On Wed, May 26, 2021 at 2:50 PM Eelco Chaudron wrote:
>
>
>
> On 26 May 2021, at 14:46, Matteo Croce wrote:
>
> > On Wed, May 26, 2021 at 2:34 PM Eelco Chaudron wrote:
> >> On 18 May 2021, at 20:15, Ilya Maximets wrote:
> >>>> +/* Tests whether 'dp
On Wed, May 26, 2021 at 2:34 PM Eelco Chaudron wrote:
> On 18 May 2021, at 20:15, Ilya Maximets wrote:
> >> +/* Tests whether 'dpif' datapath supports decrement of the IP TTL via
> >> + * OVS_ACTION_DEC_TTL. */
> >> +static bool
> >> +check_dec_ttl_action(struct dpif *dpif)
> >> +{
> >> +
tional case statements, which
> should be revisited once the OVS implementation is added.
>
>
> Co-developed-by: Matteo Croce
> Co-developed-by: Bindiya Kurle
> Signed-off-by: Eelco Chaudron
>
Acked-by: Matteo Croce
--
per aspera ad upstream
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
p
> > ip link set center-right up
> > ip -n left ip link set left0 up
> > ip -n left ip addr add 172.31.110.10/24 dev left0
> > ip -n right ip link set right0 up
> > ip -n right ip addr add 172.31.110.11/24 dev right0
> >
>
> I think the "ip
On Fri, Nov 20, 2020 at 10:12 PM Jakub Kicinski wrote:
>
> On Thu, 19 Nov 2020 04:04:04 -0500 Eelco Chaudron wrote:
> > Currently, the openvswitch module is not accepting the correctly formated
> > netlink message for the TTL decrement action. For both setting and getting
> > the dec_ttl action,
On Fri, Nov 13, 2020 at 3:28 PM Eelco Chaudron wrote:
>
> It is already in the mainline kernel, so changing it now would break the
> UAPI. Don't think this is allowed from the kernel side.
>
> > What I'm suggesting is to send a bugfix to kernel
> > to accept only format with nested
tional case statements, which
> should be revisited once the OVS implementation is added.
>
>
> Co-developed-by: Matteo Croce
> Co-developed-by: Bindiya Kurle
> Signed-off-by: Eelco Chaudron
> ---
> datapath/linux/compat/include/linux/openvswitch.h |8 ++
> lib/dpif-netdev.
On Fri, Aug 28, 2020 at 7:00 PM Mark Gray wrote:
> v3: one port latency results as per Matteo's comments
>
> Stock:
> min/avg/max/mdev = 21.5/36.5/96.5/1.0 us
> With Patch:
> min/avg/max/mdev = 5.3/9.7/98.4/0.5 us
>
Awesome, LGTM!
--
Matteo Croce
Good job!
Did you have the chance to run the same tests I did in 69c51582ff78?
I'm curious to see how much it improves.
Bye,
--
Matteo Croce
perl -e 'for($t=0;;$t++){print chr($t*($t>>8|$t>>13)&255)}' |aplay
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
.
>
> Reported-by: David Ahern
> Reported-at:
> https://mail.openvswitch.org/pipermail/ovs-dev/2019-December/365857.html
> Cc: Matteo Croce
> Cc: Flavio Leitner
> Signed-off-by: Aaron Conole
>
Great work!
Just to understand, this reverts some logic of my patch, and
tr
Extend the balance-tcp one so it tests lb-output-action too.
The test checks that that the option is shown in bond/show,
and that the lb_output action is programmed in the datapath.
