[Bug c++/15967] New: ICE: Segmentation fault

[Bug c++/15967] New: ICE: Segmentation fault

[wwieser at gmx dot de]

Compiling the following code will crash gcc with a SIGSEGV ICE:  
 
------------------------<gccbug10.cc>------------------------- 
// (c) Wolfgang Wieser 06/2004.  
 
typedef unsigned int size_t; 
 
struct A { void *operator new(size_t s){} }; 
struct B { void *operator new(size_t s){} }; 
 
struct C : A,B {}; 
 
int crash() 
{ 
	C *c=new C();     // <-- line 12, ICE 
} 
-------------------------------------------------------------- 
 
bash# gcc -c gccbug10.cc 
gccbug10.cc: In function `int crash()': 
gccbug10.cc:12: internal compiler error: Segmentation fault 
Please submit a full bug report, 
with preprocessed source if appropriate. 
See <URL:http://gcc.gnu.org/bugs.html> for instructions. 
 
Affected versions:  
gcc (GCC) 3.4.1 20040501 (prerelease) 
gcc (GCC) 3.4.1 20040612 (prerelease) 
maybe others as well.

-- 
           Summary: ICE: Segmentation fault
           Product: gcc
           Version: 3.4.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: wwieser at gmx dot de
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-linux-gnu
  GCC host triplet: i686-linux-gnu
GCC target triplet: i686-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2004-06-12 23:22 -------
Confirmed. The code is invalid, but the ICE is a regression. Up to 
3.3.x we got this: 
 
g/x> /home/bangerth/bin/gcc-3.3.4-pre/bin/c++ -c x.cc 
x.cc: In function `int crash()': 
x.cc:10: error: request for member `operator new' is ambiguous 
x.cc:4: error: candidates are: static void* B::operator new(unsigned int) 
x.cc:3: error:                 static void* A::operator new(unsigned int) 
 
W. 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
           Keywords|                            |ice-on-invalid-code
      Known to fail|                            |3.4.0 3.5.0
      Known to work|                            |3.3.4
   Last reconfirmed|0000-00-00 00:00:00         |2004-06-12 23:22:48
               date|                            |
            Summary|ICE: Segmentation fault     |[3.4/3.5 regression] ICE on
                   |                            |ambiguous operator new in
                   |                            |class hierarchy
   Target Milestone|---                         |3.4.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-06-14 02:37 -------
I'm testing a patch for this


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-06-14 10:46 -------
Patch posted, waiting for review:
http://gcc.gnu.org/ml/gcc-patches/2004-06/msg00976.html

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2004-06-14 14:15 -------
Subject: Re:  [3.4/3.5 regression] ICE on ambiguous operator
 new in class hierarchy

giovannibajo at libero dot it wrote:

> ------- Additional Comments From giovannibajo at libero dot it  2004-06-14 10:46 -------
> Patch posted, waiting for review:
> http://gcc.gnu.org/ml/gcc-patches/2004-06/msg00976.html

I'm nervous about this patch.

I agree that it seems like the right idea in the abstract, but I'm 
nervous about the impact on other fragile bits of the compiler. 
Furthermore, the function can still return NULL, so the caller that did 
not expect a NULL return value is still broken.  Is there a more local 
patch that will just deal with that problem?



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[jason at gcc dot gnu dot org]

------- Additional Comments From jason at gcc dot gnu dot org  2004-06-14 20:15 -------
My inclination would be to put Giovanni's patch on the trunk; the caller failing
to deal with NULL is an implicit assertion that the lookup will never be NULL. 
That assertion was tripped in this case, so he's fixing the lookup to conform to
the assertion.

A more limited patch to just check for null and give an error might be better
for 3.4.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-06-14 23:51 -------
Subject: Re:  [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

jason at gcc dot gnu dot org wrote:

> My inclination would be to put Giovanni's
> patch on the trunk; the caller failing
> to deal with NULL is an implicit assertion that the lookup will never
> be NULL.

