Re: [ovs-dev] [PATCH v13] Improved Packet Drop Statistics in OVS

2019-10-18 Thread Eelco Chaudron
Are you planning on sending a v14? Cheers, Eelco On 16 Sep 2019, at 16:12, Eelco Chaudron wrote: Hi Anju, For some reason, this version does not show up in patchwork! I have two small comments below but other than that it looks fine to me. If those two will be the only changes in your

Re: [ovs-dev] [PATCH v2] netdev-dpdk: Track vhost tx contention.

2019-10-15 Thread Eelco Chaudron
-by: David Marchand --- Changelog since v1: - added documentation as a bonus: not sure this is the right place, or if it really makes sense to enter into such details. But I still find it useful. Comments? Thanks for the detailed documentation! Acked-by: Eelco Chaudron

Re: [ovs-dev] [PATCHv3] netdev-linux: Detect numa node id.

2019-10-15 Thread Eelco Chaudron
See some comment below… On 2 Oct 2019, at 22:26, William Tu wrote: The patch detects the numa node id from the name of the netdev, by reading the '/sys/class/net//device/numa_node'. If not available, ex: virtual device, or any error happens, return numa id 0. Currently only the afxdp netdev

[ovs-dev] [dpdk-latest PATCH] netdev-dpdk: add custom vhost statistics to count IRQs

2019-10-15 Thread Eelco Chaudron
system-calls were called in the DPDK fast path. The custom statistics can be seen with the following command: ovs-ofctl -O OpenFlow14 dump-ports {} Signed-off-by: Eelco Chaudron --- lib/netdev-dpdk.c | 27 ++- 1 file changed, 26 insertions(+), 1 deletion(-) diff --git

[ovs-dev] [dpdk-latest PATCH v3 2/2] netdev-dpdk: Add new DPDK RFC 4115 egress policer

2019-10-01 Thread Eelco Chaudron
1000pps. Signed-off-by: Eelco Chaudron --- Documentation/topics/dpdk/qos.rst | 36 lib/netdev-dpdk.c | 321 + vswitchd/vswitch.xml | 30 +++ 3 files changed, 386 insertions(+), 1 deletion(-) diff --git a/Documentation/topics

[ovs-dev] [dpdk-latest PATCH v3 1/2] netdev-dpdk: Add support for multi-queue QoS to the DPDK datapath

2019-10-01 Thread Eelco Chaudron
This patch adds support for multi-queue QoS to the DPDK datapath. Most of the code is based on an earlier patch from a patchset sent out by zhaozhanxu. The patch was titled "[ovs-dev, v2, 1/4] netdev-dpdk.c: Support the multi-queue QoS configuration for dpdk datapath" Signed-off

[ovs-dev] [dpdk-latest PATCH v3 0/2] netdev-dpdk: Add new DPDK RFC 4115 egress policer

2019-10-01 Thread Eelco Chaudron
ier version did not result in any feedback - Rebase to the latest dpdk-latest branch and DPDK master v1 -> v2: - Rebase so it would apply to the dpdk-latest branch and latest DPDK master Eelco Chaudron (2): netdev-dpdk: Add support for multi-queue QoS to the DPDK datapath netde

[ovs-dev] [PATCH] netdev-afxdp: add afxdp specific maximum MTU check

2019-10-01 Thread Eelco Chaudron
to make sure only AF_XDP valid values are excepted. Signed-off-by: Eelco Chaudron --- lib/netdev-afxdp.c | 17 + lib/netdev-afxdp.h |1 + lib/netdev-linux.c |9 + 3 files changed, 27 insertions(+) diff --git a/lib/netdev-afxdp.c b/lib/netdev-afxdp.c index

Re: [ovs-dev] [PATCH net-next 0/9] optimize openvswitch flow looking up

2019-10-01 Thread Eelco Chaudron
Interesting performance gain, copied in the OVS development mailing list. //Eelco On 29 Sep 2019, at 19:09, xiangxia.m@gmail.com wrote: From: Tonghao Zhang This series patch optimize openvswitch. Patch 1, 2, 4: Port Pravin B Shelar patches to linux upstream with little changes. Patch

Re: [ovs-dev] [PATCHv2] netdev-afxdp: Detect numa node id.

2019-09-30 Thread Eelco Chaudron
On 27 Sep 2019, at 19:26, William Tu wrote: The patch detects the numa node id from the name of the netdev, by reading the '/sys/class/net//device/numa_node'. If not available, ex: virtual device, or any error happens, return numa id 0. Signed-off-by: William Tu Will you update the

Re: [ovs-dev] [PATCHv5] netdev-afxdp: Add need_wakeup supprt.

2019-09-27 Thread Eelco Chaudron
v3 2.4GHz system, performance of physical port to physical port improves from 6.1Mpps to 7.3Mpps. Suggested-by: Ilya Maximets Signed-off-by: William Tu Did a quick test, and all works fine. Acke’ed based on previous review and delta… Acked-by: Eelc

Re: [ovs-dev] [PATCH ovs V1 0/2] Introduce dpdkvdpa netdev

2019-09-26 Thread Eelco Chaudron
Hi Noa, Did not review the patchset yet, but it would be good to at least include some documentation around use cases, and maybe an example to be used for testing? Cheers, Eelco On 15 Sep 2019, at 7:37, Noa Ezra wrote: Introduce dpdkvdpa netdev allowing HW offloads over VirtIO network

Re: [ovs-dev] [PATCH] netdev-afxdp: Detect numa node id.

