[Bug preprocessor/14634] New: Unterminated literals not diagnosed

[Bug preprocessor/14634] New: Unterminated literals not diagnosed

[jsm28 at gcc dot gnu dot org]

Unterminated string and character constants are no longer diagnosed when
the preprocessing tokens are not converted to tokens, even with -pedantic
-Wall -Wextra.  This is a regression in quality of implementation.  For
example, the following unportable code is not diagnosed.

#if 0
#error "
'
#endif
#define foo '

Furthermore, this contradicts the cpp documentation.  A previous request
to document these extensions
<http://gcc.gnu.org/ml/gcc-patches/2003-04/msg01855.html>
does not seem to have resulted in any documentation of them, and cpp.texi
still says

# Do not use @code{#if 0} for comments which are not C code.  Use a real
# comment, instead.  The interior of @code{#if 0} must consist of complete
# tokens; in particular, single-quote characters must balance.

and

# Neither @samp{#error} nor @samp{#warning} macro-expands its argument.
# Internal whitespace sequences are each replaced with a single space.
# The line must consist of complete tokens.  It is wisest to make the
# argument of these directives be a single string constant; this avoids
# problems with apostrophes and the like.

-- 
           Summary: Unterminated literals not diagnosed
           Product: gcc
           Version: 3.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: preprocessor
        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=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-03-18 17:08 -------
I thought that this was undefined behavior and that is why 12743 was closed?

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
           Keywords|                            |documentation
   Last reconfirmed|0000-00-00 00:00:00         |2004-03-18 17:08:32
               date|                            |
            Summary|Unterminated literals not   |[3.4/3.5 Regression]
                   |diagnosed                   |Unterminated literals not
                   |                            |diagnosed
   Target Milestone|---                         |3.4.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[jsm at polyomino dot org dot uk]

------- Additional Comments From jsm at polyomino dot org dot uk  2004-03-18 17:13 -------
Subject: Re:  [3.4/3.5 Regression] Unterminated
 literals not diagnosed

On Thu, 18 Mar 2004, pinskia at gcc dot gnu dot org wrote:

> I thought that this was undefined behavior and that is why 12743 was closed?

That it's undefined is why I described it as a quality of implementation
regression rather than a conformance regression.  Quality of
implementation dictates that we should not lose the feature of diagnosing
unportable code even though a diagnostic is not required.  (I prefer a
mandatory error for all these usages - since no good justification was
ever given for these extensions - but a pedwarn if pedantic would suffice
to give portability warnings to those wanting them.)



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[zack at codesourcery dot com]

------- Additional Comments From zack at codesourcery dot com  2004-03-18 17:25 -------
Subject: Re:  [3.4/3.5 Regression] Unterminated
 literals not diagnosed


It used to be a mandatory error, and that was changed for reasons that
I no longer remember.  Can you try to dig up the discussion
surrounding the change?

zw


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[jsm at polyomino dot org dot uk]

------- Additional Comments From jsm at polyomino dot org dot uk  2004-03-18 17:35 -------
Subject: Re:  [3.4/3.5 Regression] Unterminated
 literals not diagnosed

On Thu, 18 Mar 2004, zack at codesourcery dot com wrote:

> It used to be a mandatory error, and that was changed for reasons that
> I no longer remember.  Can you try to dig up the discussion
> surrounding the change?

Neil's patch was
<http://gcc.gnu.org/ml/gcc-patches/2003-04/msg01827.html>.  I didn't and
don't see the value in the extensions listed in the message but not
documented; I don't think there was further discussion of them.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[zack at codesourcery dot com]

------- Additional Comments From zack at codesourcery dot com  2004-03-18 18:15 -------
Subject: Re:  [3.4/3.5 Regression] Unterminated
 literals not diagnosed


The value of extensions 1 and 2 (from Neil's list) was that we used to
get regular bug reports complaining about them not working.  I suspect
there wer complaints about 3 not working, too, but less frequently.

Anyhow, take this up with Neil, I have too many other things on my
plate.

zw



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-05-28 22:17 -------
Neil --

Please comment on this PR.

-- Mark

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |neil at daikokuya dot co dot
                   |                            |uk


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[neil at gcc dot gnu dot org]

------- Additional Comments From neil at gcc dot gnu dot org  2004-05-29 08:29 -------
The behaviour is as I intended.  IMO we don't need a diagnostic.  If someone
feels we should have one I won't oppose a reasonable patch.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |minor


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634

[Bug preprocessor/14634] [3.4/3.5 Regression] Unterminated literals not diagnosed

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-06-05 20:29 -------
I've removed the target milestone; adding a warning for this extension is not
release-critical now or ever.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14634