Wireshark-bugs: [Wireshark-bugs] [Bug 11152] Wireshark decodes some valid RTP packets as STUN pa
      
      
    
     Michael Mann
 changed
              bug 11152
        
          
             
          
            | What | Removed | Added | 
         
           | CC |  | mmann78@netscape.net | 
      
        
            Comment # 9
              on bug 11152
              from  Michael Mann
        (In reply to Hadriel Kaplan from comment #8)
> During testing of this issue and a fix for it, I found several other related
> issues with weak heuristic checks:
>     1) the CATTP heuristic dissector is weak and matched some packets.
>     2) the TFTP heuristic dissector matched data inside TURN ChannelData
> that it shouldn't have, and couldn't be turned off.
>     3) the RTP heuristic dissector couldn't be turned on/off for use inside
> TURN ChannelData separately from UDP.
>     4) the RTCP heuristic dissector also couldn't be turned on/off for use
> inside TURN ChannelData separately from UDP, and performed a odd-port check
> even when inside TURN ChannelData.
> 
> I'll address those in separate code changes, and separate bugs.  Mostly
> because some might need to be back-ported (or not).
I don't want to hijack this bug, but I looked at the suggested changes.  Would
it make more sense to have a "global" list of heuristic dissectors rather than
have individual preferences in individual dissectors?  We have a global list
for dissectors in general, but I guess uses may want to differentiate disabling
a protocol from disabling a heuristic.
         
      
      
      You are receiving this mail because:
      
      
          - You are watching all bug changes.