Hi, all

kernel:3.4.34 64bit
host:vsphere4/5
NIC:82545EM
NIC driver:7.3.21-k8-NAPI
memory:4G
CPU: 2 cores

I am facing this in the kernel logs:

Jul 26 00:11:32 localhost kernel: e1000 0000:02:00.0: eth0: Detected Tx
Unit Hang
Jul 26 00:11:32 localhost kernel:   Tx Queue             <0>
Jul 26 00:11:32 localhost kernel:   TDH                  <9f>
Jul 26 00:11:32 localhost kernel:   TDT                  <9f>
Jul 26 00:11:32 localhost kernel:   next_to_use          <9f>
Jul 26 00:11:32 localhost kernel:   next_to_clean        <b2>
Jul 26 00:11:32 localhost kernel: buffer_info[next_to_clean]
Jul 26 00:11:32 localhost kernel:   time_stamp <10003327b>
Jul 26 00:11:32 localhost kernel:   next_to_watch <b3>^C
Jul 26 00:11:32 localhost kernel:   jiffies <10117e8c2>
Jul 26 00:11:32 localhost kernel:   next_to_watch.status <0>

When I run "ethtool -K eth0 tx off"

This problem disappears. I know this command will shutdown tx checksum
offloading. Thus this will increase the host cpu loads.
And from 82545EM spec, this kind of nic can support tx checksum offloading.

I want to know whether it is a bug of driver? It would be great if a fix
could be found.

Best Regards!
zhuyj




------------------------------------------------------------------------------
Introducing Performance Central, a new site from SourceForge and 
AppDynamics. Performance Central is your source for news, insights, 
analysis and resources for efficient Application Performance Management. 
Visit us today!
http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
_______________________________________________
E1000-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit 
http://communities.intel.com/community/wired

Reply via email to