Re: [ovs-dev] [ovn-ipv6 07/26] ovn: Add xxreg[01] symbols.

>On Wed, Jul 13, 2016 at 11:24 AM, Justin Pettit wrote: > >> On Jul 12, 2016, at 7:53 PM, Zong Kai Li wrote: >> >>> + >>> +The regX symbols are 32-bit integers. >>> +The xxregX symbols are 128-bit integers, >>> +which overlay

Re: [ovs-dev] [ovn-ipv6 11/26] ovn-util: Add string representations to 'lport_addresses'.

> On Jul 12, 2016, at 5:16 PM, Ben Pfaff wrote: > > On Mon, Jul 11, 2016 at 11:56:41PM -0700, Justin Pettit wrote: >> A future commit will reduce the amount of conversions used by the >> existing users of 'lport_addresses'. This change will also make it >> possible to use this

[ovs-dev] [PATCH 19/24] datapath: backport: ovs: do not allocate memory from offline numa node

Upstream commit: commit 598c12d0ba6de9060f04999746eb1e015774044b Author: Konstantin Khlebnikov ovs: do not allocate memory from offline numa node When openvswitch tries allocate memory from offline numa node 0: stats =

[ovs-dev] [PATCH 15/24] datapath: backport: ovs: propagate per dp max headroom to all vports

Upstream commit: commit 3a927bc7cf9d0fbe8f4a8189dd5f8440228f64e7 Author: Paolo Abeni ovs: propagate per dp max headroom to all vports This patch implements bookkeeping support to compute the maximum headroom for all the devices in each datapath. When said

[ovs-dev] [PATCH 16/24] datapath: backport: libnl: nla_put_be64(): align on a 64-bit area

Upstream commit: commit b46f6ded906ef0be52a4881ba50a084aeca64d7e Author: Nicolas Dichtel libnl: nla_put_be64(): align on a 64-bit area nla_data() is now aligned on a 64-bit area. A temporary version (nla_put_be64_32bit()) is added for

[ovs-dev] [PATCH 14/24] datapath: backport: ovs: use nla_put_u64_64bit()

Upstream commit: commit 0238b7204b7ff1bad1d2d4489f010d670cbd89f2 Author: Nicolas Dichtel ovs: use nla_put_u64_64bit() Signed-off-by: Nicolas Dichtel Signed-off-by: David S. Miller Signed-off-by:

[ovs-dev] [PATCH 18/24] datapath: Move key memset to ovs_flow_key_extract_userspace()

Synchronize code with upstream ovs_nla_get_flow_metadata(). Signed-off-by: Pravin B Shelar --- datapath/flow.c | 2 ++ datapath/flow_netlink.c | 1 - 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/datapath/flow.c b/datapath/flow.c index 4e6dfb2..390286c

[ovs-dev] [PATCH 13/24] datapath: backport: ovs: align nlattr properly when needed

Upstream commit: commit 66c7a5ee1a6b7c69d41dfd68d207fdd54efba56a Author: Nicolas Dichtel ovs: align nlattr properly when needed I also fix commit 8b32ab9e6ef1: use nla_total_size_64bit() for OVS_FLOW_ATTR_USED in ovs_flow_cmd_msg_size().

[ovs-dev] [PATCH 17/24] datapath: use PTR_ERR_OR_ZERO

Upstream commit: commit f35423c137b0e64155f52c166db1d13834a551f2 Author: Fabian Frederick openvswitch: use PTR_ERR_OR_ZERO Signed-off-by: Fabian Frederick Acked-by: Pravin B Shelar Signed-off-by: David S. Miller

[ovs-dev] [PATCH 10/24] datapath: backport: retain parsed IPv6 header fields in flow on error skipping extension headers

Upstream commit: commit c30da497893718abc6cec4f1d34d35875200edee Author: Simon Horman openvswitch: retain parsed IPv6 header fields in flow on error skipping extension headers When an error occurs skipping IPv6 extension headers retain the already

[ovs-dev] [PATCH 11/24] datapath: backport: allow output of MPLS packets on tunnel vports

Upstream commit: commit fe3a5f6c795810edb1646a840fec3c8c350c2a4e Author: Simon Horman openvswitch: allow output of MPLS packets on tunnel vports Currently output of MPLS packets on tunnel vports is not allowed by Open vSwitch. This is because

[ovs-dev] [PATCH 12/24] datapath: backport: ovs: allow nl 'flow set' to use ufid without flow key

Upstream commit: commit 6f15cdbf8a8ac2e22767cc8b1eae225702733c95 Author: Samuel Gauthier ovs: allow nl 'flow set' to use ufid without flow key When we want to change a flow using netlink, we have to identify it to be able to perform a lookup. Both

