Wireshark · Ethereal-dev: Re: [Ethereal-dev] Re: New dissectors: H.223 and friends

[Guy Harris <gharris@xxxxxxxxx>]

Richard van der Hoff wrote:

> However, I've spotted a bug in the new desegmentation in iax2, which 
> means that higher-level PDUs aren't dissected when they ought to be, but 
> tend to be grouped together in later packets - I've attached a patch 
> against svn r16110 to fix this; it also makes the code a bit clearer...

Are those higher-level PDUs always in order, so that there's no sequence 
number in them?  If so, then you should be using "fragment_add_seq_next()".

That would, I think, mean that you might be able to avoid doing 
reassembly differently on the first pass and subsequent passes.

That would probably be a good thing, as the code currently will, on the 
first pass, set "complete" to TRUE even if the "fragment_add()" call 
returns NULL, causing a null pointer dereference crash, at least with 
some captures, when it tries to use "fd_head" to point to values to use 
as arguments to "tvb_new_real_data()".

See bugs 515 through 519.