Message ID | 1517209188-16608-7-git-send-email-yliu@fridaylinux.org |
---|---|
State | Changes Requested |
Delegated to: | Ian Stokes |
Headers | show |
Series | OVS-DPDK flow offload with rte_flow | expand |
> -----Original Message----- > From: Yuanhan Liu [mailto:yliu@fridaylinux.org] > Sent: Monday, January 29, 2018 7:00 AM > To: dev@openvswitch.org > Cc: Stokes, Ian <ian.stokes@intel.com>; Finn Christensen > <fc@napatech.com>; Darrell Ball <dball@vmware.com>; Chandran, Sugesh > <sugesh.chandran@intel.com>; Simon Horman <simon.horman@netronome.com>; > Yuanhan Liu <yliu@fridaylinux.org> > Subject: [PATCH v7 6/6] Documentation: document ovs-dpdk flow offload > > And mark it as experimental. Commit message should be a bit more detailed than this. Details on what is being added and where. > > Signed-off-by: Yuanhan Liu <yliu@fridaylinux.org> > --- > Documentation/howto/dpdk.rst | 17 +++++++++++++++++ > NEWS | 1 + > 2 files changed, 18 insertions(+) > > diff --git a/Documentation/howto/dpdk.rst b/Documentation/howto/dpdk.rst > index 40f9d96..047525c 100644 > --- a/Documentation/howto/dpdk.rst > +++ b/Documentation/howto/dpdk.rst > @@ -727,3 +727,20 @@ devices to bridge ``br0``. Once complete, follow the > below steps: > Check traffic on multiple queues:: > > $ cat /proc/interrupts | grep virtio > + > +.. _dpdk-flow-hardware-offload: > + > +Flow Hardware Offload (Experimental) > +------------------------------------ > + > +The flow hardware offload is disabled by default and can be enabled by:: > + > + $ ovs-vsctl set Open_vSwitch . other_config:hw-offload=true > + > +So far only partial flow offload is implemented. Moreover, it only > +works with PDM drivers have the rte flow action "MARK + RSS" support. > + > +The validated NICs are: > + > +- Mellanox (ConnectX-4, ConnectX-4 Lx, ConnectX-5) > +- Napatech (NT200B01) I think it would be useful to document the flow types protocols that can/can't be offloaded in a table here i.e. udp, tcp, ICMP etc. I don't think these are documented anywhere, I could see users asking what traffic they can offload when getting started. Regards Ian > diff --git a/NEWS b/NEWS > index d7d585b..d0c9f44 100644 > --- a/NEWS > +++ b/NEWS > @@ -54,6 +54,7 @@ v2.9.0 - xx xxx xxxx > * New appctl command 'dpif-netdev/pmd-rxq-rebalance' to rebalance > rxq to > pmd assignments. > * Add rxq utilization of pmd to appctl 'dpif-netdev/pmd-rxq-show'. > + * Add experimental flow hardware offload support > - Userspace datapath: > * Output packet batching support. > - vswitchd: > -- > 2.7.4
> -----Original Message----- > From: Stokes, Ian > Sent: Monday, March 12, 2018 3:28 PM > To: Yuanhan Liu <yliu@fridaylinux.org>; dev@openvswitch.org > Cc: Finn Christensen <fc@napatech.com>; Darrell Ball <dball@vmware.com>; > Chandran, Sugesh <sugesh.chandran@intel.com>; Simon Horman > <simon.horman@netronome.com> > Subject: RE: [PATCH v7 6/6] Documentation: document ovs-dpdk flow offload > Adding Shahaf Shuler. > > -----Original Message----- > > From: Yuanhan Liu [mailto:yliu@fridaylinux.org] > > Sent: Monday, January 29, 2018 7:00 AM > > To: dev@openvswitch.org > > Cc: Stokes, Ian <ian.stokes@intel.com>; Finn Christensen > > <fc@napatech.com>; Darrell Ball <dball@vmware.com>; Chandran, Sugesh > > <sugesh.chandran@intel.com>; Simon Horman > > <simon.horman@netronome.com>; Yuanhan Liu <yliu@fridaylinux.org> > > Subject: [PATCH v7 6/6] Documentation: document ovs-dpdk