[Bug c++/14278] New: fails when calling through a function pointer

[Bug c++/14278] New: fails when calling through a function pointer

[jbrandmeyer at earthlink dot net]

Also fails on i686-pc-linux-gun.

When calling a function through a function pointer, and the return type is
passed to a constructor whose return type is also passed through a constructor,
G++ 3.4 fails to parse the expressions.

The feature this test excersizes is used by num_util, a collection of helper
functions that extends Boost.Python's interface with Numeric Python.  G++ 3.3
compiles this without error.

Thanks,
Jonathan Brandmeyer

pion:~/src jonathan$ g++-3.4 -v -c funcptr_crash.cpp 
Reading specs from /usr/local/lib/gcc/powerpc-apple-darwin7.2.0/3.4.0/specs
Configured with: ../gcc/configure --program-suffix=-3.4
--enable-version-specific-runtime-libs --disable-shared --enable-static
--enable-languages=c,c++
Thread model: posix
gcc version 3.4.0 20040223 (prerelease)
 /usr/local/libexec/gcc/powerpc-apple-darwin7.2.0/3.4.0/cc1plus -quiet -v
-D__DYNAMIC__ funcptr_crash.cpp -fPIC -quiet -dumpbase funcptr_crash.cpp
-auxbase funcptr_crash -version -o /var/tmp//ccRG1xS6.s
ignoring nonexistent directory "NONE/include"
ignoring nonexistent directory
"/usr/local/lib/gcc/powerpc-apple-darwin7.2.0/3.4.0/../../../../powerpc-apple-darwin7.2.0/include"
#include "..." search starts here:
#include <...> search starts here:
 /usr/local/lib/gcc/powerpc-apple-darwin7.2.0/3.4.0/include/c++
 /usr/local/lib/gcc/powerpc-apple-darwin7.2.0/3.4.0/include/c++/powerpc-apple-darwin7.2.0
 /usr/local/lib/gcc/powerpc-apple-darwin7.2.0/3.4.0/include/c++/backward
 /usr/local/include
 /usr/local/lib/gcc/powerpc-apple-darwin7.2.0/3.4.0/include
 /usr/include
End of search list.
GNU C++ version 3.4.0 20040223 (prerelease) (powerpc-apple-darwin7.2.0)
        compiled by GNU C version 3.4.0 20040223 (prerelease).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
funcptr_crash.cpp:38: error: expected `)' before "visual_PyArrayHandle"
funcptr_crash.cpp:38: error: expected `)' before "visual_PyArrayHandle"
funcptr_crash.cpp:38: error: expected `,' or `...' before "visual_PyArrayHandle"
funcptr_crash.cpp:38: error: function `object ret(handle<PyObject>)' is
initialized like a variable
funcptr_crash.cpp:38: error: expected `,' or `;' before ')' token
----------------------------------------------------------------------  


typedef struct _object {
 int ob_refcnt; 
 struct _typeobject *ob_type;
} PyObject;

void **visual_PyArrayHandle;


template <class T = PyObject>
class handle
{
 public:

    template <class Y>
    explicit handle(Y* p)
        : m_p(static_cast<T*>(p))
    {
    }

    T* get() { return m_p; }
 private:
    T* m_p;
};

class object {
 private:
     PyObject* obj;
 public:
    template<class T>
    object( handle<T> other) :obj(static_cast<PyObject*>(other.get()))
    {}
};

// Changing the function pointer type to have less than these two arguments
// passes.  This one fails:
object ret(handle<>
      ((*(PyObject * (*)( int, int)) visual_PyArrayHandle[0])(0, 0)));
// This passes:
handle<> ret2
      ((*(PyObject * (*)( int, int)) visual_PyArrayHandle[0])(0, 0));
// As does this:
object ret3(handle<>
      ((*(PyObject * (*)( int)) visual_PyArrayHandle[0])(0)));

-- 
           Summary: fails when calling through a function pointer
           Product: gcc
           Version: 3.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: jbrandmeyer at earthlink dot net
                CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: powerpc-apple-darwin7.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-02-24 19:15 -------
I think this is valid code, anyways this is caused by Mark's patch to have better error messages for 
missing a ",".

