* [PATCH] be2net: netpoll support
From: Ivan Vecera @ 2011-12-08 11:31 UTC (permalink / raw)
To: netdev; +Cc: sathya.perla, subbu.seetharaman, ajit.khaparde
Add missing netpoll support.
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
---
drivers/net/ethernet/emulex/benet/be_main.c | 18 +++++++++++++++++-
1 files changed, 17 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 7236280..3854fb0 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -2666,6 +2666,19 @@ err:
return status;
}
+#ifdef CONFIG_NET_POLL_CONTROLLER
+static void be_netpoll(struct net_device *netdev)
+{
+ struct be_adapter *adapter = netdev_priv(netdev);
+ struct be_rx_obj *rxo;
+ int i;
+
+ event_handle(adapter, &adapter->tx_eq, false);
+ for_all_rx_queues(adapter, rxo, i)
+ event_handle(adapter, &rxo->rx_eq, true);
+}
+#endif
+
#define FW_FILE_HDR_SIGN "ServerEngines Corp. "
static bool be_flash_redboot(struct be_adapter *adapter,
const u8 *p, u32 img_start, int image_size,
@@ -3014,7 +3027,10 @@ static struct net_device_ops be_netdev_ops = {
.ndo_set_vf_mac = be_set_vf_mac,
.ndo_set_vf_vlan = be_set_vf_vlan,
.ndo_set_vf_tx_rate = be_set_vf_tx_rate,
- .ndo_get_vf_config = be_get_vf_config
+ .ndo_get_vf_config = be_get_vf_config,
+#ifdef CONFIG_NET_POLL_CONTROLLER
+ .ndo_poll_controller = be_netpoll,
+#endif
};
static void be_netdev_init(struct net_device *netdev)
--
1.7.3.4
^ permalink raw reply related
* Re: [net-next PATCH] net: netprio_cgroup: make net_prio_subsys static
From: Neil Horman @ 2011-12-08 11:45 UTC (permalink / raw)
To: John Fastabend; +Cc: davem, netdev
In-Reply-To: <20111208051717.20236.41881.stgit@jf-dev1-dcblab>
On Wed, Dec 07, 2011 at 09:17:17PM -0800, John Fastabend wrote:
> net_prio_subsys can be made static this removes the sparse
> warning it was throwing.
>
> Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
> ---
>
> net/core/netprio_cgroup.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/net/core/netprio_cgroup.c b/net/core/netprio_cgroup.c
> index 3a9fd48..ea16c8f 100644
> --- a/net/core/netprio_cgroup.c
> +++ b/net/core/netprio_cgroup.c
> @@ -28,7 +28,7 @@ static struct cgroup_subsys_state *cgrp_create(struct cgroup_subsys *ss,
> static void cgrp_destroy(struct cgroup_subsys *ss, struct cgroup *cgrp);
> static int cgrp_populate(struct cgroup_subsys *ss, struct cgroup *cgrp);
>
> -struct cgroup_subsys net_prio_subsys = {
> +static struct cgroup_subsys net_prio_subsys = {
> .name = "net_prio",
> .create = cgrp_create,
> .destroy = cgrp_destroy,
>
>
Thanks John.
Acked-by: Neil Horman <nhorman@tuxdriver.com>
^ permalink raw reply
* Re: IPv6 defrag question ?
From: Hans Schillstrom @ 2011-12-08 13:29 UTC (permalink / raw)
To: Patrick McHardy
Cc: Hans Schillstrom, pablo@netfilter.org, jengelh@medozas.de,
netfilter-devel@vger.kernel.org, netdev@vger.kernel.org
In-Reply-To: <4EE09B39.5070401@trash.net>
On Thursday 08 December 2011 12:10:49 Patrick McHardy wrote:
> On 12/08/2011 10:12 AM, Hans Schillstrom wrote:
> > Hi
> > While testing HMARK and IPv6 with nf_defrag_ipv6 (and nf_conntrack_ipv6 loaded) I can't see the defrag ?
> >
> > From what I can see nf_conntrack_reasm goes into PREROUTING with prio -400
> > and HMARK in PREROUTING with prio -150
> >
> > I was expecting that the reasaembled packet whould reach HMARK not the fragments.
> >
> > (Debug print from hmark)
> > HMARK() mark:489, hash:4d04eaa1, frag:1, nhoffs:30 plen:1408 (2008::10 - 1000::1)
> > HMARK() mark:489, hash:4d04eaa1, frag:1, nhoffs:0 plen:86 (2008::10 - 1000::1)
> >
> > IPv4 do reassm. the packets not IPv6...
>
> Yeah, IPv6 currently only passes the defragmented packet through conntrack,
> then associates the conntrack information with the individual fragments and
> passes those on. I'll post patches for IPv6 NAT which will change this
> to behave similar to IPv4 soon.
>
OK great, current beaiviour was kind of unexpected.
BTW this piece of code looks like it's broken
or I might have missunderstod this :-)
at least /* queued */ causes some confusion .
static unsigned int ipv6_defrag(unsigned int hooknum,
...
reasm = nf_ct_frag6_gather(skb, nf_ct6_defrag_user(hooknum, skb));
/* queued */
if (reasm == NULL)
return NF_STOLEN;
NF_STOLEN will only be returned when nf_ct_frag6_reasm() returns an error.
(called by ct_frag6_gather)
--
Regards
Hans Schillstrom <hans.schillstrom@ericsson.com>
^ permalink raw reply
* IPv4/IPv6 nf_defrag on/off ?
From: Hans Schillstrom @ 2011-12-08 13:44 UTC (permalink / raw)
To: Patrick McHardy, pablo@netfilter.org, jengelh@medozas.de
Cc: Hans Schillstrom, netfilter-devel@vger.kernel.org,
netdev@vger.kernel.org
In-Reply-To: <201112081012.13065.hans.schillstrom@ericsson.com>
Hello
On Thursday 08 December 2011 10:12:11 Hans Schillstrom wrote:
> Hi
> While testing HMARK and IPv6 with nf_defrag_ipv6 (and nf_conntrack_ipv6 loaded) I can't see the defrag ?
>
> From what I can see nf_conntrack_reasm goes into PREROUTING with prio -400
> and HMARK in PREROUTING with prio -150
>
We are running the external interfaces in LXC containers
i.e. in a netns and depend on that no defragmentation is done in this stage.
Fragments can arrive on any interface on any blade so...
I had an idea of a sysctl to be able to turn off nf_defrag_ipv{4,6} per namespace
Default is of course on so excisting apps will be happy.
Any objections to that idea ?
--
Regards
Hans Schillstrom <hans.schillstrom@ericsson.com>
^ permalink raw reply
* [patch net-next 0/6] vlan: introduce per-device vlan id list
From: Jiri Pirko @ 2011-12-08 14:11 UTC (permalink / raw)
To: netdev
Cc: davem, eric.dumazet, bhutchings, shemminger, ebiederm, mirqus,
kaber, greearb, jesse
Many drivers store vlan ids in their private structures.
They are maintaining the private lists in vlan_add_vid and vlan_kill_vid.
This patchset creates an infrastructure of storing vlan ids in one place,
vlan core. Another benefit of this is that it allows to keep track of vlan
ids needed. For example if eth0 has vlan id 10 on it, then it enters bonding,
bonding gets added 10 as well, then eth0 leaves bonding. Final step would
remove vlan id 10 from eth1 vlan filter resulting in potentially unfunction
vlan device. This patchset fixes this.
Jiri Pirko (6):
vlan: rename vlan_dev_info to vlan_dev_priv
net: make vlan ndo_vlan_rx_[add/kill]_vid return error value
net: introduce vlan_vid_[add/del] and use them instead of direct
[add/kill]_vid ndo calls
vlan: introduce vid list with reference counting
vlan: introduce functions to do mass addition/deletion of vids by
another device
team: use vlan_vids_[addr/del]_by_dev
drivers/net/bonding/bond_main.c | 63 +++---
drivers/net/ethernet/adaptec/starfire.c | 8 +-
drivers/net/ethernet/brocade/bna/bnad.c | 12 +-
drivers/net/ethernet/cisco/enic/enic_dev.c | 14 +-
drivers/net/ethernet/cisco/enic/enic_dev.h | 4 +-
drivers/net/ethernet/emulex/benet/be_main.c | 12 +-
drivers/net/ethernet/ibm/ehea/ehea_main.c | 21 ++-
drivers/net/ethernet/intel/e1000/e1000_main.c | 14 +-
drivers/net/ethernet/intel/e1000e/netdev.c | 12 +-
drivers/net/ethernet/intel/igb/igb_main.c | 12 +-
drivers/net/ethernet/intel/igbvf/netdev.c | 20 +-
drivers/net/ethernet/intel/ixgb/ixgb_main.c | 12 +-
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 8 +-
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 8 +-
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 7 +-
drivers/net/ethernet/neterion/vxge/vxge-main.c | 6 +-
drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c | 10 +-
drivers/net/ethernet/qlogic/qlge/qlge_main.c | 38 ++--
drivers/net/ethernet/tehuti/tehuti.c | 6 +-
drivers/net/ethernet/via/via-rhine.c | 10 +-
drivers/net/ethernet/via/via-velocity.c | 6 +-
drivers/net/macvlan.c | 14 +-
drivers/net/team/team.c | 53 ++++--
drivers/net/virtio_net.c | 6 +-
drivers/net/vmxnet3/vmxnet3_drv.c | 8 +-
drivers/s390/net/qeth_l2_main.c | 18 +-
drivers/s390/net/qeth_l3_main.c | 9 +-
include/linux/if_vlan.h | 46 +++--
include/linux/netdevice.h | 11 +-
net/8021q/vlan.c | 126 ++++-------
net/8021q/vlan.h | 38 +++-
net/8021q/vlan_core.c | 233 ++++++++++++++++++++-
net/8021q/vlan_dev.c | 72 ++++----
net/8021q/vlan_gvrp.c | 4 +-
net/8021q/vlan_netlink.c | 10 +-
net/8021q/vlanproc.c | 42 ++--
36 files changed, 662 insertions(+), 331 deletions(-)
--
1.7.6
^ permalink raw reply
* [patch net-next 1/6] vlan: rename vlan_dev_info to vlan_dev_priv
From: Jiri Pirko @ 2011-12-08 14:11 UTC (permalink / raw)
To: netdev
Cc: davem, eric.dumazet, bhutchings, shemminger, ebiederm, mirqus,
kaber, greearb, jesse
In-Reply-To: <1323353480-1900-1-git-send-email-jpirko@redhat.com>
As this structure is priv, name it approprietely. Also for pointer to it
use name "vlan".