2019-09-26 Thread Eelco Chaudron
On 13 Sep 2019, at 19:15, William Tu wrote: The patch detects the numa node id from the name of the netdev, by reading the '/sys/class/net//device/numa_node'. If not available, ex: virtual device, or any error happens, return numa id 0. See comments below, but I think the main problem that

Re: [ovs-dev] dpdk-latest rebase

2019-09-26 Thread Eelco Chaudron
On 26 Sep 2019, at 11:08, Stokes, Ian wrote: Hi all, the dpdk-latest branch has been lagging behind ovs master for a bit. Now that the OVS 2.12 release is out it seems timely to rebase it on top of OVS master and push, note as before this will be a force push on the dpdk-latest branch as

Re: [ovs-dev] [PATCH v3] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-24 Thread Eelco Chaudron
On 12 Sep 2019, at 12:24, Ilya Maximets wrote: On 12.09.2019 13:19, Ilya Maximets wrote: On 12.09.2019 13:07, Eelco Chaudron wrote: On 12 Sep 2019, at 10:39, Ilya Maximets wrote: On 11.09.2019 16:20, Eelco Chaudron wrote: Currently, OVS does not register and therefore not handle

Re: [ovs-dev] [PATCHv4] netdev-afxdp: Add need_wakeup supprt.

2019-09-18 Thread Eelco Chaudron
> > Suggested-by: Ilya Maximets > Signed-off-by: William Tu Thanks William! Acked-by: Eelco Chaudron ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCHv3] netdev-afxdp: Add need_wakeup supprt.

2019-09-17 Thread Eelco Chaudron
Two comments below… On 11 Sep 2019, at 19:58, William Tu wrote: The patch adds support for using need_wakeup flag in AF_XDP rings. A new option, use_need_wakeup, is added. When this option is used, it means that OVS has to explicitly wake up the kernel RX, using

Re: [ovs-dev] [PATCH v13] Improved Packet Drop Statistics in OVS

2019-09-16 Thread Eelco Chaudron
Hi Anju, For some reason, this version does not show up in patchwork! I have two small comments below but other than that it looks fine to me. If those two will be the only changes in your next rev, I’ll ack it straight away :) Cheers, Eelco On 9 Sep 2019, at 13:53, Anju Thomas wrote:

Re: [ovs-dev] [PATCH] netdev-dpdk: Fix Tx queue false sharing.

2019-09-12 Thread Eelco Chaudron
pps. The patch looks good to me… Acked-by: Eelco Chaudron ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH v3] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-12 Thread Eelco Chaudron
On 12 Sep 2019, at 10:39, Ilya Maximets wrote: On 11.09.2019 16:20, Eelco Chaudron wrote: Currently, OVS does not register and therefore not handle the interface reset event from the DPDK framework. This would cause a problem in cases where a VF is used as an interface, and its configuration

Re: [ovs-dev] [PATCH v2] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-11 Thread Eelco Chaudron
On 10 Sep 2019, at 16:02, Ilya Maximets wrote: On 10.09.2019 15:12, Eelco Chaudron wrote: Currently, OVS does not register and therefore not handle the interface reset event from the DPDK framework. This would cause a problem in cases where a VF is used as an interface, and its

[ovs-dev] [PATCH v3] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-11 Thread Eelco Chaudron
52:54:00:92:d3:33 Signed-off-by: Eelco Chaudron --- v2 -> v3: v1 -> v2: Fixed Ilya's comments lib/netdev-dpdk.c | 53 +++-- 1 file changed, 51 insertions(+), 2 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c inde

[ovs-dev] [PATCH] vswitchd: update bond-rebalance-interval maxInteger in vswtich.xml

2019-09-11 Thread Eelco Chaudron
in 32bit integer. This will allow a value of ~25 days, which should be enough to cover everybody's needs. Note that a value of 0 disables the automatic rebalancing anyway. Signed-off-by: Eelco Chaudron --- vswitchd/vswitch.xml | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git

Re: [ovs-dev] [PATCH] netdev-dpdk: Track vhost tx contention.

2019-09-10 Thread Eelco Chaudron
On 26 Aug 2019, at 16:33, David Marchand wrote: > Add a coverage counter to help diagnose contention on the vhost txqs. > This is seen as dropped packets on the physical ports for rates that > are usually handled fine by OVS. > > Signed-off-by: David Marchand LGTM, Acked-by:

Re: [ovs-dev] [PATCHv2] netdev-afxdp: Add need_wakeup supprt.

