* [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes [not found] <cover.1396111568.git.decot@googlers.com> @ 2014-03-29 16:48 ` David Decotigny 2014-03-30 1:36 ` Stephen Hemminger 2014-03-31 20:25 ` David Miller 0 siblings, 2 replies; 6+ messages in thread From: David Decotigny @ 2014-03-29 16:48 UTC (permalink / raw) To: David S. Miller, Jamal Hadi Salim, netdev, linux-kernel Cc: Greg Kroah-Hartman, Eric Dumazet, Eric W. Biederman, Weilong Chen, Jiri Pirko, Amir Vadai, Michael Dalton, David Decotigny, Al Viro, Tejun Heo This allows to monitor carrier on/off transitions and detect link flapping issues: - new /sys/class/net/X/carrier_changes - new rtnetlink IFLA_CARRIER_CHANGES (getlink) Tested: - grep . /sys/class/net/*/carrier_changes + ip link set dev X down/up + plug/unplug cable - updated iproute2: prints IFLA_CARRIER_CHANGES - iproute2 20121211-2 (debian): unchanged behavior Signed-off-by: David Decotigny <decot@googlers.com> --- include/linux/netdevice.h | 3 +++ include/uapi/linux/if_link.h | 1 + net/core/net-sysfs.c | 11 +++++++++++ net/core/rtnetlink.c | 6 +++++- net/sched/sch_generic.c | 2 ++ 5 files changed, 22 insertions(+), 1 deletion(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 159c7e7..feb60c0 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1316,6 +1316,9 @@ struct net_device { atomic_long_t rx_dropped; atomic_long_t tx_dropped; + /* Stats to monitor carrier on<->off transitions */ + atomic_t carrier_changes; + #ifdef CONFIG_WIRELESS_EXT /* List of functions to handle Wireless Extensions (instead of ioctl). * See <net/iw_handler.h> for details. Jean II */ diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 16410b6..9a7f7ac 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -144,6 +144,7 @@ enum { IFLA_NUM_RX_QUEUES, IFLA_CARRIER, IFLA_PHYS_PORT_ID, + IFLA_CARRIER_CHANGES, __IFLA_MAX }; diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index daed9a6..4623962 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -253,6 +253,16 @@ static ssize_t operstate_show(struct device *dev, } static DEVICE_ATTR_RO(operstate); +static ssize_t carrier_changes_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct net_device *netdev = to_net_dev(dev); + return sprintf(buf, fmt_dec, + atomic_read(&netdev->carrier_changes)); +} +static DEVICE_ATTR_RO(carrier_changes); + /* read-write attributes */ static int change_mtu(struct net_device *net, unsigned long new_mtu) @@ -386,6 +396,7 @@ static struct attribute *net_class_attrs[] = { &dev_attr_duplex.attr, &dev_attr_dormant.attr, &dev_attr_operstate.attr, + &dev_attr_carrier_changes.attr, &dev_attr_ifalias.attr, &dev_attr_carrier.attr, &dev_attr_mtu.attr, diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index e7c6006..d4ff417 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -822,6 +822,7 @@ static noinline size_t if_nlmsg_size(const struct net_device *dev, + nla_total_size(4) /* IFLA_NUM_RX_QUEUES */ + nla_total_size(1) /* IFLA_OPERSTATE */ + nla_total_size(1) /* IFLA_LINKMODE */ + + nla_total_size(4) /* IFLA_CARRIER_CHANGES */ + nla_total_size(ext_filter_mask & RTEXT_FILTER_VF ? 4 : 0) /* IFLA_NUM_VF */ + rtnl_vfinfo_size(dev, ext_filter_mask) /* IFLA_VFINFO_LIST */ @@ -970,7 +971,9 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev, (dev->qdisc && nla_put_string(skb, IFLA_QDISC, dev->qdisc->ops->id)) || (dev->ifalias && - nla_put_string(skb, IFLA_IFALIAS, dev->ifalias))) + nla_put_string(skb, IFLA_IFALIAS, dev->ifalias)) || + nla_put_u32(skb, IFLA_CARRIER_CHANGES, + atomic_read(&dev->carrier_changes))) goto nla_put_failure; if (1) { @@ -1147,6 +1150,7 @@ static const struct nla_policy ifla_policy[IFLA_MAX+1] = { [IFLA_NUM_TX_QUEUES] = { .type = NLA_U32 }, [IFLA_NUM_RX_QUEUES] = { .type = NLA_U32 }, [IFLA_PHYS_PORT_ID] = { .type = NLA_BINARY, .len = MAX_PHYS_PORT_ID_LEN }, + [IFLA_CARRIER_CHANGES] = { .type = NLA_U32 }, /* ignored */ }; static const struct nla_policy ifla_info_policy[IFLA_INFO_MAX+1] = { diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index e82e43b..e1543b0 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -310,6 +310,7 @@ void netif_carrier_on(struct net_device *dev) if (test_and_clear_bit(__LINK_STATE_NOCARRIER, &dev->state)) { if (dev->reg_state == NETREG_UNINITIALIZED) return; + atomic_inc(&dev->carrier_changes); linkwatch_fire_event(dev); if (netif_running(dev)) __netdev_watchdog_up(dev); @@ -328,6 +329,7 @@ void netif_carrier_off(struct net_device *dev) if (!test_and_set_bit(__LINK_STATE_NOCARRIER, &dev->state)) { if (dev->reg_state == NETREG_UNINITIALIZED) return; + atomic_inc(&dev->carrier_changes); linkwatch_fire_event(dev); } } -- 1.9.1.423.g4596e3a ^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes 2014-03-29 16:48 ` [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes David Decotigny @ 2014-03-30 1:36 ` Stephen Hemminger 2014-03-30 1:50 ` Eric Dumazet 2014-03-31 20:25 ` David Miller 1 sibling, 1 reply; 6+ messages in thread From: Stephen Hemminger @ 2014-03-30 1:36 UTC (permalink / raw) To: David Decotigny Cc: David S. Miller, Jamal Hadi Salim, netdev, linux-kernel, Greg Kroah-Hartman, Eric Dumazet, Eric W. Biederman, Weilong Chen, Jiri Pirko, Amir Vadai, Michael Dalton, Al Viro, Tejun Heo On Sat, 29 Mar 2014 09:48:35 -0700 David Decotigny <decot@googlers.com> wrote: > This allows to monitor carrier on/off transitions and detect link > flapping issues: > - new /sys/class/net/X/carrier_changes > - new rtnetlink IFLA_CARRIER_CHANGES (getlink) > > Tested: > - grep . /sys/class/net/*/carrier_changes > + ip link set dev X down/up > + plug/unplug cable > - updated iproute2: prints IFLA_CARRIER_CHANGES > - iproute2 20121211-2 (debian): unchanged behavior > > Signed-off-by: David Decotigny <decot@googlers.com> Is there a related SNMP MIB variable? should there be? ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes 2014-03-30 1:36 ` Stephen Hemminger @ 2014-03-30 1:50 ` Eric Dumazet 2014-03-30 15:18 ` David Decotigny 0 siblings, 1 reply; 6+ messages in thread From: Eric Dumazet @ 2014-03-30 1:50 UTC (permalink / raw) To: Stephen Hemminger Cc: David Decotigny, David S. Miller, Jamal Hadi Salim, netdev, linux-kernel, Greg Kroah-Hartman, Eric Dumazet, Eric W. Biederman, Weilong Chen, Jiri Pirko, Amir Vadai, Michael Dalton, Al Viro, Tejun Heo On Sat, 2014-03-29 at 18:36 -0700, Stephen Hemminger wrote: > On Sat, 29 Mar 2014 09:48:35 -0700 > David Decotigny <decot@googlers.com> wrote: > > > This allows to monitor carrier on/off transitions and detect link > > flapping issues: > > - new /sys/class/net/X/carrier_changes > > - new rtnetlink IFLA_CARRIER_CHANGES (getlink) > > > > Tested: > > - grep . /sys/class/net/*/carrier_changes > > + ip link set dev X down/up > > + plug/unplug cable > > - updated iproute2: prints IFLA_CARRIER_CHANGES > > - iproute2 20121211-2 (debian): unchanged behavior > > > > Signed-off-by: David Decotigny <decot@googlers.com> > > Is there a related SNMP MIB variable? should there be? I don't think we have any SNMP MIB variable for devices ? It seems "ip link" is the facto standard. (BTW we put BusyPollRxPackets in TcpExt group, this was an error IMHO) ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes 2014-03-30 1:50 ` Eric Dumazet @ 2014-03-30 15:18 ` David Decotigny 2014-03-31 16:49 ` Florian Fainelli 0 siblings, 1 reply; 6+ messages in thread From: David Decotigny @ 2014-03-30 15:18 UTC (permalink / raw) To: Stephen Hemminger Cc: David S. Miller, Jamal Hadi Salim, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Greg Kroah-Hartman, Eric Dumazet, Eric W. Biederman, Weilong Chen, Jiri Pirko, Amir Vadai, Michael Dalton, Al Viro, Tejun Heo, Florian Fainelli, Eric Dumazet Hi all, As Eric mentioned, this is a per-link counter. I am not sure an aggregate stat would be as useful. As suggested by Florian, I'm planning to add a doc. As far as I can tell, I think it goes beyond just documenting this new counter: I believe sysfs-class-net needs to be added as a whole. Stephen: I have a patch for iproute2, I was planning to send it after we're done with this one, unless you want it now. Also, should I document this IFLA thing, where would be best? On Sat, Mar 29, 2014 at 6:50 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote: > On Sat, 2014-03-29 at 18:36 -0700, Stephen Hemminger wrote: >> On Sat, 29 Mar 2014 09:48:35 -0700 >> David Decotigny <decot@googlers.com> wrote: >> >> > This allows to monitor carrier on/off transitions and detect link >> > flapping issues: >> > - new /sys/class/net/X/carrier_changes >> > - new rtnetlink IFLA_CARRIER_CHANGES (getlink) >> > >> > Tested: >> > - grep . /sys/class/net/*/carrier_changes >> > + ip link set dev X down/up >> > + plug/unplug cable >> > - updated iproute2: prints IFLA_CARRIER_CHANGES >> > - iproute2 20121211-2 (debian): unchanged behavior >> > >> > Signed-off-by: David Decotigny <decot@googlers.com> >> >> Is there a related SNMP MIB variable? should there be? > > I don't think we have any SNMP MIB variable for devices ? > > It seems "ip link" is the facto standard. > > (BTW we put BusyPollRxPackets in TcpExt group, this was an error IMHO) > > ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes 2014-03-30 15:18 ` David Decotigny @ 2014-03-31 16:49 ` Florian Fainelli 0 siblings, 0 replies; 6+ messages in thread From: Florian Fainelli @ 2014-03-31 16:49 UTC (permalink / raw) To: David Decotigny Cc: Stephen Hemminger, David S. Miller, Jamal Hadi Salim, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Greg Kroah-Hartman, Eric Dumazet, Eric W. Biederman, Weilong Chen, Jiri Pirko, Amir Vadai, Michael Dalton, Al Viro, Tejun Heo, Eric Dumazet Hi David, 2014-03-30 8:18 GMT-07:00 David Decotigny <decot@googlers.com>: > Hi all, > > As Eric mentioned, this is a per-link counter. I am not sure an > aggregate stat would be as useful. > > As suggested by Florian, I'm planning to add a doc. As far as I can > tell, I think it goes beyond just documenting this new counter: I > believe sysfs-class-net needs to be added as a whole. Right, I sort of started to do that here: http://patchwork.ozlabs.org/patch/334888/ Once this gets accepted, I plan on documenting the other sysfs attributes in statistics/ and queues/ in particular > > Stephen: I have a patch for iproute2, I was planning to send it after > we're done with this one, unless you want it now. Also, should I > document this IFLA thing, where would be best? > > On Sat, Mar 29, 2014 at 6:50 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote: >> On Sat, 2014-03-29 at 18:36 -0700, Stephen Hemminger wrote: >>> On Sat, 29 Mar 2014 09:48:35 -0700 >>> David Decotigny <decot@googlers.com> wrote: >>> >>> > This allows to monitor carrier on/off transitions and detect link >>> > flapping issues: >>> > - new /sys/class/net/X/carrier_changes >>> > - new rtnetlink IFLA_CARRIER_CHANGES (getlink) >>> > >>> > Tested: >>> > - grep . /sys/class/net/*/carrier_changes >>> > + ip link set dev X down/up >>> > + plug/unplug cable >>> > - updated iproute2: prints IFLA_CARRIER_CHANGES >>> > - iproute2 20121211-2 (debian): unchanged behavior >>> > >>> > Signed-off-by: David Decotigny <decot@googlers.com> >>> >>> Is there a related SNMP MIB variable? should there be? >> >> I don't think we have any SNMP MIB variable for devices ? >> >> It seems "ip link" is the facto standard. >> >> (BTW we put BusyPollRxPackets in TcpExt group, this was an error IMHO) >> >> -- Florian ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes 2014-03-29 16:48 ` [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes David Decotigny 2014-03-30 1:36 ` Stephen Hemminger @ 2014-03-31 20:25 ` David Miller 1 sibling, 0 replies; 6+ messages in thread From: David Miller @ 2014-03-31 20:25 UTC (permalink / raw) To: decot Cc: jhs, netdev, linux-kernel, gregkh, edumazet, ebiederm, chenweilong, jiri, amirv, mwdalton, viro, tj From: David Decotigny <decot@googlers.com> Date: Sat, 29 Mar 2014 09:48:35 -0700 > This allows to monitor carrier on/off transitions and detect link > flapping issues: > - new /sys/class/net/X/carrier_changes > - new rtnetlink IFLA_CARRIER_CHANGES (getlink) > > Tested: > - grep . /sys/class/net/*/carrier_changes > + ip link set dev X down/up > + plug/unplug cable > - updated iproute2: prints IFLA_CARRIER_CHANGES > - iproute2 20121211-2 (debian): unchanged behavior > > Signed-off-by: David Decotigny <decot@googlers.com> Looks good, applied, thanks David. ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2014-03-31 20:25 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <cover.1396111568.git.decot@googlers.com>
2014-03-29 16:48 ` [PATCH net-next v2] net-sysfs: expose number of carrier on/off changes David Decotigny
2014-03-30 1:36 ` Stephen Hemminger
2014-03-30 1:50 ` Eric Dumazet
2014-03-30 15:18 ` David Decotigny
2014-03-31 16:49 ` Florian Fainelli
2014-03-31 20:25 ` David Miller
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).