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
---
include/linux/if_vlan.h | 2 +-
net/8021q/vlan.c | 24 +++++++-------
net/8021q/vlan.h | 8 ++--
net/8021q/vlan_core.c | 8 ++--
net/8021q/vlan_dev.c | 72 +++++++++++++++++++++++-----------------------
net/8021q/vlan_gvrp.c | 4 +-
net/8021q/vlan_netlink.c | 10 +++---
net/8021q/vlanproc.c | 42 +++++++++++++-------------
8 files changed, 85 insertions(+), 85 deletions(-)
diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h
index 070ac50..31d7c97 100644
--- a/include/linux/if_vlan.h
+++ b/include/linux/if_vlan.h
@@ -386,7 +386,7 @@ struct vlan_ioctl_args {
unsigned int skb_priority;
unsigned int name_type;
unsigned int bind_type;
- unsigned int flag; /* Matches vlan_dev_info flags */
+ unsigned int flag; /* Matches vlan_dev_priv flags */
} u;
short vlan_qos;
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index 5471628..e075625 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -99,7 +99,7 @@ static void vlan_rcu_free(struct rcu_head *rcu)
void unregister_vlan_dev(struct net_device *dev, struct list_head *head)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
struct vlan_group *grp;
@@ -167,7 +167,7 @@ int vlan_check_real_dev(struct net_device *real_dev, u16 vlan_id)
int register_vlan_dev(struct net_device *dev)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
u16 vlan_id = vlan->vlan_id;
@@ -192,7 +192,7 @@ int register_vlan_dev(struct net_device *dev)
if (err < 0)
goto out_uninit_applicant;
- /* Account for reference in struct vlan_dev_info */
+ /* Account for reference in struct vlan_dev_priv */
dev_hold(real_dev);
netif_stacked_transfer_operstate(real_dev, dev);
@@ -267,7 +267,7 @@ static int register_vlan_device(struct net_device *real_dev, u16 vlan_id)
snprintf(name, IFNAMSIZ, "vlan%.4i", vlan_id);
}
- new_dev = alloc_netdev(sizeof(struct vlan_dev_info), name, vlan_setup);
+ new_dev = alloc_netdev(sizeof(struct vlan_dev_priv), name, vlan_setup);
if (new_dev == NULL)
return -ENOBUFS;
@@ -278,10 +278,10 @@ static int register_vlan_device(struct net_device *real_dev, u16 vlan_id)
*/
new_dev->mtu = real_dev->mtu;
- vlan_dev_info(new_dev)->vlan_id = vlan_id;
- vlan_dev_info(new_dev)->real_dev = real_dev;
- vlan_dev_info(new_dev)->dent = NULL;
- vlan_dev_info(new_dev)->flags = VLAN_FLAG_REORDER_HDR;
+ vlan_dev_priv(new_dev)->vlan_id = vlan_id;
+ vlan_dev_priv(new_dev)->real_dev = real_dev;
+ vlan_dev_priv(new_dev)->dent = NULL;
+ vlan_dev_priv(new_dev)->flags = VLAN_FLAG_REORDER_HDR;
new_dev->rtnl_link_ops = &vlan_link_ops;
err = register_vlan_dev(new_dev);
@@ -298,7 +298,7 @@ out_free_newdev:
static void vlan_sync_address(struct net_device *dev,
struct net_device *vlandev)
{
- struct vlan_dev_info *vlan = vlan_dev_info(vlandev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(vlandev);
/* May be called without an actual change */
if (!compare_ether_addr(vlan->real_dev_addr, dev->dev_addr))
@@ -362,7 +362,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
struct vlan_group *grp;
int i, flgs;
struct net_device *vlandev;
- struct vlan_dev_info *vlan;
+ struct vlan_dev_priv *vlan;
LIST_HEAD(list);
if (is_vlan_dev(dev))
@@ -447,7 +447,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
if (!(flgs & IFF_UP))
continue;
- vlan = vlan_dev_info(vlandev);
+ vlan = vlan_dev_priv(vlandev);
if (!(vlan->flags & VLAN_FLAG_LOOSE_BINDING))
dev_change_flags(vlandev, flgs & ~IFF_UP);
netif_stacked_transfer_operstate(dev, vlandev);
@@ -465,7 +465,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
if (flgs & IFF_UP)
continue;
- vlan = vlan_dev_info(vlandev);
+ vlan = vlan_dev_priv(vlandev);
if (!(vlan->flags & VLAN_FLAG_LOOSE_BINDING))
dev_change_flags(vlandev, flgs | IFF_UP);
netif_stacked_transfer_operstate(dev, vlandev);
diff --git a/net/8021q/vlan.h b/net/8021q/vlan.h
index 9fd45f3..d3c4ea4 100644
--- a/net/8021q/vlan.h
+++ b/net/8021q/vlan.h
@@ -41,7 +41,7 @@ struct vlan_pcpu_stats {
};
/**
- * struct vlan_dev_info - VLAN private device data
+ * struct vlan_dev_priv - VLAN private device data
* @nr_ingress_mappings: number of ingress priority mappings
* @ingress_priority_map: ingress priority mappings
* @nr_egress_mappings: number of egress priority mappings
@@ -53,7 +53,7 @@ struct vlan_pcpu_stats {
* @dent: proc dir entry
* @vlan_pcpu_stats: ptr to percpu rx stats
*/
-struct vlan_dev_info {
+struct vlan_dev_priv {
unsigned int nr_ingress_mappings;
u32 ingress_priority_map[8];
unsigned int nr_egress_mappings;
@@ -69,7 +69,7 @@ struct vlan_dev_info {
struct vlan_pcpu_stats __percpu *vlan_pcpu_stats;
};
-static inline struct vlan_dev_info *vlan_dev_info(const struct net_device *dev)
+static inline struct vlan_dev_priv *vlan_dev_priv(const struct net_device *dev)
{
return netdev_priv(dev);
}
@@ -121,7 +121,7 @@ void unregister_vlan_dev(struct net_device *dev, struct list_head *head);
static inline u32 vlan_get_ingress_priority(struct net_device *dev,
u16 vlan_tci)
{
- struct vlan_dev_info *vip = vlan_dev_info(dev);
+ struct vlan_dev_priv *vip = vlan_dev_priv(dev);
return vip->ingress_priority_map[(vlan_tci >> VLAN_PRIO_SHIFT) & 0x7];
}
diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c
index 9c95e8e..85241f0 100644
--- a/net/8021q/vlan_core.c
+++ b/net/8021q/vlan_core.c
@@ -36,7 +36,7 @@ bool vlan_do_receive(struct sk_buff **skbp, bool last_handler)
skb->pkt_type = PACKET_HOST;
}
- if (!(vlan_dev_info(vlan_dev)->flags & VLAN_FLAG_REORDER_HDR)) {
+ if (!(vlan_dev_priv(vlan_dev)->flags & VLAN_FLAG_REORDER_HDR)) {
unsigned int offset = skb->data - skb_mac_header(skb);
/*
@@ -55,7 +55,7 @@ bool vlan_do_receive(struct sk_buff **skbp, bool last_handler)
skb->priority = vlan_get_ingress_priority(vlan_dev, skb->vlan_tci);
skb->vlan_tci = 0;
- rx_stats = this_cpu_ptr(vlan_dev_info(vlan_dev)->vlan_pcpu_stats);
+ rx_stats = this_cpu_ptr(vlan_dev_priv(vlan_dev)->vlan_pcpu_stats);
u64_stats_update_begin(&rx_stats->syncp);
rx_stats->rx_packets++;
@@ -90,13 +90,13 @@ EXPORT_SYMBOL(__vlan_find_dev_deep);
struct net_device *vlan_dev_real_dev(const struct net_device *dev)
{
- return vlan_dev_info(dev)->real_dev;
+ return vlan_dev_priv(dev)->real_dev;
}
EXPORT_SYMBOL(vlan_dev_real_dev);
u16 vlan_dev_vlan_id(const struct net_device *dev)
{
- return vlan_dev_info(dev)->vlan_id;
+ return vlan_dev_priv(dev)->vlan_id;
}
EXPORT_SYMBOL(vlan_dev_vlan_id);
diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
index 2b5fcde..3b4db82 100644
--- a/net/8021q/vlan_dev.c
+++ b/net/8021q/vlan_dev.c
@@ -72,7 +72,7 @@ vlan_dev_get_egress_qos_mask(struct net_device *dev, struct sk_buff *skb)
{
struct vlan_priority_tci_mapping *mp;
- mp = vlan_dev_info(dev)->egress_priority_map[(skb->priority & 0xF)];
+ mp = vlan_dev_priv(dev)->egress_priority_map[(skb->priority & 0xF)];
while (mp) {
if (mp->priority == skb->priority) {
return mp->vlan_qos; /* This should already be shifted
@@ -103,10 +103,10 @@ static int vlan_dev_hard_header(struct sk_buff *skb, struct net_device *dev,
u16 vlan_tci = 0;
int rc;
- if (!(vlan_dev_info(dev)->flags & VLAN_FLAG_REORDER_HDR)) {
+ if (!(vlan_dev_priv(dev)->flags & VLAN_FLAG_REORDER_HDR)) {
vhdr = (struct vlan_hdr *) skb_push(skb, VLAN_HLEN);
- vlan_tci = vlan_dev_info(dev)->vlan_id;
+ vlan_tci = vlan_dev_priv(dev)->vlan_id;
vlan_tci |= vlan_dev_get_egress_qos_mask(dev, skb);
vhdr->h_vlan_TCI = htons(vlan_tci);
@@ -129,7 +129,7 @@ static int vlan_dev_hard_header(struct sk_buff *skb, struct net_device *dev,
saddr = dev->dev_addr;
/* Now make the underlying real hard header */
- dev = vlan_dev_info(dev)->real_dev;
+ dev = vlan_dev_priv(dev)->real_dev;
rc = dev_hard_header(skb, dev, type, daddr, saddr, len + vhdrlen);
if (rc > 0)
rc += vhdrlen;
@@ -149,27 +149,27 @@ static netdev_tx_t vlan_dev_hard_start_xmit(struct sk_buff *skb,
* OTHER THINGS LIKE FDDI/TokenRing/802.3 SNAPs...
*/
if (veth->h_vlan_proto != htons(ETH_P_8021Q) ||
- vlan_dev_info(dev)->flags & VLAN_FLAG_REORDER_HDR) {
+ vlan_dev_priv(dev)->flags & VLAN_FLAG_REORDER_HDR) {
u16 vlan_tci;
- vlan_tci = vlan_dev_info(dev)->vlan_id;
+ vlan_tci = vlan_dev_priv(dev)->vlan_id;
vlan_tci |= vlan_dev_get_egress_qos_mask(dev, skb);
skb = __vlan_hwaccel_put_tag(skb, vlan_tci);
}
- skb_set_dev(skb, vlan_dev_info(dev)->real_dev);
+ skb_set_dev(skb, vlan_dev_priv(dev)->real_dev);
len = skb->len;
ret = dev_queue_xmit(skb);
if (likely(ret == NET_XMIT_SUCCESS || ret == NET_XMIT_CN)) {
struct vlan_pcpu_stats *stats;
- stats = this_cpu_ptr(vlan_dev_info(dev)->vlan_pcpu_stats);
+ stats = this_cpu_ptr(vlan_dev_priv(dev)->vlan_pcpu_stats);
u64_stats_update_begin(&stats->syncp);
stats->tx_packets++;
stats->tx_bytes += len;
u64_stats_update_end(&stats->syncp);
} else {
- this_cpu_inc(vlan_dev_info(dev)->vlan_pcpu_stats->tx_dropped);
+ this_cpu_inc(vlan_dev_priv(dev)->vlan_pcpu_stats->tx_dropped);
}
return ret;
@@ -180,7 +180,7 @@ static int vlan_dev_change_mtu(struct net_device *dev, int new_mtu)
/* TODO: gotta make sure the underlying layer can handle it,
* maybe an IFF_VLAN_CAPABLE flag for devices?
*/
- if (vlan_dev_info(dev)->real_dev->mtu < new_mtu)
+ if (vlan_dev_priv(dev)->real_dev->mtu < new_mtu)
return -ERANGE;
dev->mtu = new_mtu;
@@ -191,7 +191,7 @@ static int vlan_dev_change_mtu(struct net_device *dev, int new_mtu)
void vlan_dev_set_ingress_priority(const struct net_device *dev,
u32 skb_prio, u16 vlan_prio)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
if (vlan->ingress_priority_map[vlan_prio & 0x7] && !skb_prio)
vlan->nr_ingress_mappings--;
@@ -204,7 +204,7 @@ void vlan_dev_set_ingress_priority(const struct net_device *dev,
int vlan_dev_set_egress_priority(const struct net_device *dev,
u32 skb_prio, u16 vlan_prio)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct vlan_priority_tci_mapping *mp = NULL;
struct vlan_priority_tci_mapping *np;
u32 vlan_qos = (vlan_prio << VLAN_PRIO_SHIFT) & VLAN_PRIO_MASK;
@@ -241,7 +241,7 @@ int vlan_dev_set_egress_priority(const struct net_device *dev,
/* Flags are defined in the vlan_flags enum in include/linux/if_vlan.h file. */
int vlan_dev_change_flags(const struct net_device *dev, u32 flags, u32 mask)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
u32 old_flags = vlan->flags;
if (mask & ~(VLAN_FLAG_REORDER_HDR | VLAN_FLAG_GVRP |
@@ -261,12 +261,12 @@ int vlan_dev_change_flags(const struct net_device *dev, u32 flags, u32 mask)
void vlan_dev_get_realdev_name(const struct net_device *dev, char *result)
{
- strncpy(result, vlan_dev_info(dev)->real_dev->name, 23);
+ strncpy(result, vlan_dev_priv(dev)->real_dev->name, 23);
}
static int vlan_dev_open(struct net_device *dev)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
int err;
@@ -313,7 +313,7 @@ out:
static int vlan_dev_stop(struct net_device *dev)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
dev_mc_unsync(real_dev, dev);
@@ -332,7 +332,7 @@ static int vlan_dev_stop(struct net_device *dev)
static int vlan_dev_set_mac_address(struct net_device *dev, void *p)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
struct sockaddr *addr = p;
int err;
@@ -358,7 +358,7 @@ out:
static int vlan_dev_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
struct ifreq ifrr;
int err = -EOPNOTSUPP;
@@ -383,7 +383,7 @@ static int vlan_dev_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
static int vlan_dev_neigh_setup(struct net_device *dev, struct neigh_parms *pa)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int err = 0;
@@ -397,7 +397,7 @@ static int vlan_dev_neigh_setup(struct net_device *dev, struct neigh_parms *pa)
static int vlan_dev_fcoe_ddp_setup(struct net_device *dev, u16 xid,
struct scatterlist *sgl, unsigned int sgc)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int rc = 0;
@@ -409,7 +409,7 @@ static int vlan_dev_fcoe_ddp_setup(struct net_device *dev, u16 xid,
static int vlan_dev_fcoe_ddp_done(struct net_device *dev, u16 xid)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int len = 0;
@@ -421,7 +421,7 @@ static int vlan_dev_fcoe_ddp_done(struct net_device *dev, u16 xid)
static int vlan_dev_fcoe_enable(struct net_device *dev)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int rc = -EINVAL;
@@ -432,7 +432,7 @@ static int vlan_dev_fcoe_enable(struct net_device *dev)
static int vlan_dev_fcoe_disable(struct net_device *dev)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int rc = -EINVAL;
@@ -443,7 +443,7 @@ static int vlan_dev_fcoe_disable(struct net_device *dev)
static int vlan_dev_fcoe_get_wwn(struct net_device *dev, u64 *wwn, int type)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int rc = -EINVAL;
@@ -455,7 +455,7 @@ static int vlan_dev_fcoe_get_wwn(struct net_device *dev, u64 *wwn, int type)
static int vlan_dev_fcoe_ddp_target(struct net_device *dev, u16 xid,
struct scatterlist *sgl, unsigned int sgc)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
const struct net_device_ops *ops = real_dev->netdev_ops;
int rc = 0;
@@ -468,7 +468,7 @@ static int vlan_dev_fcoe_ddp_target(struct net_device *dev, u16 xid,
static void vlan_dev_change_rx_flags(struct net_device *dev, int change)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
if (dev->flags & IFF_UP) {
if (change & IFF_ALLMULTI)
@@ -480,8 +480,8 @@ static void vlan_dev_change_rx_flags(struct net_device *dev, int change)
static void vlan_dev_set_rx_mode(struct net_device *vlan_dev)
{
- dev_mc_sync(vlan_dev_info(vlan_dev)->real_dev, vlan_dev);
- dev_uc_sync(vlan_dev_info(vlan_dev)->real_dev, vlan_dev);
+ dev_mc_sync(vlan_dev_priv(vlan_dev)->real_dev, vlan_dev);
+ dev_uc_sync(vlan_dev_priv(vlan_dev)->real_dev, vlan_dev);
}
/*
@@ -519,7 +519,7 @@ static const struct net_device_ops vlan_netdev_ops;
static int vlan_dev_init(struct net_device *dev)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
int subclass = 0;
netif_carrier_off(dev);
@@ -568,8 +568,8 @@ static int vlan_dev_init(struct net_device *dev)
vlan_dev_set_lockdep_class(dev, subclass);
- vlan_dev_info(dev)->vlan_pcpu_stats = alloc_percpu(struct vlan_pcpu_stats);
