v8->v9:
Applied Ben's patch for fixing sparse and type errors.
Addressed review comments from Ben.
https://mail.openvswitch.org/pipermail/ovs-dev/2019-December/366101.html
v7->v8:
Removed hash action for balance-tcp mode.
Removed bond-only pmd reload action.
Rebased to OVS master.
v6->v7:
Problem:
In OVS-DPDK, flows with output over a bond interface of type “balance-tcp”
(using a hash on TCP/UDP 5-tuple) get translated by the ofproto layer into
"HASH" and "RECIRC" datapath actions. After recirculation, the packet is
forwarded to the bond member port based on 8-bits of the
Hi, William
What kernel version can support AF_PACKET v3? I can try it with your patch.
-邮件原件-
发件人: William Tu [mailto:u9012...@gmail.com]
发送时间: 2019年12月20日 8:41
收件人: d...@openvswitch.org
抄送: i.maxim...@ovn.org; Yi Yang (杨�D)-云服务集团 ;
b...@ovn.org; echau...@redhat.com
主题: [PATCH RFC]
Formations en Magntisme Thrapeutique
Le Magntisme Thrapeutique permet par simple imposition des
mains de soulager en r-harmonisant les nergies.
Dcouvrez, comprenez et contrlez ces incroyables pouvoirs qui
sont cachs en vous !
On Thu, Dec 19, 2019 at 04:41:25PM -0800, William Tu wrote:
> Currently the performance of sending packets from userspace
> ovs to kernel veth device is pretty bad as reported from YiYang[1].
> The patch adds AF_PACKET v3, tpacket v3, as another way to
> tx/rx packet to linux device, hopefully
On 12/11/2019 3:33 PM, David Ahern wrote:
On 12/11/19 12:17 PM, Gregory Rose wrote:
Hi David,
I just want to update you on this work. I've got everything fixed up
and compiling
on 5.4.2 but in the modpostsection there are a lot of bizarre errors.
I'm not sure where
this comes from so I'm
Current ovs matser has included sendmmsg declaration in
include/sparse/sys/socket.h
int sendmmsg(int, struct mmsghdr *, unsigned int, unsigned int);
I saw "+^L" in your patch.
--- a/lib/socket-util.c
+++ b/lib/socket-util.c
@@ -1283,3 +1283,59 @@ wrap_sendmmsg(int fd, struct mmsghdr *msgs,
On Thu, Dec 19, 2019 at 8:36 AM Matteo Croce wrote:
>
> On Wed, Dec 18, 2019 at 4:06 AM Pravin Shelar wrote:
> >
> > On Tue, Dec 17, 2019 at 7:51 AM Matteo Croce wrote:
> > >
> > > New action to decrement TTL instead of setting it to a fixed value.
> > > This action will decrement the TTL and,
Thanks Ben for the review. Sorry for the delay!
On Thu, Nov 21, 2019 at 4:54 PM Ben Pfaff wrote:
> On Tue, Nov 12, 2019 at 12:47:58AM -0800, amgin...@gmail.com wrote:
> > -./configure --localstatedir="/var" --sysconfdir="/etc" --prefix="/usr" \
> > +if [ $KERNEL_VERSION == "host" ]; then
>
>
From: Aliasgar Ginwala
--with-linux can be made configurable while building containers
for leveraging kernel modules installed on host.
KERNEL_VERSION=host should be used in env variable for the same.
Signed-off-by: Aliasgar Ginwala
---
utilities/docker/Makefile | 2 +-
Ben, per my understanding, buffers are allocated in
netdev_linux_batch_rxq_recv_sock, they should have been thread-local, do you
mean maintain static buffers per thread (declared as thread local storage
array)? I don't think it is feasible, maybe they aren't free on receiving next
time because
Currently the performance of sending packets from userspace
ovs to kernel veth device is pretty bad as reported from YiYang[1].
The patch adds AF_PACKET v3, tpacket v3, as another way to
tx/rx packet to linux device, hopefully showing better performance.
AF_PACKET v3 should get closed to 1Mpps,
On Thu, Dec 19, 2019 at 04:57:37PM +0200, Paul Blakey wrote:
> Each recirculation id will create a tc chain, and we translate
> the recirculation action to a tc goto chain action.
