[Bug c++/40076] g++ should not permit types to be defined in compound literals

[Bug c++/40076] g++ should not permit types to be defined in compound literals

[manu at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40076

Manuel López-Ibáñez <manu at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-05-02
                 CC|                            |manu at gcc dot gnu.org
     Ever Confirmed|0                           |1

--- Comment #4 from Manuel López-Ibáñez <manu at gcc dot gnu.org> 2012-05-02 20:04:08 UTC ---
I think this is confirmed. Clang prints:

pr40076.cc:1:17: error: 's3' can not be defined in a type specifier
int i = (struct s3 { int j; }) { 1 }.j;
                ^
1 error generated.

for both C and C++.

GCC prints:

pr40076.cc:1:17: error: initializer element is not constant
 int i = (struct s3 { int j; }) { 1 }.j;
                 ^

which is less clear.

[Bug c++/40076] g++ should not permit types to be defined in compound literals

[joseph at codesourcery dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40076

--- Comment #5 from joseph at codesourcery dot com <joseph at codesourcery dot com> 2012-05-02 20:13:02 UTC ---
On Wed, 2 May 2012, manu at gcc dot gnu.org wrote:

> I think this is confirmed. Clang prints:
> 
> pr40076.cc:1:17: error: 's3' can not be defined in a type specifier
> int i = (struct s3 { int j; }) { 1 }.j;
>                 ^
> 1 error generated.
> 
> for both C and C++.

That's correct for C++ (by the analogy discussed in this bug report), but 
not for C.

> pr40076.cc:1:17: error: initializer element is not constant
>  int i = (struct s3 { int j; }) { 1 }.j;
>                  ^
> 
> which is less clear.

That's a correct error for C, however; C allows types to be defined like 
that in sizeof, casts and compound literals, but that initializer is not 
(required to be) constant in C standard terms.

[Bug c++/40076] g++ should not permit types to be defined in compound literals

[manu at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40076

--- Comment #6 from Manuel López-Ibáñez <manu at gcc dot gnu.org> 2012-05-02 20:20:53 UTC ---
(In reply to comment #5)
> That's correct for C++ (by the analogy discussed in this bug report), but 
> not for C.

My fault. Clang pays more attention to the file extension than to the driver
used. The correct output for C is:

test.c:1:9: error: initializer element is not a compile-time constant
int i = (struct s3 { int j; }) { 1 }.j;
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated

which makes the error clearer than in gcc because of the caret location and the
range (plus the beautiful colors ;-).

[Bug c++/40076] g++ should not permit types to be defined in compound literals

[ian at airs dot com]

------- Comment #3 from ian at airs dot com  2009-05-08 22:19 -------
Yeah, in C you can define a struct type in a type cast and then use it later. 
Cool stuff.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40076

[Bug c++/40076] g++ should not permit types to be defined in compound literals

[pinskia at gcc dot gnu dot org]

------- Comment #2 from pinskia at gcc dot gnu dot org  2009-05-08 21:51 -------
Oh, types are valid to be defined in casts in C.  You did not mention that in
the bug report so I thought you could not define a type in C either.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40076

[Bug c++/40076] g++ should not permit types to be defined in compound literals

[pinskia at gcc dot gnu dot org]

------- Comment #1 from pinskia at gcc dot gnu dot org  2009-05-08 21:41 -------
But is defining a new type in a compound literal legal for C99?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40076