- if (!vlan_dev_info(dev)->vlan_pcpu_stats)
+ vlan_dev_priv(dev)->vlan_pcpu_stats = alloc_percpu(struct vlan_pcpu_stats);
+ if (!vlan_dev_priv(dev)->vlan_pcpu_stats)
return -ENOMEM;
return 0;
@@ -578,7 +578,7 @@ static int vlan_dev_init(struct net_device *dev)
static void vlan_dev_uninit(struct net_device *dev)
{
struct vlan_priority_tci_mapping *pm;
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
int i;
free_percpu(vlan->vlan_pcpu_stats);
@@ -594,7 +594,7 @@ static void vlan_dev_uninit(struct net_device *dev)
static netdev_features_t vlan_dev_fix_features(struct net_device *dev,
netdev_features_t features)
{
- struct net_device *real_dev = vlan_dev_info(dev)->real_dev;
+ struct net_device *real_dev = vlan_dev_priv(dev)->real_dev;
u32 old_features = features;
features &= real_dev->vlan_features;
@@ -610,7 +610,7 @@ static netdev_features_t vlan_dev_fix_features(struct net_device *dev,
static int vlan_ethtool_get_settings(struct net_device *dev,
struct ethtool_cmd *cmd)
{
- const struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ const struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
return __ethtool_get_settings(vlan->real_dev, cmd);
}
@@ -626,7 +626,7 @@ static void vlan_ethtool_get_drvinfo(struct net_device *dev,
static struct rtnl_link_stats64 *vlan_dev_get_stats64(struct net_device *dev, struct rtnl_link_stats64 *stats)
{
- if (vlan_dev_info(dev)->vlan_pcpu_stats) {
+ if (vlan_dev_priv(dev)->vlan_pcpu_stats) {
struct vlan_pcpu_stats *p;
u32 rx_errors = 0, tx_dropped = 0;
int i;
@@ -635,7 +635,7 @@ static struct rtnl_link_stats64 *vlan_dev_get_stats64(struct net_device *dev, st
u64 rxpackets, rxbytes, rxmulticast, txpackets, txbytes;
unsigned int start;
- p = per_cpu_ptr(vlan_dev_info(dev)->vlan_pcpu_stats, i);
+ p = per_cpu_ptr(vlan_dev_priv(dev)->vlan_pcpu_stats, i);
do {
start = u64_stats_fetch_begin_bh(&p->syncp);
rxpackets = p->rx_packets;
diff --git a/net/8021q/vlan_gvrp.c b/net/8021q/vlan_gvrp.c
index 061cece..6f97553 100644
--- a/net/8021q/vlan_gvrp.c
+++ b/net/8021q/vlan_gvrp.c
@@ -29,7 +29,7 @@ static struct garp_application vlan_gvrp_app __read_mostly = {
int vlan_gvrp_request_join(const struct net_device *dev)
{
- const struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ const struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
__be16 vlan_id = htons(vlan->vlan_id);
return garp_request_join(vlan->real_dev, &vlan_gvrp_app,
@@ -38,7 +38,7 @@ int vlan_gvrp_request_join(const struct net_device *dev)
void vlan_gvrp_request_leave(const struct net_device *dev)
{
- const struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ const struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
__be16 vlan_id = htons(vlan->vlan_id);
garp_request_leave(vlan->real_dev, &vlan_gvrp_app,
diff --git a/net/8021q/vlan_netlink.c b/net/8021q/vlan_netlink.c
index 235c219..5071136 100644
--- a/net/8021q/vlan_netlink.c
+++ b/net/8021q/vlan_netlink.c
@@ -105,7 +105,7 @@ static int vlan_changelink(struct net_device *dev,
static int vlan_newlink(struct net *src_net, struct net_device *dev,
struct nlattr *tb[], struct nlattr *data[])
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev;
int err;
@@ -149,7 +149,7 @@ static inline size_t vlan_qos_map_size(unsigned int n)
static size_t vlan_get_size(const struct net_device *dev)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
return nla_total_size(2) + /* IFLA_VLAN_ID */
sizeof(struct ifla_vlan_flags) + /* IFLA_VLAN_FLAGS */
@@ -159,14 +159,14 @@ static size_t vlan_get_size(const struct net_device *dev)
static int vlan_fill_info(struct sk_buff *skb, const struct net_device *dev)
{
- struct vlan_dev_info *vlan = vlan_dev_info(dev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct vlan_priority_tci_mapping *pm;
struct ifla_vlan_flags f;
struct ifla_vlan_qos_mapping m;
struct nlattr *nest;
unsigned int i;
- NLA_PUT_U16(skb, IFLA_VLAN_ID, vlan_dev_info(dev)->vlan_id);
+ NLA_PUT_U16(skb, IFLA_VLAN_ID, vlan_dev_priv(dev)->vlan_id);
if (vlan->flags) {
f.flags = vlan->flags;
f.mask = ~0;
@@ -218,7 +218,7 @@ struct rtnl_link_ops vlan_link_ops __read_mostly = {
.kind = "vlan",
.maxtype = IFLA_VLAN_MAX,
.policy = vlan_policy,
- .priv_size = sizeof(struct vlan_dev_info),
+ .priv_size = sizeof(struct vlan_dev_priv),
.setup = vlan_setup,
.validate = vlan_validate,
.newlink = vlan_newlink,
diff --git a/net/8021q/vlanproc.c b/net/8021q/vlanproc.c
index d34b6da..c718fd3 100644
--- a/net/8021q/vlanproc.c
+++ b/net/8021q/vlanproc.c
@@ -168,13 +168,13 @@ err:
int vlan_proc_add_dev(struct net_device *vlandev)
{
- struct vlan_dev_info *dev_info = vlan_dev_info(vlandev);
+ struct vlan_dev_priv *vlan = vlan_dev_priv(vlandev);
struct vlan_net *vn = net_generic(dev_net(vlandev), vlan_net_id);
- dev_info->dent =
+ vlan->dent =
proc_create_data(vlandev->name, S_IFREG|S_IRUSR|S_IWUSR,
vn->proc_vlan_dir, &vlandev_fops, vlandev);
- if (!dev_info->dent)
+ if (!vlan->dent)
return -ENOBUFS;
return 0;
}
@@ -187,10 +187,10 @@ int vlan_proc_rem_dev(struct net_device *vlandev)
struct vlan_net *vn = net_generic(dev_net(vlandev), vlan_net_id);
/** NOTE: This will consume the memory pointed to by dent, it seems. */
- if (vlan_dev_info(vlandev)->dent) {
- remove_proc_entry(vlan_dev_info(vlandev)->dent->name,
+ if (vlan_dev_priv(vlandev)->dent) {
+ remove_proc_entry(vlan_dev_priv(vlandev)->dent->name,
vn->proc_vlan_dir);
- vlan_dev_info(vlandev)->dent = NULL;
+ vlan_dev_priv(vlandev)->dent = NULL;
}
return 0;
}
@@ -268,10 +268,10 @@ static int vlan_seq_show(struct seq_file *seq, void *v)
nmtype ? nmtype : "UNKNOWN");
} else {
const struct net_device *vlandev = v;
- const struct vlan_dev_info *dev_info = vlan_dev_info(vlandev);
+ const struct vlan_dev_priv *vlan = vlan_dev_priv(vlandev);
seq_printf(seq, "%-15s| %d | %s\n", vlandev->name,
- dev_info->vlan_id, dev_info->real_dev->name);
+ vlan->vlan_id, vlan->real_dev->name);
}
return 0;
}
@@ -279,7 +279,7 @@ static int vlan_seq_show(struct seq_file *seq, void *v)
static int vlandev_seq_show(struct seq_file *seq, void *offset)
{
struct net_device *vlandev = (struct net_device *) seq->private;
- const struct vlan_dev_info *dev_info = vlan_dev_info(vlandev);
+ const struct vlan_dev_priv *vlan = vlan_dev_priv(vlandev);
struct rtnl_link_stats64 temp;
const struct rtnl_link_stats64 *stats;
static const char fmt64[] = "%30s %12llu\n";
@@ -291,8 +291,8 @@ static int vlandev_seq_show(struct seq_file *seq, void *offset)
stats = dev_get_stats(vlandev, &temp);
seq_printf(seq,
"%s VID: %d REORDER_HDR: %i dev->priv_flags: %hx\n",
- vlandev->name, dev_info->vlan_id,
- (int)(dev_info->flags & 1), vlandev->priv_flags);
+ vlandev->name, vlan->vlan_id,
+ (int)(vlan->flags & 1), vlandev->priv_flags);
seq_printf(seq, fmt64, "total frames received", stats->rx_packets);
seq_printf(seq, fmt64, "total bytes received", stats->rx_bytes);
@@ -300,23 +300,23 @@ static int vlandev_seq_show(struct seq_file *seq, void *offset)
seq_puts(seq, "\n");
seq_printf(seq, fmt64, "total frames transmitted", stats->tx_packets);
seq_printf(seq, fmt64, "total bytes transmitted", stats->tx_bytes);
- seq_printf(seq, "Device: %s", dev_info->real_dev->name);
+ seq_printf(seq, "Device: %s", vlan->real_dev->name);
/* now show all PRIORITY mappings relating to this VLAN */
seq_printf(seq, "\nINGRESS priority mappings: "
"0:%u 1:%u 2:%u 3:%u 4:%u 5:%u 6:%u 7:%u\n",
- dev_info->ingress_priority_map[0],
- dev_info->ingress_priority_map[1],
- dev_info->ingress_priority_map[2],
- dev_info->ingress_priority_map[3],
- dev_info->ingress_priority_map[4],
- dev_info->ingress_priority_map[5],
- dev_info->ingress_priority_map[6],
- dev_info->ingress_priority_map[7]);
+ vlan->ingress_priority_map[0],
+ vlan->ingress_priority_map[1],
+ vlan->ingress_priority_map[2],
+ vlan->ingress_priority_map[3],
+ vlan->ingress_priority_map[4],
+ vlan->ingress_priority_map[5],
+ vlan->ingress_priority_map[6],
+ vlan->ingress_priority_map[7]);
seq_printf(seq, " EGRESS priority mappings: ");
for (i = 0; i < 16; i++) {
const struct vlan_priority_tci_mapping *mp
- = dev_info->egress_priority_map[i];
+ = vlan->egress_priority_map[i];
while (mp) {
seq_printf(seq, "%u:%hu ",
mp->priority, ((mp->vlan_qos >> 13) & 0x7));
--
1.7.6
^ permalink raw reply related
* [patch net-next 4/6] vlan: introduce vid list with reference counting
From: Jiri Pirko @ 2011-12-08 14:11 UTC (permalink / raw)
To: netdev
Cc: davem, eric.dumazet, bhutchings, shemminger, ebiederm, mirqus,
kaber, greearb, jesse
In-Reply-To: <1323353480-1900-1-git-send-email-jpirko@redhat.com>
This allows to keep track of vids needed to be in rx vlan filters of
devices even if they are used in bond/team etc.
vlan_info as well as vlan_group previously was, is allocated when first
vid is added and dealocated whan last vid is deleted.
vlan_group definition is moved to private header.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
---
include/linux/if_vlan.h | 17 +----
include/linux/netdevice.h | 3 +-
net/8021q/vlan.c | 90 ++++++++----------------
net/8021q/vlan.h | 30 +++++++-
net/8021q/vlan_core.c | 168 ++++++++++++++++++++++++++++++++++++++++++--
5 files changed, 219 insertions(+), 89 deletions(-)
diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h
index 71168a6..0c96913 100644
--- a/include/linux/if_vlan.h
+++ b/include/linux/if_vlan.h
@@ -74,22 +74,7 @@ static inline struct vlan_ethhdr *vlan_eth_hdr(const struct sk_buff *skb)
/* found in socket.c */
extern void vlan_ioctl_set(int (*hook)(struct net *, void __user *));
-/* if this changes, algorithm will have to be reworked because this
- * depends on completely exhausting the VLAN identifier space. Thus
- * it gives constant time look-up, but in many cases it wastes memory.
- */
-#define VLAN_GROUP_ARRAY_SPLIT_PARTS 8
-#define VLAN_GROUP_ARRAY_PART_LEN (VLAN_N_VID/VLAN_GROUP_ARRAY_SPLIT_PARTS)
-
-struct vlan_group {
- struct net_device *real_dev; /* The ethernet(like) device
- * the vlan is attached to.
- */
- unsigned int nr_vlans;
- struct hlist_node hlist; /* linked list */
- struct net_device **vlan_devices_arrays[VLAN_GROUP_ARRAY_SPLIT_PARTS];
- struct rcu_head rcu;
-};
+struct vlan_info;
static inline int is_vlan_dev(struct net_device *dev)
{
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index f7bff96..6037308 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -55,7 +55,6 @@
#include <linux/netdev_features.h>
-struct vlan_group;
struct netpoll_info;
struct phy_device;
/* 802.11 specific */
@@ -1096,7 +1095,7 @@ struct net_device {
/* Protocol specific pointers */
#if IS_ENABLED(CONFIG_VLAN_8021Q)
- struct vlan_group __rcu *vlgrp; /* VLAN group */
+ struct vlan_info __rcu *vlan_info; /* VLAN info */
#endif
#if IS_ENABLED(CONFIG_NET_DSA)
struct dsa_switch_tree *dsa_ptr; /* dsa specific data */
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index dd9aa40..efea35b 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -51,27 +51,6 @@ const char vlan_version[] = DRV_VERSION;
/* End of global variables definitions. */
-static void vlan_group_free(struct vlan_group *grp)
-{
- int i;
-
- for (i = 0; i < VLAN_GROUP_ARRAY_SPLIT_PARTS; i++)
- kfree(grp->vlan_devices_arrays[i]);
- kfree(grp);
-}
-
-static struct vlan_group *vlan_group_alloc(struct net_device *real_dev)
-{
- struct vlan_group *grp;
-
- grp = kzalloc(sizeof(struct vlan_group), GFP_KERNEL);
- if (!grp)
- return NULL;
-
- grp->real_dev = real_dev;
- return grp;
-}
-
static int vlan_group_prealloc_vid(struct vlan_group *vg, u16 vlan_id)
{
struct net_device **array;
@@ -92,22 +71,20 @@ static int vlan_group_prealloc_vid(struct vlan_group *vg, u16 vlan_id)
return 0;
}
-static void vlan_rcu_free(struct rcu_head *rcu)
-{
- vlan_group_free(container_of(rcu, struct vlan_group, rcu));
-}
-
void unregister_vlan_dev(struct net_device *dev, struct list_head *head)
{
struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
+ struct vlan_info *vlan_info;
struct vlan_group *grp;
u16 vlan_id = vlan->vlan_id;
ASSERT_RTNL();
- grp = rtnl_dereference(real_dev->vlgrp);
- BUG_ON(!grp);
+ vlan_info = rtnl_dereference(real_dev->vlan_info);
+ BUG_ON(!vlan_info);
+
+ grp = &vlan_info->grp;
/* Take it out of our own structures, but be sure to interlock with
* HW accelerating devices or SW vlan input packet processing if
@@ -116,7 +93,7 @@ void unregister_vlan_dev(struct net_device *dev, struct list_head *head)
if (vlan_id)
vlan_vid_del(real_dev, vlan_id);
- grp->nr_vlans--;
+ grp->nr_vlan_devs--;
if (vlan->flags & VLAN_FLAG_GVRP)
vlan_gvrp_request_leave(dev);
@@ -128,16 +105,9 @@ void unregister_vlan_dev(struct net_device *dev, struct list_head *head)
*/
unregister_netdevice_queue(dev, head);
- /* If the group is now empty, kill off the group. */
- if (grp->nr_vlans == 0) {
+ if (grp->nr_vlan_devs == 0)
vlan_gvrp_uninit_applicant(real_dev);
- RCU_INIT_POINTER(real_dev->vlgrp, NULL);
-
- /* Free the group, after all cpu's are done. */
- call_rcu(&grp->rcu, vlan_rcu_free);
- }
-
/* Get rid of the vlan's reference to real_dev */
dev_put(real_dev);
}
@@ -169,17 +139,23 @@ int register_vlan_dev(struct net_device *dev)
struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
u16 vlan_id = vlan->vlan_id;
- struct vlan_group *grp, *ngrp = NULL;
+ struct vlan_info *vlan_info;
+ struct vlan_group *grp;
int err;
- grp = rtnl_dereference(real_dev->vlgrp);
- if (!grp) {
- ngrp = grp = vlan_group_alloc(real_dev);
- if (!grp)
- return -ENOBUFS;
+ err = vlan_vid_add(real_dev, vlan_id);
+ if (err)
+ return err;
+
+ vlan_info = rtnl_dereference(real_dev->vlan_info);
+ /* vlan_info should be there now. vlan_vid_add took care of it */
+ BUG_ON(!vlan_info);
+
+ grp = &vlan_info->grp;
+ if (grp->nr_vlan_devs == 0) {
err = vlan_gvrp_init_applicant(real_dev);
if (err < 0)
- goto out_free_group;
+ goto out_vid_del;
}
err = vlan_group_prealloc_vid(grp, vlan_id);
@@ -200,23 +176,15 @@ int register_vlan_dev(struct net_device *dev)
* it into our local structure.
