On Tuesday, November 06, 2012 04:25:27 Sven Eckelmann wrote:
All packet headers in front of an ethernet header have to be completely divisible by 2 but not by 4 to make the payload after the ethernet header again 4 bytes boundary aligned.
A packing of 2 is necessary to avoid extra padding at the end of the struct caused by a structure member which is larger than two bytes. Otherwise the structure would not fulfill the previously mentioned rule to avoid the misalignment of the payload after the ethernet header. It may also lead to leakage of information when the padding it not initialized before sending.
Signed-off-by: Sven Eckelmann sven@narfation.org
Because we had protests against the approach of splitting the 32 bit seqno... here is a different approach. Now get your dices to find out what David may will like.
packet.h | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-)
Applied in revision 52f79c4.
Thanks, Marek