TCP differences in 7.2 vs 7.1

[Pyun YongHyeon]

On Thu, May 14, 2009 at 11:28:43AM +0300, Lars Eggert wrote:
> Hi,
> 
> On 2009-5-14, at 11:27, Pyun YongHyeon wrote:
> >Then you're seeing different problem on em(4). Last time I checked
> >em(4) TSO code in em(4) didn't use m_pullup and just returned
> >ENXIO to caller. I'm not sure that is related with your issue but
> >would you tell us your network configuration?
> 
> this box is a Dell 2950 server/router running 7.2-STABLE. It has an  
> onboard bce interface and four dual-port Intel PRO/1000 NICs, giving  
> it 8 em interfaces. (Let me know if you want the boot dmesg.)
> 
> >If you can easily
> >reproduce the issue would you let us know?
> 
> Reproducing the issue is as easy as setting net.inet.tcp.tso=1.
> 
> What's interesting is that I only see the issue on one of the eight em  
> interfaces. That interface is connected to a D-Link DIR-655 WLAN  
> router. When I tcpdump on the other interfaces with TSO enabled, I see  
> no "IP bad-len 0" messages.
> 

Would you try attached patch? I'm using the patch on my development
box. Originally the patch was written to address checksum offload
breakage on multicast packets(r182463).
However I didn't encounter TSO issue without the patch. Note, the
patch was not heavily tested so it may have uncovered bugs.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: em.csum_tso.patch
Type: text/x-diff
Size: 6312 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20090515/be903cc6/em.csum_tso.bin