[ovs-dev] [PATCH v4 3/6] tests/ofproto-dpif.at: Avoid testing with incomplete TCP packets.

[Ben Pfaff]

No, I'd just like that example added to the commit message.  I
understand the issue now.

Acked-by: Ben Pfaff <blp at nicira.com>

On Tue, Dec 03, 2013 at 01:45:04PM -0800, Jarno Rajahalme wrote:
> Ben,
> 
> Do you want changes to this patch?
> 
>   Jarno
> 
> On Dec 3, 2013, at 1:25 PM, Jarno Rajahalme <jrajahalme at nicira.com> wrote:
> 
> > 
> > On Dec 3, 2013, at 8:25 AM, Ben Pfaff <blp at nicira.com> wrote:
> > 
> >> On Tue, Nov 26, 2013 at 11:24:49AM -0800, Jarno Rajahalme wrote:
> >>> Use nw_proto values that generate less debug logging due to parsing
> >>> errors.
> >>> 
> >>> Signed-off-by: Jarno Rajahalme <jrajahalme at nicira.com>
> >> 
> >> This seems reasonable, but what kind of debug logging was generated?
> > 
> > 
> > "expected but not present?, due to nw_proto being e.g. ICMP, but no ICMP payload in the packet, or the same for TCP:
> > 
> > 2013-12-03T21:18:49Z|00313|odp_util|DBG|expected but not present: tcp: in_port(1),eth(src=40:44:44:44:44:41,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no)
> > 
> > I tried changing this to an error to catch them all, but I recall there were cases where getting rid of the logging would have been non-trivial.
> > 
> > This logging is triggered only by netdev-dummy/receive that takes the packet in odp syntax, so I was a bit over-jealous in fixing the traces, too.
> > 
> >  Jarno
> > 
>