Date: Sat, 26 Apr 2003 20:25:09 +0300 From: "Eli Zaretskii" Sender: halo1 AT zahav DOT net DOT il To: djgpp-workers AT delorie DOT com Message-Id: <7263-Sat26Apr2003202509+0300-eliz@elta.co.il> X-Mailer: emacs 21.3.50 (via feedmail 8 I) and Blat ver 1.8.9 In-reply-to: <3EAA7D27.E14BD900@yahoo.com> (message from CBFalconer on Sat, 26 Apr 2003 08:35:51 -0400) Subject: Re: nmalloc revisited References: <10304252113 DOT AA21892 AT clio DOT rice DOT edu> <3EA9EBED DOT E8ADB007 AT yahoo DOT com> <9628-Sat26Apr2003113504+0300-eliz AT elta DOT co DOT il> <3EAA7D27 DOT E14BD900 AT yahoo DOT com> Reply-To: djgpp-workers AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk > Date: Sat, 26 Apr 2003 08:35:51 -0400 > From: CBFalconer > > In my particular case it wouldn't work cleanly. I have gcc set up > to automatically use -W -Wall -ansi -pedantic (via a 4dos alias). > However I can still access those things with a #include > for compilation from the command line, thus overriding > the -pedantic for those things alone. What would be the purpose of such a trick? A program compiled with strict ANSI compliance cannot possibly use a non-ANSI library function, or else it invades the user's namespace. If you do that to get a better compiler diagnostics, simply add more "-Wfoo" warning switches to the command line. > It is no big deal as long as someone else does the integration > into stdlib.h, but I consider it easier to keep them separate. Dropping the necessary additions into the non-ANSI, non-Posix portion of stdlib.h is not hard. I'd suggest that you do that, it will probably make the code be accepted faster. > Can you make that .txh I sent work correctly and send it back, so > I can edit it again? I asked for diffs, not the entire file. The file is large, so I didn't want to spend my time finding the differences between the CVS version and the one you produced. If you send diffs, I promose to review them and get back to you. > I plan to split the malldbg section off from > the malloc section, so only the new stuff gets altered. It doesn't matter: even if the malldbg section is split, we still ask for diffs, not a full new version each time. > I need the proper outline of things, so I just fill in the blanks. ??? The outline is already there: you see it in the original malloc.txh. Or maybe I don't understand what outline you need.