Ethereal-users: [Ethereal-users] Re: Ethereal crashes when Diameter spans multipleTCP packets
Hi,
I have checked in a change to not display a time before 00:00:00 UTC,
January 1, 1970,
Could you download the latest automated build to see if it fixes the crash.
Best regards
Anders
Hi,
Yes Ethereal shouldn't crash if an invalid Time format is used, the problem
is ( I think) in the way the
Diameter dissector calculates the time to be able to present it.
Unfortunatly I don't have any Diameter
traces with "time" in them to verify any code change, also I'm a bit busy at
the moment but I'll try to have a
look at it as soon as I can. A quick Fix would be to not present a time
where the value is less than
2208988800 ( which is the factor used to convert between "unix time(?)" and
NTP time).
Best regards
Anders
-----Original Message-----
From: ethereal-users-bounces@xxxxxxxxxxxx
[mailto:ethereal-users-bounces@xxxxxxxxxxxx]On Behalf Of Michael Kopp
Sent: den 23 februari 2005 10:09
To: ethereal-users@xxxxxxxxxxxx
Subject: [Ethereal-users] Re: Ethereal crashes when Diameter spans
multipleTCP packets
Hi Anders,
>It looks to me like your time format is wrong
You were right, the time format of the application was/is wrong :-(
But this does not explain why ethereal crashes, when the type of the event
timestamp AVP will be changed to "Time" ?
><avp name="Event-Timestamp" code="55" mandatory="must">
> <type type-name="Time"/>
></avp>
if I change this to Unsigned32 or Interger32
<avp name="Event-Timestamp" code="55" mandatory="must">
<type type-name="Unsigned32"/>
</avp>
Everything works fine, except that the time will not get decoded.
I think also the format is wrong, "1108118735" (see also ethereal
-diameter.gif) is a valid NTP time (ok a little bit early in time -
somewhere around 1935 :-) ) So I suspect ethereal should not crash in this
situation, or?
So what is responsible for decoding Time Typdefinition .
libxml2 or ethereal diameter dissector, so I could have a deeper look (also
I am not very good in programming.)
As you said that the XML files are all outdated I will have a look and will
update them with the newest drafts and RFCs regarding Diameter and
contribute this to the devel list. )
Regards
Michael