2019-09-10 Thread Eelco Chaudron
Hi William, One comment below, but I’m also wondering if we should warn/log a message if OVS is compiled without this feature and we explicitly configure it? Same for the reverse what if the kernel does not support this option, will it fail? //Eelco On 5 Sep 2019, at 22:51, William Tu

[ovs-dev] [PATCH v2] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-10 Thread Eelco Chaudron
52:54:00:92:d3:33 Signed-off-by: Eelco Chaudron --- v1 -> v2: - Fixed Ilya's comments lib/netdev-dpdk.c | 53 +++-- 1 file changed, 51 insertions(+), 2 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index 4805783..37a431e

Re: [ovs-dev] [PATCH] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-10 Thread Eelco Chaudron
On 5 Sep 2019, at 14:40, Ilya Maximets wrote: Hi Eelco, , 2 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index bc20d6843..a23150387 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev-dpdk.c @@ -362,6 +362,7 @@ struct netdev_dpdk { bool attached; /* If

Re: [ovs-dev] [PATCH] dpdk: Remove unneeded log message copy.

2019-09-06 Thread Eelco Chaudron
On 6 Sep 2019, at 13:26, David Marchand wrote: > No need to duplicate and null-terminate the passed buffer. > We can directly give it to the vlog subsystem using a dynamic precision > in the format string. > > Signed-off-by: David Marchand Change looks good to me! Acked-by:

Re: [ovs-dev] [PATCH 2/2] netdev-afxdp: Add need_wakeup supprt.

2019-09-06 Thread Eelco Chaudron
On 5 Sep 2019, at 18:21, William Tu wrote: Did some more testing and with PVP I see a performance decrease, with physical to physical I see an increase. Tests are performed with a port redirect open flow rule on an ixgbe (Xeon E5-2690 v4 2.60GHz):

Re: [ovs-dev] [PATCH] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-05 Thread Eelco Chaudron
, but this might be fixed too. More comments inline. Best regards, Ilya Maximets. On 05.09.2019 14:48, Eelco Chaudron wrote: Currently, OVS does not register and therefore not handle the interface reset event from the DPDK framework. This would cause a problem in cases where a VF is used

[ovs-dev] [PATCH] netdev-dpdk: add support for the RTE_ETH_EVENT_INTR_RESET event

2019-09-05 Thread Eelco Chaudron
:92:d3:33 Signed-off-by: Eelco Chaudron --- lib/netdev-dpdk.c | 49 +-- 1 file changed, 47 insertions(+), 2 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index bc20d6843..a23150387 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev-dpd

Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS

2019-09-05 Thread Eelco Chaudron
Message- From: Eelco Chaudron Sent: Tuesday, September 3, 2019 7:27 PM To: Anju Thomas Cc: d...@openvswitch.org; Keshav Gupta Subject: Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS On 20 Aug 2019, at 11:57, Anju Thomas wrote: Thanks for the comments Eelco. I will address

Re: [ovs-dev] [PATCH 2/2] netdev-afxdp: Add need_wakeup supprt.

2019-09-04 Thread Eelco Chaudron
| 3448908 | 2349593 | 1586276 | 1197304 | 814854 | +---+-+-+-+-+-+-+-++ Did not research why PVP is slower, maybe related to the TAP interface with AF_XDP? See one small comment inline… Suggested-by: Eelco Chaudron

Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS

2019-09-03 Thread Eelco Chaudron
-Original Message- From: Eelco Chaudron Sent: Monday, August 12, 2019 5:48 PM To: Anju Thomas Cc: d...@openvswitch.org; Keshav Gupta Subject: Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS Hi Anju, See comments inline… //Eelco On 25 Jul 2019, at 14:16, Anju

Re: [ovs-dev] [PATCH] Docs: Add DPDK AF_XDP PMD use case.

2019-08-22 Thread Eelco Chaudron
On 22 Aug 2019, at 0:48, William Tu wrote: Add command lines for using DPDK's AF_XDP PMD driver. Signed-off-by: William Tu --- Documentation/intro/install/afxdp.rst | 36 +++ 1 file changed, 36 insertions(+) diff --git

Re: [ovs-dev] [PATCHv18] netdev-afxdp: add new netdev type for AF_XDP.

2019-08-21 Thread Eelco Chaudron
On 20 Aug 2019, at 17:20, Ilya Maximets wrote: On 20.08.2019 14:19, Eelco Chaudron wrote: On 20 Aug 2019, at 12:10, Ilya Maximets wrote: On 14.08.2019 19:16, William Tu wrote: On Wed, Aug 14, 2019 at 7:58 AM William Tu wrote: On Wed, Aug 14, 2019 at 5:09 AM Eelco Chaudron wrote

Re: [ovs-dev] [PATCHv18] netdev-afxdp: add new netdev type for AF_XDP.

2019-08-20 Thread Eelco Chaudron
On 20 Aug 2019, at 12:10, Ilya Maximets wrote: On 14.08.2019 19:16, William Tu wrote: On Wed, Aug 14, 2019 at 7:58 AM William Tu wrote: On Wed, Aug 14, 2019 at 5:09 AM Eelco Chaudron wrote: On 8 Aug 2019, at 17:38, Ilya Maximets wrote: I see a rather high number of afxdp_cq_skip

Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS

2019-08-20 Thread Eelco Chaudron
- From: Eelco Chaudron Sent: Monday, August 12, 2019 5:48 PM To: Anju Thomas Cc: d...@openvswitch.org; Keshav Gupta Subject: Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS Hi Anju, See comments inline… //Eelco On 25 Jul 2019, at 14:16, Anju Thomas wrote: Currently OVS

Re: [ovs-dev] [PATCHv18] netdev-afxdp: add new netdev type for AF_XDP.

