net/e1000: fix buffer overrun while i219 processing DMA
authorXiao Zhang <xiao.zhang@intel.com>
Sat, 20 Jul 2019 17:01:37 +0000 (01:01 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Tue, 23 Jul 2019 12:31:36 +0000 (14:31 +0200)
commitf0027fa182c7f3badea96a1e993eb74aae53309d
tree00a48b7d1955dec522853fe95a953da0582cbc66
parenta8cbae89da782997631ed27160dd61686664ea00
net/e1000: fix buffer overrun while i219 processing DMA

IntelĀ® 100/200 Series Chipset platforms reduced the round-trip
latency for the LAN Controller DMA accesses, causing in some high
performance cases a buffer overrun while the I219 LAN Connected
Device is processing the DMA transactions. I219LM and I219V devices
can fall into unrecovered Tx hang under very stressfully UDP traffic
and multiple reconnection of Ethernet cable. This Tx hang of the LAN
Controller is only recovered if the system is rebooted. Slightly slow
down DMA access by reducing the number of outstanding requests.
This workaround could have an impact on TCP traffic performance
on the platform. Disabling TSO eliminates performance loss for TCP
traffic without a noticeable impact on CPU performance.

Please, refer to I218/I219 specification update:
https://www.intel.com/content/www/us/en/embedded/products/networking/
ethernet-connection-i218-family-documentation.html

Cc: stable@dpdk.org
Signed-off-by: Xiao Zhang <xiao.zhang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
drivers/net/e1000/base/e1000_ich8lan.h
drivers/net/e1000/em_rxtx.c