[Bug tree-optimization/22279] New: [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[Bug tree-optimization/22279] New: [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[jsm28 at gcc dot gnu dot org]

On various platforms, the libstdc++ testsuite fails early with an ICE building
testsuite_abi.cc.  This includes at least hppa64-hpux and ia64-hpux (compilers
as of 20050702 07:00 UTC) and gcc-testresults shows it on x86_64-linux; it
didn't happen with 20050630 compilers.  The dates show this is a different bug
from bug 22071.

/scratch/gcc/nightly-2005-07-02-mainline/hppa64-hp-hpux11.23/build_gcc/install/lib/gcc/hppa64-hp-hpux11.23/4.1.0/../../../../include/c++/4.1.0/bits/vector.tcc:
In member function 'void std::vector<_Tp,
_Alloc>::_M_insert_aux(__gnu_cxx::__normal_iterator<typename
std::_Vector_base<_Tp, _Alloc>::_Tp_alloc_type::pointer, 
std::vector<_Tp, _Alloc> >, const _Tp&) [with _Tp = compare_symbols(const char*,
const char*, bool)::symbol_pair, _Alloc = std::allocator<compare_symbols(const
char*, const char*,
bool)::symbol_pair>]':/scratch/gcc/nightly-2005-07-02-mainline/hppa64-hp-hpux11.23/build_gcc/install/l
ib/gcc/hppa64-hp-hpux11.23/4.1.0/../../../../include/c++/4.1.0/bits/vector.tcc:249:
internal compiler error: in first_vi_for_offset, at tree-ssa-structalias.c:2566

-- 
           Summary: [4.1 Regression] ICE in first_vi_for_offset, at tree-
                    ssa-structalias.c:2566
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: tree-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: jsm28 at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE 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
           Keywords|                            |ice-on-valid-code
   Target Milestone|---                         |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE 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 17:01 -------
Looks like it only effects 64bit targets:
http://gcc.gnu.org/ml/gcc-testresults/2005-07/msg00058.html

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
 GCC target triplet|                            |64bit targets


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
OtherBugsDependingO|                            |22277
              nThis|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2005-07-02 19:10 -------
Can we get a preprocessed/reduced source?

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[jsm28 at gcc dot gnu dot org]

------- Additional Comments From jsm28 at gcc dot gnu dot org  2005-07-03 12:51 -------
Created an attachment (id=9197)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9197&action=view)
preprocessed source on x86_64-unknown-linux-gnu

Here is the preprocessed version of testsuite_abi.cc on
x86_64-unknown-linux-gnu.  Compile with -O2.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE 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  2005-07-03 13:20 -------
Subject: Re:  [4.1 Regression] ICE in
	first_vi_for_offset, at tree-ssa-structalias.c:2566


> Here is the preprocessed version of testsuite_abi.cc on
> x86_64-unknown-linux-gnu.  Compile with -O2.
> 

Thanks, i'm on it.




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE 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  2005-07-03 14:46 -------
Confirmed, reduced testcase for powerpc64-darwin at least:
template<class _T1, class _T2>
struct pair
{
  typedef _T1 first_type;
  typedef _T2 second_type;
  _T1 first;
  _T2 second;
  pair(const _T1& __a, const _T2& __b) : first(__a), second(__b) { }
};
struct string
{
  char* _M_p;
  int i;
  string();
  int begin();
  int end();
  string(int, int);
};
struct symbol
{
  int type;
  string name;
  string raw_name;
  string demangled_name;
  string version_name;
  int version_status;
  int status;
  void init(string& data);
};
void symbol::init(string& data)
{
  name = string(data.begin(), data.end());
}
typedef pair<symbol, symbol> symbol_pair;
struct vector
{
  void push_back(const symbol_pair& __x);
};
void f(vector incompatible)
{
  symbol base;
  incompatible.push_back(symbol_pair(base, base));
}


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2005-07-03 14:46:42
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE 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  2005-07-03 14:56 -------
And here is a testcase which fails in both 32 and 64bit mode:
struct string
{
  long long _M_p;
  long long i;
  string();
  int begin();
  int end();
  string(int, int);
};
struct symbol
{
  int type;
  string name;
  long long raw_name;
  long long demangled_name;
  long long version_name;
  int version_status;
  int status;
  void init();
};
void symbol::init() { name = string(); }
struct pair
{
  symbol first;
  symbol second;
  pair(const symbol& __a, const symbol& __b) : first(__a), second(__b) { }
};
struct vector
{
  void push_back(const pair& __x);
};
void f(vector incompatible)
{
  symbol base;
  incompatible.push_back(pair(base, base));
}


Though only on powerpc-darwin.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[dberlin at gcc dot gnu dot org]

------- Additional Comments From dberlin at gcc dot gnu dot org  2005-07-04 02:21 -------
The bug you have indentified (which is related but not the same) reduces to:

struct string
{
  long long _M_p;
  long long i;
  string();
  int begin();
  int end();
  string(int, int);
};
struct symbol
{
  int type;
  string name;
  long long raw_name;
  long long demangled_name;
  long long version_name;
  int version_status;
  int status;
  void init();
};
void symbol::init() { name = string(); }
struct pair
{
  symbol first;
  symbol second;
  pair(const symbol& __a, const symbol& __b) : first(__a), second(__b) { }
};
struct vector
{
  void push_back(const pair& __x);
};
void f(vector incompatible)
{
  symbol base;
  incompatible.push_back(pair(base, base));
}




However, this is not the same actual backtrace as the other bug, which comes
from solution_set_add.
Related, but not exactly the same.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE in first_vi_for_offset, at tree-ssa-structalias.c:2566

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-07-04 17:49 -------
Subject: Bug 22279

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	dberlin@gcc.gnu.org	2005-07-04 17:49:17

Modified files:
	gcc            : ChangeLog tree-ssa-structalias.c 
Added files:
	gcc/testsuite/g++.dg/tree-ssa: pr22279.C 

Log message:
	2005-07-04  Daniel Berlin  <dberlin@dberlin.org>
	
	Fix PR tree-optimization/22279
	
	* tree-ssa-structalias.c (offset_overlaps_with_access): Use
	correct operator.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.9333&r2=2.9334
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-ssa-structalias.c.diff?cvsroot=gcc&r1=2.9&r2=2.10
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/tree-ssa/pr22279.C.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279

[Bug tree-optimization/22279] [4.1 Regression] ICE 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 19:15 -------
Fixed.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22279