Hello "d",
On Mon, Jul 20, 2020 at 6:23 PM d via openssl-users <
openssl-users@openssl.org> wrote:
> No memory leakage, if I comment out all SSL related constructs from my
> code.
> Thus, the SSL related code parts seem to be the cause of this leak.
> What is the issue here?
> Makefile:
>
SSL_
On 03/12/14 10:34, T@Run..! Polisetty wrote:
> Hai All,
>
>We are using Openssl for DTLS Negotiations. When we run the Valgrind
> with this setup. We are finding some major loss of memory at one place.
>
>
> ==23871== 4,224 (1,056 direct, 3,168 indirect) bytes in 3 blocks are
On Wed, Dec 03, 2014 at 04:04:16PM +0530, T@Run..! Polisetty wrote:
> Hai All,
>
>We are using Openssl for DTLS Negotiations. When we run the Valgrind
> with this setup. We are finding some major loss of memory at one place.
Can you check with a current git version? There have b
I found my memory leak origin. I didn't free an X509 object when I called
SSL_get_peer_certificate function (missing X509_free() call). Hope it will be
useful for others
Cheers!!
From: hicham55...@hotmail.com
To: openssl-users@openssl.org
Subject: RE: Memory leak in ssl3_get_server_certif
Well, I found out where it's defined. It's a MACRO definition (
IMPLEMENT_ASN1_ENCODE_FUNCTIONS_fname ). I'm still investigating the leak
Regards
From: hicham55...@hotmail.com
To: openssl-users@openssl.org
Subject: Memory leak in ssl3_get_server_certificate
Date: Thu, 3 Apr 2014 14:4
On Wed, Dec 21, 2011 at 1:26 PM, nandan shantharaj wrote:
> Hi All,
> SSL_CTX_load_verify_locations() is causing memory leak in my
> application. Folowing is the function trace.
>
> 262 1072 bytes leaked in 4 blocks (2.25% of all bytes leaked)
> 263 These range in size from 268 to 268
>>OpenSSL artifacts that are created by the application, then handed to
>>some other part of the API for use, so who is now responsible for the
>>destructions of them? That kind of misunderstanding.
>>Darryl
Hi Darryl,
To an extent you are correct. We have fixed the issue and it is due to o
Please provide a simple testcase program exhibiting the memory leak.
Please try with the most recent OpenSSL release. There are no known
issue with normal usage, however some recent patches have been added for
theoretical failure path memory leaks.
It is possible for memory leaks to still
I was able to solve the problem, the session cache was activated and
stored all the sessions and a bug in DTLS caused that twice as much
sessions as necessary have been created. The bug will be fixed with an
updated version of bugfix #2028.
- Robin
Am 02.09.2009 um 10:47 schrieb Robin Seg
: openssl-users@openssl.org
Subject: Re: memory leak in openssl
It looks like you did not do an SSL_free when the connection was shut down.
-Kyle H
On Mon, Jun 22, 2009 at 1:16 AM, Vivek Subbarao wrote:
> Hi,
>
>
>
> I am using openssl 0.9.8e that comes along with the linux rhel 5 u3
It looks like you did not do an SSL_free when the connection was shut down.
-Kyle H
On Mon, Jun 22, 2009 at 1:16 AM, Vivek Subbarao wrote:
> Hi,
>
>
>
> I am using openssl 0.9.8e that comes along with the linux rhel 5 u3 OS. We
> have a client server architecture where both the client and server
enssl-users@openssl.org
Subject: Re: Memory Leak Creating a CSR
On Sat, May 30, 2009, Paul Suhler wrote:
> Hi.
>
> Using OpenSSL 0.9.8i, I'm getting a memory leak when I create a CSR.
> My process is taken more-or-less from the Viega, et al. book:
>
No idea what that versio
On Sat, May 30, 2009, Paul Suhler wrote:
> Hi.
>
> Using OpenSSL 0.9.8i, I'm getting a memory leak when I create a CSR. My
> process is taken more-or-less from the Viega, et al. book:
>
No idea what that version is but the one in demos/x509/mkreq.c doesn't leak
memory.
Steve.
--
Dr Stephen N.
Commenting below.
> For better clarification, I have attached the trace of Valgrind on the
> Pastebin:
> http://pastebin.com/f1e222abd
> Here is the last lines
>
>1. ==3290== LEAK SUMMARY:
>2. ==3290==definitely lost: 268 bytes in 1 blocks.
>3. ==3290==indirectly lost: 66
Hello,
> I have been testing my SSL server for 2 days after 30k connections
> and memory remains calm.I assume the probelm has been solved by just adding
> ERR_free_strings();
> ERR_remove_state(0);
> to each children thread before pth
pthread_exit()
Thanks everyone for the help!
Lei.C
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Marek Marcola
Sent: Friday, May 26, 2006 2:48 PM
To: openssl-users@openssl.org
Subject: RE: Memory leak with multiple threads running with
asingleSSLserver
> After accepting many incoming SSL connections hence many threads creation
> and deletion
> the memory usage of the program will increase up to a higher percentage
> after some time.
