Bugzilla – Bug 102961
VUL-0: CVE-2005-21: gaim problems
Last modified: 2021-11-04 16:26:15 UTC
Date: Mon, 8 Aug 2005 17:57:42 -0400 From: Josh Bressers <bressers@redhat.com> To: vendor-sec@lst.de Subject: [vendor-sec] [mark@kingant.net: [Gaim-packagers] Gaim 1.5.0 and security problems] Parts/Attachments: 1 Shown 83 lines Text 2 OK 79 lines Text 3 OK 27 lines Text ---------------------------------------- This came form the gaim folk: ----- Forwarded message from Mark Doliner <mark@kingant.net> ----- From: "Mark Doliner" <mark@kingant.net> To: gaim-packagers@lists.sourceforge.net In-Reply-To: <e61d333f0508071549132be364@mail.gmail.com> X-Mailer: Open WebMail 2.51 20050228 X-OriginatingIP: 24.136.241.99 (mark@kingant.net) X-Spam-Report: Spam Filtering performed by sourceforge.net. See http://spamassassin.org/tag/ for more details. Report problems to http://sf.net/tracker/?func=add&group_id=1&atid=200001 0.0 SF_CHICKENPOX_PARATHESES_OPEN BODY: Text interparsed with ( 0.0 SF_CHICKENPOX_PERIOD BODY: Text interparsed with . 0.0 SF_CHICKENPOX_COMMA BODY: Text interparsed with , 0.0 SF_CHICKENPOX_SLASH BODY: Text interparsed with / 0.0 SF_CHICKENPOX_MINUS BODY: Text interparsed with - 0.0 SF_CHICKENPOX_BRACKET_OPEN BODY: Text interparsed with [ 0.0 SF_CHICKENPOX_UNDERSCORE BODY: Text interparsed with _ 0.0 SF_CHICKENPOX_QUESTION BODY: Text interparsed with ? 0.0 SF_CHICKENPOX_APOSTROPHE BODY: Text interparsed with ' -0.1 AWL AWL: From: address is in the auto white-list Subject: [Gaim-packagers] Gaim 1.5.0 and security problems Errors-To: gaim-packagers-admin@lists.sourceforge.net X-BeenThere: gaim-packagers@lists.sourceforge.net X-Mailman-Version: 2.0.9-sf.net Precedence: bulk List-Unsubscribe: <https://lists.sourceforge.net/lists/listinfo/gaim-packagers>, <mailto:gaim-packagers-request@lists.sourceforge.net?subject=unsubscribe> List-Id: <gaim-packagers.lists.sourceforge.net> List-Post: <mailto:gaim-packagers@lists.sourceforge.net> List-Help: <mailto:gaim-packagers-request@lists.sourceforge.net?subject=help> List-Subscribe: <https://lists.sourceforge.net/lists/listinfo/gaim-packagers>, <mailto:gaim-packagers-request@lists.sourceforge.net?subject=subscribe> List-Archive: <http://sourceforge.net/mailarchive/forum.php?forum=gaim-packagers> X-RedHat-Spam-Score: 0 Date: Sun, 7 Aug 2005 21:59:40 -0500 On Sun, 7 Aug 2005 18:49:32 -0400, Sean Egan wrote > All, > > Let's release 1.5.0 this week. This comes from oldstatus, and may be > the last release from there. > > -s. First off, I'll be out of down Monday, Tuesday and Wednesday. In light of our pending release on Thursday, we should discuss some security problems (yep--again). I don't know if we'll need CVE numbers for this, I can never remember quite how that works. But if we DO need a CVE number or three, then Warren and Josh, we would all very much appreciate your assisstance, yet again. 1. The gadu-gadu PRPL. There was a memory alignment bug that apparently only affected can not be exploited on x86. I do not believe the ekg/libgadu project was issued a CAN for this. You can grab a patch for Gaim from: http://cvs.sourceforge.net/viewcvs.py/gaim/gaim/src/protocols/gg/libgg.c?r1=1.21 .2.1&r2=1.21.2.2 2. There is an oscar remotely exploitable crash bug. A remote AIM or ICQ user would need to compile their own client and send a specially crafted IM (basically an IM containing lots of %s, and flag it as an away message). It results in a buffer overflow. A fix for this has not yet been commited to CVS, and the issue should be fairly unknown. A patch for this is attached (I would appreciate if some Gaim devs could double-check it--it's small). 3. Daniel Atallah fixed a crash in oscar dealing with invalid file names. It is remotely exploitable by anyone sending you a file with a non-utf8 filename. It sometimes causes a crash in pango. It might be depending on the version of gtk you're using. And seems to crash Linux machines less than Windows machines (if ever). This has not yet been fixed in CVS, but it IS in the wild and you can get a client from http://www.sevenz.net/ to exploit it. A patch is attached. I haven't actually tested it, so hopefully it still applies. The logic is kosher. Other Gaim developers: If you know of other crashes, please follow-up this email. Thanks! -Mark
Created attachment 45235 [details] One of the attchments ...
Created attachment 45236 [details] the other attachment ...
Gadu-Gadu patch mentioned in comment #1 is mentioned in bug 97408 comment #18.
Updated for STABLE, 9.3, 9.2, SLES9-SLD: - Fixed memory alignment bug in libgadu (#102961). - Fixed AIM/ICQ malformed filename crash (#102961). - Fixed AIM/ICQ away message buffer overflow (#102961). Updated for STABLE, 9.1, 9.0: - Fixed AIM/ICQ malformed filename crash (#102961). - Fixed AIM/ICQ away message buffer overflow (#102961). SLES8-SLEC: - Fixed AIM/ICQ away message buffer overflow (#102961). Not fixed for 8.2.
Thanks. I will submit patchinfos. SM-Tracker-2003.
Submitted patchinfos. Go ahead :-)
Created attachment 45469 [details] patchinfo for box ...
Created attachment 45470 [details] patchinfo for maintained ...
The 9.0 - 9.2 gaims use "gaim_utf8_salvage" in oscar_malformed_filename_crash_fix.patch This function does not exist in the 9.0 ... 9.2 versions yet.
For 9.0, 9.1, 9.2 gaim_utf8_salvage() was added as static function to oscar.c. I hope it will fix the problem.
need to re-submit patchinfos?
no. they are checked in already and waiting for QA.
packages released
CAN-2005-2103 (under review) Buffer overflow in the AIM and ICQ module in Gaim before 1.5.0 allows remote attackers to cause a denial of service (application crash) and possibly execute arbitrary code via an away message with a large number of AIM substitution strings, such as %t or %n. CAN-2005-2102 (under review) The AIM/ICQ module in Gaim before 1.5.0 allows remote attackers to cause a denial of service (application crash) via a filename that contains invalid UTF-8 characters. CAN-2005-1852 (under review) Multiple integer overflows in libgadu, as used in Kopete in KDE 3.2.3 to 3.4.1, ekg before 1.6rc3, and other packages, allows remote attackers to cause a denial of service (crash) and possibly execute arbitrary code via an incoming message.
CVE-2005-2103: CVSS v2 Base Score: 7.5 (AV:N/AC:L/Au:N/C:P/I:P/A:P)