*/
vlan_group_set_device(grp, vlan_id, dev);
- grp->nr_vlans++;
-
- if (ngrp) {
- rcu_assign_pointer(real_dev->vlgrp, ngrp);
- }
- vlan_vid_add(real_dev, vlan_id);
+ grp->nr_vlan_devs++;
return 0;
out_uninit_applicant:
- if (ngrp)
+ if (grp->nr_vlan_devs == 0)
vlan_gvrp_uninit_applicant(real_dev);
-out_free_group:
- if (ngrp) {
- /* Free the group, after all cpu's are done. */
- call_rcu(&ngrp->rcu, vlan_rcu_free);
- }
+out_vid_del:
+ vlan_vid_del(real_dev, vlan_id);
return err;
}
@@ -357,6 +325,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
{
struct net_device *dev = ptr;
struct vlan_group *grp;
+ struct vlan_info *vlan_info;
int i, flgs;
struct net_device *vlandev;
struct vlan_dev_priv *vlan;
@@ -372,9 +341,10 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
vlan_vid_add(dev, 0);
}
- grp = rtnl_dereference(dev->vlgrp);
- if (!grp)
+ vlan_info = rtnl_dereference(dev->vlan_info);
+ if (!vlan_info)
goto out;
+ grp = &vlan_info->grp;
/* It is OK that we do not hold the group lock right now,
* as we run under the RTNL lock.
@@ -478,9 +448,9 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
if (!vlandev)
continue;
- /* unregistration of last vlan destroys group, abort
+ /* removal of last vid destroys vlan_info, abort
* afterwards */
- if (grp->nr_vlans == 1)
+ if (vlan_info->nr_vids == 1)
i = VLAN_N_VID;
unregister_vlan_dev(vlandev, &list);
diff --git a/net/8021q/vlan.h b/net/8021q/vlan.h
index d3c4ea4..28d8dc2 100644
--- a/net/8021q/vlan.h
+++ b/net/8021q/vlan.h
@@ -3,6 +3,7 @@
#include <linux/if_vlan.h>
#include <linux/u64_stats_sync.h>
+#include <linux/list.h>
/**
@@ -74,6 +75,29 @@ static inline struct vlan_dev_priv *vlan_dev_priv(const struct net_device *dev)
return netdev_priv(dev);
}
+/* if this changes, algorithm will have to be reworked because this
+ * depends on completely exhausting the VLAN identifier space. Thus
+ * it gives constant time look-up, but in many cases it wastes memory.
+ */
+#define VLAN_GROUP_ARRAY_SPLIT_PARTS 8
+#define VLAN_GROUP_ARRAY_PART_LEN (VLAN_N_VID/VLAN_GROUP_ARRAY_SPLIT_PARTS)
+
+struct vlan_group {
+ unsigned int nr_vlan_devs;
+ struct hlist_node hlist; /* linked list */
+ struct net_device **vlan_devices_arrays[VLAN_GROUP_ARRAY_SPLIT_PARTS];
+};
+
+struct vlan_info {
+ struct net_device *real_dev; /* The ethernet(like) device
+ * the vlan is attached to.
+ */
+ struct vlan_group grp;
+ struct list_head vid_list;
+ unsigned int nr_vids;
+ struct rcu_head rcu;
+};
+
static inline struct net_device *vlan_group_get_device(struct vlan_group *vg,
u16 vlan_id)
{
@@ -97,10 +121,10 @@ static inline void vlan_group_set_device(struct vlan_group *vg,
static inline struct net_device *vlan_find_dev(struct net_device *real_dev,
u16 vlan_id)
{
- struct vlan_group *grp = rcu_dereference_rtnl(real_dev->vlgrp);
+ struct vlan_info *vlan_info = rcu_dereference_rtnl(real_dev->vlan_info);
- if (grp)
- return vlan_group_get_device(grp, vlan_id);
+ if (vlan_info)
+ return vlan_group_get_device(&vlan_info->grp, vlan_id);
return NULL;
}
diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c
index 544f9cb..329e031 100644
--- a/net/8021q/vlan_core.c
+++ b/net/8021q/vlan_core.c
@@ -71,10 +71,10 @@ bool vlan_do_receive(struct sk_buff **skbp, bool last_handler)
struct net_device *__vlan_find_dev_deep(struct net_device *real_dev,
u16 vlan_id)
{
- struct vlan_group *grp = rcu_dereference_rtnl(real_dev->vlgrp);
+ struct vlan_info *vlan_info = rcu_dereference_rtnl(real_dev->vlan_info);
- if (grp) {
- return vlan_group_get_device(grp, vlan_id);
+ if (vlan_info) {
+ return vlan_group_get_device(&vlan_info->grp, vlan_id);
} else {
/*
* Bonding slaves do not have grp assigned to themselves.
@@ -147,25 +147,177 @@ err_free:
return NULL;
}
-int vlan_vid_add(struct net_device *dev, unsigned short vid)
+
+/*
+ * vlan info and vid list
+ */
+
+static void vlan_group_free(struct vlan_group *grp)
+{
+ int i;
+
+ for (i = 0; i < VLAN_GROUP_ARRAY_SPLIT_PARTS; i++)
+ kfree(grp->vlan_devices_arrays[i]);
+}
+
+static void vlan_info_free(struct vlan_info *vlan_info)
+{
+ vlan_group_free(&vlan_info->grp);
+ kfree(vlan_info);
+}
+
+static void vlan_info_rcu_free(struct rcu_head *rcu)
+{
+ vlan_info_free(container_of(rcu, struct vlan_info, rcu));
+}
+
+static struct vlan_info *vlan_info_alloc(struct net_device *dev)
+{
+ struct vlan_info *vlan_info;
+
+ vlan_info = kzalloc(sizeof(struct vlan_info), GFP_KERNEL);
+ if (!vlan_info)
+ return NULL;
+
+ vlan_info->real_dev = dev;
+ INIT_LIST_HEAD(&vlan_info->vid_list);
+ return vlan_info;
+}
+
+struct vlan_vid_info {
+ struct list_head list;
+ unsigned short vid;
+ int refcount;
+};
+
+static struct vlan_vid_info *vlan_vid_info_get(struct vlan_info *vlan_info,
+ unsigned short vid)
+{
+ struct vlan_vid_info *vid_info;
+
+ list_for_each_entry(vid_info, &vlan_info->vid_list, list) {
+ if (vid_info->vid == vid)
+ return vid_info;
+ }
+ return NULL;
+}
+
+static struct vlan_vid_info *vlan_vid_info_alloc(unsigned short vid)
+{
+ struct vlan_vid_info *vid_info;
+
+ vid_info = kzalloc(sizeof(struct vlan_vid_info), GFP_KERNEL);
+ if (!vid_info)
+ return NULL;
+ vid_info->vid = vid;
+
+ return vid_info;
+}
+
+static int __vlan_vid_add(struct vlan_info *vlan_info, unsigned short vid,
+ struct vlan_vid_info **pvid_info)
{
+ struct net_device *dev = vlan_info->real_dev;
const struct net_device_ops *ops = dev->netdev_ops;
+ struct vlan_vid_info *vid_info;
+ int err;
+
+ vid_info = vlan_vid_info_alloc(vid);
+ if (!vid_info)
+ return -ENOMEM;
if ((dev->features & NETIF_F_HW_VLAN_FILTER) &&
- ops->ndo_vlan_rx_add_vid) {
- return ops->ndo_vlan_rx_add_vid(dev, vid);
+ ops->ndo_vlan_rx_add_vid) {
+ err = ops->ndo_vlan_rx_add_vid(dev, vid);
+ if (err) {
+ kfree(vid_info);
+ return err;
+ }
}
+ list_add(&vid_info->list, &vlan_info->vid_list);
+ vlan_info->nr_vids++;
+ *pvid_info = vid_info;
return 0;
}
+
+int vlan_vid_add(struct net_device *dev, unsigned short vid)
+{
+ struct vlan_info *vlan_info;
+ struct vlan_vid_info *vid_info;
+ bool vlan_info_created = false;
+ int err;
+
+ ASSERT_RTNL();
+
+ vlan_info = rtnl_dereference(dev->vlan_info);
+ if (!vlan_info) {
+ vlan_info = vlan_info_alloc(dev);
+ if (!vlan_info)
+ return -ENOMEM;
+ vlan_info_created = true;
+ }
+ vid_info = vlan_vid_info_get(vlan_info, vid);
+ if (!vid_info) {
+ err = __vlan_vid_add(vlan_info, vid, &vid_info);
+ if (err)
+ goto out_free_vlan_info;
+ }
+ vid_info->refcount++;
+
+ if (vlan_info_created)
+ rcu_assign_pointer(dev->vlan_info, vlan_info);
+
+ return 0;
+
+out_free_vlan_info:
+ if (vlan_info_created)
+ kfree(vlan_info);
+ return err;
+}
EXPORT_SYMBOL(vlan_vid_add);
-void vlan_vid_del(struct net_device *dev, unsigned short vid)
+static void __vlan_vid_del(struct vlan_info *vlan_info,
+ struct vlan_vid_info *vid_info)
{
+ struct net_device *dev = vlan_info->real_dev;
const struct net_device_ops *ops = dev->netdev_ops;
+ unsigned short vid = vid_info->vid;
+ int err;
if ((dev->features & NETIF_F_HW_VLAN_FILTER) &&
ops->ndo_vlan_rx_kill_vid) {
- ops->ndo_vlan_rx_kill_vid(dev, vid);
+ err = ops->ndo_vlan_rx_kill_vid(dev, vid);
+ if (err) {
+ pr_warn("failed to kill vid %d for device %s\n",
+ vid, dev->name);
+ }
+ }
+ list_del(&vid_info->list);
+ kfree(vid_info);
+ vlan_info->nr_vids--;
+}
+
+void vlan_vid_del(struct net_device *dev, unsigned short vid)
+{
+ struct vlan_info *vlan_info;
+ struct vlan_vid_info *vid_info;
+
+ ASSERT_RTNL();
+
+ vlan_info = rtnl_dereference(dev->vlan_info);
+ if (!vlan_info)
+ return;
+
+ vid_info = vlan_vid_info_get(vlan_info, vid);
+ if (!vid_info)
+ return;
+ vid_info->refcount--;
+ if (vid_info->refcount == 0) {
+ __vlan_vid_del(vlan_info, vid_info);
+ if (vlan_info->nr_vids == 0) {
+ RCU_INIT_POINTER(dev->vlan_info, NULL);
+ call_rcu(&vlan_info->rcu, vlan_info_rcu_free);
+ }
}
}
EXPORT_SYMBOL(vlan_vid_del);
--
1.7.6
^ permalink raw reply related
* [patch net-next 3/6] net: introduce vlan_vid_[add/del] and use them instead of direct [add/kill]_vid ndo calls
From: Jiri Pirko @ 2011-12-08 14:11 UTC (permalink / raw)
To: netdev
Cc: davem, eric.dumazet, bhutchings, shemminger, ebiederm, mirqus,
kaber, greearb, jesse, fubar, andy
In-Reply-To: <1323353480-1900-1-git-send-email-jpirko@redhat.com>
This patch adds wrapper for ndo_vlan_rx_add_vid/ndo_vlan_rx_kill_vid
functions. Check for NETIF_F_HW_VLAN_FILTER feature is done in this
wrapper.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
---
drivers/net/bonding/bond_main.c | 53 +++++++++++++++++----------------------
drivers/net/macvlan.c | 10 ++-----
drivers/net/team/team.c | 34 +++++++++++++++---------
include/linux/if_vlan.h | 12 +++++++++
net/8021q/vlan.c | 14 +++------
net/8021q/vlan_core.c | 23 +++++++++++++++++
6 files changed, 87 insertions(+), 59 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index d72c37f..0c0dacb 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -431,17 +431,13 @@ int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb,
static int bond_vlan_rx_add_vid(struct net_device *bond_dev, uint16_t vid)
{
struct bonding *bond = netdev_priv(bond_dev);
- struct slave *slave;
+ struct slave *slave, *stop_at;
int i, res;
bond_for_each_slave(bond, slave, i) {
- struct net_device *slave_dev = slave->dev;
- const struct net_device_ops *slave_ops = slave_dev->netdev_ops;
-
- if ((slave_dev->features & NETIF_F_HW_VLAN_FILTER) &&
- slave_ops->ndo_vlan_rx_add_vid) {
- slave_ops->ndo_vlan_rx_add_vid(slave_dev, vid);
- }
+ res = vlan_vid_add(slave->dev, vid);
+ if (res)
+ goto unwind;
}
res = bond_add_vlan(bond, vid);
@@ -452,6 +448,14 @@ static int bond_vlan_rx_add_vid(struct net_device *bond_dev, uint16_t vid)
}
return 0;
+
+unwind:
+ /* unwind from head to the slave that failed */
+ stop_at = slave;
+ bond_for_each_slave_from_to(bond, slave, i, bond->first_slave, stop_at)
+ vlan_vid_del(slave->dev, vid);
+
+ return res;
}
/**
@@ -465,15 +469,8 @@ static int bond_vlan_rx_kill_vid(struct net_device *bond_dev, uint16_t vid)
struct slave *slave;
int i, res;
- bond_for_each_slave(bond, slave, i) {
- struct net_device *slave_dev = slave->dev;
- const struct net_device_ops *slave_ops = slave_dev->netdev_ops;
-
- if ((slave_dev->features & NETIF_F_HW_VLAN_FILTER) &&
- slave_ops->ndo_vlan_rx_kill_vid) {
- slave_ops->ndo_vlan_rx_kill_vid(slave_dev, vid);
- }
- }
+ bond_for_each_slave(bond, slave, i)
+ vlan_vid_del(slave->dev, vid);
res = bond_del_vlan(bond, vid);
if (res) {
@@ -488,30 +485,26 @@ static int bond_vlan_rx_kill_vid(struct net_device *bond_dev, uint16_t vid)
static void bond_add_vlans_on_slave(struct bonding *bond, struct net_device *slave_dev)
{
struct vlan_entry *vlan;
- const struct net_device_ops *slave_ops = slave_dev->netdev_ops;
-
- if (!(slave_dev->features & NETIF_F_HW_VLAN_FILTER) ||
- !(slave_ops->ndo_vlan_rx_add_vid))
- return;
+ int res;
- list_for_each_entry(vlan, &bond->vlan_list, vlan_list)
- slave_ops->ndo_vlan_rx_add_vid(slave_dev, vlan->vlan_id);
+ list_for_each_entry(vlan, &bond->vlan_list, vlan_list) {
+ res = vlan_vid_add(slave_dev, vlan->vlan_id);
+ if (res)
+ pr_warning("%s: Failed to add vlan id %d to device %s\n",
+ bond->dev->name, vlan->vlan_id,
+ slave_dev->name);
+ }
}
static void bond_del_vlans_from_slave(struct bonding *bond,
struct net_device *slave_dev)
{
- const struct net_device_ops *slave_ops = slave_dev->netdev_ops;
struct vlan_entry *vlan;
- if (!(slave_dev->features & NETIF_F_HW_VLAN_FILTER) ||
- !(slave_ops->ndo_vlan_rx_kill_vid))
- return;
-
list_for_each_entry(vlan, &bond->vlan_list, vlan_list) {
if (!vlan->vlan_id)
continue;
- slave_ops->ndo_vlan_rx_kill_vid(slave_dev, vlan->vlan_id);