It cannot be NULL because at that point we already checked that
TYPE_HAS_ARRAY_NEW_OPERATOR, or TYPE_HAS_NEW_OPERATOR. If the lookup finds no
operator new, there is something very broken somewhere, so an ICE is ok.

Right now, we were getting NULL because lookup_fnfields was returning NULL in
case of ambiguity, which is against its description: when protect=2, if there
is ambiguity, we want an ambiguity.

> That assertion was tripped in this case, so he's fixing the lookup to
> conform to the assertion.

Exactly.

> A more limited patch to just check for null and give an error might
> be better for 3.4.

And what kind of error message can we emit? "Cannot find operator new" is
wrong, because there are two operator new. "Ambiguous operator new" could be
emitted, but then we have a small regression from 3.3 that used to emit the
list of ambiguous oveloads (with correct line information), plus it's totally
against lookup_fnfields/lookup_member API which says that NULL means no symbol
found.

Giovanni Bajo




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2004-06-15 00:04 -------
Subject: Re:  [3.4/3.5 regression] ICE on ambiguous operator
 new in class hierarchy

giovannibajo at libero dot it wrote:

>>A more limited patch to just check for null and give an error might
>>be better for 3.4.
> 
> 
> And what kind of error message can we emit? "Cannot find operator new" is
> wrong, because there are two operator new. "Ambiguous operator new" could be
> emitted, but then we have a small regression from 3.3 that used to emit the
> list of ambiguous oveloads (with correct line information)

That's not a regression worth worrying about.

, plus it's totally
> against lookup_fnfields/lookup_member API which says that NULL means no symbol
> found.

And so? :-)

Unless you can prove that returning a TREE_LIST will not result in a 
problem anywhere else -- including other ice-on-invalid cases -- your 
patch is not a win on the release branch.  If we have code that's 
failing to handle a TREE_LIST as the return value -- and we probably do, 
since the function is never returning that -- then we risk problems if 
we do that.

For 3.4, please just have the caller say "no suitable 'operator new'" 
and leave it at that.  That's sufficiently ambiguous that it will not be 
a lie, but still gives the user a good clue.  Much better than an ICE.

For mainline, your patch is OK.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-06-15 00:51 -------
Subject: Re:  [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

mark at codesourcery dot com wrote:

> Unless you can prove that returning a TREE_LIST will not result in a
> problem anywhere else -- including other ice-on-invalid cases -- your
> patch is not a win on the release branch.

There is only one other place where lookup_fnfields is called with protect=2,
which is in build_dtor_call(): the value returned is passed directly to
build_new_method_call, which checks for error_operand_p at the very beginning
and returns error_mark_node. lookup_field is *never* called with protect=2.

I think this is a prove.

If you are still not convinced, I will prepare the easy patch for 3.4.

Giovanni Bajo




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2004-06-15 00:59 -------
Subject: Re:  [3.4/3.5 regression] ICE on ambiguous operator
 new in class hierarchy

giovannibajo at libero dot it wrote:

> If you are still not convinced, I will prepare the easy patch for 3.4.

I don't doubt you, but I'd still be more comfortable with the easy patch 
for 3.4.  Why be smart when we can be lazy? :-)



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-06-15 01:13 -------
Subject: Re:  [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

mark at codesourcery dot com wrote:

>> If you are still not convinced, I will prepare the easy patch for
>> 3.4.
>
> I don't doubt you, but I'd still be more comfortable with the easy
> patch  for 3.4.  Why be smart when we can be lazy? :-)

In fact, I *was* being lazy: a new patch means a new testing cycle for my poor
computer, so pushing my already tested patch was the laziest thing I could do
:)

No problem though, I'll prepare the easy fix for 3.4 and commit it after
testing.

Giovanni Bajo




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-06-15 01:46 -------
Subject: Bug 15967

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	giovannibajo@gcc.gnu.org	2004-06-15 01:46:21

Modified files:
	gcc/cp         : ChangeLog search.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/lookup: crash3.C 

Log message:
	PR c++/15967
	* search.c (lookup_field): Propagate the ambiguity list.
	(lookup_fnfields): Likewise.
	