>
> We check for kernel support for this by probing OvS Datapath for the
> tc recirc id sharing feature. If
On Fri, Dec 13, 2019 at 03:15:36PM -0500, num...@ovn.org wrote:
> From: Numan Siddique
>
> When ovsdb-server is in backup mode and connects to the active
> ovsdb-server for replication, and if takes more than 5 seconds to
> get the dump of the whole database, it will drop the connection
> soon
On Mon, Dec 9, 2019 at 12:12 AM Dumitru Ceara wrote:
>
> If the last ovn-controller main loop run started a transaction to the
> Southbound DB and the transaction is still in progress, ovn-controller
> will not call engine_run(). In case there were changes to the DB,
> engine_need_run() would
From: venu iyer
If one creates a port group and a MAC address set, and an
ACL that prevents packets being output to a port in that Port Group from
any MAC address in that address set, the outcome is not consistent.
The outcome depends on whether there is a stateful rule on the switch or not.
Hallo,
Ich bin Azim Hashim Premji, ein indischer Wirtschaftsmagnat, Investor
und Philanthrop. Ich bin der Vorsitzende von Wipro Limited. Ich gab 25
Prozent meines persönlichen Vermögens für wohltätige Zwecke ab. Ich bin
eine Witwerin, die an Krebs und Schlaganfall leidet, was mir ein Kind
On Wed, Dec 18, 2019 at 4:06 AM Pravin Shelar wrote:
>
> On Tue, Dec 17, 2019 at 7:51 AM Matteo Croce wrote:
> >
> > New action to decrement TTL instead of setting it to a fixed value.
> > This action will decrement the TTL and, in case of expired TTL, drop it
> > or execute an action passed via
Bleep bloop. Greetings Paul Blakey, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
build:
depbase=`echo lib/dpif-netlink-rtnl.lo | sed 's|[^/]*$|.deps/&|;s|\.lo$||'`;\
/bin/sh ./libtool
On Wed, Dec 18, 2019 at 02:01:47AM +, Yi Yang (杨�D)-云服务集团 wrote:
> Ben, thank for your review, for recvmmsg, we have to prepare some buffers
> for it, but we have no way to know how many packets are there for socket, so
> these mallocs are must-have overhead, maybe self-adaptive malloc
On Thu, Dec 19, 2019 at 11:59:00AM +0100, Timothy Redaelli wrote:
> On Wed, 18 Dec 2019 16:50:05 -0800
> Ben Pfaff wrote:
>
> > On Wed, Dec 18, 2019 at 04:53:10PM +0100, Timothy Redaelli wrote:
> > > Currently pip is used to install the python packages on linux by travis,
> > > but pip3 should
sent v6
From: Ilya Maximets
Sent: Wednesday, December 18, 2019 11:38 PM
To: Paul Blakey ; Roi Dayan ; Simon
Horman ; Oz Shlomo ; Marcelo
Ricardo Leitner ; Justin Pettit ; Ilya
Maximets ; Ben Pfaff ; d...@openvswitch.org
Subject: Re: [ovs-dev][PATCH v5 07/10]
Zone and ct_state first.
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
---
Changelog:
V4->V5:
Removed unused int err in parse_put_flow_ct_action
---
lib/dpif-netlink.c | 2 +
lib/netdev-offload-tc.c | 135
lib/tc.c
Each recirculation id will create a tc chain, and we translate
the recirculation action to a tc goto chain action.
We check for kernel support for this by probing OvS Datapath for the
tc recirc id sharing feature. If supported, we can offload rules
that match on recirc_id, and recirculation
To be consistent with our tc-ufid mapping after flush, and to support tc
chains flushing in the next commit, implement flush operation via
deleting all the filters we actually added and delete their mappings.
This will also not delete the configured qos policing via matchall filters,
while old
Since OvS datapath gets packets already decapsulated from tunnel devices,
it doesn't explicitly decapsulate them. So in a recirculation setup,
the tunnel matching continues in the recirculation as the tunnel
metadata still exists on the SKB.
