Re: [ovs-dev] [Single DP 02/15] flow: Set ttl in flow_compose().

2012-10-22 Thread Simon Horman
On Thu, Oct 18, 2012 at 12:51:47PM -0700, Justin Pettit wrote: > Thanks to Ben Pfaff for immediately pinpointing the likely location of > the issue. > > Signed-off-by: Justin Pettit Tested-by: Simon Horman ___ dev mailing list dev@openvswitch.org htt

Re: [ovs-dev] [PATCH] flow: Set ip_ttl in flow_compose()

2012-10-22 Thread Simon Horman
On Mon, Oct 22, 2012 at 07:53:12PM -0700, Ben Pfaff wrote: > On Tue, Oct 23, 2012 at 11:33:32AM +0900, Simon Horman wrote: > > I am unsure if this is intentional or an oversight, but flow_compose() does > > not set ip->ip_ttl and thus always produces packets with TTL 0. > > > > This seems to be in

Re: [ovs-dev] [PATCH] flow: Set ip_ttl in flow_compose()

2012-10-22 Thread Ben Pfaff
On Tue, Oct 23, 2012 at 11:33:32AM +0900, Simon Horman wrote: > I am unsure if this is intentional or an oversight, but flow_compose() does > not set ip->ip_ttl and thus always produces packets with TTL 0. > > This seems to be incorrect in the case of using ovs-appctl > netdev-dummy/receive with a

[ovs-dev] [PATCH] flow: Set ip_ttl in flow_compose()

2012-10-22 Thread Simon Horman
I am unsure if this is intentional or an oversight, but flow_compose() does not set ip->ip_ttl and thus always produces packets with TTL 0. This seems to be incorrect in the case of using ovs-appctl netdev-dummy/receive with a non-zero IP TTL. Signed-off-by: Simon Horman --- lib/flow.c

Re: [ovs-dev] [PATCH 3/3] User-Space MPLS actions and matches

2012-10-22 Thread Simon Horman
On Tue, Oct 16, 2012 at 01:33:03PM +0900, Simon Horman wrote: > This patch implements use-space datapath and non-datapath code > to match and use the datapath API set out in Leo Alterman's patch > "user-space datapath: Add basic MPLS support to kernel". > > The resulting MPLS implementation suppor

Re: [ovs-dev] [Single DP 14/15] dpif: Add new dpif_port_exists() function.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:59PM -0700, Justin Pettit wrote: > Provide the ability to determine whether a port exists in a datapath > without having to deal with a "dpif_port" structure as with > dpif_port_query_by_name(). A future patch will use this function. > > Signed-off-by: Justin Pettit

Re: [ovs-dev] [Single DP 13/15] ofproto: Add type "run" and "run_fast" provider methods.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:58PM -0700, Justin Pettit wrote: > Add the ability for ofproto providers to have top-level "run" and > "run_fast" methods, similar to the current ofproto ones. There are no > current users, but this will be useful in a future commit. > > Signed-off-by: Justin Pettit

Re: [ovs-dev] [Single DP 12/15] ovs-dpctl: Make specifying datapath optional for some commands.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:57PM -0700, Justin Pettit wrote: > A future commit will make all bridges use a single backing datapath. > This commit makes the "dp" argument for "dump-flows" and "del-flows" > optional, since there will typically only be one actual datapath. > > Signed-off-by: Justin

Re: [ovs-dev] [Single DP 11/15] ofproto-dpif: Introduce subfacet_destroy_batch().