+ vlan_vid_del(slave_dev, vlan->vlan_id);
}
}
diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c
index 2511bc5..f2f820c 100644
--- a/drivers/net/macvlan.c
+++ b/drivers/net/macvlan.c
@@ -26,6 +26,7 @@
#include <linux/etherdevice.h>
#include <linux/ethtool.h>
#include <linux/if_arp.h>
+#include <linux/if_vlan.h>
#include <linux/if_link.h>
#include <linux/if_macvlan.h>
#include <net/rtnetlink.h>
@@ -525,11 +526,8 @@ static int macvlan_vlan_rx_add_vid(struct net_device *dev,
{
struct macvlan_dev *vlan = netdev_priv(dev);
struct net_device *lowerdev = vlan->lowerdev;
- const struct net_device_ops *ops = lowerdev->netdev_ops;
- if (ops->ndo_vlan_rx_add_vid)
- return ops->ndo_vlan_rx_add_vid(lowerdev, vid);
- return 0;
+ return vlan_vid_add(lowerdev, vid);
}
static int macvlan_vlan_rx_kill_vid(struct net_device *dev,
@@ -537,10 +535,8 @@ static int macvlan_vlan_rx_kill_vid(struct net_device *dev,
{
struct macvlan_dev *vlan = netdev_priv(dev);
struct net_device *lowerdev = vlan->lowerdev;
- const struct net_device_ops *ops = lowerdev->netdev_ops;
- if (ops->ndo_vlan_rx_kill_vid)
- return ops->ndo_vlan_rx_kill_vid(lowerdev, vid);
+ vlan_vid_del(lowerdev, vid);
return 0;
}
diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c
index 8e8bf95..79c2d1b 100644
--- a/drivers/net/team/team.c
+++ b/drivers/net/team/team.c
@@ -18,6 +18,7 @@
#include <linux/ctype.h>
#include <linux/notifier.h>
#include <linux/netdevice.h>
+#include <linux/if_vlan.h>
#include <linux/if_arp.h>
#include <linux/socket.h>
#include <linux/etherdevice.h>
@@ -906,17 +907,28 @@ static int team_vlan_rx_add_vid(struct net_device *dev, uint16_t vid)
{
struct team *team = netdev_priv(dev);
struct team_port *port;
+ int err;
- rcu_read_lock();
- list_for_each_entry_rcu(port, &team->port_list, list) {
- const struct net_device_ops *ops = port->dev->netdev_ops;
-
- if (ops->ndo_vlan_rx_add_vid)
- ops->ndo_vlan_rx_add_vid(port->dev, vid);
+ /*
+ * Alhough this is reader, it's guarded by team lock. It's not possible
+ * to traverse list in reverse under rcu_read_lock
+ */
+ mutex_lock(&team->lock);
+ list_for_each_entry(port, &team->port_list, list) {
+ err = vlan_vid_add(port->dev, vid);
+ if (err)
+ goto unwind;
}
- rcu_read_unlock();
+ mutex_unlock(&team->lock);
return 0;
+
+unwind:
+ list_for_each_entry_continue_reverse(port, &team->port_list, list)
+ vlan_vid_del(port->dev, vid);
+ mutex_unlock(&team->lock);
+
+ return err;
}
static int team_vlan_rx_kill_vid(struct net_device *dev, uint16_t vid)
@@ -925,12 +937,8 @@ static int team_vlan_rx_kill_vid(struct net_device *dev, uint16_t vid)
struct team_port *port;
rcu_read_lock();
- list_for_each_entry_rcu(port, &team->port_list, list) {
- const struct net_device_ops *ops = port->dev->netdev_ops;
-
- if (ops->ndo_vlan_rx_kill_vid)
- ops->ndo_vlan_rx_kill_vid(port->dev, vid);
- }
+ list_for_each_entry_rcu(port, &team->port_list, list)
+ vlan_vid_del(port->dev, vid);
rcu_read_unlock();
return 0;
diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h
index 31d7c97..71168a6 100644
--- a/include/linux/if_vlan.h
+++ b/include/linux/if_vlan.h
@@ -109,6 +109,9 @@ extern u16 vlan_dev_vlan_id(const struct net_device *dev);
extern bool vlan_do_receive(struct sk_buff **skb, bool last_handler);
extern struct sk_buff *vlan_untag(struct sk_buff *skb);
+extern int vlan_vid_add(struct net_device *dev, unsigned short vid);
+extern void vlan_vid_del(struct net_device *dev, unsigned short vid);
+
#else
static inline struct net_device *
__vlan_find_dev_deep(struct net_device *real_dev, u16 vlan_id)
@@ -139,6 +142,15 @@ static inline struct sk_buff *vlan_untag(struct sk_buff *skb)
{
return skb;
}
+
+static inline int vlan_vid_add(struct net_device *dev, unsigned short vid)
+{
+ return 0;
+}
+
+static inline void vlan_vid_del(struct net_device *dev, unsigned short vid)
+{
+}
#endif
/**
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index e075625..dd9aa40 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -101,7 +101,6 @@ void unregister_vlan_dev(struct net_device *dev, struct list_head *head)
{
struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
- const struct net_device_ops *ops = real_dev->netdev_ops;
struct vlan_group *grp;
u16 vlan_id = vlan->vlan_id;
@@ -114,8 +113,8 @@ void unregister_vlan_dev(struct net_device *dev, struct list_head *head)
* HW accelerating devices or SW vlan input packet processing if
* VLAN is not 0 (leave it there for 802.1p).
*/
- if (vlan_id && (real_dev->features & NETIF_F_HW_VLAN_FILTER))
- ops->ndo_vlan_rx_kill_vid(real_dev, vlan_id);
+ if (vlan_id)
+ vlan_vid_del(real_dev, vlan_id);
grp->nr_vlans--;
@@ -169,7 +168,6 @@ int register_vlan_dev(struct net_device *dev)
{
struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
struct net_device *real_dev = vlan->real_dev;
- const struct net_device_ops *ops = real_dev->netdev_ops;
u16 vlan_id = vlan->vlan_id;
struct vlan_group *grp, *ngrp = NULL;
int err;
@@ -207,8 +205,7 @@ int register_vlan_dev(struct net_device *dev)
if (ngrp) {
rcu_assign_pointer(real_dev->vlgrp, ngrp);
}
- if (real_dev->features & NETIF_F_HW_VLAN_FILTER)
- ops->ndo_vlan_rx_add_vid(real_dev, vlan_id);
+ vlan_vid_add(real_dev, vlan_id);
return 0;
@@ -369,11 +366,10 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
__vlan_device_event(dev, event);
if ((event == NETDEV_UP) &&
- (dev->features & NETIF_F_HW_VLAN_FILTER) &&
- dev->netdev_ops->ndo_vlan_rx_add_vid) {
+ (dev->features & NETIF_F_HW_VLAN_FILTER)) {
pr_info("adding VLAN 0 to HW filter on device %s\n",
dev->name);
- dev->netdev_ops->ndo_vlan_rx_add_vid(dev, 0);
+ vlan_vid_add(dev, 0);
}
grp = rtnl_dereference(dev->vlgrp);
diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c
index 85241f0..544f9cb 100644
--- a/net/8021q/vlan_core.c
+++ b/net/8021q/vlan_core.c
@@ -146,3 +146,26 @@ err_free:
kfree_skb(skb);
return NULL;
}
+
+int vlan_vid_add(struct net_device *dev, unsigned short vid)
+{
+ const struct net_device_ops *ops = dev->netdev_ops;
+
+ if ((dev->features & NETIF_F_HW_VLAN_FILTER) &&
+ ops->ndo_vlan_rx_add_vid) {
+ return ops->ndo_vlan_rx_add_vid(dev, vid);
+ }
+ return 0;
+}
+EXPORT_SYMBOL(vlan_vid_add);
+
+void vlan_vid_del(struct net_device *dev, unsigned short vid)
+{
+ const struct net_device_ops *ops = dev->netdev_ops;
+
+ if ((dev->features & NETIF_F_HW_VLAN_FILTER) &&
+ ops->ndo_vlan_rx_kill_vid) {
+ ops->ndo_vlan_rx_kill_vid(dev, vid);
+ }
+}
+EXPORT_SYMBOL(vlan_vid_del);
--
1.7.6
^ permalink raw reply related
* [patch net-next 6/6] team: use vlan_vids_[addr/del]_by_dev
From: Jiri Pirko @ 2011-12-08 14:11 UTC (permalink / raw)
To: netdev
Cc: davem, eric.dumazet, bhutchings, shemminger, ebiederm, mirqus,
kaber, greearb, jesse
In-Reply-To: <1323353480-1900-1-git-send-email-jpirko@redhat.com>
So far when vlan id was added to team device befor port was added, this
vid was not added to port's vlan filter. Also after removal, vid stayed
in port device's vlan filter. Benefit of new vlan functions to handle
this work.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
---
drivers/net/team/team.c | 11 +++++++++++
1 files changed, 11 insertions(+), 0 deletions(-)
diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c
index 79c2d1b..ed2a862 100644
--- a/drivers/net/team/team.c
+++ b/drivers/net/team/team.c
@@ -588,6 +588,13 @@ static int team_port_add(struct team *team, struct net_device *port_dev)
goto err_dev_open;
}
+ err = vlan_vids_add_by_dev(port_dev, dev);
+ if (err) {
+ netdev_err(dev, "Failed to add vlan ids to device %s\n",
+ portname);
+ goto err_vids_add;
+ }
+
err = netdev_set_master(port_dev, dev);
if (err) {
netdev_err(dev, "Device %s failed to set master\n", portname);
@@ -615,6 +622,9 @@ err_handler_register:
netdev_set_master(port_dev, NULL);
err_set_master:
+ vlan_vids_del_by_dev(port_dev, dev);
+
+err_vids_add:
dev_close(port_dev);
err_dev_open:
@@ -648,6 +658,7 @@ static int team_port_del(struct team *team, struct net_device *port_dev)
team_adjust_ops(team);
netdev_rx_handler_unregister(port_dev);
netdev_set_master(port_dev, NULL);
+ vlan_vids_del_by_dev(port_dev, dev);
dev_close(port_dev);
team_port_leave(team, port);
team_port_set_orig_mac(port);
--
1.7.6
^ permalink raw reply related
* [patch net-next 5/6] vlan: introduce functions to do mass addition/deletion of vids by another device
From: Jiri Pirko @ 2011-12-08 14:11 UTC (permalink / raw)
To: netdev
Cc: davem, eric.dumazet, bhutchings, shemminger, ebiederm, mirqus,
kaber, greearb, jesse
In-Reply-To: <1323353480-1900-1-git-send-email-jpirko@redhat.com>
Introduce functions handy to copy vlan ids from one driver's list to
another.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
---
include/linux/if_vlan.h | 15 +++++++++++++++
net/8021q/vlan_core.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 59 insertions(+), 0 deletions(-)
diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h
index 0c96913..13aff1e 100644
--- a/include/linux/if_vlan.h
+++ b/include/linux/if_vlan.h
@@ -97,6 +97,10 @@ extern struct sk_buff *vlan_untag(struct sk_buff *skb);
extern int vlan_vid_add(struct net_device *dev, unsigned short vid);
extern void vlan_vid_del(struct net_device *dev, unsigned short vid);
+extern int vlan_vids_add_by_dev(struct net_device *dev,
+ const struct net_device *by_dev);
+extern void vlan_vids_del_by_dev(struct net_device *dev,
+ const struct net_device *by_dev);
#else
static inline struct net_device *
__vlan_find_dev_deep(struct net_device *real_dev, u16 vlan_id)
@@ -136,6 +140,17 @@ static inline int vlan_vid_add(struct net_device *dev, unsigned short vid)
static inline void vlan_vid_del(struct net_device *dev, unsigned short vid)
{
}
+
+static inline int vlan_vids_add_by_dev(struct net_device *dev,
+ const struct net_device *by_dev)
+{
+ return 0;
+}
+
+static inline void vlan_vids_del_by_dev(struct net_device *dev,
+ const struct net_device *by_dev)
+{
+}
#endif
/**
diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c
index 329e031..1414c93 100644
--- a/net/8021q/vlan_core.c
+++ b/net/8021q/vlan_core.c
@@ -321,3 +321,47 @@ void vlan_vid_del(struct net_device *dev, unsigned short vid)
}
}
EXPORT_SYMBOL(vlan_vid_del);
+
+int vlan_vids_add_by_dev(struct net_device *dev,
+ const struct net_device *by_dev)
+{
+ struct vlan_vid_info *vid_info;
+ int err;
+
+ ASSERT_RTNL();
+
+ if (!by_dev->vlan_info)
+ return 0;
+
+ list_for_each_entry(vid_info, &by_dev->vlan_info->vid_list, list) {
+ err = vlan_vid_add(dev, vid_info->vid);
+ if (err)
+ goto unwind;
+ }
+ return 0;
+
+unwind:
+ list_for_each_entry_continue_reverse(vid_info,
+ &by_dev->vlan_info->vid_list,
+ list) {
+ vlan_vid_del(dev, vid_info->vid);
+ }
+
+ return err;
+}
+EXPORT_SYMBOL(vlan_vids_add_by_dev);
+
+void vlan_vids_del_by_dev(struct net_device *dev,
+ const struct net_device *by_dev)
+{
+ struct vlan_vid_info *vid_info;
+
+ ASSERT_RTNL();
+
+ if (!by_dev->vlan_info)
+ return;
+
+ list_for_each_entry(vid_info, &by_dev->vlan_info->vid_list, list)
+ vlan_vid_del(dev, vid_info->vid);
+}
+EXPORT_SYMBOL(vlan_vids_del_by_dev);
--
1.7.6
^ permalink raw reply related
* [BUG] bnx2x: crash at boot
From: Eric Dumazet @ 2011-12-08 14:13 UTC (permalink / raw)
To: David Miller; +Cc: barak, netdev, eilong
In-Reply-To: <20111207.133903.465696982570558665.davem@davemloft.net>
Hi guys
Latest net-next crashes in bnx2x_stats_init+0x678/0xc90
CR2: 000000000000000004da
# CONFIG_SCSI_BNX2X_FCOE is not set
Oh well...
bnx2x_nic_init()
bnx2x_nic_load()
bnx2x_open()
dev_change_flags()
FCOE support is good, but could we get back Ethernet support ?