[ovs-dev] [PATCH 23/24] datapath: openvswitch: update checksum in {push, pop}_mpls

Upstream commit: commit bc7cc5999fd392cc799630d7e375b2f4e29cc398 Author: Simon Horman openvswitch: update checksum in {push,pop}_mpls In the case of CHECKSUM_COMPLETE the skb checksum should be updated in {push,pop}_mpls() as they the type in the

[ovs-dev] [PATCH 07/24] datapath: remove get_pcpu_ptr

There is no need to support old kernel so now we can use newer api to access per cpu data. Signed-off-by: Pravin B Shelar --- datapath/actions.c | 4 ++-- datapath/linux/compat/include/linux/percpu.h | 6 -- 2 files changed, 2 insertions(+), 8

[ovs-dev] [PATCH 20/24] datapath: backport: ovs: set name assign type of internal port

upstream commit: commit 684ff4ef5edd758c47929b852b4ea79be56f8bc0 Author: Zhang Shengju ovs: set name assign type of internal port Set name_assign_type of internal port to NET_NAME_USER. Signed-off-by: Zhang Shengju

[ovs-dev] [PATCH 09/24] datapath: backport: openvswitch: allow management from inside user namespaces

Upstream commit: commit 4a92602aa1cd5bbaeedbd9536ff992f7d26fe9d1 Author: Tycho Andersen openvswitch: allow management from inside user namespaces Operations with the GENL_ADMIN_PERM flag fail permissions checks because this flag means we call

[ovs-dev] [PATCH 06/24] datapath: Sync OVS recursive loop counter with upstream.

Signed-off-by: Pravin B Shelar --- datapath/actions.c | 31 --- 1 file changed, 12 insertions(+), 19 deletions(-) diff --git a/datapath/actions.c b/datapath/actions.c index ed44ead..3cd2922 100644 --- a/datapath/actions.c +++ b/datapath/actions.c @@

[ovs-dev] [PATCH 21/24] datapath: internal-dev: set IFF_NO_QUEUE

Introduced by upstream commit 4272cc51a6d ("openvswitch: Convert to using IFF_NO_QUEUE"). Signed-off-by: Pravin B Shelar --- datapath/linux/compat/include/linux/netdevice.h | 6 ++ datapath/vport-internal_dev.c | 5 - 2 files changed, 10 insertions(+),

[ovs-dev] [PATCH 00/24] datapath: various fixes and updates.

From: Pravin B Shelar Following patches brings in couple of fixes to recent tunneling changes. along with various other improvement and fixes. Pravin B Shelar (24): datapath: compat: clear skb IPCB block. datapath: tnl-segmentation: clone dst for each segment.

[ovs-dev] [PATCH 08/24] datapath: genlmsg_new_unicast to genlmsg_new

Use newer API which is consistent with upstream. Signed-off-by: Pravin B Shelar --- acinclude.m4 | 1 - datapath/datapath.c | 22 -- datapath/linux/compat/include/net/genetlink.h | 9 - 3

[ovs-dev] [PATCH 22/24] datapath: vxlan: enable zero checksum on recv.

partialy backport commit c868ee7063bdb5 ("lwt: fix rx checksum setting for lwt devices tunneling over ipv6"). Signed-off-by: Pravin B Shelar --- datapath/vport-vxlan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/datapath/vport-vxlan.c

[ovs-dev] [PATCH 02/24] datapath: tnl-segmentation: clone dst for each segment.

Each skb segment needs a dst entry. Signed-off-by: Pravin B Shelar --- datapath/linux/compat/gso.c | 5 + 1 file changed, 5 insertions(+) diff --git a/datapath/linux/compat/gso.c b/datapath/linux/compat/gso.c index 8812253..9d478f8 100644 --- a/datapath/linux/compat/gso.c

[ovs-dev] [PATCH 04/24] datapath: compat: reset md_dst on recieve.

On transmit metadata dst is cleared, do same on recieve. Signed-off-by: Pravin B Shelar --- datapath/linux/compat/geneve.c | 2 +- datapath/linux/compat/include/net/dst_metadata.h | 20 ++-- datapath/linux/compat/include/net/udp_tunnel.h |

[ovs-dev] [PATCH 01/24] datapath: compat: clear skb IPCB block.

OVS compat ip_local_out/ip6_local_out clear out skb IPCB, we also need to do it for non compat version. Signed-off-by: Pravin B Shelar --- datapath/linux/compat/gso.h | 16 1 file changed, 16 insertions(+) diff --git a/datapath/linux/compat/gso.h

[ovs-dev] [PATCH 03/24] datapath: compat: Set OVS-GSO segment handler only for GSO packet.

