};

A a;

struct B
{
  char c, d;
};

union C
{
  A *p;
  B *q;

  C() : p(&a) {}
  char& foo() { return q->d; }
};

void bar() { C().foo() = 0; }
======================================


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |reichelt at gcc dot gnu dot
                   |                            |org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[dberlin at dberlin dot org]

------- Additional Comments From dberlin at gcc dot gnu dot org  2005-07-11 12:33 -------
Subject: Re:  [4.1 Regression] ACATS ICE
	cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

On Mon, 2005-07-11 at 08:48 +0000, reichelt at gcc dot gnu dot org
wrote:
> ------- Additional Comments From reichelt at gcc dot gnu dot org  2005-07-11 08:48 -------
> More compact C++ testcase:
> 
> ======================================
> struct A
> {
>   char c;
>   int i;
> };
> 
> A a;
> 
> struct B
> {
>   char c, d;
> };
> 
> union C
> {
>   A *p;
>   B *q;
> 
>   C() : p(&a) {}
>   char& foo() { return q->d; }
> };
> 
> void bar() { C().foo() = 0; }

Yes, that will work too.

Note the part that bugs me severely is that you are accessing padding in
A. :)  This is true in the original testcase as well
I'm not sure whether i should bother trying to do something sane in this
case.




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-07-11 16:37 -------
Can people start filing new bugs as the orginal problem has been fixed now.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2005-07-11 16:37:19
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277

[Bug tree-optimization/22277] [4.1 Regression] ACATS ICE cc40001 in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-07-22 18:53 -------
Fixed, I filed the new bug (which should have been filed seperately) as PR 22615.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22277