Message-ID: <3AB2BCD3.4020002@operamail.com>
From: Sahab Yazdani <sahaby AT operamail DOT com>
Organization: PheonixSoft Interactive
User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; m18) Gecko/20001204
X-Accept-Language: en
MIME-Version: 1.0
Newsgroups: comp.os.msdos.djgpp
Subject: the follies of new...
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Lines: 57
Date: Fri, 16 Mar 2001 20:24:35 -0500
NNTP-Posting-Host: 149.99.118.73
X-Complaints-To: abuse AT sprint DOT ca
X-Trace: newscontent-01.sprint.ca 984792645 149.99.118.73 (Fri, 16 Mar 2001 20:30:45 EST)
NNTP-Posting-Date: Fri, 16 Mar 2001 20:30:45 EST
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

my program keeps crashing to a call to delete every second time my 
function is ran.

so something like this:

void foo() {
	bar = new bar();

	delete bar;
}

void main() {
	foo(); // This runs fine
	foo(); // This will crash
}

I'm checking out the pointers now as this probably where the error 
origintaes.  the problem is that using fprintf to print the address of 
the pointer, i get a rather strange result:

void foo() {
	bar = new bar();
	fprintf( stderr, "bar[afterinit]: %x\n", bar );

	fprintf( stderr, "bar[beforedestruct]: %x\n", bar );
	delete bar;
}

the strangeness is that the first time this function is called I get 
this (these are actual results of my program, obvoisly using function 
foo as an example):

bar[afterinit]: 0x45eb8
bar[ondestruct]: 0x45eb8
bar[oninit]: 0xff4f31f0
bar[ondestruct]: 0xff4f31f0
SIGSEGV afterwards

this shows that the pointer doesn't shift, so the code is okay, but the 
address of the second set of pointers points to memory in the 4GB 
region, which my computer doesn't have (although I wish it did :-)).

now when I run symify on the traceback I get a function called 
___builtin_delete as the final function to be called before the program 
crashes.  could it be that the memory allocation technique in djgpp is a 
bit messed up?? or much more likely that my code is messed up.

hope somebody can help me with this.

-- 
***********************************************************
* Sahab Yazdani * "Wizard's Third Rule: Passion rules     *
* Thornhill S.S * Reason" - Kolo's Journal                *
***********************************************************
* http://pheonixsoft.virtualave.net/                      *
***********************************************************