flow offload > > > > And mark it as experimental. > > Commit message should be a bit more detailed than this. Details on what is > being added and where. > > > > > Signed-off-by: Yuanhan Liu <yliu@fridaylinux.org> > > --- > > Documentation/howto/dpdk.rst | 17 +++++++++++++++++ > > NEWS | 1 + > > 2 files changed, 18 insertions(+) > > > > diff --git a/Documentation/howto/dpdk.rst > > b/Documentation/howto/dpdk.rst index 40f9d96..047525c 100644 > > --- a/Documentation/howto/dpdk.rst > > +++ b/Documentation/howto/dpdk.rst > > @@ -727,3 +727,20 @@ devices to bridge ``br0``. Once complete, follow > > the below steps: > > Check traffic on multiple queues:: > > > > $ cat /proc/interrupts | grep virtio > > + > > +.. _dpdk-flow-hardware-offload: > > + > > +Flow Hardware Offload (Experimental) > > +------------------------------------ > > + > > +The flow hardware offload is disabled by default and can be enabled > by:: > > + > > + $ ovs-vsctl set Open_vSwitch . other_config:hw-offload=true > > + > > +So far only partial flow offload is implemented. Moreover, it only > > +works with PDM drivers have the rte flow action "MARK + RSS" support. > > + > > +The validated NICs are: > > + > > +- Mellanox (ConnectX-4, ConnectX-4 Lx, ConnectX-5) > > +- Napatech (NT200B01) > > I think it would be useful to document the flow types protocols that > can/can't be offloaded in a table here i.e. udp, tcp, ICMP etc. > I don't think these are documented anywhere, I could see users asking what > traffic they can offload when getting started. > > Regards > Ian > > > diff --git a/NEWS b/NEWS > > index d7d585b..d0c9f44 100644 > > --- a/NEWS > > +++ b/NEWS > > @@ -54,6 +54,7 @@ v2.9.0 - xx xxx xxxx > > * New appctl command 'dpif-netdev/pmd-rxq-rebalance' to > > rebalance rxq to > > pmd assignments. > > * Add rxq utilization of pmd to appctl 'dpif-netdev/pmd-rxq-show'. > > + * Add experimental flow hardware offload support > > - Userspace datapath: > > * Output packet batching support. > > - vswitchd: > > -- > > 2.7.4
diff --git a/Documentation/howto/dpdk.rst b/Documentation/howto/dpdk.rst index 40f9d96..047525c 100644 --- a/Documentation/howto/dpdk.rst +++ b/Documentation/howto/dpdk.rst @@ -727,3 +727,20 @@ devices to bridge ``br0``. Once complete, follow the below steps: Check traffic on multiple queues:: $ cat /proc/interrupts | grep virtio + +.. _dpdk-flow-hardware-offload: + +Flow Hardware Offload (Experimental) +------------------------------------ + +The flow hardware offload is disabled by default and can be enabled by:: + + $ ovs-vsctl set Open_vSwitch . other_config:hw-offload=true + +So far only partial flow offload is implemented. Moreover, it only works +with PDM drivers have the rte flow action "MARK + RSS" support. + +The validated NICs are: + +- Mellanox (ConnectX-4, ConnectX-4 Lx, ConnectX-5) +- Napatech (NT200B01) diff --git a/NEWS b/NEWS index d7d585b..d0c9f44 100644 --- a/NEWS +++ b/NEWS @@ -54,6 +54,7 @@ v2.9.0 - xx xxx xxxx * New appctl command 'dpif-netdev/pmd-rxq-rebalance' to rebalance rxq to pmd assignments. * Add rxq utilization of pmd to appctl 'dpif-netdev/pmd-rxq-show'. + * Add experimental flow hardware offload support - Userspace datapath: * Output packet batching support. - vswitchd:
And mark it as experimental. Signed-off-by: Yuanhan Liu <yliu@fridaylinux.org> --- Documentation/howto/dpdk.rst | 17 +++++++++++++++++ NEWS | 1 + 2 files changed, 18 insertions(+)