[ovs-dev] 答复: Re: [PATCH] netdev-dpdk: Assign socket id according to device's numa id
Thank you for your suggestion, i will reformat the message and resubmit the patch ASAP. Aaron Conole写于 2017/01/11 22:27:06: > 发件人: Aaron Conole > 收件人: Binbin Xu , > 抄送: d...@openvswitch.org, Ciara Loftus , > Kevin Traynor > 日期: 2017/01/11 22:27 > 主题: Re: [ovs-dev] [PATCH] netdev-dpdk: Assign socket id according > to device's numa id > > Binbin Xu writes: > > > After the commit "55e075e65ef9ecbd70e5e0fada2704c3d73724d8 > > netdev-dpdk: Arbitrary 'dpdk' port naming", we could hotplug > > attach DPDK ports specified via the 'dpdk-devargs' option. > > > > But the socket id of DPDK ports can't be assigned correctly, > > it is always 0. The socket id of DPDK ports should be assigned > > according to the numa id of the device. > > > > Signed-off-by: Binbin Xu > > --- > > I haven't reviewed the patch yet, but can you reformat this message with > a fixes: tag? That makes backporting swathes of commits easier. > > It would be: > Fixes: 55e075e65ef9 ("netdev-dpdk: Arbitrary 'dpdk' port naming") > > I've also CC'd some folks involved in testing / using this patch > > Thanks, > Aaron > Thanks, Binbin Xu ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev
[ovs-dev] 答复: Re: [PATCH] netdev: Initialize netdev's features before getting them
Hi Aaron, Thanks for your suggestion, I'll resubmit a patch for DPDK, and take a look at lib/netdev-bsd.c to confirm that the additional patch is needed or not. Thanks, Aaron Conole写于 2016/10/12 21:26:14: > 发件人: Aaron Conole > 收件人: Binbin Xu , > 抄送: dev@openvswitch.org > 日期: 2016/10/12 21:26 > 主题: Re: [ovs-dev] [PATCH] netdev: Initialize netdev's features > before getting them > > Hi Binbin, > > Binbin Xu writes: > > > When OVS is used, DPDK doesn't support features 'advertised', > > 'supported' and 'peer'. If a physical port added to bridge, features > > descirbed above can't be assigned, and the values are random. > > > > Signed-off-by: Binbin Xu > > --- > > Thanks for reporting this. I consider this a bug in dpdk, not > something that requires changing the netdev framework. A look at other > netdev classes confirms this. > > Please re-submit something that fixes DPDK, like the following. If I > read it correctly, you might also submit an additional patch to cleanup > the interface in lib/netdev-bsd.c (which seems to be 'mismatched'). > > NOTE patch is *completely* untested, not even compile tested. I have tested the patch in my environment, you means that i didn't make the unit test? > > --- > diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c > index 39bf930..6f5ec43 100644 > --- a/lib/netdev-dpdk.c > +++ b/lib/netdev-dpdk.c > @@ -1978,13 +1978,15 @@ out: > static int > netdev_dpdk_get_features(const struct netdev *netdev, > enum netdev_features *current, > - enum netdev_features *advertised OVS_UNUSED, > - enum netdev_features *supported OVS_UNUSED, > - enum netdev_features *peer OVS_UNUSED) > + enum netdev_features *advertised, > + enum netdev_features *supported, > + enum netdev_features *peer) > { > struct netdev_dpdk *dev = netdev_dpdk_cast(netdev); > struct rte_eth_link link; > > +*advertised = *peer = *supported = 0; > + > ovs_mutex_lock(>mutex); > link = dev->link; > ovs_mutex_unlock(>mutex); ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
Re: [ovs-dev] [PATCH] netdev-dpdk: Optimise the initialization of "port_conf"