This simplify OVS compat code where non gso packet can be directly sent to network stack. Signed-off-by: Pravin B Shelar --- datapath/linux/compat/ip_tunnels_core.c | 13 +++-- 1 file changed, 7 insertions(+), 6 deletions(-) diff --git

[ovs-dev] [PATCH 05/24] datapath: Use skb_postpush_rcsum()

Use kernel function to update checksum. Signed-off-by: Pravin B Shelar --- acinclude.m4 | 1 + datapath/actions.c | 8 +++- datapath/linux/compat/include/linux/skbuff.h | 20

Re: [ovs-dev] [question] How to invalidate a dp-flow which was generated by of-flow ?

On Wed, Jul 06, 2016 at 11:09:47AM +0800, Zhang Haoyu wrote: > > > On 7/6/16 01:11, Ben Pfaff wrote: > > On Tue, Jul 05, 2016 at 02:57:11PM +0800, Zhang Haoyu wrote: > >> > >> > >> On 7/5/16 13:26, Ben Pfaff wrote: > >>> On Tue, Jul 05, 2016 at 11:09:26AM +0800, Zhang Haoyu wrote: > How to

Re: [ovs-dev] Mask generation for megaflow.

On Fri, Jul 08, 2016 at 04:05:34PM +, Subrata Paul wrote: > Hi, > I am using ovs 2.4.0 and I have added 3 flows as follows: > ovs-ofctl add-flow br0 "table=0, in_port=1, tcp, nw_src=10.0.0.2/24, > tcp_src=45466, action=output:2" > ovs-ofctl add-flow br0 "table=0, in_port=1, tcp,

Re: [ovs-dev] [ovn-ipv6 13/26] ovn-northd: Use strings from extract_lsp_addresses().

On Tue, Jul 12, 2016 at 08:23:06PM -0700, Ben Pfaff wrote: > This bit of build_lswitch_flows() would probably benefit from {...} > notation: > if (op->lsp_addrs[i].n_ipv6_addrs == 1) { > ds_put_format(, "nd.target == %s", >

Re: [ovs-dev] [ovn-ipv6 05/26] Introduce 128-bit xxregs.

On Tue, Jul 12, 2016 at 05:16:32PM -0700, Justin Pettit wrote: > > > On Jul 12, 2016, at 4:15 PM, Ben Pfaff wrote: > > > > On Mon, Jul 11, 2016 at 11:56:35PM -0700, Justin Pettit wrote: > >> These are needed to handle IPv6 addresses. > >> > >> Signed-off-by: Justin Pettit

Re: [ovs-dev] [ovn-ipv6 15/26] ovn-nbctl: Print router port networks in "show".

On Mon, Jul 11, 2016 at 11:56:45PM -0700, Justin Pettit wrote: > Signed-off-by: Justin Pettit Seems reasonable but can you add a test? Acked-by: Ben Pfaff Thanks, Ben. ___ dev mailing list dev@openvswitch.org

Re: [ovs-dev] [ovn-ipv6 14/26] ovn: Support multiple addresses on a single logical router port.

On Mon, Jul 11, 2016 at 11:56:44PM -0700, Justin Pettit wrote: > Supporting multiple addresses is only mildly interesting for IPv4. > However, it is a requirement for IPv6, which will arrive in a future > commit. > > This commit introduces the extract_lrp_networks() function to ovn-util.[ch]. >

[ovs-dev] [PATCH v2 2/3] Windows: Drop TCP sockets and switch to named pipe

This patch drops the TCP socket implementation behind command line arguments punix/unix and switches to the local named pipe implementation. Acked-by: Paul Boca Signed-off-by: Alin Gabriel Serdean --- v2: Add acked ---

[ovs-dev] [PATCH v2 1/3] Windows: Local named pipe implementation

Currently in the case of command line arguments punix/unix, on Windows we create a file, write a TCP port number to connect. This is a security concern. This patch adds support for the command line arguments punix/unix trying to mimic AF_UNIX behind a local named pipe. Signed-off-by: Alin

[ovs-dev] [PATCH v2 3/3] Windows tests: daemon specific tests

We do not write anything to the file created by the punix/unix arguments. Switch tests to plain file existence. Acked-by: Paul Boca Signed-off-by: Alin Gabriel Serdean --- v2: Add acked --- tests/ovsdb-server.at | 6 +++--- 1 file

[ovs-dev] [PATCH v2 0/3] Local named pipe implementation

This patch series is intended to switch the current implementation of TCP sockets to a named pipe implementation. One test had to be changed in order for the current state of unit tests to be the same. Alin Gabriel Serdean (3): Windows: Local named pipe implementation Windows: Drop TCP

Re: [ovs-dev] [ovn-ipv6 07/26] ovn: Add xxreg[01] symbols.

