Bug List: (This bug is not in your last search results)   Show last search results      Search page      Enter new bug
Bug#: 89976
Alias:
Product:
Component:
Status: RESOLVED
Resolution: FIXED
Assigned To: Gentoo Security <security@gentoo.org>
Hardware:
OS:
Version:
Priority:
Severity:
Reporter: Carsten Lohrke <carlo@gentoo.org>
Add CC:
CC:
Remove selected CCs
URL:
Summary:
Status Whiteboard:
Keywords:
Flags: Requestee:
 
 
  ()

Filename Description Type Creator Created Size Actions
Create a New Attachment (proposed patch, testcase, etc.) View All

Bug 89976 depends on: Show dependency tree
Bug 89976 blocks:

Additional Comments: (this is where you put emerge --info)


Not eligible to see or edit group visibility for this bug.






View Bug Activity   |   Format For Printing   |   XML   |   Clone This Bug


Description:   Opened: 2005-04-21 14:22 0000
Announcement-ID: XSA-2004-8

Summary:
By a user receiving data from a malicious network streaming server, an
attacker can overrun a heap buffer, which can, on some systems, lead to
or help in executing attacker-chosen malicious code with the permissions
of the user running a xine-lib based media application.

Description:
Both the MMS and Real RTSP streaming client code made some too strong
assumptions on the transfered data. Several critical bounds checks were
missing, resulting in the possibility of heap overflows, should the remote
server not adhere to these assumptions. In the MMS case, a remote server
could present content with too many individual streams, in the RTSP case,
a remote server's reply could have too many lines.
An attacker can setup a server delivering malicious data to the users. This
can be used to overflow a heap buffer, which can, with certain implementations
of heap management, lead to attacker chosen data written to the stack. This
can cause attacker-chosen code being executed with the permissions of the
user running the application. By tricking users to retrieve a stream, which
can be as easy as providing a link on a website, this vulnerability can be
exploited remotely.

Severity:
This is difficult to exploit remotely, because the indirection involved
requires precision and knowledge of the target machine: The heap overlow
needs to alter heap management information in a way so that a return adress
on the stack is modified. This adress must lead to some malicious code to be
executed, which needs to be injected somehow.
The involved xine plugin is part of the standard xine installation, so we
consider this problem to be moderately severe.

Affected versions:
All 0.9 releases starting with and including 0.9.9
All 1-alpha releases.
All 1-beta releases.
All 1-rc releases.
The 1.0 release.

Unaffected versions:
All releases older than 0.9.9.
1.0.1 or newer.

Solution:
The enclosed patches which have been applied to xine-lib CVS fix the problem
but should only be used by distributors who do not want to upgrade.
Otherwise, we strongly advise everyone to upgrade to the soon to arrive 1.0.1
release of xine-lib.
As a temporary workaround, you may delete the files "xineplug_inp_mms.so" and
"xineplug_inp_rtsp.so" from the xine-lib plugin directory, losing the ability
to use MMS or Real RTSP streaming content.

Patches:
http://cvs.sourceforge.net/viewcvs.py/xine/xine-lib/src/input/librtsp/rtsp.c?r1=1.18&r2=1.19&diff_format=u
http://cvs.sourceforge.net/viewcvs.py/xine/xine-lib/src/input/mms.c?r1=1.55&r2=1.56&diff_format=u

For further information and in case of questions, please contact the xine
team. Our website is http://xinehq.de/

------- Comment #1 From Diego E. 'Flameeyes' Pettenò 2005-04-21 15:06:57 0000 -------
xine-lib-1.0-r2 committed with combined patch fixing the problem.

------- Comment #2 From Sune Kloppenborg Jeppesen 2005-04-21 22:47:42 0000 -------
Arches please test and mark stable.

------- Comment #3 From Michael Hanselmann (hansmi) (RETIRED) 2005-04-22 13:55:16 0000 -------
Stable on ppc.

------- Comment #4 From Simon Stelling (RETIRED) 2005-04-22 14:20:23 0000 -------
amd64 done

------- Comment #5 From Jan Brinkmann (RETIRED) 2005-04-23 01:55:01 0000 -------
stable on x86

------- Comment #6 From Markus Rothe 2005-04-23 23:04:46 0000 -------
stable on ppc64

------- Comment #7 From SpanKY 2005-04-24 01:12:26 0000 -------
arm/ia64 stable

------- Comment #8 From Bryan Østergaard (RETIRED) 2005-04-24 02:21:37 0000 -------
Stable on alpha.

------- Comment #9 From Michael Hanselmann (hansmi) (RETIRED) 2005-04-24 11:42:58 0000 -------
Stable on hppa.

------- Comment #10 From Gustavo Zacarias (RETIRED) 2005-04-25 08:50:03 0000 -------
sparc will need a backport of this fix.
Against our current stable 1_rc6-r1 rtsp applies cleanly but mms doesn't.
1.0 doesn't work for sparc as it is, and i haven't tested 1.0.1 yet, which doesn't seem to be present in portage yet.

------- Comment #11 From Diego E. 'Flameeyes' Pettenò 2005-04-25 09:33:17 0000 -------
I've backported the fix to 1_rc6 and I'm building the package to be sure it
works fine at least on amd64. I'll submit 1_rc6-r2 asap.

------- Comment #12 From Gustavo Zacarias (RETIRED) 2005-04-26 10:02:05 0000 -------
rc6-rc2 sparc happy.
Thanks Diego.

------- Comment #13 From Gustavo Zacarias (RETIRED) 2005-04-26 10:02:50 0000 -------
ehmm that would be rc6-r2, but i think you've got the idea :)

------- Comment #14 From Thierry Carrez (RETIRED) 2005-04-26 13:57:54 0000 -------
GLSA 200504-27

Bug List: (This bug is not in your last search results)   Show last search results      Search page      Enter new bug