	PR c++/15967
	* g++.dg/lookup/crash3.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4096&r2=1.4097
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/search.c.diff?cvsroot=gcc&r1=1.291&r2=1.292
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.3858&r2=1.3859
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/lookup/crash3.C.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-06-15 03:14 -------
Subject: Bug 15967

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	gcc-3_4-branch
Changes by:	giovannibajo@gcc.gnu.org	2004-06-15 03:14:16

Modified files:
	gcc/cp         : ChangeLog init.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/lookup: crash3.C 

Log message:
	PR c++/15967
	* search.c (build_new_1): Robustify.
	
	PR c++/15967
	* g++.dg/lookup/crash3.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.3892.2.124&r2=1.3892.2.125
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/init.c.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.356.2.12&r2=1.356.2.13
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.3389.2.208&r2=1.3389.2.209
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/lookup/crash3.C.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=NONE&r2=1.1.2.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-06-15 03:15 -------
Fixed in GCC 3.4.1 and GCC 3.5.0. Thank you for your report!

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[danglin at gcc dot gnu dot org]

------- Additional Comments From danglin at gcc dot gnu dot org  2004-06-19 16:16 -------
This new test fails on hppa2.0-hp-hpux11.00:

Executing on host: /xxx/gnu/gcc-3.4/objdir/gcc/testsuite/../g++ -B/xxx/gnu/gcc-3
.4/objdir/gcc/testsuite/../ /xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/cra
sh3.C  -nostdinc++ -I/xxx/gnu/gcc-3.4/objdir/hppa2.0w-hp-hpux11.00/libstdc++-v3/
include/hppa2.0w-hp-hpux11.00 -I/xxx/gnu/gcc-3.4/objdir/hppa2.0w-hp-hpux11.00/li
bstdc++-v3/include -I/xxx/gnu/gcc-3.4/gcc/libstdc++-v3/libsupc++ -I/xxx/gnu/gcc-
3.4/gcc/libstdc++-v3/libsupc++ -I/xxx/gnu/gcc-3.4/gcc/libstdc++-v3/include/backw
ard -I/xxx/gnu/gcc-3.4/gcc/libstdc++-v3/testsuite -fmessage-length=0   -ansi -pe
dantic-errors -Wno-long-long  -S  -o crash3.s    (timeout = 300)
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:7: error: `operator ne
w' takes type `size_t' (`long unsigned int') as first parameter
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:8: error: `operator ne
w' takes type `size_t' (`long unsigned int') as first parameter
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C: In function `int cras
h()':
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:14: error: no suitable
 or ambiguous `operator new' found in class `C'
compiler exited with status 1
output is:
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:7: error: `operator ne
w' takes type `size_t' (`long unsigned int') as first parameter
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:8: error: `operator ne
w' takes type `size_t' (`long unsigned int') as first parameter
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C: In function `int cras
h()':
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:14: error: no suitable
 or ambiguous `operator new' found in class `C'

PASS: g++.dg/lookup/crash3.C  (test for errors, line 14)
FAIL: g++.dg/lookup/crash3.C (test for excess errors)
Excess errors:
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:7: error: `operator ne
w' takes type `size_t' (`long unsigned int') as first parameter
/xxx/gnu/gcc-3.4/gcc/gcc/testsuite/g++.dg/lookup/crash3.C:8: error: `operator ne
w' takes type `size_t' (`long unsigned int') as first parameter


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-06-19 16:25 -------
I am seeing these also on powerpc-apple-darwin. I will apply a patch which fixes them.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[andreas dot meier_ at gmx dot de]

------- Additional Comments From andreas dot meier_ at gmx dot de  2004-07-01 20:58 -------
Please adjust the target milestone to 3.4.1

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967

[Bug c++/15967] [3.4/3.5 regression] ICE on ambiguous operator new in class hierarchy

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.4.2                       |3.4.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15967