2019-08-14 Thread Eelco Chaudron
On 8 Aug 2019, at 17:38, Ilya Maximets wrote: I see a rather high number of afxdp_cq_skip, which should to my knowledge never happen? I tried to investigate this previously, but didn't find anything suspicious. So, for my knowledge, this should never happen too. However, I only looked

Re: [ovs-dev] [PATCH] dpif-netdev: Add core id in the PMD thread name.

2019-08-14 Thread Eelco Chaudron
On 14 Aug 2019, at 9:45, Ilya Maximets wrote: On 13.08.2019 19:46, Eelco Chaudron wrote: On 13 Aug 2019, at 18:37, Ilya Maximets wrote: This is highly useful to see on which core PMD is running by only looking at the thread name. Thread Id still allows to distinguish different threads

Re: [ovs-dev] [PATCH] dpif-netdev: Add core id in the PMD thread name.

2019-08-13 Thread Eelco Chaudron
On 13 Aug 2019, at 18:37, Ilya Maximets wrote: This is highly useful to see on which core PMD is running by only looking at the thread name. Thread Id still allows to distinguish different threads running on the same core over the time: |dpif_netdev(pmd-c10/id:53)|DBG|Creating 2.

Re: [ovs-dev] [PATCH v12] Improved Packet Drop Statistics in OVS

2019-08-12 Thread Eelco Chaudron
Hi Anju, See comments inline… //Eelco On 25 Jul 2019, at 14:16, Anju Thomas wrote: Currently OVS maintains explicit packet drop/error counters only on port level. Packets that are dropped as part of normal OpenFlow processing are counted in flow stats of “drop” flows or as table misses in

Re: [ovs-dev] [PATCHv18] netdev-afxdp: add new netdev type for AF_XDP.

2019-08-08 Thread Eelco Chaudron
On 8 Aug 2019, at 14:09, Ilya Maximets wrote: On 08.08.2019 14:42, Eelco Chaudron wrote: On 19 Jul 2019, at 16:54, Ilya Maximets wrote: On 18.07.2019 23:11, William Tu wrote: The patch introduces experimental AF_XDP support for OVS netdev. AF_XDP, the Address Family of the eXpress Data

[ovs-dev] [PATCH 1/1] netdev-afxdp: fix corner case where umem entries were not released

2019-08-08 Thread Eelco Chaudron
If for some reason the last element in the batch was already pushed on the stack, none of the elements where pushed. This was leading to buffer starvation. Signed-off-by: Eelco Chaudron --- lib/netdev-afxdp.c | 12 ++-- 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/lib

Re: [ovs-dev] [PATCHv18] netdev-afxdp: add new netdev type for AF_XDP.

2019-08-08 Thread Eelco Chaudron
On 19 Jul 2019, at 16:54, Ilya Maximets wrote: On 18.07.2019 23:11, William Tu wrote: The patch introduces experimental AF_XDP support for OVS netdev. AF_XDP, the Address Family of the eXpress Data Path, is a new Linux socket type built upon the eBPF and XDP technology. It is aims to have

Re: [ovs-dev] [PATCH v4 2/5] dpif-netdev: Trigger parallel pmd reloads.

2019-07-10 Thread Eelco Chaudron
Ack the v4 also… Acked-by: Eelco Chaudron On 9 Jul 2019, at 18:19, David Marchand wrote: pmd reloads are currently serialised in each steps calling reload_affected_pmds. Any pmd processing packets, waiting on a mutex etc... will make other pmd threads wait for a delay that can

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

2019-07-09 Thread Eelco Chaudron
On 9 Jul 2019, at 12:21, Ilya Maximets wrote: On 09.07.2019 12:56, Eelco Chaudron wrote: Hi William, I tried running my PVP test, and with this patchset, I could not see the deadlock I used to see with all the previous patches. However there is still the issue that the maximum packet

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

2019-07-09 Thread Eelco Chaudron
Hi William, I tried running my PVP test, and with this patchset, I could not see the deadlock I used to see with all the previous patches. However there is still the issue that the maximum packet size supported is 1028 bytes, all larger packets are dropped!! Unfortunately, I’m going on PTO

Re: [ovs-dev] [PATCH] netdev-dpdk: Add a statistic on vhost tx contention.

2019-07-08 Thread Eelco Chaudron
On 7 Jul 2019, at 18:13, David Marchand wrote: Add a statistic to help diagnose contention on the vhost txqs. This is usually seen as dropped packets on the physical ports for rates that are usually handled fine by OVS. Signed-off-by: David Marchand --- This patch applies on top of Kevin

Re: [ovs-dev] [PATCH v3 0/5] Quicker pmd threads reloads

2019-07-08 Thread Eelco Chaudron
Reviewed the patchset, and the changes look fine to me. Ack for the series… Acked-by: Eelco Chaudron On 4 Jul 2019, at 13:59, David Marchand wrote: We have been testing the rebalance code in different situations while having traffic going through OVS. Those tests have shown that part

Re: [ovs-dev] [PATCH v4 3/3] netdev-dpdk: Enable vhost-tx-retries config.

2019-07-04 Thread Eelco Chaudron
ges are fine by me! Signed-off-by: Kevin Traynor Acked-by: Eelco Chaudron Acked-by: Flavio Leitner --- Documentation/topics/dpdk/vhost-user.rst | 28 + lib/netdev-dpdk.c| 39 +--- vswitchd/vswitch.xml |

Re: [ovs-dev] [PATCH v4 0/3] vhost tx retry updates

2019-07-02 Thread Eelco Chaudron
es-max" >in order to have a generic user facing name in case it is ever extended >to other interfaces > -- Other minor edits > -- Kept acks as the operation didn't really change Changes look good to me, so ack for the series Acked-by: Eelco Chaudron > v3: > - 2/4 real

Re: [ovs-dev] [PATCH v3 0/4] vhost tx retry updates

2019-06-27 Thread Eelco Chaudron
| 10 > 5 files changed, 115 insertions(+), 6 deletions(-) Series look fine to me, so ACK below is for the series Acked-by: Eelco Chaudron ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH v2 4/4] netdev-dpdk: Enable vhost-tx-retries config.

2019-06-27 Thread Eelco Chaudron
On 25 Jun 2019, at 16:57, Kevin Traynor wrote: vhost tx retries can provide some mitigation against dropped packets due to a temporarily slow guest/limited queue size for an interface, but on the other hand when a system is fully loaded those extra cycles retrying could mean packets are

Re: [ovs-dev] [PATCH v2 3/4] netdev-dpdk: Add stats for vhost tx retries.

2019-06-27 Thread Eelco Chaudron
ed. > > Signed-off-by: Kevin Traynor Changed look good to me… Acked-by: Eelco Chaudron ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [dpdk-latest PATCH RFCv2 2/2] netdev-dpdk: Add new DPDK RFC 4115 egress policer

2019-06-25 Thread Eelco Chaudron
1000pps. Signed-off-by: Eelco Chaudron --- lib/netdev-dpdk.c | 321 + 1 file changed, 320 insertions(+), 1 deletion(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index 37284c708..f9954a3b3 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev

[ovs-dev] [dpdk-latest PATCH RFCv2 1/2] netdev-dpdk: Add support for multi-queue QoS to the DPDK datapath

2019-06-25 Thread Eelco Chaudron
This patch adds support for multi-queue QoS to the DPDK datapath. Most of the code is based on an earlier patch from a patchset sent out by zhaozhanxu. The patch was titled "[ovs-dev, v2, 1/4] netdev-dpdk.c: Support the multi-queue QoS configuration for dpdk datapath" Signed-off

[ovs-dev] [dpdk-latest PATCH RFCv2 0/2] netdev-dpdk: Add new DPDK RFC 4115 egress policer

2019-06-25 Thread Eelco Chaudron
, will add this when this patch moves from an RFC patch to a real one. Note this patch needs to be applied to dpdk-latest with the appropriates latest DPDK version. v1 -> v2: Rebase so it would apply to the dpdk-latest branch and latest DPDK master Eelco Chaudron (2): netdev-dpdk:

Re: [ovs-dev] [PATCH 3/3] netdev-dpdk: Enable vhost-tx-retries config.

2019-06-25 Thread Eelco Chaudron
On 25 Jun 2019, at 12:09, Kevin Traynor wrote: On 25/06/2019 09:31, Eelco Chaudron wrote: On 21 Jun 2019, at 15:41, Kevin Traynor wrote: vhost tx retries can provide some mitigation against dropped packets due to a temporarily slow guest/limited queue size for an interface

Re: [ovs-dev] [PATCH] vswitchd: Always cleanup userspace datapath.

2019-06-25 Thread Eelco Chaudron
an alternative approach I now understand (i.e. kernel interfaces being part of the netdev bridge). So the change looks good to me. Acked-by: Eelco Chaudron ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH 3/3] netdev-dpdk: Enable vhost-tx-retries config.

2019-06-25 Thread Eelco Chaudron
On 21 Jun 2019, at 15:41, Kevin Traynor wrote: vhost tx retries can provide some mitigation against dropped packets due to a temporarily slow guest/limited queue size for an interface, but on the other hand when a system is fully loaded those extra cycles retrying could mean packets are

Re: [ovs-dev] [PATCH 2/3] netdev-dpdk: Add stats for vhost tx retries.

2019-06-25 Thread Eelco Chaudron
On 21 Jun 2019, at 15:41, Kevin Traynor wrote: vhost tx retries may occur, and it can be a sign that the guest is not optimally configured. Add some stats so a user will know if vhost tx retries are occurring and hence give a hint that guest config should be examined. Signed-off-by: Kevin

Re: [ovs-dev] [PATCH 1/3] netdev-dpdk: Fix and document vhost tx retries.