That doesn't mean that the memory is leaked.
> Before i applied those cleanup functions,mem usage will p
Hello,
> ENGINE_cleanup();
> CONF_modules_unload();
> CRYPTO_cleanup_all_ex_data();
> ERR_free_strings();
> ERR_remove_state(0);
This should by removed from your code.
Best regards,
--
Ma
suggestions,please?
Thanks!
Lei
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of David Schwartz
Sent: Friday, May 26, 2006 2:16 PM
To: openssl-users@openssl.org
Subject: RE: Memory leak with multiple threads running with a
singleSSLserver serving incoming
] Behalf Of Chong Peng
Sent: Friday, May 26, 2006 1:53 PM
To: openssl-users@openssl.org
Subject: RE: Memory leak with multiple threads running with a
singleSSLserver serving incoming requests...
lei:
have you enabled session cache? if you enabled session cache and you have a
lot of ssl connections
> But i still have memory leaks after large amount of
> connections(simulated by
> my own crazy test program which connects to the server recursively with
> different
> messages)
How did you establish that the memory was leaked?
DS
_
o: openssl-users@openssl.org
Subject: RE: Memory leak with multiple threads running with a
singleSSLserver serving incoming requests...
Thanks for pointing that out for me about ERR_remove_state(0);
But i still have memory leaks after large amount of connections(simulated by
my own crazy test pro
Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Marek Marcola
Sent: Friday, May 26, 2006 1:32 PM
To: openssl-users@openssl.org
Subject: RE: Memory leak with multiple threads running with a
singleSSLserver serving incoming requests...
Hello
What version of OpenSSL you use ?
I
Hello
What version of OpenSSL you use ?
I do not remember exactly but in one version
there was memory leak after general release.
(I do not remember version).
Best regards,
--
Marek Marcola <[EMAIL PROTECTED]>
__
OpenSSL Projec
Hello,
> For each of the worker-thread:
> by adding ERR_remove_state(0) before the pthread_exit() will
> free up
> the error string mem allocations.
This will free error queue from this thread - should be used.
> For the Server thread:
> after each SSL con
: openssl-users@openssl.org
Subject: Re: Memory leak with multiple threads running with a single
SSLserver serving incoming requests...
Hello,
> when it ends:
> /*cleanup the BIO chain*/
>BIO_flush(_io);
>_io->references = 0;
>BIO_free(_io);
>
Hello,
> when it ends:
> /*cleanup the BIO chain*/
>BIO_flush(_io);
>_io->references = 0;
>BIO_free(_io);
>_io = NULL;
>_my_struct->bio->references = 0;
>BIO_free_all(_my_struct->bio);
>_my_struct->bio = NULL;
>_my_struct->ssl
Hi All
the best thing to track openSSL memory leak (as per my little
understanding) is to have a debug file with Boundschecker and than
provide some traffic and shutdown ur app. In case of leak its caught
by the boundschecker.
Now use some memory allocator (like Buddy Alloc etc...) and than have
> I've been doing some research into this and so far the main
> cause seems
> to be that we were not calling SSL_free after each connection
> closed.
> From what I have gathered, the correct way to cleanup after a
> connection is to call the following functions in this order:
>
> SSL_sh
Do you have session cache enabled? I have had the same suspicion in the
past.
Following links may be usefull.
http://www.openssl.org/docs/ssl/SSL_CTX_set_session_cache_mode.html
http://www.openssl.org/docs/ssl/SSL_CTX_set_timeout.html#
-VJ
-Original Message-
From: [EMAIL PROTECTED]
[mailt
Calling Err_remove_state(0) before the thread exits
fixed it :)
--- Calista <[EMAIL PROTECTED]> wrote:
> Hi,
>
> if(!(x509 = d2i_X509_fp(fpin, NULL))) {
>fseek(fpin, 0, SEEK_SET);
>x509=PEM_read_X509(fpin, NULL, NULL, NULL);
> }
>
> CRYPTO_MDEBUG's output is:
> [09:36:15] 1661 file=lha
On Mon, Nov 22, 2004, Jason Davis wrote:
>
> I am getting a large number of memory leaks,
>
>
[snip]
>
>
> This causes a large memory leak an since this occuors frequently This is a
> big problem. I seam to have freed averything, what Have I missed
>
>
Check the FAQ for some hints.
>
>
On Tue, Aug 31, 2004, Carlos Roberto Zainos H wrote:
> Hi guys and hi Eric !
>
> I have been tested my app again and again and following the Dr Stephen
> recommendations, I discovered that the BIO's are not the problem, all they points to
> NULL after them has been freed.
>
> I think the pro
Can you run your server for thousands of iterations to see if the
memory continues to be consumed? Generally memory that has been
allocated by the C library is not returned to the OS. Instead those
pages are cached to handle future allocations without needing to
request them from the OS.
If yo
Hawk wrote:
Hi
I wrote a multithreaded program using openssl
But with every connection it grows up about 8kb
I think I freed all SSL* and SSL_CTX* I used
I used this functions to make my program threadsafe:
Perhaps anyone can give me a hint if there is an error?