> On Jul 12, 2016, at 7:53 PM, Zong Kai Li wrote: > >> + >> +The regX symbols are 32-bit integers. >> +The xxregX symbols are 128-bit integers, >> +which overlay four of the 32-bit registers: xxreg0 >> +overlays reg0 through reg3, with

Re: [ovs-dev] [ovn-ipv6 13/26] ovn-northd: Use strings from extract_lsp_addresses().

On Mon, Jul 11, 2016 at 11:56:43PM -0700, Justin Pettit wrote: > Extract port security and logical switch port addresses once and store > them as part of the ovn_port structure. Use the string representations > from the extracted addresses. > > Signed-off-by: Justin Pettit In

Re: [ovs-dev] SFC ACL and Flow Classifier

John, see my comments inline. Pls take some time to look at my comments about https://review.openstack.org/#/c/333172/, if not right, pls correct me, thanks. Regards, Juno Zhu IBM China Development Labs (CDL) Cloud IaaS Lab Email: na...@cn.ibm.com 5F, Building 10, 399 Keyuan Road,

Re: [ovs-dev] SFC-Summary: MultiTenant

>I was thinking this could be handled with child / sub-ports.  We do this >today for containers in VMs.  We can have a single VIF for a VM that is >connected to multiple networks that are owned by separate tenants.  Some >sort of encapsulation (VLAN ID, MPLS header, whatever) would be used to

[ovs-dev] SFC ACL and Flow Classifier

Hi John: Please see inline for FS: Message: 3 Date: Tue, 12 Jul 2016 22:32:58 + From: John McDowall To: "dev@openvswitch.org" Subject: [ovs-dev] SFC ACL and Flow Classifier Message-ID:

Re: [ovs-dev] [ovn-ipv6 07/26] ovn: Add xxreg[01] symbols.

> + > +The regX symbols are 32-bit integers. > +The xxregX symbols are 128-bit integers, > +which overlay four of the 32-bit registers: xxreg0 > +overlays reg0 through reg3, with > +reg0 supplying the most-significant bits of > +xxreg0 and reg3

Re: [ovs-dev] [ovn-ipv6 12/26] ovn-util: Remove 'store_ipv6' argument from extract_lsp_addresses().

On Mon, Jul 11, 2016 at 11:56:42PM -0700, Justin Pettit wrote: > With the addition of IPv6 routing, there won't be much need to > special-case not parsing IPv6 addresses. > > Signed-off-by: Justin Pettit Acked-by: Ben Pfaff

Re: [ovs-dev] [PATCH 2/3] dpctl: Fix shadowed iterator in show_dpif().

Thanks. Do you mind looking at the other patches, too? I think they're pretty obviously correct. --Justin > On Jul 12, 2016, at 6:36 PM, Daniele Di Proietto wrote: > > Looks good to me, thanks! > > Acked-by: Daniele Di Proietto > > 2016-07-12

Re: [ovs-dev] [PATCH 2/3] dpctl: Fix shadowed iterator in show_dpif().

Looks good to me, thanks! Acked-by: Daniele Di Proietto 2016-07-12 12:19 GMT-07:00 Justin Pettit : > Signed-off-by: Justin Pettit > --- > lib/dpctl.c | 11 --- > 1 file changed, 4 insertions(+), 7 deletions(-) > > diff --git

Re: [ovs-dev] [PATCH v3 3/3] dpif-netdev: XPS (Transmit Packet Steering) implementation.

Thanks for the patch. This is not a complete review, but I have some preliminary comments. If I understand correctly 'port_mutex' is converted to rwlock because we want the pmd threads in dpif_netdev_xps_get_tx_qid() to be able to grab it concurrently. I think that we can add a pointer from

Re: [ovs-dev] [PATCH 3/4] system-ovn.at: Add a OVN NAT test using OVN gateway.

On 11 July 2016 at 23:35, Gurucharan Shetty wrote: > This unit test adds a basic OVN NAT test that tests north-south > DNAT, south-north SNAT and east-west DNAT and SNAT. It uses network > namespaces connected to br-int using veth pairs to act as logical > ports. This test does not

Re: [ovs-dev] [PATCH 4/4] system-ovn.at: Add a OVN load-balancing unit test.

On 11 July 2016 at 23:35, Gurucharan Shetty wrote: > Signed-off-by: Gurucharan Shetty > --- Looks good, thanks. Only comment is if you add this file to check-system-userspace then it should also get a CHECK_CONNTRACK() at the start. Acked-by: Joe Stringer

Re: [ovs-dev] [PATCH 3/4] system-ovn.at: Add a OVN NAT test using OVN gateway.