^ permalink raw reply
* Re: bridge: HSR support
From: Arvid Brodin @ 2011-12-08 14:45 UTC (permalink / raw)
To: Jay Vosburgh; +Cc: Stephen Hemminger, netdev
In-Reply-To: <14653.1323287989@death>
Jay Vosburgh wrote:
> Arvid Brodin <arvid.brodin@enea.com> wrote:
>> * I don't know the meaning of the IFF_SLAVE flag. It's referenced all over the place
>> (core, vlan, bonding, ipv6, eql). Do I need to/want to set this?
>
> Only if you actually need to for some reason. There are a few
> tests that make actual use of IFF_SLAVE, e.g., IPv6 won't run addrconf
> on an interface with IFF_SLAVE set (which prevents bonding slaves from
> having an IPv6 address distinct from the master). Netpoll also treats
> interfaces with IFF_SLAVE in a special way. Bonding uses it internally
> for various tests.
>
>> * I don't know the effects of setting dev->master. Do I need/want this?
>
> Maybe. One effect of netdev_set_master is that a reference is
> acquired on the master, on behalf of the slave, so the master cannot
> simply vanish until the slave releases that reference. This predates
> the notifier facility, and careful use of notifiers (handling
> NETDEV_UNREGISTER) can get around the need for dev->master, but, e.g.,
> vlan still acquires a reference to the real_dev without using
> dev->master.
>
> It used to be that dev->master was used in netif_receive_skb for
> packet handling purposes (for bonding, mostly; bridge and I think
> macvlan had separate hooks). That special sauce is now done by the
> rx_handler, so there's really no requirement to use dev->master if you
> have no need.
>
>> * I don't want to forward all ingress frames on the slave devices to my master
>> device; I only want the ones with protocol 0x88FB to be forwarded (other
>> frames should be received by the slaves as normal). I think I already have this
>> covered by registering a protocol handler (using dev_add_pack(packet_type)).
>> So perhaps calling netdev_rx_handler_register() is not necessary in my case?
>
> You may want to use the rx_handler, and have it set skb->dev
> appropriately for the frames that should forward to the master, and
> leave skb->dev alone for the ones that should stick with the slave.
> Both of those need the appropriate return from the rx_handler, which is
> documented in netdevice.h.
>
> I'm not sure that you need a dev_add_pack at all; bonding
> doesn't use one anymore, since everything it needs can now be done via
> rx_handler. The dev_add_pack approach may work, but rx_handler is
> probably more efficient.
>
>> * As far as I can see, neither bridging nor bonding is handled by the ip program
>> (iproute2 suite)? I.e. no examples of binding more than one interface to a
>> virtual interface when it comes to which messages to send, etc. VLAN uses
>> IFLA_IFNAME (name of the vlan link), IFLA_LINK (physical link behind the vlan
>> link), and some IFLA_VLAN-specific messages.
>
> In current versions of iproute, something like "ip link set
> device eth0 master bond0" would add a slave to a bond. You are correct,
> though, that ip does not permit changing the bonding options, and I
> don't believe it will create new master devices, either, so the bonding
> support is limited.
>
> -J
>
> ---
> -Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com
>
Lots of great info there, many thanks!
--
Arvid Brodin
Enea Services Stockholm AB
^ permalink raw reply
* Re: [PATCH] add netpoll support for 802.1q vlans
From: Benjamin LaHaise @ 2011-12-08 15:35 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20111208.000730.63159506168958262.davem@davemloft.net>
On Thu, Dec 08, 2011 at 12:07:30AM -0500, David Miller wrote:
> From: Benjamin LaHaise <bcrl@kvack.org>
> Date: Tue, 6 Dec 2011 20:04:24 -0500
>
> > Add netpoll support to 802.1q vlan devices. Based on the netpoll support
> > in the bridging code. Tested on a forced_eth device with netconsole.
> >
> > Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
>
> Bridging doesn't try to invoke a downstream netpoll operation at all,
> and neither should you.
Please have a look at net/bridge/br_device.c:br_netpoll_setup() which is
where the approach came from.
Getting netconsole over vlans working is a requirement for those of us
with machines plugged into ethernet switches that can't mix tagged and
untagged packets on their ethernet ports. I bumped into this while
trying to capture a kernel crash on such a machine which doesn't have any
external serial ports to use either.
> Also, please format your subject line correctly so I don't have to edit
> it when applying your patch. You need to add an appropriate, lowercase,
> subsystem prefix after the [PATCH] tag otherwise people scanning the
> shortlog in GIT can't figure out what area your patch is in.
>
> In this situation "vlan: " might be appropriate.
Sorry about that oversight, I'll try to remember for the future.
-ben
--
"Thought is the essence of where you are now."
^ permalink raw reply
* Re: [PATCH] IPVS: Modify the SH scheduler to use weights
From: Mike Maxim @ 2011-12-08 15:50 UTC (permalink / raw)
To: Simon Horman
Cc: Pablo Neira Ayuso, Patrick McHardy, lvs-devel, netdev,
netfilter-devel, Wensong Zhang, Julian Anastasov
In-Reply-To: <20111208011108.GL4586@verge.net.au>
No problem, I'll send out the new patch now. Thanks!
On Wed, Dec 7, 2011 at 8:11 PM, Simon Horman <horms@verge.net.au> wrote:
> On Thu, Dec 08, 2011 at 09:40:09AM +0900, Simon Horman wrote:
>> On Wed, Dec 07, 2011 at 10:24:09AM -0500, Mike Maxim wrote:
>> > The reason I put that in is because the size of that table becomes
>> > more relevant/important if you decide to use the weights in the manner
>> > the patch lets you. It would be conceivable that someone might need to
>> > increase the size of that table to accommodate their configuration, so
>> > I thought it would be handy to be able to do that through the regular
>> > configuration system instead of editing the source.
>>
>> Hi Mike,
>>
>> I agree, but I think its also reasonable to consider IP_VS_SH_TAB_BITS as
>> a separate change.
>>
>> Could you
>>
>> 1. Split the patch into a Kconfig patch and a ip_vs_sh.c patch
>> 2. Remove spurious line additions from the resulting patches?
>
> Scratch that, please follow Pablo's request.
> --
> To unsubscribe from this list: send the line "unsubscribe lvs-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* [PATCH 2/2] IPVS: Modify the SH scheduler to use weights
From: Michael Maxim @ 2011-12-08 15:55 UTC (permalink / raw)
To: Wensong Zhang, Simon Horman, Julian Anastasov, Pablo Neira Ayuso,
Patrick McHardy, David S. Miller, netdev, lvs-devel
Cc: Mike Maxim
Modify the algorithm to build the source hashing hash table to add
extra slots for destinations with higher weight. This has the effect
of allowing an IPVS SH user to give more connections to hosts that
have been configured to have a higher weight.
The reason for the Kconfig change is because the size of the hash table
becomes more relevant/important if you decide to use the weights in the
manner this patch lets you. It would be conceivable that someone might
need to increase the size of that table to accommodate their
configuration, so it will be handy to be able to do that through the
regular configuration system instead of editing the source.
Signed-off-by: Michael Maxim <mike@okcupid.com>
---
net/netfilter/ipvs/Kconfig | 15 +++++++++++++++
net/netfilter/ipvs/ip_vs_sh.c | 18 +++++++++++++++++-
2 files changed, 32 insertions(+), 1 deletions(-)
diff --git a/net/netfilter/ipvs/Kconfig b/net/netfilter/ipvs/Kconfig
index 70bd1d0..af4c0b8 100644
--- a/net/netfilter/ipvs/Kconfig
+++ b/net/netfilter/ipvs/Kconfig
@@ -232,6 +232,21 @@ config IP_VS_NQ
If you want to compile it in kernel, say Y. To compile it as a
module, choose M here. If unsure, say N.
+comment 'IPVS SH scheduler'
+
+config IP_VS_SH_TAB_BITS
+ int "IPVS source hashing table size (the Nth power of 2)"
+ range 4 20
+ default 8
+ ---help---
+ The source hashing scheduler maps source IPs to destinations
+ stored in a hash table. This table is tiled by each destination
+ until all slots in the table are filled. When using weights to
+ allow destinations to receive more connections, the table is
+ tiled an amount proportional to the weights specified. The table
+ needs to be large enough to effectively fit all the destinations
+ multiplied by their respective weights.
+
comment 'IPVS application helper'
config IP_VS_FTP
diff --git a/net/netfilter/ipvs/ip_vs_sh.c b/net/netfilter/ipvs/ip_vs_sh.c
index 33815f4..069e8d4 100644
--- a/net/netfilter/ipvs/ip_vs_sh.c
+++ b/net/netfilter/ipvs/ip_vs_sh.c
@@ -30,6 +30,11 @@
* server is dead or overloaded, the load balancer can bypass the cache
* server and send requests to the original server directly.
*
+ * The weight destination attribute can be used to control the
+ * distribution of connections to the destinations in servernode. The
+ * greater the weight, the more connections the destination
+ * will receive.
+ *
*/
#define KMSG_COMPONENT "IPVS"
@@ -99,9 +104,11 @@ ip_vs_sh_assign(struct ip_vs_sh_bucket *tbl, struct ip_vs_service *svc)
struct ip_vs_sh_bucket *b;
struct list_head *p;
struct ip_vs_dest *dest;
+ int d_count;
b = tbl;
p = &svc->destinations;
+ d_count = 0;
for (i=0; i<IP_VS_SH_TAB_SIZE; i++) {
if (list_empty(p)) {
b->dest = NULL;
@@ -113,7 +120,16 @@ ip_vs_sh_assign(struct ip_vs_sh_bucket *tbl, struct ip_vs_service *svc)
atomic_inc(&dest->refcnt);
b->dest = dest;
- p = p->next;
+ IP_VS_DBG_BUF(6, "assigned i: %d dest: %s weight: %d\n",
+ i, IP_VS_DBG_ADDR(svc->af, &dest->addr),
+ atomic_read(&dest->weight));
+
+ /* Don't move to next dest until filling weight */
+ if (++d_count >= atomic_read(&dest->weight)) {
+ p = p->next;
+ d_count = 0;
+ }
+
}
b++;
}
--
1.7.4.1
^ permalink raw reply related
* [PATCH net-next] sch_red: Adaptative RED AQM
From: Eric Dumazet @ 2011-12-08 16:06 UTC (permalink / raw)
To: Dave Taht, David Miller
Cc: bloat, bloat-devel, netdev-u79uwXL29TY76Z2rM5mHXA, linux-wireless,
Hagen Paul Pfeifer
In-Reply-To: <1323082774.2670.40.camel@edumazet-HP-Compaq-6005-Pro-SFF-PC>
Adaptative RED AQM for linux, based on paper from Sally FLoyd,
Ramakrishna Gummadi, and Scott Shenker, August 2001 :
http://icir.org/floyd/papers/adaptiveRed.pdf
Goal of Adaptative RED is to make max_p a dynamic value between 1% and
50% to reach the target average queue : (max_th - min_th) / 2
Every 500 ms:
if (avg > target and max_p <= 0.5)
increase max_p : max_p += alpha;
else if (avg < target and max_p >= 0.01)
decrease max_p : max_p *= beta;
target :[min_th + 0.4*(min_th - max_th),
min_th + 0.6*(min_th - max_th)].
alpha : min(0.01, max_p / 4)
beta : 0.9
max_P is a Q0.32 fixed point number (unsigned, with 32 bits mantissa)
Changes against our RED implementation are :
max_p is no longer a negative power of two (1/(2^Plog)), but a Q0.32
fixed point number, to allow full range described in Adatative paper.
To deliver a random number, we now use a reciprocal divide (thats really
a multiply), but this operation is done once per marked/droped packet
when in RED_BETWEEN_TRESH window, so added cost (compared to previous
AND operation) is near zero.
dump operation gives current max_p value in a new TCA_RED_MAX_P
attribute.