With best regards
Stephan
void CRYPT
In message <[EMAIL PROTECTED]> on Wed, 13 Aug 2003 18:30:12 +0530, "Chandrasekhar R S"
<[EMAIL PROTECTED]> said:
rsr> Hello
rsr>I have tried to use OpenSSL 0.9.7b with Apache 1.3.28 (with
rsr> associated mod_ssl) and have detected significant memory leak.
Do you have the possibility to analy
> Is there a reason you are running mod_ssl and openssl.
> One is a complete replacement for the other
You are very confused.
mod_ssl is the "glue" that adds in OpenSSL to Apache 1.3.x.
In Apache 2.0 mod_ssl is integrated into the apache core.
/r$
--
Rich Salz Chief Secu
> Hello
> I have tried to use OpenSSL 0.9.7b with Apache 1.3.28
> (with associated
> mod_ssl) and have detected significant memory leak.
>
> With the same Apache Server (1.3.28) and with OpenSSL
> 0.9.6j there is no
> memory leak.
>
> Did anyone else also found this observat
If you are using Linux I stroingly recommend to try Valgrind:
http://developer.kde.org/~sewardj/
a very nice memory leaks/memory access check tool.
Aleksey
Krishnaswamy R. wrote:
>Hi all,
>
>I have written a TLS client using the OpenSSL library. It uses memory
>BIOs as the input/output B
]
==
- Original Message -
From: "Scott Frazor" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Monday, February 04, 2002 10:20 AM
Subject: RE: Memory leak
> I have also noticed a leak under Windows/32 using 9.6a. I have started
the
> ta
]
Subject: RE: Memory leak
I have also noticed a leak under Windows/32 using 9.6a. I have started the
task of locating the source of the leak. My memory "tester" says it is
OpenSSL. I have done everything I can to insure that everything is closed
and freed. The best I can get is down to
I have also noticed a leak under Windows/32 using 9.6a. I have started the
task of locating the source of the leak. My memory "tester" says it is
OpenSSL. I have done everything I can to insure that everything is closed
and freed. The best I can get is down to a 4K leak.
I'm not convicned it
TED]
Cc: [EMAIL PROTECTED]
Subject: Re: Memory Leak?? I can't find it.
Make sure to call
ERR_remove_state(0);
in all threads after the transfer
ERR_remove_state() frees the error queue associated with
thread pid. If pid == 0, the current thread will have its
error queue r
[EMAIL PROTECTED] wrote:
>
> I was wondering if this is a known issue with this particular release of
> OPEN SSL Crypto library. The dll I ran BoundsChecker against is
> Lib32eay.dll, and
> the reported memory leak indicates 16 bytes of allocated memory through
> Crypto_malloc function in \crypt
From: "Martin Skinner" <[EMAIL PROTECTED]>
N8Schwaermer> Thank you for the informative answer.
YW
N8Schwaermer> Since MS VC++ and boundschecker notices these leaks, it
N8Schwaermer> would be nice to have a "cleanup" routine which could be
N8Schwaermer> called at the end of the app.
I've been t
PROTECTED]>
To: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Sent: Wednesday, September 13, 2000 11:03 PM
Subject: Re: Memory leak in SSL_CTX_new / SSL_CTX_free thru a call to
CRYPTO_get_ex_new_index()
> From: "Martin Skinner" <[EMAIL PROTECTED]>
>
> N8Schwa
Actually, I saw the same leaks, but since they were fixed in size, I did
not bother about them. But now that someone has raised this question,
can't the memory be freed at some point of time instead of leaking?
I must say though, that very mysteriously this leak has disappeared from
my code, now t
From: "Martin Skinner" <[EMAIL PROTECTED]>
N8Schwaermer> So, a single call to CRYPTO_get_ex_new_index() allocated
N8Schwaermer> these memory blocks and nothing frees them.
N8Schwaermer> Am I missing something?
This is perfectly normal. What happens is that a number of functions
are registered f
Hi Marko,
Summary:
- IO::Socket::SSL has problems with broken connections
- Crypt::SSL (0.17, includes Net::SSL) still has a memory leak
- Net::SSLeay::get_https works fine
(with own eval()/die() timeout wrapper)
> you mentioned having both Crypt-SSLeay and IO-Socket-SSL in your
On Wed, 6 Sep 2000 [EMAIL PROTECTED] wrote:
> when using the following combinations I got memory leaks
> by sending SSL-requests via LWP objects. I do not
> got such problems when using e.g. Net::SSLeay::get_https
> directly.
you mentioned having both Crypt-SSLeay and IO-Socket-SSL in your
envi
encap2.ond
On Tue, Jul 18, 2000 at 12:09:24PM +0200, Berki Lukacs Tamas wrote:
> I have just the same problem as Bill Rebey. I have an SSL server which
> verifies client certificates by setting SSL_CTX_set_verify to
> SSL_VERIFY_PEER|SSL_FAIL_IF_NO_PEER_CERT, and every time an SSL_accept
> succeeds, there is
52 matches
Mail list logo