"port_conf" is a global const variables, and in function dpdk_eth_dev_queue_setup it is asigned to the local variables "conf". The jumbo_frame bit is set in the local varables, previous configuration shouldn't influence on the other port. Thanks "Kavanagh, Mark B"写于 2016/10/11 21:46:00: > 发件人: "Kavanagh, Mark B" > 收件人: Binbin Xu , "dev@openvswitch.org" > , > 日期: 2016/10/11 21:46 > 主题: RE: [ovs-dev] [PATCH] netdev-dpdk: Optimise the > initialization of "port_conf" > > > > >The member "max_rx_pkt_len" of "port_conf" is only used if > >jumbo_frame enabled, so it can be initialized with value '0'. > > > >Signed-off-by: Binbin Xu > >--- > > lib/netdev-dpdk.c | 4 +--- > > 1 file changed, 1 insertion(+), 3 deletions(-) > > > >diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c > >index 39bf930..c4a0cc0 100644 > >--- a/lib/netdev-dpdk.c > >+++ b/lib/netdev-dpdk.c > >@@ -154,6 +154,7 @@ static char *vhost_sock_dir = NULL; /* > Location of vhost-user sockets > >*/ > > static const struct rte_eth_conf port_conf = { > > .rxmode = { > > .mq_mode = ETH_MQ_RX_RSS, > >+.max_rx_pkt_len = 0, > > .split_hdr_size = 0, > > .header_split = 0, /* Header Split disabled */ > > .hw_ip_checksum = 0, /* IP checksum offload disabled */ > >@@ -648,9 +649,6 @@ dpdk_eth_dev_queue_setup(struct netdev_dpdk > *dev, int n_rxq, int n_txq) > > if (dev->mtu > ETHER_MTU) { > > conf.rxmode.jumbo_frame = 1; > > conf.rxmode.max_rx_pkt_len = dev->max_packet_len; > >-} else { > >-conf.rxmode.jumbo_frame = 0; > >-conf.rxmode.max_rx_pkt_len = 0; > > } > > NACK: if a previous configuration already set the jumbo_frame bit, a > subsequent non-jumbo port could inherit this attribute - that's why > it's reset explicitly here. > > > /* A device may report more queues than it makes available (this has > > * been observed for Intel xl710, which reserves some of them for > >-- > >2.9.3 > > > >___ > >dev mailing list > >dev@openvswitch.org > >http://openvswitch.org/mailman/listinfo/dev ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
Re: [ovs-dev] [CudaMailTagged] [PATCH] bond: Fix the issue of negative number of "next rebalance"
I just think the negative number may puzzle someone. The positive number may be better, it shows us the left time to next rebalance. Ben Pfaff写于 2016/09/20 23:15:22: > 发件人: Ben Pfaff > 收件人: Binbin Xu , > 抄送: dev@openvswitch.org > 日期: 2016/09/20 23:15 > 主题: Re: [ovs-dev] [CudaMailTagged] [PATCH] bond: Fix the issue of > negative number of "next rebalance" > > On Tue, Sep 13, 2016 at 07:23:41AM +0800, Binbin Xu wrote: > > The command "ovs-appctl bond/show" is used for display information > > of bond ports. > > > > When bond mode is "balance_slb" or "balance_tcp", we can find that > > the value followed behind "next rebalance" is negative somtimes. > > > > Signed-off-by: Binbin Xu > > I don't think this is a big deal. It just means that we're a little > behind on rebalancing. ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
Re: [ovs-dev] [PATCH 1/2] netdev-vport: remove unused function
Can you tell me the reaseon? We don't support the kernel space tunnel and user space tunnel at the same time?? Ben Pfaff写于 2016/08/11 00:57:04: > 发件人: Ben Pfaff > 收件人: Binbin Xu , > 抄送: dev@openvswitch.org > 日期: 2016/08/11 00:57 > 主题: Re: [ovs-dev] [PATCH 1/2] netdev-vport: remove unused function > > On Sat, Aug 06, 2016 at 04:55:14AM +0800, Binbin Xu wrote: > > The function netdev_vport_get_dpif_port_strdup is not > > used anymore. So we can remove it now. > > > > Signed-off-by: Binbin Xu > > Thanks, applied. > > I don't plan to review patch 2. ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
[ovs-dev] 答复: Re: 答复: Re: 答复: Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink socket file when constructing vhostuser.