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:56PM -0700, Justin Pettit wrote: > A future commit will introduce another caller that wants to destroy > batches of subfacets. Repurpose expire_batch() to be a more generic > method for destroying batches of subfacets. > > Signed-off-by: Justin Pettit Looks good, t

Re: [ovs-dev] [Single DP 10/15] ofproto-dpif: Add ovs-appctl commands for ovs-dpctl functions.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:55PM -0700, Justin Pettit wrote: > These commands will be useful in a future commit that makes multiple > bridges share a single backing datapath. The ovs-dpctl commands will > show information about the backing datapath, so it will be difficult to > determine which i

Re: [ovs-dev] [Single DP 04/15] Use ODP ports in dpif layer and below.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:49PM -0700, Justin Pettit wrote: > @@ -89,7 +91,7 @@ struct flow { > uint8_t arp_tha[6]; /* ARP/ND target hardware address. */ > uint8_t nw_ttl; /* IP TTL/Hop Limit. */ > uint8_t nw_frag;/* FLOW_FRAG_* flags. */ > -uin

Re: [ovs-dev] [Single DP 09/15] tests: Rewrite unit tests to not expect bridge with odp zero.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:54PM -0700, Justin Pettit wrote: > A future commit will make all bridges of a particular type share a > single backing datapath. That backing datapath will have a datapath > port number of zero and bridges will be assigned other numbers. This > commit modifies the te

Re: [ovs-dev] [Single DP 08/15] Separate OpenFlow port numbers from datapath ones.

2012-10-22 Thread Ben Pfaff
On Thu, Oct 18, 2012 at 12:51:53PM -0700, Justin Pettit wrote: > In a future commit, we will make multiple bridges share a single backing > datapath. Our simple mapping from datapath to OpenFlow port numbers > won't work, since we'll want the same OpenFlow port numbers on different > bridges. For

[ovs-dev] Hi friend

2012-10-22 Thread justina
Hi friend how are you today?well i believe you are doing great Greetings from Justina ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

[ovs-dev] ARP corner case in ovs_flow_extract()

2012-10-22 Thread Ben Pfaff
There's a corner case in ovs_flow_extract() that seems a bit troublesome. For an ARP with an opcode other than 1 or 2, we fill in key->ip.proto with the opcode but we do not adjust key_len to include the ARP info (so it won't get hashed). However, in ovs_flow_to_nlattrs() we do include the opcode

Re: [ovs-dev] [PATCH 1/2 v2] ovs-ctl.in: Ability to save flows and kernel datapath config.

2012-10-22 Thread Ben Pfaff
On Fri, Oct 19, 2012 at 06:32:14PM -0700, Gurucharan Shetty wrote: > Add a new command - "restart" to ovs-ctl. Calling this command > will save and restore the Openflow flows on each bridge while > stopping and starting the userspace daemons respectively. > > Also, during a force-reload-kmod, save

Re: [ovs-dev] [PATCH] ovs-pki: Unique subject DNs for certificate requests.

2012-10-22 Thread Gurucharan Shetty
On Mon, Oct 22, 2012 at 12:55 PM, Ben Pfaff wrote: > On Mon, Oct 22, 2012 at 12:43:11PM -0700, Gurucharan Shetty wrote: > > Some applications expect subject DNs to be unique. > > > > Bug #13609. > > Signed-off-by: Gurucharan Shetty > > Looks good. > > I think that > > +if unique=`uuidgen

[ovs-dev] [PATCH] Update INSTALL.Fedora to mention Fedora 17 instead of Fedora 15

2012-10-22 Thread Kyle Mestery
Reflect the fact Fedora RPM support was recently implemented on Fedora 17 instead of Fedora 15. Signed-off-by: Kyle Mestery --- INSTALL.Fedora | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/INSTALL.Fedora b/INSTALL.Fedora index d023127..708cae6 100644 --- a/INSTALL.Fedor

Re: [ovs-dev] [PATCH 2/3] Do not print ICMP fields for non IP packets.

2012-10-22 Thread Ethan Jackson
> A lot of people disagree with this so some time ago we removed the > text that required binary operators at the beginning of a line Ah didn't realize we had done that. Most of the code uses the convention, but if it's no longer required that's fine. Ethan __

Re: [ovs-dev] [PATCH] ovs-pki: Unique subject DNs for certificate requests.