2019-06-25 Thread Eelco Chaudron
On 21 Jun 2019, at 15:41, Kevin Traynor wrote: > Fix minor issue of one additional retry and add > documentation about vhost tx retries and ways to > reduce/remove them. > > Signed-off-by: Kevin Traynor > Change looks good to me. Acked-

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-21 Thread Eelco Chaudron
On 20 Jun 2019, at 23:58, William Tu wrote: On Thu, Jun 20, 2019 at 1:26 AM Eelco Chaudron wrote: On 20 Jun 2019, at 4:15, William Tu wrote: The PVP test seems to work fine however after a while it stops forwarding: $ ovs-ofctl dump-flows ovs_pvp_br0 cookie=0x0, duration=8.510s

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-20 Thread Eelco Chaudron
On 20 Jun 2019, at 4:15, William Tu wrote: >> The PVP test seems to work fine however after a while it stops >> forwarding: >> >> $ ovs-ofctl dump-flows ovs_pvp_br0 >> cookie=0x0, duration=8.510s, table=0, n_packets=1, n_bytes=1020, >> in_port=eno1 actions=output:tapVM >> cookie=0x0,

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-19 Thread Eelco Chaudron
On 19 Jun 2019, at 0:31, William Tu wrote: The PVP test seems to work fine however after a while it stops forwarding: $ ovs-ofctl dump-flows ovs_pvp_br0 cookie=0x0, duration=8.510s, table=0, n_packets=1, n_bytes=1020, in_port=eno1 actions=output:tapVM cookie=0x0, duration=8.504s,

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-19 Thread Eelco Chaudron
On 19 Jun 2019, at 0:28, William Tu wrote: I guess, this crash caused by trying to destroy unallocated queue. Following change could help: --- diff --git a/lib/netdev-afxdp.c b/lib/netdev-afxdp.c index a6543e8f5..6e1431dce 100644 --- a/lib/netdev-afxdp.c +++ b/lib/netdev-afxdp.c @@ -249,7

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-18 Thread Eelco Chaudron
On 18 Jun 2019, at 12:17, Ilya Maximets wrote: On 18.06.2019 12:45, Eelco Chaudron wrote: On 17 Jun 2019, at 22:32, William Tu wrote: 1000,463875,419996,402010 However I can not restart OVS (see other email on how I restart), even if I clear the XDP programs before a restart

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-18 Thread Eelco Chaudron
On 17 Jun 2019, at 22:32, William Tu wrote: On Mon, Jun 17, 2019 at 11:23 AM William Tu wrote: Hi Eelco, On Mon, Jun 17, 2019 at 3:12 AM Eelco Chaudron wrote: Hi William, See below parts of an offline email discussion I had with Magnus before, and some research I did in the end

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-18 Thread Eelco Chaudron
On 17 Jun 2019, at 20:23, William Tu wrote: Hi Eelco, On Mon, Jun 17, 2019 at 3:12 AM Eelco Chaudron wrote: Hi William, See below parts of an offline email discussion I had with Magnus before, and some research I did in the end, which explains that by design you might not get all

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-17 Thread Eelco Chaudron
:5449 #4 0x0095e94d in ovsthread_wrapper (aux_=) at lib/ovs-thread.c:352 #5 0x7fe1633872de in start_thread () from /lib64/libpthread.so.0 #6 0x7fe162b2ca63 in clone () from /lib64/libc.so.6 Cheers, Eelco On 2019-05-22 15:20, Eelco Chaudron wrote: Hi Magnus, at all, I

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-13 Thread Eelco Chaudron
On 13 Jun 2019, at 2:37, William Tu wrote: Hi Eelco, #0 0x7fbc6a78193f in raise () from /lib64/libc.so.6 #1 0x7fbc6a76bc95 in abort () from /lib64/libc.so.6 #2 0x004ed1a1 in __umem_elem_push_n (addrs=0x7fbc40f2ec50, n=32, umemp=0x24cc790) at lib/xdpsock.c:32 #3

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-12 Thread Eelco Chaudron
Tue, Jun 11, 2019 at 8:02 AM William Tu wrote: Hi Eelco, Thanks for the trace. On Tue, Jun 11, 2019 at 6:52 AM Eelco Chaudron wrote: Hi William, Here are some more details, this is a port to port test (same port in as out) using the following rule: ovs-ofctl add-flow ovs_pvp_br0 "

Re: [ovs-dev] [PATCH v10] Improved Packet Drop Statistics in OVS

2019-06-11 Thread Eelco Chaudron
On 7 Jun 2019, at 7:27, Anju Thomas wrote: Did a quick review, and I noticed some styling issues (mainly indentation), the rest seems ok. One small issue with applying the patchset to master, but nothing a manual inspection could quickly fix. Did some basic testing by forcing some of the

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-11 Thread Eelco Chaudron
On 8 Jun 2019, at 6:48, William Tu wrote: + ethtool -L enp2s0 combined 1 + ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=0x10 + ovs-vsctl add-port br0 enp2s0 -- set interface enp2s0 type="afxdp" \ +options:n_rxq=1 options:xdpmode=drv \ +

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-08 Thread Eelco Chaudron
. On Fri, Jun 7, 2019 at 8:43 AM Eelco Chaudron wrote: Hi William, No review or full test yet, just some observations… We run OVS as a non root user, which is causing OVS with XDP to fail: Right, XDP requires using root privilege. I will add this in the documentation. Is this a hard

Re: [ovs-dev] [PATCHv11] netdev-afxdp: add new netdev type for AF_XDP.