My opinion is that, the 'kill -9' signal can't be captured. The patch described in http://openvswitch.org/pipermail/dev/2016-February/065470.html can't fix the issue perfectly. Do we allow sb use 'kill -9' to kill OVS?? Ilya Maximets写于 2016/08/04 20:01:50: > 发件人: Ilya Maximets > 收件人: xu.binb...@zte.com.cn, > 抄送: Ben Pfaff , dev@openvswitch.org, Heetae Ahn > , Dyasly Sergey > 日期: 2016/08/04 20:02 > 主题: Re: 答复: Re: 答复: Re: [ovs-dev, CudaMailTagged] netdev- > dpdk: unlink socket file when constructing vhostuser. > > On 04.08.2016 14:52, xu.binb...@zte.com.cn wrote: > > Thank you. > > > > I saw the patch in the link: http://openvswitch.org/pipermail/dev/ > 2016-February/065556.html > > > > But in the latest code of OVS, I can't find this path?? > > Did you read the whole discussion? > > > By the way, should I keep the mode of 'lib/netdev-dpdk.c' file? > > Definitely. But I don't think you should send second version if > you don't have very strong arguments for such solution. > > > Ilya Maximets 写于 2016/08/04 19:33:58: > > > >> 发件人: Ilya Maximets > >> 收件人: xu.binb...@zte.com.cn, > >> 抄送: Ben Pfaff , dev@openvswitch.org, Heetae Ahn > >> , Dyasly Sergey > >> 日期: 2016/08/04 19:34 > >> 主题: Re: 答复: Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink > >> socket file when constructing vhostuser. > >> > >> On 04.08.2016 14:05, xu.binb...@zte.com.cn wrote: > >> > I get the source code clone from branch origin/master in the github. > >> > >> Sorry. All is OK, but you changed the mode of 'lib/netdev-dpdk.c' file. > >> > >> > In fact, I killed ovs-vswitchd process and then start it again. > >> > In this case, sockets left on the filesystem. > >> > >> That is normal. Why you using hard untrappable signals? There will be > >> no issues in normal case (even SIGTERM or SIGINT). > >> > >> There are few discussions about this in mail-list. That's an old topic. > >> Some links: > >> http://openvswitch.org/pipermail/dev/2016-February/065556.html > >> http://openvswitch.org/pipermail/dev/2016-February/065470.html > >> > >> > >> > Sorry for my description in the commit msg, it may be a puzzle to you. > >> > > >> > > >> > > >> > > >> > 发件人: Ilya Maximets > >> > 收件人: xu.binb...@zte.com.cn, dev@openvswitch.org, > >> > 抄送:Ben Pfaff , Dyasly Sergey > >> , Heetae Ahn > >> > 日期: 2016/08/04 18:42 > >> > 主题:Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink > >> socket file when constructing vhostuser. > >> > > >> > --- > - > >> -- > >> > > >> > > >> > > >> > Oh, again... > >> > > >> > 1. This patch not for upstream version of OVS. (I guess your OVS patched > >> > at least with vHost client support). > >> > > >> > 2. If you will restart OVS properly (even by SIGTERM or SIGINT) socket > >> > will be removed inside 'rte_vhost_driver_unregister()' or inside > >> > fatal_signal handler. Are you sure that it's the issue of upstream > >> > OVS and not your local patches? > >> > > >> > 3. Segmentation fault or another hard failure is the only reason to > >> > sockets left on the filesystem. > >> > > >> > Best regards, Ilya Maximets. > >> > > >> > On 04.08.2016 23:31, xu.binb...@zte.com.cn wrote: > >> >> Work with DPDK 16.07, a UNIX socket will be created when we > >> >> add vhostuser port. After that, the restarting of ovs-vswitchd > >> >> leads to the failure of socket binding, so the vhostuser port > >> >> can't be created successfully. > >> >> > >> >> This commit unlink socket file before creating UNIX socket to > >> >> avoid failure of socket binding. > >> >> > >> >> Signed-off-by: Binbin Xu > >> >> --- > >> >> lib/netdev-dpdk.c | 5 - > >> >> 1 file
[ovs-dev] 答复: Re: 答复: Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink socket file when constructing vhostuser.