On 11 July 2016 at 23:35, Gurucharan Shetty wrote: > This unit test adds a basic OVN NAT test that tests north-south > DNAT, south-north SNAT and east-west DNAT and SNAT. It uses network > namespaces connected to br-int using veth pairs to act as logical > ports. This test does not

Re: [ovs-dev] [ovn-ipv6 05/26] Introduce 128-bit xxregs.

> On Jul 12, 2016, at 4:15 PM, Ben Pfaff wrote: > > On Mon, Jul 11, 2016 at 11:56:35PM -0700, Justin Pettit wrote: >> These are needed to handle IPv6 addresses. >> >> Signed-off-by: Justin Pettit > > The comment in meta-flow.h talks about "Nicira extension"

Re: [ovs-dev] [ovn-ipv6 11/26] ovn-util: Add string representations to 'lport_addresses'.

On Mon, Jul 11, 2016 at 11:56:41PM -0700, Justin Pettit wrote: > A future commit will reduce the amount of conversions used by the > existing users of 'lport_addresses'. This change will also make it > possible to use this structure for logical router port networks. > > Signed-off-by: Justin

Re: [ovs-dev] [PATCH 2/4] system-common-macros.at: Add a iface-id to interfaces.

On 11 July 2016 at 23:35, Gurucharan Shetty wrote: > IntegrationGuide.md explains the meaning of the key 'iface-id'. Adding > it as part of ADD_VETH is useful for OVN related tests. For non-OVN > tests, it should have no effect. Upcoming commits take advantage of this. > >

Re: [ovs-dev] [PATCH 1/4] system-common-macros: Enhance ADD_VETH to take mac and gateway.

On 11 July 2016 at 23:35, Gurucharan Shetty wrote: > It is useful to set the mac address and gateway while using ADD_VETH > to connect a namespace to a OVN logical topology. Upcoming commits > use this enhancement. > > Signed-off-by: Gurucharan Shetty Acked-by: Joe

Re: [ovs-dev] SFC-Summary: MultiTenant

Hi Kyle/Russell, -Original Message- From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Kyle Mestery Sent: Tuesday, July 12, 2016 9:19 AM To: Russell Bryant Cc: dev@openvswitch.org; John McDowall Subject: Re: [ovs-dev] SFC-Summary: MultiTenant On Tue, Jul 12, 2016 at 9:52 AM,

Re: [ovs-dev] [ovn-ipv6 10/26] ovn-northd: Use strings for IP addresses in router instances of "ovn_port".

On Mon, Jul 11, 2016 at 11:56:40PM -0700, Justin Pettit wrote: > Reduce the amount of string conversions necessary. This will be > convenient when we want to start using IPv6 addresses, too. > > Signed-off-by: Justin Pettit It's getting weirder and weirder to use C as a string

Re: [ovs-dev] [ovn-ipv6 09/26] ovn-northd: Use dynamic strings when building router and switch flows.

On Mon, Jul 11, 2016 at 11:56:39PM -0700, Justin Pettit wrote: > Reduce the number of memory allocations and risk of introducing shadow > variables. > > Signed-off-by: Justin Pettit Acked-by: Ben Pfaff ___ dev mailing

Re: [ovs-dev] [ovn-ipv6 08/26] ovn: Remove 'default_gw' from logical router table.

On Mon, Jul 11, 2016 at 11:56:38PM -0700, Justin Pettit wrote: > With static routes, it's not necessary to have a separate default > gateway parameter. This also makes configuring router ports clearer > when IPv6 and IPv4 addresses may be assigned to the same port. > > Signed-off-by: Justin

Re: [ovs-dev] [ovn-ipv6 07/26] ovn: Add xxreg[01] symbols.

On Mon, Jul 11, 2016 at 11:56:37PM -0700, Justin Pettit wrote: > Signed-off-by: Justin Pettit Acked-by: Ben Pfaff ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [ovn-ipv6 06/26] ovn: Renumber logical field registers to the newly extended registers.

On Mon, Jul 11, 2016 at 11:56:36PM -0700, Justin Pettit wrote: > IPv6 addresses use four standard OVS registers, so move the existing > named registers back to make room for using more logical registers. > > Signed-off-by: Justin Pettit Acked-by: Ben Pfaff

Re: [ovs-dev] SFC-Summary: MultiTenant

It isn't. For networking-sfc, at least for OVS driver (the default), the MPLS label is popped before the packet reaches the VNF. But Russell did bring up a great solution for supporting multi-tenancy on VNF... it does open the can of worm where we need to distinguish between VNF that supports

Re: [ovs-dev] [ovn-ipv6 05/26] Introduce 128-bit xxregs.

