[Bug preprocessor/11064] New: Preprocessor tries to tokenize code that should be unconditionally skipped

[Bug preprocessor/11064] New: Preprocessor tries to tokenize code that should be unconditionally skipped

[rschiele@uni-mannheim.de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11064

           Summary: Preprocessor tries to tokenize code that should be
                    unconditionally skipped
           Product: gcc
           Version: 3.3.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: preprocessor
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: rschiele@uni-mannheim.de
                CC: gcc-bugs@gcc.gnu.org

gcc tries to tokenize code that is wrapped around with #if 0 ... #endif 
 
Try compiling the following code: 
 
#if 0 
" 
#endif 
 
Although the quotes are no valid C token, to my opinion the compiler should not care about it, 
as the section should be completely skipped by the preprocessor. But in fact, the compiler 
teminates by the message 
 
missing terminating " character 
 
gcc 3.4 behaves in a sane way here.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug preprocessor/11064] Preprocessor tries to tokenize code that should be unconditionally skipped

[pinskia@physics.uc.edu]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11064


pinskia@physics.uc.edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


------- Additional Comments From pinskia@physics.uc.edu  2003-06-02 14:04 -------
This is not a bug, the ISO/ANSI C says they have be tokens inside the #if ... #endif blocks, 
even if it is #if 0 ... #endif.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug preprocessor/11064] Preprocessor tries to tokenize code that should be unconditionally skipped

[neil@daikokuya.co.uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11064



------- Additional Comments From neil@daikokuya.co.uk  2003-06-02 19:44 -------
Subject: Re:  Preprocessor tries to tokenize code that should be unconditionally skipped

pinskia@physics.uc.edu wrote:-

> PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.
> 
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11064
> 
> 
> pinskia@physics.uc.edu changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>              Status|UNCONFIRMED                 |RESOLVED
>          Resolution|                            |INVALID
> 
> 
> ------- Additional Comments From pinskia@physics.uc.edu  2003-06-02 14:04 -------
> This is not a bug, the ISO/ANSI C says they have be tokens inside the #if ... #endif blocks, 
> even if it is #if 0 ... #endif.

Yes, but it leaves tokenization in the case of unterminated quotes to
the implementation, so we can do either.  As the reporter ovserved, 3.4
accepts the code.  3.3 won't change.

Neil.




------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug preprocessor/11064] Preprocessor tries to tokenize code that should be unconditionally skipped

[jsm28@cam.ac.uk]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11064



------- Additional Comments From jsm28@cam.ac.uk  2003-06-02 19:51 -------
Subject: Re:  Preprocessor tries to tokenize code
 that should be unconditionally skipped

On Mon, 2 Jun 2003, neil@daikokuya.co.uk wrote:

> Yes, but it leaves tokenization in the case of unterminated quotes to
> the implementation, so we can do either.  As the reporter ovserved, 3.4
> accepts the code.  3.3 won't change.

Does 3.4 (a) properly warn for unportable code that has unterminated
quotes (in preprocessing tokens that are not converted to tokens) and (b)  
document these extensions (I didn't see any response to
<http://gcc.gnu.org/ml/gcc-patches/2003-04/msg01855.html>)?





------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug preprocessor/11064] Preprocessor tries to tokenize code that should be unconditionally skipped

[rschiele@uni-mannheim.de]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11064



------- Additional Comments From rschiele@uni-mannheim.de  2003-06-02 20:03 -------
Neil: I do not have a strong opinion about the issue. As the standard forces to have valid 
tokens there, it is ok for me. Just didn't know about this constraint before. 
 
Joseph: No, 3.4 does not warn, even not with -W -Wall -pedantic. 



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.