<div dir="ltr"><div><div><div><div><div>We will continue to allow missing eth_type in the netlink attribute to imply Ethernet II type. 802.3 frames requires a specific eth_type attribute. <br><br></div>With Mega flows, we further require a missing eth_type in the key attribute to have a exact match (oxffff) in the eth_type of the mask attribute (if present). <br>
<br></div>In this patch, dpctl will output eth_type(0/0xffff) to indicate mssing eth_type key attribute with a exact match mask, and corresponding parsing logic to support this <br>special case. <br><br></div>IMHO, it will be better to be able to drop the '0' in the output (e.g. eth_type(/0xffff)). I plan to address this when we change how odp parser away from using sscanf(). Other suggestions are welcome. <br>
<br></div>Please let me know if you need more information. <br><br></div>--andy<br></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Wed, Jun 19, 2013 at 10:20 AM, Ben Pfaff <span dir="ltr"><<a href="mailto:blp@nicira.com" target="_blank">blp@nicira.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On Tue, Jun 18, 2013 at 04:15:10PM -0700, Andy Zhou wrote:<br>
> Added support to allow mega flow specified and displayed. ovs-dpctl tool<br>
> is mainly used as debugging tool.<br>
><br>
> This patch also implements the low level user space routines to send<br>
> and receive mega flow netlink messages. Those netlink suppor<br>
> routines are required for forthcoming user space mega flow patches.<br>
><br>
> Added a unit test to test parsing and display of mega flows.<br>
><br>
> Ethan contributed the ovs-dpctl mega flow output function.<br>
><br>
> Co-authored-by: Ethan Jackson <<a href="mailto:ethan@nicira.com">ethan@nicira.com</a>><br>
> Signed-off-by: Ethan Jackson <<a href="mailto:ethan@nicira.com">ethan@nicira.com</a>><br>
> Signed-off-by: Andy Zhou <<a href="mailto:azhou@nicira.com">azhou@nicira.com</a>><br>
<br>
</div>I know there was an offline discussion of how to handle eth_type and<br>
specifically its absence when Ethernet II is not in use. I wasn't<br>
there for that discussion, can you explain the conclusions?<br>
</blockquote></div><br></div>