K 10
svn:author
V 5
andre
K 8
svn:date
V 27
2012-11-09T17:56:49.492798Z
K 7
svn:log
V 326
Move per packet dequeueing from bge_rxeof() into a new function
bge_rx_packet() and merge bge_rxcsum() into it.

To avoid the unlock/lock pair for injection into the stack dequeue
all packets from the rx DMA ring into an m_nextpkt chain and then
inject that chain one by one into the stack w/o any locking.

Work in progress.

END