2012-10-22 Thread Ben Pfaff
On Mon, Oct 22, 2012 at 12:43:11PM -0700, Gurucharan Shetty wrote: > Some applications expect subject DNs to be unique. > > Bug #13609. > Signed-off-by: Gurucharan Shetty Looks good. I think that > +if unique=`uuidgen 2>/dev/null`; then > +: > +else > +uni

[ovs-dev] [PATCH] ovs-pki: Unique subject DNs for certificate requests.

2012-10-22 Thread Gurucharan Shetty
Some applications expect subject DNs to be unique. Bug #13609. Signed-off-by: Gurucharan Shetty --- utilities/ovs-pki.in |8 +++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/utilities/ovs-pki.in b/utilities/ovs-pki.in index bf40c29..cfbcdea 100755 --- a/utilities/ovs-pki.

Re: [ovs-dev] [PATCH] netdev-vport: Disable Path MTU Discovery by default

2012-10-22 Thread Ben Pfaff
On Mon, Oct 22, 2012 at 10:10:10AM +0300, Ansis Atteka wrote: > This patch changes the default path MTU discovery value to > disabled. > > Signed-off-by: Ansis Atteka This seems reasonable (modulo my question to Jesse). As you mentioned yourself in IM, adding a NEWS item is a good item. ___

Re: [ovs-dev] [PATCH] netdev-vport: Disable Path MTU Discovery by default

2012-10-22 Thread Ben Pfaff
On Mon, Oct 22, 2012 at 10:10:10AM +0300, Ansis Atteka wrote: > This patch changes the default path MTU discovery value to > disabled. > > Signed-off-by: Ansis Atteka Jesse, Ansis says that this is at your request. Would you mind supplying a few words for the commit message explaining why we're

Re: [ovs-dev] [PATCH 2/3] Do not print ICMP fields for non IP packets.

2012-10-22 Thread Ben Pfaff
On Sat, Oct 20, 2012 at 12:49:12PM -0500, Ethan Jackson wrote: > > -if (f->nw_proto == IPPROTO_ICMP) { > > +if (f->dl_type == htons(ETH_TYPE_IP) && > > +f->nw_proto == IPPROTO_ICMP) { > > format_be16_masked(s, "icmp_type", f->tp_src, wc->masks.tp_src); > > format_b

Re: [ovs-dev] [PATCH] v8: datapath: Add support for tun_key to Open vSwitch datapath

2012-10-22 Thread Kyle Mestery (kmestery)
On Oct 22, 2012, at 1:07 PM, Ben Pfaff wrote: > On Sun, Oct 21, 2012 at 07:35:55PM -0700, Pravin Shelar wrote: >> On Sat, Oct 20, 2012 at 3:36 PM, Kyle Mestery (kmestery) >> wrote: >>> On Oct 20, 2012, at 2:36 PM, Pravin Shelar wrote: On Thu, Oct 18, 2012 at 3:03 PM, Ben Pfaff wrote: >

Re: [ovs-dev] [PATCH] ovs-pki: Unique subject DNs for certificate requests.

2012-10-22 Thread Ben Pfaff
On Mon, Oct 22, 2012 at 11:09:27AM -0700, Gurucharan Shetty wrote: > Some applications expect subject DNs to be unique. > > Bug #13609. > Signed-off-by: Gurucharan Shetty I know that I suggested uuidgen in conversation, but now I'm thinking that it isn't the best choice because it's not entirely

[ovs-dev] [PATCH] ovs-pki: Unique subject DNs for certificate requests.

2012-10-22 Thread Gurucharan Shetty
Some applications expect subject DNs to be unique. Bug #13609. Signed-off-by: Gurucharan Shetty --- utilities/ovs-pki.in |3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utilities/ovs-pki.in b/utilities/ovs-pki.in index bf40c29..e816341 100755 --- a/utilities/ovs-pki.in ++

Re: [ovs-dev] [PATCH 1/2] ovs-ctl.in: Ability to save flows and kernel datapath config.

