Within LabVIEW memory you will have to know *exactly* what you are
doing. Mistakes are not allowed. It is much easier to capture packets
than memory.
I think LabVIEW (your vi) will stop crashing after removing the sizeof
line to the GetMemoryByPtr(). Take a look inside GetMemoryByPtr() and
see what you have done
🙂If the data to be expected is not properly being delivered by
GetMemoryByPtr(), what I would do is write a few simple routines, put
those routines inside a dll together with testdata. Then start testing
GetMemoryByPtr() inside LabVIEW by positioning that dll in a fast
running while loop, to check if that vi really works the way you think
it works and keeps on working after some time. Taking a look inside
extcode.h will help too.
Good luck !
Jac.
On Fri, 7 Feb 2003 02:10:42 -0600 (CST), Nuno Tavares
wrote:
>Well, just tried to interface LabView with WinPCap (BTW, this WinPCap
>thing is Windows port of Unix-flavours libcap, thus it's OpenSource
>too). I've implemented the C structures with clusters and I think I
>got the idea on DLL importing. But..
>
>I have to work with pointers and many kind of C structures. I'm just
>trying to do the interface promisc-uous "open", but LabView crashes
>(using WinXP, I'll try Win98 tomorrow). I don't know if it's my
>mistake, so I'd like to check the code if you can, it's available at
>http://bars2.gaesi.com/~nmct/labview/
>You will need all the files, as I use (from the examples section)
>href="http://sine.ni.com/apps/we/niepd_web_display.display_epd4?p_guid=B123AE0CBB45111EE034080020E7...
.
>
>I'd like that someone pointed me my mistake, because I've spent some
>time building this little thing, otherwise I'll have to try Ethereal
>(as Underflow) interface, though it's not WinPCap fault, for
>certain...
>
>Technically, what this VI does is:
>
>I assumed C pointers are treated as I32 (in C, pointers are usually
>cast'ed to "long int", so I referred to
>href="http://zone.ni.com/devzone/conceptd.nsf/webmain/B26A875ACA51C567862567CA0055FF24?opendocument...),
>then I get the memory position returned by pcap_open_live(). I also
>assume that GetMemoryByPtr() stops on NULL (otherwise it should be a
>valid pointer). Then I've instructed the control FlattenFromString to
>split that memory block in a cluster I've designed (which represents C
>"struct pcap" - and its composed by other structs/clusters).
>
>Note: Use TestApp.EXE to test your device (after installing
>href="http://winpcap.polito.it/">WinPCap) to get to WPCAP device.
