Moving Right window edge to the left on sender side without additional check
can lead to the TCP deadlock, when receiver ACKs proper segment, while sender
discards it for future ACK. To prevent this add check if usable window (or
shrink amount in this case) is bigger then receiver's window scale factor.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Andrey Tepin <atepin@kraftway.ru>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
Tcp driver need to use EFI_D_NET to log DEBUG message,
So it becomes easy to separate/filter out debug messages
from network stack versus generic EFI_D_INFO debugs.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Zhang Lubo <lubo.zhang@intel.com>
Cc: Hegde Nagaraj P <nagaraj-p.hegde@hpe.com>
Cc: Subramanian Sriram <sriram-s@hpe.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Cc: Ye Ting <ting.ye@intel.com>
Reviewed-by: Sriram Subramanian <sriram-s@hpe.com>
Consider the situation as shown in below chart. The last ACK message has
acknowledged the Tcb->RcvWl2, and all the segments until Tcb->RcvNxt have
been received by TCP driver. The Tcb->RcvNxt is not acknowledged due to the
delayed ACK. In this case an incoming segment (Seg->Seq, Seg->End) should
not be accepted by TCP driver, and an immediate ACK is required.
Current TcpSeqAcceptable() thought it’s an acceptable segment incorrectly, it
continues the TcpInput() process instead of sending out an ACK and droping the
segment immediately.
Tcb->RcvWl2 Tcb->RcvNxt Tcb->RcvWl2 + Tcb->RcvWnd
Seg->Seq Seg->End | |
| | | | |
---+-----+---------------+-------------+--------------------------+-----------
<income segment> <----Acceptable Range--- -->
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-By: Eugene Cohen <eugene@hp.com>
Reviewed-By: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-By: Ye Ting <ting.ye@intel.com>
Arbitrary length of packet may be received from network, including the
packets with zero payload data or malformed protocol header. So the code
much check the actually received data size before using it. For example, in
current edk2 network stack, an zero payload UDP packet may cause the
platform ASSERT in NetbufFromExt() because of the zero fragment number.
This patch update the IpIoLib and UdpIoLib to check and discard the zero
payload data packet to avoid above assert. Some other network drivers are
also updated to check the packet size to guarantee the minimum length of
protocol header is received from upper layer driver.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: Sriram Subramanian <sriram-s@hpe.com>
Reviewed-by: Wu Jiaxin <jiaxin.wu@intel.com>
Fu Siyuan
Zhang Lubo
BALATON Zoltan
hhtian