On Mon, Jul 11, 2016 at 11:56:35PM -0700, Justin Pettit wrote: > These are needed to handle IPv6 addresses. > > Signed-off-by: Justin Pettit The comment in meta-flow.h talks about "Nicira extension" registers. I think this made sense a few years ago, but Nicira isn't very

Re: [ovs-dev] [PATCH] ovn-controller-vtep: occasional failure in "binding 1" test case

On Fri, Jul 8, 2016 at 7:16 AM, Lance Richardson wrote: > The ovn-controller-vtep "binding 1" test case fails occasionally > due to a race with the ovs-vtep daemon. If ovs-vtep happens to > execute handle_physical() after "ovs-vsctl del-port p0", but before > the test

[ovs-dev] SFC ACL and Flow Classifier

Based on Ryan's suggestions, here is my proposal for implementing the SFC flowClassifier as ACLs. I wanted to get some feedback before I get into another round of changes in the code. Currently I have implemented SFC by adding a new table (ls_in_chain) in the ova-stage of ovn-northd.c as

[ovs-dev] [PATCH 5/6] compat: Remove inet_frag_evictor backport.

Kernel 3.7 and lower are now unsupported, remove this fragment. Signed-off-by: Joe Stringer --- datapath/linux/compat/include/net/inet_frag.h | 9 - 1 file changed, 9 deletions(-) diff --git a/datapath/linux/compat/include/net/inet_frag.h

[ovs-dev] [PATCH 6/6] compat: Only call nf_defrag_ipv[46]_enable() once.

This function is just a dummy to ensure that the corresponding netfilter fragment module is loaded, to initialize the shared structures. But it doesn't need to be invoked once per namespace; one call per protocol should do the trick. Signed-off-by: Joe Stringer ---

[ovs-dev] [PATCH 2/6] compat: Simplify inet_fragment backports.

The core fragmentation handling logic is exported on all supported kernels, so it's not necessary to backport the latest version of this. This greatly simplifies the code due to inconsistencies between the old per-lookup garbage collection and the newer workqueue based garbage collection. As a

[ovs-dev] [PATCH 4/6] compat: IPv6 fragmentation backport cleanups.

Remove a couple of functions that are available on all supported kernel versions. Signed-off-by: Joe Stringer --- datapath/linux/compat/nf_conntrack_reasm.c | 39 +++--- 1 file changed, 3 insertions(+), 36 deletions(-) diff --git

[ovs-dev] [PATCH 0/6] compat: Fix various fragmentation-related crashes.

This series fixes a couple of crashes that may occur due to the backports of the inet_fragment.c code from upstream. These are primarily related to cases where a fragment is received and cached in the OVS fragment handling code with a timer attached to expire it, then the OVS kernel module is

[ovs-dev] [PATCH 3/6] compat: Remove ip6_expire_frag_queue().

This was previously backported to fix issues with our inet_fragment backport; with that largely gone, we can get rid of this too. Signed-off-by: Joe Stringer --- This is effectively a revert of commit bf1f0d35ceb9 ("datapath: Fix IPv6 fragment expiry crash."). It should be paired

[ovs-dev] [PATCH 1/6] compat: Fix IPv6 frag expiry crash.

If a user sends some fragments of an IPv6 message through OVS, but OVS fails to assemble the IPv6 message and the OVS module is then unloaded before the fragments expire, it could lead to a kernel panic like the following: Call Trace: [] ? call_timer_fn+0x39/0x130 []

Re: [ovs-dev] [PATCH v3 2/2] Windows: Use NETLINK_NETFILTER instead of NETLINK_GENERIC

Acked-by: Nithin Raju One another change that comes to mind is the validation that a socket of type NETLINK_GENERIC does not send a CT command and vice-versa. It should be a small change within Datapath.c and can be done as a incremental patch. -Original Message-

[ovs-dev] [PATCH] ovn-northd: Combine two NAT loops into one.

It is wasteful to have two loops that go through all the NAT entries. Signed-off-by: Gurucharan Shetty --- ovn/northd/ovn-northd.c | 45 + 1 file changed, 13 insertions(+), 32 deletions(-) diff --git a/ovn/northd/ovn-northd.c

Re: [ovs-dev] [PATCH v3 1/2] Windows: Add support for handling protocol (netlink family)

Acked-by: Nithin Raju -Original Message- From: Sairam Venugopal Date: Tuesday, July 12, 2016 at 2:41 PM To: Nithin Raju Subject: FW: [PATCH v3 1/2] Windows: Add support for handling protocol (netlink family) > > >On 7/11/16,

[ovs-dev] [PATCH 1/2] Move lib/json.h to include/openvswitch

To easily allow both in- and out-of-tree building of the Python wrapper for the OVS JSON parser (e.g. w/ pip), move json.h to include/openvswitch. This also requires moving lib/{hmap,shash}.h. Both hmap.h and shash.h were #include-ing "util.h" even though the headers themselves did not use

[ovs-dev] [PATCH 2/2] JSON serialization via Python's json lib

There is no particularly good reason to use our own Python JSON serialization implementation when serialization can be done faster with Python's built-in JSON library. A few tests were changed due to Python's default JSON library returning slightly more precise floating point numbers.

[ovs-dev] [PATCH 0/2] Python JSON parser improvements

These patches should address the previous issue with building the Python wrapper when builddir != srcdir. It also ensures that the wrapper can be properly built by pip from out-of-tree by ensuring that json.h is added to include/openvswitch/json.h. Terry Wilson (2): Move lib/json.h to

Re: [ovs-dev] Question about ovs-vtep implementation

On Fri, Jul 8, 2016 at 12:20 PM, Lance Richardson wrote: > The "ovn-controller-vtep - vtep-macs 1" test case fails occasionally, > with ovs-vswitchd logs similar to these: > > bridge|INFO|bridge br-vtep_vtep_ls1: added interface vx1 on port 2 >

[ovs-dev] Profile

Dear dev, Please find attached the profile of Mr.Leon for a suitable role in your Organisation King regards, Annmarie Cox Business Director USA Job ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

[ovs-dev] Profile

Dear dev, Please find attached the profile of Mr.Lamb for a suitable role in your Organisation King regards, Harland Gillespie Vice President of Operations ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [ovn-ipv6 02/26] ovn-controller: Log OpenFlow errors at "info" level.

> On Jul 12, 2016, at 1:28 PM, Ben Pfaff wrote: > > On Mon, Jul 11, 2016 at 11:56:32PM -0700, Justin Pettit wrote: >> Otherwise, errors are logged at "debug" level. Errors when pushing >> flows can then seemingly be silently lost. >> >> Signed-off-by: Justin Pettit

[ovs-dev] Profile

Dear dev, Please find attached the profile of Mr.Mercer for a suitable role in your Organisation King regards, Nell Gross Chief Executive Officer - Food Packaging Company ___ dev mailing list dev@openvswitch.org

Re: [ovs-dev] [ovn-ipv6 04/26] Increase number of registers to 16.

On Mon, Jul 11, 2016 at 11:56:34PM -0700, Justin Pettit wrote: > With eight 32-bit registers, we can only store two IPv6 addresses, which is > pretty tight. > > Signed-off-by: Justin Pettit We need a different long-term solution. Acked-by: Ben Pfaff

[ovs-dev] Profile

Dear dev, Please find attached the profile of Mr.Navarro for a suitable role in your Organisation King regards, Drew Black Managing Director - Property Advisory Industry ___ dev mailing list dev@openvswitch.org

Re: [ovs-dev] [ovn-ipv6 03/26] ovn-northd: Fix comments about the flows.

On Mon, Jul 11, 2016 at 11:56:33PM -0700, Justin Pettit wrote: > References to the specifc tables should probably be dropped, since > they'll continue to drift towards wrong. In the meantime, correct the > ones that are there. > > Signed-off-by: Justin Pettit Acked-by: Ben

Re: [ovs-dev] [ovn-ipv6 02/26] ovn-controller: Log OpenFlow errors at "info" level.

On Mon, Jul 11, 2016 at 11:56:32PM -0700, Justin Pettit wrote: > Otherwise, errors are logged at "debug" level. Errors when pushing > flows can then seemingly be silently lost. > > Signed-off-by: Justin Pettit It'd be a little cleaner like this: diff --git

[ovs-dev] [PATCH 2/3] dpctl: Fix shadowed iterator in show_dpif().

Signed-off-by: Justin Pettit --- lib/dpctl.c | 11 --- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/lib/dpctl.c b/lib/dpctl.c index b870e30..003602a 100644 --- a/lib/dpctl.c +++ b/lib/dpctl.c @@ -577,13 +577,10 @@ show_dpif(struct dpif *dpif, struct

[ovs-dev] [PATCH 3/3] ovsdb-client: Move iterator definition to prevent shadowing.

This doesn't fix a bug, but is more standard. Signed-off-by: Justin Pettit --- ovsdb/ovsdb-client.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/ovsdb/ovsdb-client.c b/ovsdb/ovsdb-client.c index 80819a8..25fe903 100644 --- a/ovsdb/ovsdb-client.c +++

[ovs-dev] [PATCH 1/3] ovn-northd: Fix shadowed iterators in port security functions.

Signed-off-by: Justin Pettit --- ovn/northd/ovn-northd.c | 24 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/ovn/northd/ovn-northd.c b/ovn/northd/ovn-northd.c index f4b4435..3fb4063 100644 --- a/ovn/northd/ovn-northd.c +++

Re: [ovs-dev] [PATCH 1/3] Windows: Local named pipe implementation

Thanks for the review! > -Mesaj original- > De la: Paul Boca > Trimis: Tuesday, July 12, 2016 10:29 PM > Către: Alin Serdean ; > dev@openvswitch.org > Subiect: RE: [PATCH 1/3] Windows: Local named pipe implementation > > Hi Alin! > > As discussed

Re: [ovs-dev] [ovn-ipv6 01/26] ofctrl: Fix non-sensical comment.

On Mon, Jul 11, 2016 at 11:56:31PM -0700, Justin Pettit wrote: > Signed-off-by: Justin Pettit Acked-by: Ben Pfaff ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH V6 12/17] python tests: Ported Python daemon to Windows

I am wondering if it is better if we could just import the shared library and call the detach function (https://github.com/openvswitch/ovs/blob/master/lib/daemon-windows.c#L342) instead of duplicating the effort. I wonder if the same could be applied un the *unix side as well. Thanks, Alin.

Re: [ovs-dev] [PATCH 2/3] Windows: Drop TCP sockets and switch to named pipe

Acked-by: Paul Boca > -Original Message- > From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Alin Serdean > Sent: Tuesday, July 12, 2016 6:55 PM > To: dev@openvswitch.org > Subject: [ovs-dev] [PATCH 2/3] Windows: Drop TCP sockets and switch to >

Re: [ovs-dev] [PATCH 3/3] Windows tests: daemon specific tests

Acked-by: Paul Boca > -Original Message- > From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Alin Serdean > Sent: Tuesday, July 12, 2016 6:55 PM > To: dev@openvswitch.org > Subject: [ovs-dev] [PATCH 3/3] Windows tests: daemon specific tests > > We

Re: [ovs-dev] [PATCH 1/3] Windows: Local named pipe implementation

Hi Alin! As discussed offline, please see my comments inline. Thanks, Paul > -Original Message- > From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Alin Serdean > Sent: Tuesday, July 12, 2016 6:55 PM > To: dev@openvswitch.org > Subject: [ovs-dev] [PATCH 1/3] Windows: Local

Re: [ovs-dev] [PATCH] tests: daemon specific tests

Seems like "daemon --detach --monitor startup errors" is disabled on Windows. Disregard my last comment. Thanks, Paul > -Original Message- > From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Paul Boca > Sent: Tuesday, July 12, 2016 8:55 PM > To: Alin Serdean;

Re: [ovs-dev] [PATCH V6 11/17] python tests: Ported UNIX sockets to Windows

Thanks for the patch. I will change the test after the named pipe implementation gets in. Acked-by: Alin Gabriel Serdean > -Mesaj original- > De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca > Trimis: Wednesday, July 6, 2016 3:38 PM

Re: [ovs-dev] [PATCH V6 09/17] python tests: Fixed OSError not iterable on Windows

I am wondering why you get an OSError instead of a select error? The order of the exception should be inverted (first the select one and after the OSerror one), or add a general exception at the end. Thanks, Alin. > -Mesaj original- > De la: dev [mailto:dev-boun...@openvswitch.org] În

Re: [ovs-dev] [PATCH V6 01/17] python tests: Implemented signal.alarm for Windows

Thanks for the patch. Overall it looks good my only suggestion would be to move it to a common ground rather than multiplying the code wherever it is needed. Thanks, Alin. > -Mesaj original- > De la: dev [mailto:dev-boun...@openvswitch.org] În numele Paul Boca > Trimis: Wednesday, July

Re: [ovs-dev] [PATCH] tests: daemon specific tests

Hi Alin! I think that the same fix can be applied to " daemon --detach --monitor startup errors" Acked-by: Paul Boca > -Original Message- > From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Alin Serdean > Sent: Tuesday, July 12, 2016 6:36 PM >

Re: [ovs-dev] [RFC PATCH 1/1] netdev-dpdk: vHost client mode and reconnect

Ciara Loftus writes: > A new other_config DB option has been added called 'vhost_driver_mode'. > By default this is set to 'server' which is the mode of operation OVS > with DPDK has used up until this point - whereby OVS creates and manages > vHost user sockets. > > If

  1   2   >