Signed-off-by: Matteo Croce
---
tests/ofproto-dpif.at | 31 ---
1 file changed, 28
GE_ADD(datapath_drop_invalid_port,
+ dp_packet_batch_size(_pkts));
+dp_packet_delete_batch(_pkts, should_steal);
+
+return true;
}
static void
Regards,
--
Matteo Croce
per aspera ad upstream
___
dev mailin
4
# ping -c1 192.168.0.2 -t 1
#
Co-developed-by: Bindiya Kurle
Signed-off-by: Bindiya Kurle
Signed-off-by: Matteo Croce
---
include/uapi/linux/openvswitch.h | 7
net/openvswitch/actions.c| 67 ++
net/openvswitch/flow_netl
4
# ping -c1 192.168.0.2 -t 1
#
Co-developed-by: Bindiya Kurle
Signed-off-by: Bindiya Kurle
Signed-off-by: Matteo Croce
---
include/uapi/linux/openvswitch.h | 7
net/openvswitch/actions.c| 67 ++
net/openvswitch/flow_netl
4
# ping -c1 192.168.0.2 -t 1
#
Co-developed-by: Bindiya Kurle
Signed-off-by: Bindiya Kurle
Signed-off-by: Matteo Croce
---
include/uapi/linux/openvswitch.h | 2 +
net/openvswitch/actions.c| 67 ++
net/openvswitch/flow_netl
On Tue, Dec 17, 2019 at 5:30 PM Nikolay Aleksandrov
wrote:
>
> On 17/12/2019 17:51, 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 ac
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
> > o
tes from the
commit message? They shows in git log as:
<80><9C>balance-tcp<80><9D>
Regards,
--
Matteo Croce
per aspera ad upstream
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
1, length 64
# ping -c1 192.168.0.2 -t 1
#
Co-authored-by: Bindiya Kurle
Signed-off-by: Bindiya Kurle
Signed-off-by: Matteo Croce
---
include/uapi/linux/openvswitch.h | 22 +++
net/openvswitch/actions.c| 71 +
net/openvswitch/flow_netli
On Tue, Dec 10, 2019 at 10:00 PM David Ahern wrote:
>
> [ adding Jason as author of the patch that added the epoll exclusive flag ]
>
> On 12/10/19 12:37 PM, Matteo Croce wrote:
> > On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
> >>
> >> Hi Matte
On Tue, Dec 10, 2019 at 8:41 PM David Ahern wrote:
>
> On 12/10/19 12:37 PM, Matteo Croce wrote:
> > On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
> >>
> >> Hi Matteo:
> >>
> >> On a hypervisor running a 4.14.91 kernel and OVS 2.11 I am seei
commits it appears
> that the change in behavior comes from this commit:
>
> commit 69c51582ff786a68fc325c1c50624715482bc460
> Author: Matteo Croce
> Date: Tue Sep 25 10:51:05 2018 +0200
>
> dpif-netlink: don't allocate per thread netlink sockets
>
>
> Is this a kno
On Mon, Nov 18, 2019 at 5:20 PM Ben Pfaff wrote:
>
> On Tue, Nov 12, 2019 at 04:46:12PM +0100, Matteo Croce wrote:
> > On Tue, Nov 12, 2019 at 4:00 PM Simon Horman
> > wrote:
> > >
> > > On Tue, Nov 12, 2019 at 11:25:18AM +0100, Matteo Croce wrote:
> &g
On Tue, Nov 12, 2019 at 4:00 PM Simon Horman wrote:
>
> On Tue, Nov 12, 2019 at 11:25:18AM +0100, 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, send the
> >
indiya Kurle
Signed-off-by: Matteo Croce
---
include/uapi/linux/openvswitch.h | 2 ++
net/openvswitch/actions.c| 46
net/openvswitch/flow_netlink.c | 6 +
3 files changed, 54 insertions(+)
diff --git a/include/uapi/linux/openvswitch.h b/include/uapi
On Fri, Nov 8, 2019 at 11:04 PM William Tu wrote:
>
> On Fri, Nov 08, 2019 at 05:12:55PM +0100, Matteo Croce wrote:
> > Hi,
> >
> > I need to add a field to enum ovs_action_attr, but I see that the
> > definition between the upstream header[1] and the one in compa
nel/git/torvalds/linux.git/tree/include/uapi/linux/openvswitch.h?h=v5.3#n899
[2]
https://github.com/openvswitch/ovs/blob/v2.12.0/datapath/linux/compat/include/linux/openvswitch.h#L962
Regards,
--
Matteo Croce
per aspera ad upstream
___
de
t; Rebased to OVS master.
> Fixed git merge issue.
>
> v1->v2:
> Updated datapath action to hash + lb-output.
> Updated throughput test observations.
> Rebased to OVS master.
>
All unit test passed, plus others I did.
Tested-by: Matteo Croce
--
Matt
On Wed, Sep 11, 2019 at 8:37 PM Matteo Croce wrote:
>
> On Tue, Sep 10, 2019 at 11:53 AM Vishal Deep Ajmera
> wrote:
> >
> > v5->v6:
> > Addressed comments from Ilya Maximets.
> > https://mail.openvswitch.org/pipermail/ovs-dev/2019-August/362001.html
> &
.9.1
>
Hi,
I confirm a decent performance improvement with DPDK and balance-tcp bonding:
lb-output-action=false
rx: 740 Mbps 1446 kpps
lb-output-action=true
rx: 860 Mbps 1680 kpps
I'm running a very simple test with a tweaked version of testpmd which
generates 256 L4 flows, I guess t
>
> CC: Jan Scheurich
> CC: Venkatesan Pradeep
> CC: Nitin Katiyar
Hi Vishal,
I quickly tested your patch on two servers with 2 ixgbe cards each
linked via a juniper switch with LACP.
With testpmd using a single core the switching rate raised from ~2.0
Mpps to ~2.4+ Mpps, so I read at least a +20% gain.