2019-06-07 Thread Eelco Chaudron
Hi William, No review or full test yet, just some observations… We run OVS as a non root user, which is causing OVS with XDP to fail: 2019-06-07T09:14:20.628Z|00023|ofproto_dpif|INFO|netdev@ovs-netdev: Datapath supports ct_orig_tuple

Re: [ovs-dev] [PATCH v9] Improved Packet Drop Statistics in OVS

2019-06-04 Thread Eelco Chaudron
. Regards Anju -Original Message- From: Eelco Chaudron Sent: Monday, June 3, 2019 6:23 PM To: Anju Thomas Cc: d...@openvswitch.org Subject: Re: [ovs-dev] [PATCH v9] Improved Packet Drop Statistics in OVS Hi Anju, Wondering if you got the email below… Cheers, Eelco On 24 May 2019, at 11

Re: [ovs-dev] [PATCH v9] Improved Packet Drop Statistics in OVS

2019-06-03 Thread Eelco Chaudron
Hi Anju, Wondering if you got the email below… Cheers, Eelco On 24 May 2019, at 11:46, Eelco Chaudron wrote: Hi Anju, Was there ever a follow up on this patch? I only see one response from Ilya on this asking about his v8 comments. Thanks, Eelco On 27 Feb 2019, at 10:22, Anju Thomas

Re: [ovs-dev] [PATCH v9] Improved Packet Drop Statistics in OVS

2019-05-24 Thread Eelco Chaudron
Hi Anju, Was there ever a follow up on this patch? I only see one response from Ilya on this asking about his v8 comments. Thanks, Eelco On 27 Feb 2019, at 10:22, Anju Thomas wrote: Currently OVS maintains explicit packet drop/error counters only on port level. Packets that are dropped

Re: [ovs-dev] [PATCHv8] netdev-afxdp: add new netdev type for AF_XDP.

2019-05-20 Thread Eelco Chaudron
On 18 May 2019, at 3:16, William Tu wrote: Hi Eelco, Thanks for all the feedbacks. There are some issues in driver, some in libbpf, and some in my implementation. I will work on it ASAP. My pleasure, see answers to your questions below… On Fri, May 17, 2019 at 3:23 AM Eelco Chaudron

Re: [ovs-dev] [PATCHv8] netdev-afxdp: add new netdev type for AF_XDP.

2019-05-20 Thread Eelco Chaudron
On 17 May 2019, at 14:39, Ilya Maximets wrote: Hi. Just a few comments to the issues you're listed. Best regards, Ilya Maximets. On 17.05.2019 13:23, Eelco Chaudron wrote: Hi William, First a list of issues I found during some basic testing... - When I restart or stop OVS (using

Re: [ovs-dev] [RFC v2 0/5] Quicker pmd threads reloads

2019-05-20 Thread Eelco Chaudron
David this patch set looks fine by me, guess a none-RFC patch would be next? Acked-by: Eelco Chaudron On 14 May 2019, at 18:33, David Marchand wrote: We have been testing the rebalance code in different situations while having traffic going through OVS. Those tests have shown that part

Re: [ovs-dev] [PATCHv8] netdev-afxdp: add new netdev type for AF_XDP.

2019-05-17 Thread Eelco Chaudron
Hi William, First a list of issues I found during some basic testing... - When I restart or stop OVS (using the systemctl interface as found in RHEL) it does not clean up the BFP program causing the restart to fail: 2019-05-10T09:12:11.384Z|00042|netdev_afxdp|ERR|AF_XDP device eno1

Re: [ovs-dev] [PATCH] stopwatch: Free stopwatch packets after processing

2019-05-08 Thread Eelco Chaudron
On 8 May 2019, at 14:56, Dumitru Ceara wrote: > The free(pkt) call was missing inside the stopwatch_thread processing > loop. > > Signed-off-by: Dumitru Ceara > --- > lib/stopwatch.c | 1 + > 1 file changed, 1 insertion(+) > Change looks good to me…

Re: [ovs-dev] [RFC 4/5] dpif-netdev: Only reload static tx qid when needed.

2019-05-08 Thread Eelco Chaudron
Patch looks good to me, however need to fix the relaxed atomic operations. On 30 Apr 2019, at 14:18, David Marchand wrote: pmd->static_tx_qid is allocated under a mutex by the different pmd threads. Unconditionally reallocating it will make those pmd threads sleep when contention occurs.

Re: [ovs-dev] [RFC 5/5] dpif-netdev: Catch reloads faster.

2019-05-08 Thread Eelco Chaudron
On 30 Apr 2019, at 14:18, David Marchand wrote: Looking at the reload flag only every 1024 loops can be a long time under load, since we might be handling 32 packets per iteration, which means 32k packets. Look at the flag every loop, no major performance impact seen. Signed-off-by: David

Re: [ovs-dev] [RFC 2/5] dpif-netdev: Trigger parallel pmd reloads.