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mmitchel at gcc dot gnu dot
                   |                            |org
           Keywords|                            |rejects-valid
            Summary|fails when calling through a|[3.4/3.5 Regression] fails
                   |function pointer            |when calling through a
                   |                            |function pointer
   Target Milestone|---                         |3.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2004-02-24 20:15 -------
Here's something smaller: 
---------------------- 
struct S {}; 
 
struct X { 
    X(S* p); 
}; 
 
struct A { 
    A(X); 
}; 
 
void *p_fun; 
A ret(X ((*(S* (*)(int, int)) p_fun)(0, 0))); 
------------------ 
 
W. 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2004-02-24 20:20 -------
OK, this is the cleanest I can come up with: 
---------------- 
struct X { 
    X(int p); 
}; 
 
struct A { 
    A(X); 
}; 
 
void *p_fun; 
A a(X ((*(int (*)(int, int)) p_fun)(0, 0))); 
-------------------- 
 
I think the code is valid. It fails with 3.4 and mainline, though, 
while it compiles with 3.3.4: 
 
g/x> /home/bangerth/bin/gcc-3.3.4-pre/bin/c++ -c x.cc 
 
g/x> /home/bangerth/bin/gcc-3.4-pre/bin/c++ -c x.cc 
x.cc:10: error: expected `)' before "p_fun" 
x.cc:10: error: expected `)' before "p_fun" 
x.cc:10: error: expected `,' or `...' before "p_fun" 
x.cc:10: error: function `A a(X)' is initialized like a variable 
x.cc:10: error: expected `,' or `;' before ')' token 
 
g/x> /home/bangerth/bin/gcc-3.5-pre/bin/c++ -c x.cc 
x.cc:10: error: expected `)' before "p_fun" 
x.cc:10: error: expected `)' before "p_fun" 
x.cc:10: error: expected `,' or `...' before "p_fun" 
x.cc:10: error: function `A a(X)' is initialized like a variable 
x.cc:10: error: expected `,' or `;' before ')' token 
 
W. 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
      Known to fail|                            |3.4.0 3.5.0
      Known to work|                            |3.3.4
   Last reconfirmed|0000-00-00 00:00:00         |2004-02-24 20:20:07
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-02-26 18:07 -------
Actually, you can't cast a void* to a function pointer, but (besides the parse 
error) we don't enforce this right now.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2004-02-26 19:30 -------
I don't think that's the problem here. This code should 
be valid, then: 
------------------------ 
struct X {  
    X(int p);  
};  
  
struct A {  
    A(X);  
};  
 
int (*p_fun)(int, int); 
A a(X ((*(int (*)(int, int)) p_fun)(0, 0))); 
--------------------- 
since we only cast p_fun to its exact same type. It fails with the 
same error, though. 
 
W. 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[giovannibajo at libero dot it]

------- Additional Comments From giovannibajo at libero dot it  2004-02-26 19:59 -------
Wolfgang, 

I was not clear. I was saying that there is indeed a parse error (which is the 
bug covered in this PR), but the testcase you proposed was invalid because of 
the conversion issue (even if GCC doesn't currently enforce it, even for 
straightforward "void* p; (void (*)(void))p;"). 

Your new testcase is of course valid C++, so now it's a rejects-valid by all 
means.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-02-27 00:30 -------
Subject: Bug 14278

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	gcc-3_4-branch
Changes by:	mmitchel@gcc.gnu.org	2004-02-27 00:30:47

Modified files:
	gcc/testsuite  : ChangeLog 
	gcc/cp         : ChangeLog parser.c 
Added files:
	gcc/testsuite/g++.dg/parse: comma1.C 

Log message:
	PR c++/14278
	* parser.c (cp_parser_parameter_declaration_list): Commit
	to fewer tentative parses.
	
	PR c++/14278
	* g++.dg/parse/comma1.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.3389.2.110&r2=1.3389.2.111
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/parse/comma1.C.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=NONE&r2=1.1.2.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.3892.2.60&r2=1.3892.2.61
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/parser.c.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.157.2.14&r2=1.157.2.15



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-02-27 00:31 -------
Subject: Bug 14278

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	mmitchel@gcc.gnu.org	2004-02-27 00:31:50

Modified files:
	gcc/cp         : ChangeLog config-lang.in parser.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/parse: comma1.C 

Log message:
	PR c++/14278
	* parser.c (cp_parser_parameter_declaration_list): Commit
	to fewer tentative parses.
	
	PR c++/14278
	* g++.dg/parse/comma1.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.3972&r2=1.3973
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/config-lang.in.diff?cvsroot=gcc&r1=1.20&r2=1.21
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/parser.c.diff?cvsroot=gcc&r1=1.174&r2=1.175
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.3545&r2=1.3546
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/parse/comma1.C.diff?cvsroot=gcc&r1=1.1&r2=1.2



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278

[Bug c++/14278] [3.4/3.5 Regression] fails when calling through a function pointer

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-02-27 00:32 -------
Fixed in GCC 3.4.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14278