Thank you. I saw the patch in the link: http://openvswitch.org/pipermail/dev/2016-February/065556.html But in the latest code of OVS, I can't find this path?? By the way, should I keep the mode of 'lib/netdev-dpdk.c' file? Ilya Maximets写于 2016/08/04 19:33:58: > 发件人: Ilya Maximets > 收件人: xu.binb...@zte.com.cn, > 抄送: Ben Pfaff , dev@openvswitch.org, Heetae Ahn > , Dyasly Sergey > 日期: 2016/08/04 19:34 > 主题: Re: 答复: Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink > socket file when constructing vhostuser. > > On 04.08.2016 14:05, xu.binb...@zte.com.cn wrote: > > I get the source code clone from branch origin/master in the github. > > Sorry. All is OK, but you changed the mode of 'lib/netdev-dpdk.c' file. > > > In fact, I killed ovs-vswitchd process and then start it again. > > In this case, sockets left on the filesystem. > > That is normal. Why you using hard untrappable signals? There will be > no issues in normal case (even SIGTERM or SIGINT). > > There are few discussions about this in mail-list. That's an old topic. > Some links: > http://openvswitch.org/pipermail/dev/2016-February/065556.html > http://openvswitch.org/pipermail/dev/2016-February/065470.html > > > > Sorry for my description in the commit msg, it may be a puzzle to you. > > > > > > > > > > 发件人: Ilya Maximets > > 收件人: xu.binb...@zte.com.cn, dev@openvswitch.org, > > 抄送:Ben Pfaff , Dyasly Sergey > , Heetae Ahn > > 日期: 2016/08/04 18:42 > > 主题:Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink > socket file when constructing vhostuser. > > > > -- > > > > > > > > Oh, again... > > > > 1. This patch not for upstream version of OVS. (I guess your OVS patched > > at least with vHost client support). > > > > 2. If you will restart OVS properly (even by SIGTERM or SIGINT) socket > > will be removed inside 'rte_vhost_driver_unregister()' or inside > > fatal_signal handler. Are you sure that it's the issue of upstream > > OVS and not your local patches? > > > > 3. Segmentation fault or another hard failure is the only reason to > > sockets left on the filesystem. > > > > Best regards, Ilya Maximets. > > > > On 04.08.2016 23:31, xu.binb...@zte.com.cn wrote: > >> Work with DPDK 16.07, a UNIX socket will be created when we > >> add vhostuser port. After that, the restarting of ovs-vswitchd > >> leads to the failure of socket binding, so the vhostuser port > >> can't be created successfully. > >> > >> This commit unlink socket file before creating UNIX socket to > >> avoid failure of socket binding. > >> > >> Signed-off-by: Binbin Xu > >> --- > >> lib/netdev-dpdk.c | 5 - > >> 1 file changed, 4 insertions(+), 1 deletion(-) > >> mode change 100644 => 100755 lib/netdev-dpdk.c > >> > >> diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c > >> old mode 100644 > >> new mode 100755 > >> index aaac0d1..95cf7c3 > >> --- a/lib/netdev-dpdk.c > >> +++ b/lib/netdev-dpdk.c > >> @@ -885,7 +885,10 @@ netdev_dpdk_vhost_user_construct(struct > netdev *netdev) > >> */ > >> snprintf(dev->vhost_id, sizeof(dev->vhost_id), "%s/%s", > >> vhost_sock_dir, name); > >> - > >> + > >> +if (!(flags & RTE_VHOST_USER_CLIENT)) { > >> +unlink(dev->vhost_id); > >> +} > >> err = rte_vhost_driver_register(dev->vhost_id, flags); > >> if (err) { > >> VLOG_ERR("vhost-user socket device setup failure for socket %s\n", > >> > > ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
[ovs-dev] 答复: Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink socket file when constructing vhostuser.