Tunnel key unset action unsets this metadata. Some
Update kernel UAPI to support conntrack matches, and the
tc actions ct and goto chain.
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
---
include/linux/automake.mk| 3 ++-
include/linux/pkt_cls.h | 46 +---
include/linux/tc_act/tc_ct.h | 41
Move all that is needed to identify a tc filter to a
new structure, tcf_id. This removes a lot of duplication
in accessing/creating tc filters.
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
---
Changelog:
V3->V4:
Fix accidently removed Block_id in flow_put()
V2->V3:
Renamed
The following patchset introduces hardware offload of OVS connection
tracking datapath rules.
OVS uses ct() and recirc() (recirculation) actions and recirc_id()/ct_state()
matches to support connection tracking.
The datapath rules are in the form of:
Sets zone in match.
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
---
include/openvswitch/match.h | 2 ++
lib/match.c | 10 --
2 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/include/openvswitch/match.h b/include/openvswitch/match.h
index
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
---
Changelog:
V2->V3:
Added missing match on ct_mark.
---
lib/dpif-netlink.c | 2 ++
lib/netdev-offload-tc.c | 66 +
lib/tc.c| 53
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
---
Changelog:
V4->V5:
Moved the unused 'int err' from patch 4/10 to it's first usage here
V2->V3:
Ipv6 nat dump fix (missing unspec type in policy)
Renamed tc range struct to be consistent with NL ATTRIBUTES and ovs
V1->V2:
This enables user features on the kernel datapath via the DP_CMD_SET
command, and also retrieves them to check for actual support and
not just an older kernel ignoring the requested features.
This will be used in next patch to enable recirc_id sharing with tc.
Signed-off-by: Paul Blakey
Bleep bloop. Greetings Eli Britstein, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
checkpatch:
WARNING: Line is 96 characters long (recommended limit is 79)
#195 FILE: lib/dpctl.man:130:
On 12/19/2019 1:11 AM, Ilya Maximets wrote:
> I really roughly looked through the series and made a couple of comments.
> Also, there are few checkpatch warnings about line lengths. And it's
> still unclear what is the profit from dynamic allocation of flow patters
> (I didn't decide if I like
On 10/12/2019 10:10, David Marchand wrote:
> Most DPDK components make the assumption that rte_lcore_id() returns
> a valid lcore_id in [0..RTE_MAX_LCORE[ range (with the exception of
> the LCORE_ID_ANY special value).
>
> OVS does not currently check which value is set in
>
Implement the flow get method for DPDK, to get the statistics of the
provided ufid, towards reading statistics of fully offloaded flows.
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
lib/netdev-offload-dpdk.c | 47 +++
1 file changed, 47
Currently, netdev datapath offload only accelerates the flow match
sequence by associating a mark per flow. This series introduces the full
offload of netdev datapath flows by having the HW also perform the flow
actions.
This series adds HW offload for output, drop, set MAC, set IPv4, set
TCP/UDP
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
Documentation/howto/dpdk.rst | 1 +
NEWS | 4 ++--
lib/netdev-offload-dpdk.c| 43 +++
3 files changed, 46 insertions(+), 2 deletions(-)
diff --git
Refactor the flow patterns code to a helper function for better
readability and towards supporting more matches.
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
lib/netdev-offload-dpdk.c | 207 +-
1 file changed, 111 insertions(+), 96
Add a getter function for using the dpdk port id outside the scope of
netdev-dpdk.c to be used for HW offload.
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
lib/netdev-dpdk.c | 18 ++
lib/netdev-dpdk.h | 2 ++
2 files changed, 20 insertions(+)
diff --git
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
Documentation/howto/dpdk.rst | 1 +
NEWS | 3 +-
lib/netdev-offload-dpdk.c| 99
3 files changed, 102 insertions(+), 1 deletion(-)
diff --git
Currently HW offload is accelerating only the rule matching sequence.
Introduce a framework for offloading rule actions as a pre-step for
processing the rule actions in HW. In case of a failure, fallback to the
legacy partial offload scheme.
Note: a flow will be fully offloaded only if it can
Populate dpif class field in offload struct to be used in offloading
flow put.
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
lib/dpif-netdev.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index e59a75ef0..892b2d0f5 100644
---
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
Documentation/howto/dpdk.rst | 1 +
NEWS | 2 +-
lib/netdev-offload-dpdk.c| 5 +++--
3 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/Documentation/howto/dpdk.rst b/Documentation/howto/dpdk.rst
Action item data structures are pointed by rte_flow_action items.
Refactor the code to free the data structures when freeing the
rte_flow_action items, allowing simpler future actions simpler to add to
the code.
Signed-off-by: Eli Britstein
---
lib/netdev-offload-dpdk.c | 92
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
Documentation/howto/dpdk.rst | 1 +
NEWS | 4 ++--
lib/netdev-offload-dpdk.c| 41 +
3 files changed, 44 insertions(+), 2 deletions(-)
diff --git
Signed-off-by: Eli Britstein
---
lib/netdev-offload-provider.h | 1 +
lib/netdev-offload.c | 12
2 files changed, 13 insertions(+)
diff --git a/lib/netdev-offload-provider.h b/lib/netdev-offload-provider.h
index 4e1c4251d..5a809c0cd 100644
---
Flows that are offloaded via DPDK can be partially offloaded (matches
only) or fully offloaded (matches and actions). Set partially offloaded
display to (offloaded=partial, dp:ovs), and fully offloaded to
(offloaded=yes, dp:dpdk). Also support filter types "dpdk" and
"partially-offloaded".
From: Ophir Munk
In case a flow is HW offloaded, packets do not reach the SW, thus not
counted for statistics. Use netdev flow get API in order to update the
statistics of flows by the HW statistics.
Co-authored-by: Eli Britstein
Signed-off-by: Ophir Munk
Reviewed-by: Oz Shlomo
Add debug prints when creating and destroying rte flows, with all the
flow details (attributes, patterns, actions).
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
lib/netdev-offload-dpdk.c | 209 ++
1 file changed, 138 insertions(+), 71
Instead of statically allocated pattern items on the stack, dynamically
allocate only the required items while parsing the matches, to simplify
the parsing and make it self-contained, without need of external types,
making it easier to support more matches in the future.
Signed-off-by: Eli
Introduce a rte flow query function as a pre-step towards reading HW
statistics of fully offloaded flows.
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
lib/netdev-dpdk.c | 30 ++
lib/netdev-dpdk.h | 6 ++
2 files changed, 36 insertions(+)
diff --git
Support offload of output action, also configuring count action for
allowing query statistics of HW offloaded flows.
Signed-off-by: Eli Britstein
Reviewed-by: Oz Shlomo
---
Documentation/howto/dpdk.rst | 17 ++---
NEWS | 1 +
lib/netdev-offload-dpdk.c| 86
Change the find method to return the whole entry of UFID-rte_flow
association instead of only the rte_flow field in it, as a pre-step
towards adding and using more fields into that map entry.
Signed-off-by: Eli Britstein
---
lib/netdev-offload-dpdk.c | 29 ++---
1 file
On Wed, 18 Dec 2019 16:50:05 -0800
Ben Pfaff wrote:
> On Wed, Dec 18, 2019 at 04:53:10PM +0100, Timothy Redaelli wrote:
> > Currently pip is used to install the python packages on linux by travis,
> > but pip3 should be used since pip is a symlink of pip2.
> >
> > Fixes: 1ca0323e7c29 ("Require
On Tue, Dec 17, 2019 at 11:11 PM venugopal iyer via dev
wrote:
>
> Would appreciate any thoughts on this.
> thanks,
> -venu
Hi Venu,
The patch doesn't apply cleanly. Can you please rebase and submit v3.
Thanks
Numan
>
> On Tuesday, November 19, 2019, 12:40:47 PM PST,
> wrote:
>
>
On 16.12.2019 19:21, William Tu wrote:
> On Mon, Dec 16, 2019 at 03:22:00PM +0100, Ilya Maximets wrote:
>> Marking the structure as 64 bytes aligned forces compiler to produce
>> big holes in the containing structures in order to fulfill this
>> requirement. Also, any structure that contains this
58 matches
Mail list logo