Re: "double free or corruption" - how to solve this?

["Steve Graegert" <graegerts@gmail.com>]

On 5/12/06, Nate Jenkins <nate@uniwest.com> wrote:
>
> >> Hello Shriramana,
> >>
> >>
> >> On Fri, 12 May 2006 13:41:12 +0530 "Shriramana Sharma"
> >> <samjnaa@gmail.com> wrote:
> >>
> >> > One of my programs, which was working quite well till now, suddenly
> >> > gives me the error:
> >> >
> >> > *** glibc detected *** double free or corruption (top): 0x0808a338 ***
> >> > Aborted
> >> >
> >> > It is a pure C program compiled with GCC 4.02 -- I do not understand
> >> > why it does not work suddenly. Please tell me what the above error can
> >> > be.
> >>
> >> Would be nice to run it from gdb, in order to get the backtrace when it
> >> crashes. This should help you understand where and why :).
> >
> > Additionally, you can try valgrind, which reports typical programming
> > errors like calling free() twice on the same object.
> >
> > \Steve
> > -
>
> Steve,
>
> I just saw that utility the other day.  It looks interesting.  How do you
> typically use it?  or grindcall-grindval?  Do you know of a good tutorial
> you could point me/us to?

Nate,

The easiest way to use valgrind is to pass it the "apparently faulty"
program including all arguments using the --tool switch:

graegerts@kenji:~/vmware/haiku> valgrind --tool=memcheck ls -la

Afterwards a summary is printed:

==11048== Memcheck, a memory error detector for x86-linux.
==11048== Copyright (C) 2002-2004, and GNU GPL'd, by Julian Seward et al.
==11048== Using valgrind-2.2.0, a program supervision framework for x86-linux.
==11048== Copyright (C) 2000-2004, and GNU GPL'd, by Julian Seward et al.
==11048== For more details, rerun with: -v
==11048==

[lengthy file listing omitted]

==11048==
==11048== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 23 from 1)
==11048== malloc/free: in use at exit: 14901 bytes in 21 blocks.
==11048== malloc/free: 446 allocs, 425 frees, 48474 bytes allocated.
==11048== For a detailed leak analysis,  rerun with: --leak-check=yes
==11048== For counts of detected errors, rerun with: -v
graegerts@kenji:~/vmware/haiku>

By entering "valgrind" in the console a list of available tools is
listed, all of them obeying the --help switch:

% valgrind
valgrind: Missing --tool option
Available tools:
        helgrind
        addrcheck
        cachegrind
        memcheck
        callgrind
        corecheck
        lackey
        none
        massif
% valgrind --tool=memcheck --help

The valgrind homepage (http://valgrind.org) is a good starting point
for documentation and description of related concepts.  Hope that was
helpful.

Bye

	\Steve