Please add an example command on how to enable it in the commit message, e.g.
ovs-vsctl set port bond0 other_config:opt-bond-tcp=true
Thanks,
--
Matteo Croce
per aspera ad upstream
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
e_mode(const struct bond *bond)
> +{
> +return bond->use_bond_cache;
> +}
Hi,
why not a static function in the header file? So it gets inlined.
Regards,
--
Matteo Croce
per aspera ad upstream
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
old behaviour of using a zero pid when dpif->handlers is 0.
Fixes: 69c51582f ("dpif-netlink: don't allocate per thread netlink sockets")
Reported-by: Flavio Leitner
Reported-by: Guru Shetty
Tested-by: Guru Shetty
Signed-off-by: Matteo Croce
---
v2: fix checkpatch.py error about coding
On Mon, Oct 8, 2018 at 3:11 PM Guru Shetty wrote:
>
>
>
> On Sat, 6 Oct 2018 at 09:20, Matteo Croce wrote:
>>
>> In dpif_netlink_port_add__(), socksp could be NULL, because
>> vport_socksp_to_pids() would allocate a new array and return a single
>> zero elemen
On Fri, Oct 5, 2018 at 9:58 AM Markos Chandras wrote:
>
> On 05/10/2018 09:55, Matteo Croce wrote:
> > On Fri, Oct 5, 2018 at 8:32 AM Markos Chandras wrote:
> >>
> >> On 25/09/2018 22:14, Ben Pfaff wrote:
> >>>
> >>> Applied to maste
old behaviour of using a zero pid when dpif->handlers is 0.
Fixes: 69c51582f ("dpif-netlink: don't allocate per thread netlink sockets")
Reported-by: Flavio Leitner
Reported-by: Guru Shetty
Signed-off-by: Matteo Croce
---
lib/dpif-netlink.c | 5 +++--
1 file changed, 3 insertions
vswitchd/bridge.c:3023
> #13 0x004080c5 in main (argc=2, argv=0x7fffe5e8) at
> vswitchd/ovs-vswitchd.c:125
>
>
Hi Guru,
I received a similar report on IRC and I have already investigated it.
Can you try if the following patch fixes it?
Fixes: 69c51582f ("dpif-netlink
it issue was first reported on OVS 2.8, so it's worth it.
If you backport it, please consider backporting the following commit from Ben,
which removes some code which become unused after my patch:
commit 769b50349f28c5f9e4bff102bc61dadcb9b99c37
Author: Ben Pfaff
Date: Tue Sep 25 15:14:13 2018 -0700
On Wed, Sep 26, 2018 at 10:29 PM Yifeng Sun wrote:
>
> It looks good to me, and testing shows no problem. Thanks.
>
> Tested-by: Yifeng Sun
> Reviewed-by: Yifeng Sun
>
Works fine here too.
Regards,
--
Matteo Croce
per aspera ad upstream
_
with almost 99% CPU usage.
[1] https://github.com/teknoraver/network-tools/blob/master/weed.c
Signed-off-by: Matteo Croce
---
v1 -> v2:
- define EPOLLEXCLUSIVE on systems with older kernel headers
- explain the thread safety test in the commit message
lib/dpif-netlink.c |
On Fri, Sep 21, 2018 at 8:41 PM Ben Pfaff wrote:
>
> On Wed, Sep 19, 2018 at 02:47:03PM +0200, Matteo Croce wrote:
> > When using the kernel datapath, OVS allocates a pool of sockets to handle
> > netlink events. The number of sockets is: ports * n-handler-threads, where
>
:
# ps u $(pidof ovs-vswitchd)
USER PID %CPU %MEMVSZ RSS TTY STAT START TIME COMMAND
openvsw+ 5430 54.3 0.3 4263964 510968 pts/1 RLl+ 16:20 0:50 ovs-vswitchd
[1] https://github.com/teknoraver/network-tools/blob/master/weed.c
Signed-off-by: Matteo Croce
---
lib/dpif
On Mon, Jul 16, 2018 at 4:54 PM Matteo Croce wrote:
>
> On Tue, Jul 10, 2018 at 6:31 PM Pravin Shelar wrote:
> >
> > On Wed, Jul 4, 2018 at 7:23 AM, Matteo Croce wrote:
> > > From: Stefano Brivio
> > >
> > > Open vSwitch sends to userspace all r
On Tue, Jul 10, 2018 at 6:31 PM Pravin Shelar wrote:
>
> On Wed, Jul 4, 2018 at 7:23 AM, Matteo Croce wrote:
> > From: Stefano Brivio
> >
> > Open vSwitch sends to userspace all received packets that have
> > no associated flow (thus doing an "upcall")
ation (i.e. upcalls rate) otherwise.