I get the source code clone from branch origin/master in the github. In fact, I killed ovs-vswitchd process and then start it again. In this case, sockets left on the filesystem. Sorry for my description in the commit msg, it may be a puzzle to you. 发件人: Ilya Maximets收件人: xu.binb...@zte.com.cn, dev@openvswitch.org, 抄送: Ben Pfaff , Dyasly Sergey , Heetae Ahn 日期: 2016/08/04 18:42 主题: Re: [ovs-dev, CudaMailTagged] netdev-dpdk: unlink socket file when constructing vhostuser. Oh, again... 1. This patch not for upstream version of OVS. (I guess your OVS patched at least with vHost client support). 2. If you will restart OVS properly (even by SIGTERM or SIGINT) socket will be removed inside 'rte_vhost_driver_unregister()' or inside fatal_signal handler. Are you sure that it's the issue of upstream OVS and not your local patches? 3. Segmentation fault or another hard failure is the only reason to sockets left on the filesystem. Best regards, Ilya Maximets. On 04.08.2016 23:31, xu.binb...@zte.com.cn wrote: > Work with DPDK 16.07, a UNIX socket will be created when we > add vhostuser port. After that, the restarting of ovs-vswitchd > leads to the failure of socket binding, so the vhostuser port > can't be created successfully. > > This commit unlink socket file before creating UNIX socket to > avoid failure of socket binding. > > Signed-off-by: Binbin Xu > --- > lib/netdev-dpdk.c | 5 - > 1 file changed, 4 insertions(+), 1 deletion(-) > mode change 100644 => 100755 lib/netdev-dpdk.c > > diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c > old mode 100644 > new mode 100755 > index aaac0d1..95cf7c3 > --- a/lib/netdev-dpdk.c > +++ b/lib/netdev-dpdk.c > @@ -885,7 +885,10 @@ netdev_dpdk_vhost_user_construct(struct netdev *netdev) > */ > snprintf(dev->vhost_id, sizeof(dev->vhost_id), "%s/%s", > vhost_sock_dir, name); > - > + > +if (!(flags & RTE_VHOST_USER_CLIENT)) { > +unlink(dev->vhost_id); > +} > err = rte_vhost_driver_register(dev->vhost_id, flags); > if (err) { > VLOG_ERR("vhost-user socket device setup failure for socket %s\n", > ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
[ovs-dev] [PATACH]netdev-dpdk: remove duplicated code in netdev_dpdk_get_status
From caeb84217c38ccd0b2076689fd36b578c00678ad Mon Sep 17 00:00:00 2001 From: xubinbinDate: Thu, 21 Jul 2016 21:52:29 +0800 Subject: [PATCH] netdev-dpdk: remove duplicated code in netdev_dpdk_get_status Put "driver_name" into "args" twice, that's meaninglessness. So need to remove duplicated code. Signed-off-by: Binbin Xu --- lib/netdev-dpdk.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index 85b18fd..b515bee 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev-dpdk.c @@ -2091,8 +2091,6 @@ netdev_dpdk_get_status(const struct netdev *netdev, struct smap *args) rte_eth_dev_info_get(dev->port_id, _info); ovs_mutex_unlock(>mutex); -smap_add_format(args, "driver_name", "%s", dev_info.driver_name); - smap_add_format(args, "port_no", "%d", dev->port_id); smap_add_format(args, "numa_id", "%d", rte_eth_dev_socket_id(dev->port_id)); smap_add_format(args, "driver_name", "%s", dev_info.driver_name); -- 1.8.3.1 ZTE Information Security Notice: The information contained in this mail (and any attachment transmitted herewith) is privileged and confidential and is intended for the exclusive use of the addressee(s). If you are not an intended recipient, any disclosure, reproduction, distribution or other dissemination or use of the information contained is strictly prohibited. If you have received this mail in error, please delete it and notify us immediately. ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
[ovs-dev] [PATACH]netdev-dpdk: remove duplicated code in netdev_dpdk_get_status
From caeb84217c38ccd0b2076689fd36b578c00678ad Mon Sep 17 00:00:00 2001 From: xubinbinDate: Thu, 21 Jul 2016 21:52:29 +0800 Subject: [PATCH] netdev-dpdk: remove duplicated code in netdev_dpdk_get_status Put "driver_name" into "args" twice, that's meaninglessness. So need to remove duplicated code. Signed-off-by: Binbin Xu --- lib/netdev-dpdk.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index 85b18fd..b515bee 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev-dpdk.c @@ -2091,8 +2091,6 @@ netdev_dpdk_get_status(const struct netdev *netdev, struct smap *args) rte_eth_dev_info_get(dev->port_id, _info); ovs_mutex_unlock(>mutex); -smap_add_format(args, "driver_name", "%s", dev_info.driver_name); - smap_add_format(args, "port_no", "%d", dev->port_id); smap_add_format(args, "numa_id", "%d", rte_eth_dev_socket_id(dev->port_id)); smap_add_format(args, "driver_name", "%s", dev_info.driver_name); -- 1.8.3.1 ZTE Information Security Notice: The information contained in this mail (and any attachment transmitted herewith) is privileged and confidential and is intended for the exclusive use of the addressee(s). If you are not an intended recipient, any disclosure, reproduction, distribution or other dissemination or use of the information contained is strictly prohibited. If you have received this mail in error, please delete it and notify us immediately. ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev