On Aug 15, 2012, at 12:10 PM, Guy Harris wrote:
> On Aug 15, 2012, at 8:12 AM, Evan Huus wrote:
>
>> I'm a fan of a macro like Jakub mentioned as part of the old conversation:
>>
>> http://www.wireshark.org/lists/wireshark-dev/201105/msg00205.html
>
> ...but let's have the macro report a dissector bug if it sees a null pointer. The dissector in question should probably say something other than just "(null)" if whatever routine supplied the pointer failed for some reason.
...or the non-dissector:
http://anonsvn.wireshark.org/viewvc?revision=44512&view=revision
"unspecified" is better than "(null)" in that context.
Perhaps, instead, the macro should take *two* arguments:
#define NULL_CHECK(p, substitute) ((p) != NULL ? (p) : (substitute))
and force everybody to decide what the protocol tree item/message/whatever should say if the pointer *is* null.