[ovs-dev] [PATCH 2/2] datapath: Disable LRO from userspace instead of the kernel.

Justin Pettit jpettit at nicira.com
Sat Aug 27 06:56:16 UTC 2011


On Aug 19, 2011, at 8:39 PM, Jesse Gross wrote:

> The final piece that I just thought of, is the relationship of the MTU
> check in vport_send() and skb_warn_if_lro().  The former will catch
> hardware LRO and interfaces with mismatched MTU and the later the
> deprecated (but still extant software LRO, precursor to GRO).  When
> disabling LRO, we will turn off both the software and hardware
> versions so there is some overlap.  The checks are useful but they'll
> trigger warnings if we start using LRO.  That's not strictly a
> backwards compatibility problem but it will still create a little bit
> of a versioning mess.  What if we just move all of them to
> netdev_send(), which is where they are actually useful?


If I understand what you're proposing, I'm not entirely clear on the benefit.  What's the benefit of putting those checks in netdev_send() as opposed to vport_send() and netdev_port_receive()?  I may be misunderstanding what you're suggesting.  Were you thinking something along the lines of the attached patch?

--Justin


-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-datapath-Move-LRO-checks-to-netdev_send.patch
Type: application/octet-stream
Size: 3577 bytes
Desc: not available
URL: <http://mail.openvswitch.org/pipermail/ovs-dev/attachments/20110826/73d1f9be/attachment-0004.obj>
-------------- next part --------------





More information about the dev mailing list