2012-10-22 Thread Ben Pfaff
On Fri, Oct 19, 2012 at 05:33:47PM -0700, Gurucharan Shetty wrote: > > > > I think one could argue for saving a copy if restoring flows fails. > > However: the most likely source of problems is "dump-flows" output > > that "add-flows" fails to accept. In such a case, I would hope that > > ovs-ofct

Re: [ovs-dev] [PATCH] v8: datapath: Add support for tun_key to Open vSwitch datapath

2012-10-22 Thread Ben Pfaff
On Sun, Oct 21, 2012 at 07:35:55PM -0700, Pravin Shelar wrote: > On Sat, Oct 20, 2012 at 3:36 PM, Kyle Mestery (kmestery) > wrote: > > On Oct 20, 2012, at 2:36 PM, Pravin Shelar wrote: > >> On Thu, Oct 18, 2012 at 3:03 PM, Ben Pfaff wrote: > >>> On Thu, Oct 18, 2012 at 02:57:52PM -0700, Pravin S

Re: [ovs-dev] [PATCH] nicira-ext: Clarify behavior of overlapping src and dst for NXAST_REG_MOVE.

2012-10-22 Thread Ethan Jackson
Sounds good. Ethan On Mon, Oct 22, 2012 at 11:51 AM, Ben Pfaff wrote: > Thanks. > > It seems that I pushed this last week after Pankaj gave me an ack > off-list. > > On Sat, Oct 20, 2012 at 12:51:35PM -0500, Ethan Jackson wrote: >> Acked-by: Ethan Jackson >> >> >> On Tue, Oct 16, 2012 at 1:01 P

Re: [ovs-dev] [PATCH] nicira-ext: Clarify behavior of overlapping src and dst for NXAST_REG_MOVE.

2012-10-22 Thread Ben Pfaff
Thanks. It seems that I pushed this last week after Pankaj gave me an ack off-list. On Sat, Oct 20, 2012 at 12:51:35PM -0500, Ethan Jackson wrote: > Acked-by: Ethan Jackson > > > On Tue, Oct 16, 2012 at 1:01 PM, Ben Pfaff wrote: > > Reported-by: Pankaj Thakkar > > Signed-off-by: Ben Pfaff >

Re: [ovs-dev] [PATCH] v8: datapath: Add support for tun_key to Open vSwitch datapath

2012-10-22 Thread Kyle Mestery (kmestery)
On Oct 21, 2012, at 9:35 PM, Pravin Shelar wrote: > On Sat, Oct 20, 2012 at 3:36 PM, Kyle Mestery (kmestery) > wrote: >> On Oct 20, 2012, at 2:36 PM, Pravin Shelar wrote: >>> On Thu, Oct 18, 2012 at 3:03 PM, Ben Pfaff wrote: On Thu, Oct 18, 2012 at 02:57:52PM -0700, Pravin Shelar wrote: >>

[ovs-dev] [RFC 5/4] Add support for set_mpls_ttl action

2012-10-22 Thread Simon Horman
This adds support for the OpenFlow 1.1+ set_mpls_ttl action. And also adds an NX set_mpls_ttl action. The handling of the TTL decrement is entirely handled in userspace. Signed-off-by: Simon Horman --- This is a sumplemental RFC patch that adds support for set_mpls_ttl. It is modeled on Patch

Re: [ovs-dev] [PATCH 4/4] Add support for set_mpls_ttl action

2012-10-22 Thread Simon Horman
On Fri, Oct 19, 2012 at 04:54:41PM +0900, Simon Horman wrote: > On Thu, Oct 18, 2012 at 06:02:21PM +0900, Simon Horman wrote: > > This adds support for the OpenFlow 1.1+ set_mpls_ttl action. > > And also adds an NX set_mpls_ttl action. > > > > The handling of the TTL decrement is entirely handled

[ovs-dev] [PATCH] netdev-vport: Disable Path MTU Discovery by default

2012-10-22 Thread Ansis Atteka
This patch changes the default path MTU discovery value to disabled. Signed-off-by: Ansis Atteka --- lib/netdev-vport.c | 10 +- vswitchd/vswitch.xml |2 +- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/lib/netdev-vport.c b/lib/netdev-vport.c index 621abd1..41689