As stated on the official mplayer homepage there are two head overflows. Fixes are available for download on the homepage. I will attach the ported patches. Interestingly, they write about a pre7 version on the homepage, but it's not officially been released yet. Reproducible: Always Steps to Reproduce:
Created attachment 56428 [details, diff] Real RTSP heap overflow patch
Created attachment 56429 [details, diff] MMST heap overflow patch
2005.04.16, Saturday :: MMST heap overflow posted by Roberto Summary A potential buffer overflow was found and fixed in code used to handle MMST streams. Severity High (arbitrary remote code execution under the user ID running the player) when streaming MMS/TCP data from a malicious server, null if you do not use this feature. At this time there is no known exploit. Description While enumerating streams from a server, MMST code stores stream IDs in a fixed length array, but there is no check to stop the process if too many stream IDs are received. A malicious server could announce more than 20 streams and overflow the array. Solution A fix for the vulnerability was checked into MPlayer CVS on Fri Apr 15 23:31:57 2005 UTC. Users of affected MPlayer versions should upgrade to an unaffected MPlayer version. Alternatively a patch is available that can be applied to the MPlayer source tree. Affected versions MPlayer 1.0pre6 and before (including pre6a) Unaffected versions MPlayer 1.0pre7 and after CVS HEAD after Fri Apr 15 23:31:57 2005 UTC ===================== 2005.04.16, Saturday :: Real RTSP heap overflow posted by Roberto Summary A potential buffer overflow was found and fixed in code used to handle RealMedia RTSP streams. Severity High (arbitrary remote code execution under the user ID running the player) when streaming RTSP data from a malicious server, null if you do not use this feature. At this time there is no known exploit. Description While getting lines from a server, Real RTSP code stores them in a fixed size array of MAX_FIELDS elements, but there is no check to stop the process if too many lines are received. A malicious server could send more than MAX_FIELDS lines and overflow the array. Since the array holds pointers to answer strings, an attacker cannot write arbitrary data into it, making an exploit more difficult. Solution A fix for the vulnerability was checked into MPlayer CVS on Fri Apr 15 23:30:44 2005 UTC. Users of affected MPlayer versions should upgrade to an unaffected MPlayer version. Alternatively a patch is available that can be applied to the MPlayer source tree. Affected versions MPlayer 1.0pre6 and before (including pre6a) Unaffected versions MPlayer 1.0pre7 and after CVS HEAD after Fri Apr 15 23:30:44 2005 UTC
media-video, please bump/patch/comment
mplayer-1.0_pre5-r5 is the latest stable version for most architectures. we could apply the patches to mplayer-1.0_pre6* and then we should aim to mark mplayer-1.0_pre6-r1 stable. mplayer related problems which came in lately were related to -r2. when -pre7 will be released we could integrate it via the normal procedur into the tree (with the 30 day testing period etc) before going into stable.
patches are now in the tree, mplayer-1.0_pre6-r{1,2,3} apply them. we should mark -r1 stable, afterwards _pre5 can be removed.
ok, done. now we should mark mplayer-1.0_pre6-r4 stable. changed the following things: mplayer-1.0_pre6-r1 -> mplayer-1.0_pre6-r4 mplayer-1.0_pre6-r2 -> mplayer-1.0_pre6-r5 mplayer-1.0_pre6-r3 -> mplayer-1.0_pre6-r6
Arches, please test mplayer-1.0_pre6-r4 and mark stable
done for amd64 and x86
Stable on ppc.
*** Bug 89384 has been marked as a duplicate of this bug. ***
stable on ppc64
Stable on SPARC.
GLSA drafted, security pls review alpha, pls mark test and mark stable if possible
Stable on alpha.
GLSA 200504-19 thanks everyone