2019-05-08 Thread Eelco Chaudron
On 7 May 2019, at 16:00, Ilya Maximets wrote: On 07.05.2019 16:46, Eelco Chaudron wrote: On 30 Apr 2019, at 14:17, David Marchand wrote: pmd reloads are currently serialised in each steps calling reload_affected_pmds. Any pmd processing packets, waiting on a mutex etc... will make other

Re: [ovs-dev] [RFC 3/5] dpif-netdev: Do not sleep when swapping queues.

2019-05-07 Thread Eelco Chaudron
This patch look fine to me assuming Ilya’s comments in patch 1 are taking care of. Acked-by: Eelco Chaudron On 30 Apr 2019, at 14:18, David Marchand wrote: When swapping queues from a pmd thread to another (q0 polled by pmd0/q1 polled by pmd1 -> q1 polled by pmd0/q0 polled by p

Re: [ovs-dev] [RFC 2/5] dpif-netdev: Trigger parallel pmd reloads.

2019-05-07 Thread Eelco Chaudron
On 30 Apr 2019, at 14:17, David Marchand wrote: pmd reloads are currently serialised in each steps calling reload_affected_pmds. Any pmd processing packets, waiting on a mutex etc... will make other pmd threads wait for a delay that can be undeterministic when syscalls adds up. Switch to a

[ovs-dev] [PATCH net-next] net: openvswitch: return an error instead of doing BUG_ON()

2019-05-02 Thread Eelco Chaudron
For all other error cases in queue_userspace_packet() the error is returned, so it makes sense to do the same for these two error cases. Reported-by: Davide Caratti Signed-off-by: Eelco Chaudron --- net/openvswitch/datapath.c |7 +-- 1 file changed, 5 insertions(+), 2 deletions

Re: [ovs-dev] [ovs-dev, PATCHv6] netdev-afxdp: add new netdev type for AF_XDP.

2019-04-30 Thread Eelco Chaudron
On 30 Apr 2019, at 18:10, Ben Pfaff wrote: > On 25.04.2019 2:47, William Tu wrote: >> +#define unlikely OVS_UNLIKELY >> +#define likely OVS_LIKELY >> +#define barrier() __asm__ __volatile__("": : :"memory") >> +#define smp_rmb() barrier() >> +#define smp_wmb() barrier() > > Does any of this

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-19 Thread Eelco Chaudron
On 19 Apr 2019, at 0:11, William Tu wrote: Hi Eelco, Thanks for your feedbacks! Not necessary all ports. On a OVS switch, you can have some ports supporting AF_XDP, and some ports are other types, ex: DPDK vhost, or tap. But I’m wondering how would you deal with ports not supporting

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-18 Thread Eelco Chaudron
On 2 Apr 2019, at 0:46, William Tu wrote: The patch series introduces AF_XDP support for OVS netdev. AF_XDP is a new address family working together with eBPF. In short, a socket with AF_XDP family can receive and send packets from an eBPF/XDP program attached to the netdev. For more details

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-18 Thread Eelco Chaudron
On 17 Apr 2019, at 19:09, William Tu wrote: On Wed, Apr 17, 2019 at 1:09 AM Eelco Chaudron wrote: On 16 Apr 2019, at 21:55, Ben Pfaff wrote: On Mon, Apr 01, 2019 at 03:46:48PM -0700, William Tu wrote: The patch series introduces AF_XDP support for OVS netdev. AF_XDP is a new address

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-18 Thread Eelco Chaudron
On 17 Apr 2019, at 18:47, Ben Pfaff wrote: On Wed, Apr 17, 2019 at 10:09:53AM +0200, Eelco Chaudron wrote: On 16 Apr 2019, at 21:55, Ben Pfaff wrote: AF_XDP is a faster way to access the existing kernel devices. If we take that point of view, then it would be ideal if AF_XDP were

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-18 Thread Eelco Chaudron
On 17 Apr 2019, at 19:16, William Tu wrote: Hi Eelco, Thanks for trying this patchset! In addition you need to do “make install_headers” from kernel libbpf and copy the libbpf_util.h manually. I was able to do a simple physical port in same physical port out test without crashing, but

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-18 Thread Eelco Chaudron
On 17 Apr 2019, at 19:39, William Tu wrote: On Wed, Apr 17, 2019 at 9:48 AM Ben Pfaff wrote: On Wed, Apr 17, 2019 at 12:16:59PM +0200, Eelco Chaudron wrote: One other thing that popped up in my head is how (will) it work together with DPDK enabled on the same system? Why not? I’m

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-17 Thread Eelco Chaudron
On 17 Apr 2019, at 14:01, Eelco Chaudron wrote: Hi William, I think you applied the following patch to get it to compile? Or did you copy in the kernel headers? https://www.spinics.net/lists/netdev/msg563507.html I noticed you duplicated the macros, which resulted in all kind

Re: [ovs-dev] [PATCH RFCv4 0/4] AF_XDP netdev support for OVS

2019-04-17 Thread Eelco Chaudron
Hi William, I think you applied the following patch to get it to compile? Or did you copy in the kernel headers? https://www.spinics.net/lists/netdev/msg563507.html //Eelco On 2 Apr 2019, at 0:46, William Tu wrote: The patch series introduces AF_XDP support for OVS netdev. AF_XDP is a new

  1   2   3   4   >