This ensures fairness among ports under load and with few
netlink sockets.
Signed-off-by: Matteo Croce
Co-authored-by: Stefano Brivio
---
net/openvswitch/datapath.c | 143 ++---
net/openvswitch/datapath.h | 27 ++-
2 fil
Documentation for both n-handler-threads and n-revalidator-threads was
wrong, the spawned threads are per port instead of per datapath.
Update the docs according to vswitchd/ovs-vswitchd.8.in and code behaviour.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
vswitchd/vswitch.xm
On Mon, Jan 15, 2018 at 7:21 PM, Matteo Croce <mcr...@redhat.com> wrote:
> Show DPDK version if Open vSwitch is compiled with DPDK support.
> Version can be retrieved with `ovs-switchd --version` or from OVS logs.
> Small change in ovs-ctl to avoid breakage on output change.
with a single dpif_channel instance
and edit the code accordingly. By doing so, the port idx information is lost
in upcall event code, so the call to report_loss() must be commented, as it
already is in the Windows code path.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
lib/dpif-netlink.c
Show DPDK version if Open vSwitch is compiled with DPDK support.
Version can be retrieved with `ovs-switchd --version` or from OVS logs.
Small change in ovs-ctl to avoid breakage on output change.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
v3:
print version in OVS logs too
don'
Show DPDK version if Open vSwitch is compiled with DPDK support.
Version can be retrieved with `ovs-switchd --version` or from OVS logs.
Small change in ovs-ctl to avoid breakage on output change.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
v3:
print version in OVS logs too
don'
version from the log files while debugging issues.
>
> Warm Regards,
> Vishal Ajmera
>
Sure, makes sense.
Thanks,
--
Matteo Croce
per aspera ad upstream
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On Tue, Jan 2, 2018 at 6:08 PM, Aaron Conole <acon...@redhat.com> wrote:
> Hi Matteo,
>
> Matteo Croce <mcr...@redhat.com> writes:
>
>> Show DPDK version if Open vSwitch is compiled with DPDK support.
>> Add a `dpdk_version` column in the datamodel, change o
On Tue, Dec 26, 2017 at 6:38 PM, Ben Pfaff <b...@ovn.org> wrote:
> On Mon, Dec 25, 2017 at 03:31:02PM +0100, Matteo Croce wrote:
>> Show DPDK version if Open vSwitch is compiled with DPDK support.
>> Add a `dpdk_version` column in the datamodel, change ovs-ctl and ovs-
Show DPDK version if Open vSwitch is compiled with DPDK support.
Add a `dpdk_version` column in the datamodel, change ovs-ctl and ovs-dev.py
to populate the field with the right value.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
utilities/ovs-ctl.in | 9 -
utiliti
On Sun, Dec 24, 2017 at 8:02 PM, Ben Pfaff <b...@ovn.org> wrote:
> On Sun, Dec 24, 2017 at 02:58:37PM +0100, Matteo Croce wrote:
>> Show DPDK version if Open vSwitch is compiled with DPDK support.
>> Add a `dpdk_version` column in the datamodel, change ovs-ctl and ovs-
Show DPDK version if Open vSwitch is compiled with DPDK support.
Add a `dpdk_version` column in the datamodel, change ovs-ctl and ovs-dev.py
to populate the field with the right value.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
v2: edit ovs-ctl as well
utilities/ovs-ctl.in
Show DPDK version if Open vSwitch is compiled with DPDK support.
Add a `dpdk_version` column in the datamodel and also edit ovs-dev.py to
populate the field with the right value.
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
utilities/ovs-dev.py | 15 ---
utiliti
>> as the call to rte_eth_dev_set_mtu() was in a loop.
>>
>> CC: Mark Kavanagh <mark.b.kavan...@intel.com>
>> Fixes: 67fe6d635193 ("netdev-dpdk: use rte_eth_dev_set_mtu.")
>> Signed-off-by: Matteo Croce <mcr...@redhat.com>
>> ---
>>
to rte_eth_dev_set_mtu() was in a loop.
CC: Mark Kavanagh <mark.b.kavan...@intel.com>
Fixes: 67fe6d635193 ("netdev-dpdk: use rte_eth_dev_set_mtu.")
Signed-off-by: Matteo Croce <mcr...@redhat.com>
---
lib/netdev-dpdk.c | 14 +++---
1 file changed, 7 insertions(+), 7 deletion
o indicate if the DB needs upgrading,
but if the DB is corrupted it returns a list of errors.
Change a condition from "!= no" to "= yes" because in case of DB
corruption upgrade_db would purge the existing DB without writing
anything in the logs.
Signed-off-by: Matteo Croc
60 matches
Mail list logo