About:
Valgrind is a tool that helps you find memory management problems in programs. When a program is run under Valgrind's supervision, all reads and writes of memory are checked, and calls to malloc/new/free/delete are intercepted. As a result, Valgrind can detect problems such as use of uninitialized memory, reading/writing of memory after it has been freed, reading/writing off the end of malloced blocks, reading/writing inappropriate areas on the stack, memory leaks in which pointers to malloced blocks are lost forever, passing of uninitialized and/or unaddressable memory to system calls, and mismatched use of malloc/new/new [] vs. free/delete/delete [].

Author:
Julian Seward <jseward __at__ acm __dot__ org> [contact developer]

Rating:	reset 1 2 3 4 5 6 7 8 9 10
8.71/10.00 (89 votes)

Homepage:
http://www.valgrind.org
Tar/BZ2:
http://www.valgrind.org/downloads/valgrind-3.3.1.tar.bz2
Changelog:
http://www.valgrind.org/docs/manual/dist.news.html
Bug tracker:
http://bugs.kde.org
Mailing list archive:
http://sourceforge.net/[..]hive/forum.php?forum=valgrind-developers

Trove categories: [change]

[Development Status]		5 - Production/Stable
[Environment]		Console (Text Based)
[Intended Audience]		Developers
[License]		OSI Approved :: GNU General Public License (GPL)
[Operating System]		POSIX :: Linux
[Programming Language]		C++
[Topic]		Software Development :: Testing

Dependencies: [change]
No dependencies filed

 Branches

Branch	Version	Last release	License	URLs
Default	3.3.1	20-Jun-2008	GNU General Public License (GPL)
Unstable	2.1.1	16-May-2004	GNU General Public License (GPL)

 Comments

[»] Just Awesome
by kodgehopper - Jan 8th 2004 06:20:14

There's just no other way to describe it.

This program is an absolute must for any C/C++ developer. My compliments to the developers, you've really done an outstanding job.

And thanks for saving me from a long day of bug tracking.

--
kodge

[reply] [top]

[»] I like it a lot. Any plans for stack variable checking?
by David Trusty - Jan 7th 2003 12:01:38

I like this tool a lot!

Are there any plans for checking stack variables?

For example:
int arr[10];
arr[11] = ...;

David

[reply] [top]

[»] valgrind is my new favorite toy!
by Adam Sah - Oct 10th 2002 21:42:44

We love valgrind at Addamark, and have already
spread the religion throughout our development
process. It's *way better* than Purify ever was,
(a) fewer false-positives and (b) easier to run,
i.e. don't have to 'purify' the executable.

Julian,Nick-- hats off!! and btw I love the ideas
behind the nextgen architecture.

[reply] [top]

[»] great tool!
by sin281 - Aug 25th 2002 02:02:15

Valgrind is an excellent memory debugging tool. I
have used it effectively to squish a memory leak
where many others had failed. Excellent work,
developers! This software is a godsend to any
programmer.

[reply] [top]

[»] Excellent tool
by Xavier Roche - Aug 15th 2002 05:17:40

Wow, at last we have an efficient memory leak debugger on Linux/i386! Thanks for this tool, which is the only way to cleanup complex code, detect leaks/buffer overflows and unitialized areas.
For information, similar tools (like purify) are rare, and generally costs thousands of dollars.

[reply] [top]

[»] Problems running it, author doesn't answer E-mail...
by erl - Aug 14th 2002 11:21:48

I built valgrind 1.0.0 ok, but when I run the test "valgrind ls -l" according to the README, I get errors starting with:

==3478== pthread_mutex_unlock: mutex is not locked
==3478== at 0x403818C6: __pthread_mutex_unlock (vg_libpthread.c:829)
==3478== by 0x40378B01: __register_frame_info_bases (in /mnt/hde1/usr/lib/libgcc_s.so.1)
==3478== by 0x40046D3E: (within /mnt/hde1/usr/lib/valgrind/valgrind.so)
==3478== by 0x40046BE1: (within /mnt/hde1/usr/lib/valgrind/valgrind.so)

followed by the ls -l listing, followed by more errors, and finally valgrind crashes with a segmentation fault.

I sent an E-mail to the author, but received no reply :-(

[reply] [top]

[»] Re: Problems running it, author doesn't answer E-mail...
by Jan Willem - Aug 14th 2002 17:57:25


> I built valgrind 1.0.0 ok, but when I
> run the test "valgrind ls -l"
> according to the README, I get errors
> starting with:
>
> ==3478== pthread_mutex_unlock: mutex is
> not locked
> ==3478== at 0x403818C6:
> __pthread_mutex_unlock
> (vg_libpthread.c:829)
> ==3478== by 0x40378B01:
> __register_frame_info_bases (in
> /mnt/hde1/usr/lib/libgcc_s.so.1)
> ==3478== by 0x40046D3E: (within
> /mnt/hde1/usr/lib/valgrind/valgrind.so)
> ==3478== by 0x40046BE1: (within
> /mnt/hde1/usr/lib/valgrind/valgrind.so)
>
> followed by the ls -l listing, followed
> by more errors, and finally valgrind
> crashes with a segmentation fault.
>
> I sent an E-mail to the author, but
> received no reply :-(
>
Hmmm, I've also been successful in building Valgrind, and the example just works here. Could you be more specific about the ``kind of system you use'': gcc version, kernel version, etc.?


--
Greetings, -- JaWi

[reply] [top]

---

[»] Re: Problems running it, author doesn't answer E-mail...
by erl - Aug 15th 2002 03:22:08


>
> Hmmm, I've also been successful in
> building Valgrind, and the example just
> works here. Could you be more specific
> about the ``kind of system you use'':
> gcc version, kernel version, etc.?
>
Sure (I was in the mail to the author, btw).
Actually, Valgrind seems to work despite on my own code despite these errors.
Here are my specs:

Kernel 2.4.19-pre10 on an SMP machine
ls (fileutils) 4.1
gcc 3.0.2 (for building valgrind, I dunno what ls was built with)
glibc-2.2.2

Are there any other specs you want?
I can give you the complete error output from valgrind if you want.


[reply] [top]

---

[»] Re: Problems running it, author doesn't answer E-mail...
by Jan Willem - Aug 15th 2002 04:53:37


>
> %
> % Hmmm, I've also been successful in
> % building Valgrind, and the example
> just
> % works here. Could you be more
> specific
> % about the ``kind of system you use'':
> % gcc version, kernel version, etc.?
> %
>
>
> Sure (I was in the mail to the author,
> btw).
> Actually, Valgrind seems to work
> despite on my own code despite these
> errors.
> Here are my specs:
>
> Kernel 2.4.19-pre10 on an SMP machine
> ls (fileutils) 4.1
> gcc 3.0.2 (for building valgrind, I
> dunno what ls was built with)
> glibc-2.2.2
>
> Are there any other specs you want?
> I can give you the complete error
> output from valgrind if you want.
>
From the info you gave me, I only can presume that it's GCC which is causing troubles. Its 3.0 release was quite buggy (the 3.1 release is 'much' better/stable) and therefore I think that you should try to build Valgrind to another (2.95.xx or 3.1.x) version of GCC to see if this works...

Success,




--
Greetings, -- JaWi

[reply] [top]

---

[»] Thanks to the developers!
by Pavel Roskin - Jul 31st 2002 00:02:15

Using Valgrind on Midnight Commander, I could find bugs that were sitting in the sources for many years. Yes, it were bugs, one of them had 1/256th probability of occurring. The internal memory debugger could not find those bugs. Valgrind is an excellent tool. In fact, it's a reason to keep an Intel-based machine around!

[reply] [top]