Hi, > At 12:55 03/05/2002, Loris Degioanni wrote: > >Hi, > > > > > Hi, > > > > > > I have a small app that sits on my laptop listening for some specific > > > Ethernet packets. Works great (Thanks for the great job!), except when I > > > put my laptop to sleep with the app running, and then resume. After that, > > > my app doesn't seem to receive anything anymore, > > > >I think this is normal with current version of WinPcap. > >NDIS closes the connections when the machine goes to sleep, so WinPcap > >releases them, but the ability to reopen the connections when the machine is > >restarted is not implemented yet. > > Yep, going around the sources shows that there are no PnpEvent handlers, > and the dynamic adapter binding is a bit short still :-) Is there any way I > can help by working on some of those issues? Not a pro of Windows driver > development (yet), but I can try...
Well, you are welcome. As you can see, the code is available and if you need help you can use this mailing list. I will be happy to include any patch that improves the PnP support. :-) > > > and other applications using TCP/IP become sluggish. > > > >This is not normal instaead. At least, I never obeserver such a problem, > >both on laptops and on desktop machines. Are you sure the problem is > >directly related to winpcap? > > Well, as soon as I abort the app that is listening (with an infinite loop > around a pcap_next), performance gets back to normal. Actually, it might > just be that the pcap_next returns immediately and I thus am just eating up > a lot of CPU for no reason. Need to add a bit of error checking, I guess :-) Yes, this is probably the cause of the slowdown. I noticed a similar problem some time ago: PacketReceivePacket() returns immediately, so the application brings the CPU to 100% issuing a huge number of reads. You should probably check the return value. Loris > Jacques. > >
