Hi.
Is it necessary to queue a DPC from a WFP-Classify-Routine (Packet Layer)
before cloning a NetBufferList? Why not clone the NetBufferList directly
inside WFP-Classify-Routine?
The reason I’m asking that question: Here’s a sample of “Out-of-band Packet
Modification from Inbound Datagram Data Layers”
http://msdn2.microsoft.com/en-gb/library/aa938500.aspx
The sample states:
“To prevent WFP from making a deep clone (deep-copying MDLs,
net buffers, net buffer lists, structures, and data mapped by MDLs,
DatagramDataReceiveWorker should be run by a DPC targeting the
processor to which the referenced net buffer list was initially
classified. See KeSetTargetProcessorDpc for DPC targeting.”
So, why to queue a DPC anyway?
Thanks
Frank