Example on a 10Mbit link :
tc qdisc add dev $DEV parent 1:1 handle 10: est 1sec 8sec red \
limit 400000 min 30000 max 90000 avpkt 1000 \
burst 55 ecn adaptative bandwidth 10Mbit
# tc -s -d qdisc show dev eth3
...
qdisc red 10: parent 1:1 limit 400000b min 30000b max 90000b ecn
adaptative ewma 5 max_p=0.113335 Scell_log 15
Sent 50414282 bytes 34504 pkt (dropped 35, overlimits 1392 requeues 0)
rate 9749Kbit 831pps backlog 72056b 16p requeues 0
marked 1357 early 35 pdrop 0 other 0
Signed-off-by: Eric Dumazet <eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
include/linux/pkt_sched.h | 6 +-
include/net/red.h | 101 +++++++++++++++++++++++++++++-------
lib/reciprocal_div.c | 2
net/sched/sch_red.c | 21 +++++++
4 files changed, 111 insertions(+), 19 deletions(-)
diff --git a/include/linux/pkt_sched.h b/include/linux/pkt_sched.h
index fb556dc..e41e0d4 100644
--- a/include/linux/pkt_sched.h
+++ b/include/linux/pkt_sched.h
@@ -181,6 +181,7 @@ enum {
TCA_RED_UNSPEC,
TCA_RED_PARMS,
TCA_RED_STAB,
+ TCA_RED_MAX_P,
__TCA_RED_MAX,
};
@@ -194,8 +195,9 @@ struct tc_red_qopt {
unsigned char Plog; /* log(P_max/(qth_max-qth_min)) */
unsigned char Scell_log; /* cell size for idle damping */
unsigned char flags;
-#define TC_RED_ECN 1
-#define TC_RED_HARDDROP 2
+#define TC_RED_ECN 1
+#define TC_RED_HARDDROP 2
+#define TC_RED_ADAPTATIVE 4
};
struct tc_red_xstats {
diff --git a/include/net/red.h b/include/net/red.h
index b72a3b8..f4e9533 100644
--- a/include/net/red.h
+++ b/include/net/red.h
@@ -5,6 +5,7 @@
#include <net/pkt_sched.h>
#include <net/inet_ecn.h>
#include <net/dsfield.h>
+#include <linux/reciprocal_div.h>
/* Random Early Detection (RED) algorithm.
=======================================
@@ -87,6 +88,29 @@
etc.
*/
+/*
+ * Adaptative RED : An Algorithm for Increasing the Robustness of RED's AQM
+ * (Sally FLoyd, Ramakrishna Gummadi, and Scott Shenker) August 2001
+ *
+ * Every 500 ms:
+ * if (avg > target and max_p <= 0.5)
+ * increase max_p : max_p += alpha;
+ * else if (avg < target and max_p >= 0.01)
+ * decrease max_p : max_p *= beta;
+ *
+ * target :[qth_min + 0.4*(qth_min - qth_max),
+ * qth_min + 0.6*(qth_min - qth_max)].
+ * alpha : min(0.01, max_p / 4)
+ * beta : 0.9
+ * max_P is a Q0.32 fixed point number (with 32 bits mantissa)
+ * max_P between 0.01 and 0.5 (1% - 50%) [ Its no longer a negative power of two ]
+ */
+#define RED_ONE_PERCENT ((u32)DIV_ROUND_CLOSEST(1ULL<<32, 100))
+
+#define MAX_P_MIN (1 * RED_ONE_PERCENT)
+#define MAX_P_MAX (50 * RED_ONE_PERCENT)
+#define MAX_P_ALPHA(val) min(MAX_P_MIN, val / 4)
+
#define RED_STAB_SIZE 256
#define RED_STAB_MASK (RED_STAB_SIZE - 1)
@@ -101,10 +125,14 @@ struct red_stats {
struct red_parms {
/* Parameters */
- u32 qth_min; /* Min avg length threshold: A scaled */
- u32 qth_max; /* Max avg length threshold: A scaled */
+ u32 qth_min; /* Min avg length threshold: Wlog scaled */
+ u32 qth_max; /* Max avg length threshold: Wlog scaled */
u32 Scell_max;
- u32 Rmask; /* Cached random mask, see red_rmask */
+ u32 max_P; /* probability, [0 .. 1.0] 32 scaled */
+ u32 max_P_reciprocal; /* reciprocal_value(max_P / qth_delta) */
+ u32 qth_delta; /* max_th - min_th */
+ u32 target_min; /* min_th + 0.4*(max_th - min_th) */
+ u32 target_max; /* min_th + 0.6*(max_th - min_th) */
u8 Scell_log;
u8 Wlog; /* log(W) */
u8 Plog; /* random number bits */
@@ -115,19 +143,22 @@ struct red_parms {
number generation */
u32 qR; /* Cached random number */
- unsigned long qavg; /* Average queue length: A scaled */
+ unsigned long qavg; /* Average queue length: Wlog scaled */
ktime_t qidlestart; /* Start of current idle period */
};
-static inline u32 red_rmask(u8 Plog)
+static inline u32 red_maxp(u8 Plog)
{
- return Plog < 32 ? ((1 << Plog) - 1) : ~0UL;
+ return Plog < 32 ? (~0U >> Plog) : ~0U;
}
+
static inline void red_set_parms(struct red_parms *p,
u32 qth_min, u32 qth_max, u8 Wlog, u8 Plog,
u8 Scell_log, u8 *stab)
{
+ int delta = qth_max - qth_min;
+
/* Reset average queue length, the value is strictly bound
* to the parameters below, reseting hurts a bit but leaving
* it might result in an unreasonable qavg for a while. --TGR
@@ -139,14 +170,29 @@ static inline void red_set_parms(struct red_parms *p,
p->qth_max = qth_max << Wlog;
p->Wlog = Wlog;
p->Plog = Plog;
- p->Rmask = red_rmask(Plog);
+ if (delta < 0)
+ delta = 1;
+ p->qth_delta = delta;
+ p->max_P = red_maxp(Plog);
+ p->max_P *= delta; /* max_P = (qth_max-qth_min)/2^Plog */
+
+ p->max_P_reciprocal = reciprocal_value(p->max_P / delta);
+
+ /* RED Adaptative target :
+ * [min_th + 0.4*(min_th - max_th),
+ * min_th + 0.6*(min_th - max_th)].
+ */
+ delta /= 5;
+ p->target_min = qth_min + 2*delta;
+ p->target_max = qth_min + 3*delta;
+
p->Scell_log = Scell_log;
p->Scell_max = (255 << Scell_log);
memcpy(p->Stab, stab, sizeof(p->Stab));
}
-static inline int red_is_idling(struct red_parms *p)
+static inline int red_is_idling(const struct red_parms *p)
{
return p->qidlestart.tv64 != 0;
}
@@ -168,7 +214,7 @@ static inline void red_restart(struct red_parms *p)
p->qcount = -1;
}
-static inline unsigned long red_calc_qavg_from_idle_time(struct red_parms *p)
+static inline unsigned long red_calc_qavg_from_idle_time(const struct red_parms *p)
{
s64 delta = ktime_us_delta(ktime_get(), p->qidlestart);
long us_idle = min_t(s64, delta, p->Scell_max);
@@ -215,7 +261,7 @@ static inline unsigned long red_calc_qavg_from_idle_time(struct red_parms *p)
}
}
-static inline unsigned long red_calc_qavg_no_idle_time(struct red_parms *p,
+static inline unsigned long red_calc_qavg_no_idle_time(const struct red_parms *p,
unsigned int backlog)
{
/*
@@ -230,7 +276,7 @@ static inline unsigned long red_calc_qavg_no_idle_time(struct red_parms *p,
return p->qavg + (backlog - (p->qavg >> p->Wlog));
}
-static inline unsigned long red_calc_qavg(struct red_parms *p,
+static inline unsigned long red_calc_qavg(const struct red_parms *p,
unsigned int backlog)
{
if (!red_is_idling(p))
@@ -239,23 +285,24 @@ static inline unsigned long red_calc_qavg(struct red_parms *p,
return red_calc_qavg_from_idle_time(p);
}
-static inline u32 red_random(struct red_parms *p)
+
+static inline u32 red_random(const struct red_parms *p)
{
- return net_random() & p->Rmask;
+ return reciprocal_divide(net_random(), p->max_P_reciprocal);
}
-static inline int red_mark_probability(struct red_parms *p, unsigned long qavg)
+static inline int red_mark_probability(const struct red_parms *p, unsigned long qavg)
{
/* The formula used below causes questions.
- OK. qR is random number in the interval 0..Rmask
+ OK. qR is random number in the interval
+ (0..1/max_P)*(qth_max-qth_min)
i.e. 0..(2^Plog). If we used floating point
arithmetics, it would be: (2^Plog)*rnd_num,
where rnd_num is less 1.
Taking into account, that qavg have fixed
- point at Wlog, and Plog is related to max_P by
- max_P = (qth_max-qth_min)/2^Plog; two lines
+ point at Wlog, two lines
below have the following floating point equivalent:
max_P*(qavg - qth_min)/(qth_max-qth_min) < rnd/qcount
@@ -315,4 +362,24 @@ static inline int red_action(struct red_parms *p, unsigned long qavg)
return RED_DONT_MARK;
}
+static inline void red_adaptative_algo(struct red_parms *p)
+{
+ unsigned long qavg;
+ u32 max_p_delta;
+
+ qavg = p->qavg;
+ if (red_is_idling(p))
+ qavg = red_calc_qavg_from_idle_time(p);
+
+ /* p->qavg is fixed point number with point at Wlog */
+ qavg >>= p->Wlog;
+
+ if (qavg > p->target_max && p->max_P <= MAX_P_MAX)
+ p->max_P += MAX_P_ALPHA(p->max_P); /* maxp = maxp + alpha */
+ else if (qavg < p->target_min && p->max_P >= MAX_P_MIN)
+ p->max_P = (p->max_P/10)*9; /* maxp = maxp * Beta */
+
+ max_p_delta = DIV_ROUND_CLOSEST(p->max_P, p->qth_delta);
+ p->max_P_reciprocal = reciprocal_value(max_p_delta);
+}
#endif
diff --git a/lib/reciprocal_div.c b/lib/reciprocal_div.c
index 6a3bd48..75510e9 100644
--- a/lib/reciprocal_div.c
+++ b/lib/reciprocal_div.c
@@ -1,5 +1,6 @@
#include <asm/div64.h>
#include <linux/reciprocal_div.h>
+#include <linux/export.h>
u32 reciprocal_value(u32 k)
{
@@ -7,3 +8,4 @@ u32 reciprocal_value(u32 k)
do_div(val, k);
return (u32)val;
}
+EXPORT_SYMBOL(reciprocal_value);
diff --git a/net/sched/sch_red.c b/net/sched/sch_red.c
index d617161..8f5a85b 100644
--- a/net/sched/sch_red.c
+++ b/net/sched/sch_red.c
@@ -39,6 +39,7 @@
struct red_sched_data {
u32 limit; /* HARD maximal queue length */
unsigned char flags;
+ struct timer_list adapt_timer;
struct red_parms parms;
struct red_stats stats;
struct Qdisc *qdisc;
@@ -161,6 +162,8 @@ static void red_reset(struct Qdisc *sch)
static void red_destroy(struct Qdisc *sch)
{
struct red_sched_data *q = qdisc_priv(sch);
+
+ del_timer_sync(&q->adapt_timer);
qdisc_destroy(q->qdisc);
}
@@ -209,6 +212,10 @@ static int red_change(struct Qdisc *sch, struct nlattr *opt)
ctl->Plog, ctl->Scell_log,
nla_data(tb[TCA_RED_STAB]));
+ del_timer(&q->adapt_timer);
+ if (ctl->flags & TC_RED_ADAPTATIVE)
+ mod_timer(&q->adapt_timer, jiffies + HZ/2);
+
if (!q->qdisc->q.qlen)
red_start_of_idle_period(&q->parms);
@@ -216,11 +223,24 @@ static int red_change(struct Qdisc *sch, struct nlattr *opt)
return 0;
}
+static inline void red_adaptative_timer(unsigned long arg)
+{
+ struct Qdisc *sch = (struct Qdisc *)arg;
+ struct red_sched_data *q = qdisc_priv(sch);
+ spinlock_t *root_lock = qdisc_lock(qdisc_root_sleeping(sch));
+
+ spin_lock(root_lock);
+ red_adaptative_algo(&q->parms);
+ mod_timer(&q->adapt_timer, jiffies + HZ/2);
+ spin_unlock(root_lock);
+}
+
static int red_init(struct Qdisc *sch, struct nlattr *opt)
{
struct red_sched_data *q = qdisc_priv(sch);
q->qdisc = &noop_qdisc;
+ setup_timer(&q->adapt_timer, red_adaptative_timer, (unsigned long)sch);
return red_change(sch, opt);
}
@@ -243,6 +263,7 @@ static int red_dump(struct Qdisc *sch, struct sk_buff *skb)
if (opts == NULL)
goto nla_put_failure;
NLA_PUT(skb, TCA_RED_PARMS, sizeof(opt), &opt);
+ NLA_PUT_U32(skb, TCA_RED_MAX_P, q->parms.max_P);
return nla_nest_end(skb, opts);
nla_put_failure:
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH net-next v2] vlan: add 802.1q netpoll support
From: Benjamin LaHaise @ 2011-12-08 16:20 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20111208153558.GE30652@kvack.org>
On Thu, Dec 08, 2011 at 10:35:58AM -0500, Benjamin LaHaise wrote:
> Please have a look at net/bridge/br_device.c:br_netpoll_setup() which is
> where the approach came from.
And I was blind. I see what you mean about not invoking the
poll_controller of the underlying device. Here's a respin. Boot tested
on the same forced_eth system.
