Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-10-03 Thread Ben Pfaff
OK, thanks. I'll mark this series as "superseded" in patchwork. On Tue, Oct 04, 2016 at 12:42:19AM +, Daniele Di Proietto wrote: > I'm rebasing the series and considering a different approach for this. > > I'll send something shortly. > > Thanks, > > Daniele > > On 03/10/2016 17:38, "Ben

Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-10-03 Thread Daniele Di Proietto
I'm rebasing the series and considering a different approach for this. I'll send something shortly. Thanks, Daniele On 03/10/2016 17:38, "Ben Pfaff" wrote: >Does this patch need anything more? I don't see it on master and it's >still in patchwork. Same thing for patch 4/5 in

Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-10-03 Thread Ben Pfaff
Does this patch need anything more? I don't see it on master and it's still in patchwork. Same thing for patch 4/5 in this series. On Thu, Sep 01, 2016 at 01:31:06AM +, Daniele Di Proietto wrote: > > On 31/08/2016 11:32, "Jarno Rajahalme" wrote: > > >I’d put the registers

Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-08-31 Thread Daniele Di Proietto
On 31/08/2016 11:32, "Jarno Rajahalme" wrote: >I’d put the registers and metadata field to the ‘false’ and also maybe >non-writeable fields (ether type, frags, nw_proto, etc.) in to >OVS_NOT_REACHED() case, just in case. > > Jarno Agreed, done Thanks, Daniele > >> On Aug

Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-08-31 Thread Daniele Di Proietto
On 31/08/2016 10:38, "Jesse Gross" wrote: >On Tue, Aug 30, 2016 at 6:47 PM, Daniele Di Proietto > wrote: >> When translating a set action we also unwildcard the field in question. >> This is done to correctly translate set actions with the value

Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-08-31 Thread Jarno Rajahalme
I’d put the registers and metadata field to the ‘false’ and also maybe non-writeable fields (ether type, frags, nw_proto, etc.) in to OVS_NOT_REACHED() case, just in case. Jarno > On Aug 31, 2016, at 10:38 AM, Jesse Gross wrote: > > On Tue, Aug 30, 2016 at 6:47 PM,

Re: [ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-08-31 Thread Jesse Gross
On Tue, Aug 30, 2016 at 6:47 PM, Daniele Di Proietto wrote: > When translating a set action we also unwildcard the field in question. > This is done to correctly translate set actions with the value identical > to the ingress flow, like in the following example: > > flow

[ovs-dev] [PATCH v2 1/5] ofproto-dpif-xlate: Don't unwildcard tunnel attributes on set.

2016-08-30 Thread Daniele Di Proietto
When translating a set action we also unwildcard the field in question. This is done to correctly translate set actions with the value identical to the ingress flow, like in the following example: flow table: tcp,actions=set_field:80->tcp_dst,output:5 ingress packet: ...,tcp,tcp_dst=80