-ben
---snip---
[PATCH net-next v2] vlan: add 802.1q netpoll support
Add netpoll support to 802.1q vlan devices. Based on the netpoll support
in the bridging code. Tested on a forced_eth device with netconsole.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
---
vlan.h | 5 +++++
vlan_dev.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 64 insertions(+)
diff --git a/net/8021q/vlan.h b/net/8021q/vlan.h
index 9fd45f3..df562b5 100644
--- a/net/8021q/vlan.h
+++ b/net/8021q/vlan.h
@@ -40,6 +40,8 @@ struct vlan_pcpu_stats {
u32 tx_dropped;
};
+struct netpoll;
+
/**
* struct vlan_dev_info - VLAN private device data
* @nr_ingress_mappings: number of ingress priority mappings
@@ -67,6 +69,9 @@ struct vlan_dev_info {
struct proc_dir_entry *dent;
struct vlan_pcpu_stats __percpu *vlan_pcpu_stats;
+#ifdef CONFIG_NET_POLL_CONTROLLER
+ struct netpoll *netpoll;
+#endif
};
static inline struct vlan_dev_info *vlan_dev_info(const struct net_device *dev)
diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
index 2b5fcde..0071829 100644
--- a/net/8021q/vlan_dev.c
+++ b/net/8021q/vlan_dev.c
@@ -33,6 +33,7 @@
#include "vlan.h"
#include "vlanproc.h"
#include <linux/if_vlan.h>
+#include <linux/netpoll.h>
/*
* Rebuild the Ethernet MAC header. This is called after an ARP
@@ -158,6 +159,8 @@ static netdev_tx_t vlan_dev_hard_start_xmit(struct sk_buff *skb,
skb_set_dev(skb, vlan_dev_info(dev)->real_dev);
len = skb->len;
+ if (netpoll_tx_running(dev))
+ return skb->dev->netdev_ops->ndo_start_xmit(skb, skb->dev);
ret = dev_queue_xmit(skb);
if (likely(ret == NET_XMIT_SUCCESS || ret == NET_XMIT_CN)) {
@@ -660,6 +663,57 @@ static struct rtnl_link_stats64 *vlan_dev_get_stats64(struct net_device *dev, st
return stats;
}
+#ifdef CONFIG_NET_POLL_CONTROLLER
+void vlan_dev_poll_controller(struct net_device *dev)
+{
+ return;
+}
+
+int vlan_dev_netpoll_setup(struct net_device *dev, struct netpoll_info *npinfo)
+{
+ struct vlan_dev_info *info = vlan_dev_info(dev);
+ struct net_device *real_dev = info->real_dev;
+ struct netpoll *netpoll;
+ int err = 0;
+
+ netpoll = kzalloc(sizeof(*netpoll), GFP_KERNEL);
+ err = -ENOMEM;
+ if (!netpoll)
+ goto out;
+
+ netpoll->dev = real_dev;
+ strlcpy(netpoll->dev_name, real_dev->name, IFNAMSIZ);
+
+ err = __netpoll_setup(netpoll);
+ if (err) {
+ kfree(netpoll);
+ goto out;
+ }
+
+ info->netpoll = netpoll;
+
+out:
+ return err;
+}
+
+void vlan_dev_netpoll_cleanup(struct net_device *dev)
+{
+ struct vlan_dev_info *info = vlan_dev_info(dev);
+ struct netpoll *netpoll = info->netpoll;
+
+ if (!netpoll)
+ return;
+
+ info->netpoll = NULL;
+
+ /* Wait for transmitting packets to finish before freeing. */
+ synchronize_rcu_bh();
+
+ __netpoll_cleanup(netpoll);
+ kfree(netpoll);
+}
+#endif /* CONFIG_NET_POLL_CONTROLLER */
+
static const struct ethtool_ops vlan_ethtool_ops = {
.get_settings = vlan_ethtool_get_settings,
.get_drvinfo = vlan_ethtool_get_drvinfo,
@@ -688,6 +742,11 @@ static const struct net_device_ops vlan_netdev_ops = {
.ndo_fcoe_get_wwn = vlan_dev_fcoe_get_wwn,
.ndo_fcoe_ddp_target = vlan_dev_fcoe_ddp_target,
#endif
+#ifdef CONFIG_NET_POLL_CONTROLLER
+ .ndo_poll_controller = vlan_dev_poll_controller,
+ .ndo_netpoll_setup = vlan_dev_netpoll_setup,
+ .ndo_netpoll_cleanup = vlan_dev_netpoll_cleanup,
+#endif
.ndo_fix_features = vlan_dev_fix_features,
};
^ permalink raw reply related
* Re: [PATCH] add netpoll support for 802.1q vlans
From: David Miller @ 2011-12-08 16:50 UTC (permalink / raw)
To: bcrl; +Cc: netdev
In-Reply-To: <20111208153558.GE30652@kvack.org>
From: Benjamin LaHaise <bcrl@kvack.org>
Date: Thu, 8 Dec 2011 10:35:58 -0500
> On Thu, Dec 08, 2011 at 12:07:30AM -0500, David Miller wrote:
>> From: Benjamin LaHaise <bcrl@kvack.org>
>> Date: Tue, 6 Dec 2011 20:04:24 -0500
>>
>> > Add netpoll support to 802.1q vlan devices. Based on the netpoll support
>> > in the bridging code. Tested on a forced_eth device with netconsole.
>> >
>> > Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
>>
>> Bridging doesn't try to invoke a downstream netpoll operation at all,
>> and neither should you.
>
> Please have a look at net/bridge/br_device.c:br_netpoll_setup() which is
> where the approach came from.
I did and I'm not talking about that part, I'm talking about the
part where you invoke the downstream device's netpoll operation.
Did you even read my entire sentence, or did you just let the
knee jerk reaction set in before you got to the end?
> Getting netconsole over vlans working is a requirement for those of us
What does this have to do with anything? I never made any even remote
argument that this change wasn't useful. I said only that you
implemented it incorrectly.
Take a deep breath and just implement the patch correctly.
^ permalink raw reply
* Re: More understanding HFSC
From: John A. Sullivan III @ 2011-12-08 18:08 UTC (permalink / raw)
To: netdev
In-Reply-To: <1322979139.3347.147.camel@denise.theartistscloset.com>
----- Original Message -----
From: "John A. Sullivan III" <jsullivan@opensourcedevel.com>
To: netdev@vger.kernel.org
Sent: Sunday, December 4, 2011 1:12:19 AM
Subject: More understanding HFSC
<snip>
A great thanks to all those who took a great deal of time to help me understand HFSC. It was particularly helpful to gain a better understanding of the relationship between fit time, virtual time, eligible time, and deadline time. I would still like to make sure I've got it in a practical way. If any one does have the time to see if the below illustration is valid, I'd appreciate it. And I also appreciate that no one may have that time :) Thanks again - John
If I've got that right, does the following illustration hold:
Let's assume we have a T1 at 1.544 Mbps (if I've remembered that
correctly). So we create the highest level child class with a ul of 1.5
Mbps to keep it slightly under the link speed to ensure we do not trip
QoS on the upstream router. Hmm . . . I suppose that principle would
not hold true in a shared environment like cable or DSL. In any event,
I think that means 1536kbits in tc-speak.
We then create five child classes named Interactive, VoIP, Video, Web,
and Bulk with the following characteristics and settings:
Bulk is basically everything that has no latency sensitivity and can
tolerate dropped packets. We assign it rt=100kbits just to keep it from
being starved when other classes are highly active but we'd like it to
access excess bandwidth a little more aggressively so we set
ls=300kbits. Does this seem reasonable?
Interactive must nowadays include VDI so it is no longer small ssh or
telnet packets. It needs to accommodate full sized Ethernet packets in
order to transfer screens with a minimum of latency. It is moderately
latency sensitive and cannot tolerate dropped packets. We want
this router to add no more than 30ms to any interactive session. Thus
we define it with:
rt umax 1514b dmax 30ms rate 300kbits ls rate 500kbits
We defined ls because we want this type of traffic to aggressively
acquire excess bandwidth if we need more than we have guaranteed.
1514b/30ms~=400kbits so we have a concave service curve.
VoIP is very latency sensitive and cannot tolerate drops. We want to
add no more than 10ms latency to the traffic flow. We are figuring that
ulaw/alaw will produce the largest packets at 234b so we define it as:
rt umax 234b dmax 10ms rate 200kbits ls rate 500kbits
234b/10ms~=203kbits so we have a slightly concave service curve. This
raises an interesting question. What if we had set dmax to 20 ms? This
would have given us a convex service curve where m1 != 0. Is that
illegal? I thought the specification said any convex curves must have m1
= 0. If it is illegal and we did this by accident, what would happen?
Video is always a problem :) We need to guarantee a large amount of
bandwidth but we also do not want to be eaten alive by video. We
characterize it as very latency sensitive but we would rather tolerate
drops than queueing and unwanted latency. Hmm . . . is queue depth
induced latency even an issue with HFSC?
In any event, we want to introduce no more than 10ms latency. The
typical frame size is 16Kb and there is no difference between the rt and
ls rate so we define the class with:
sc umax 16kbits dmax 10ms rate 400kbits
Interestingly, m1 (1.6Mbps) exceeds ul. As asked previously, is this a
problem?
Web follows the example cited in my previous email. We want the text
and css of the served web pages to load quickly. Larger, non-text data
can take a back seat. We will guess that the average text is 10KB and
we want to introduce no more than 200ms to start loading the page text.
We thus define it as:
rt umax 80kbits dmax 200ms rate 200kbits ls 400 kbits
Is this setup reasonable, practical, and reflecting a proper
understanding of HFSC or have I missed the boat entirely and need to go
to the back of the class? Thanks to anyone who has taken the time to
read this novel :) - John
^ permalink raw reply
* [PATCH net-next 2/3] bridge: rearrange fdb notifications (v2)
From: Stephen Hemminger @ 2011-12-08 17:17 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20111206230316.026471682@vyatta.com>
Pass bridge to fdb_notify so it can determine correct namespace based
on namespace of bridge rather than namespace of destination port.
Also makes next patch easier.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
---
v2 - fix compile also pass bridge to fill_info (for next patch)
--- a/net/bridge/br_fdb.c 2011-12-07 11:24:42.048686670 -0800
+++ b/net/bridge/br_fdb.c 2011-12-07 11:27:05.233784054 -0800
@@ -28,7 +28,8 @@
static struct kmem_cache *br_fdb_cache __read_mostly;
static int fdb_insert(struct net_bridge *br, struct net_bridge_port *source,
const unsigned char *addr);
-static void fdb_notify(const struct net_bridge_fdb_entry *, int);
+static void fdb_notify(struct net_bridge *br,
+ const struct net_bridge_fdb_entry *, int);
static u32 fdb_salt __read_mostly;
@@ -80,10 +81,10 @@ static void fdb_rcu_free(struct rcu_head
kmem_cache_free(br_fdb_cache, ent);
}
-static inline void fdb_delete(struct net_bridge_fdb_entry *f)
+static void fdb_delete(struct net_bridge *br, struct net_bridge_fdb_entry *f)
{
- fdb_notify(f, RTM_DELNEIGH);
hlist_del_rcu(&f->hlist);
+ fdb_notify(br, f, RTM_DELNEIGH);
call_rcu(&f->rcu, fdb_rcu_free);
}
@@ -114,7 +115,7 @@ void br_fdb_changeaddr(struct net_bridge
}
/* delete old one */
- fdb_delete(f);
+ fdb_delete(br, f);
goto insert;
}
}
@@ -144,7 +145,7 @@ void br_fdb_cleanup(unsigned long _data)
continue;
this_timer = f->updated + delay;
if (time_before_eq(this_timer, jiffies))
- fdb_delete(f);
+ fdb_delete(br, f);
else if (time_before(this_timer, next_timer))
next_timer = this_timer;
}
@@ -165,7 +166,7 @@ void br_fdb_flush(struct net_bridge *br)
struct hlist_node *h, *n;
hlist_for_each_entry_safe(f, h, n, &br->hash[i], hlist) {
if (!f->is_static)
- fdb_delete(f);
+ fdb_delete(br, f);
}
}
spin_unlock_bh(&br->hash_lock);
@@ -209,7 +210,7 @@ void br_fdb_delete_by_port(struct net_br
}
}
- fdb_delete(f);
+ fdb_delete(br, f);
skip_delete: ;
}
}
@@ -370,7 +371,7 @@ static int fdb_insert(struct net_bridge
br_warn(br, "adding interface %s with same address "
"as a received packet\n",
source->dev->name);
- fdb_delete(fdb);
+ fdb_delete(br, fdb);
}
fdb = fdb_create(head, source, addr);
@@ -378,7 +379,7 @@ static int fdb_insert(struct net_bridge
return -ENOMEM;
fdb->is_local = fdb->is_static = 1;
- fdb_notify(fdb, RTM_NEWNEIGH);
+ fdb_notify(br, fdb, RTM_NEWNEIGH);
return 0;
}
@@ -427,7 +428,7 @@ void br_fdb_update(struct net_bridge *br
if (likely(!fdb_find(head, addr))) {
fdb = fdb_create(head, source, addr);
if (fdb)
- fdb_notify(fdb, RTM_NEWNEIGH);
+ fdb_notify(br, fdb, RTM_NEWNEIGH);
}
/* else we lose race and someone else inserts
* it first, don't bother updating
@@ -448,7 +449,7 @@ static int fdb_to_nud(const struct net_b
return NUD_REACHABLE;
}
-static int fdb_fill_info(struct sk_buff *skb,
+static int fdb_fill_info(struct sk_buff *skb, const struct net_bridge *br,
const struct net_bridge_fdb_entry *fdb,
u32 pid, u32 seq, int type, unsigned int flags)
{
@@ -461,7 +462,6 @@ static int fdb_fill_info(struct sk_buff
if (nlh == NULL)
return -EMSGSIZE;
-
ndm = nlmsg_data(nlh);
ndm->ndm_family = AF_BRIDGE;
ndm->ndm_pad1 = 0;
@@ -493,9 +493,10 @@ static inline size_t fdb_nlmsg_size(void
+ nla_total_size(sizeof(struct nda_cacheinfo));
}
-static void fdb_notify(const struct net_bridge_fdb_entry *fdb, int type)
+static void fdb_notify(struct net_bridge *br,
+ const struct net_bridge_fdb_entry *fdb, int type)
{
- struct net *net = dev_net(fdb->dst->dev);
+ struct net *net = dev_net(br->dev);
struct sk_buff *skb;
int err = -ENOBUFS;
@@ -503,7 +504,7 @@ static void fdb_notify(const struct net_
if (skb == NULL)
goto errout;
- err = fdb_fill_info(skb, fdb, 0, 0, type, 0);
+ err = fdb_fill_info(skb, br, fdb, 0, 0, type, 0);
if (err < 0) {
/* -EMSGSIZE implies BUG in fdb_nlmsg_size() */
WARN_ON(err == -EMSGSIZE);
@@ -540,7 +541,7 @@ int br_fdb_dump(struct sk_buff *skb, str
if (idx < cb->args[0])
goto skip;
- if (fdb_fill_info(skb, f,
+ if (fdb_fill_info(skb, br, f,
NETLINK_CB(cb->skb).pid,
cb->nlh->nlmsg_seq,
RTM_NEWNEIGH,
@@ -574,7 +575,7 @@ static int fdb_add_entry(struct net_brid
fdb = fdb_create(head, source, addr);
if (!fdb)
return -ENOMEM;
- fdb_notify(fdb, RTM_NEWNEIGH);
+ fdb_notify(br, fdb, RTM_NEWNEIGH);
} else {
if (flags & NLM_F_EXCL)
return -EEXIST;
@@ -590,7 +591,7 @@ static int fdb_add_entry(struct net_brid
fdb->is_local = fdb->is_static = 0;
fdb->updated = fdb->used = jiffies;
- fdb_notify(fdb, RTM_NEWNEIGH);
+ fdb_notify(br, fdb, RTM_NEWNEIGH);
}
return 0;
@@ -670,7 +671,7 @@ static int fdb_delete_by_addr(struct net
if (!fdb)
return -ENOENT;
- fdb_delete(fdb);
+ fdb_delete(p->br, fdb);
return 0;
}
^ permalink raw reply
* [PATCH net-next 3/3] bridge: add local MAC address to forwarding table (v2)
From: Stephen Hemminger @ 2011-12-08 17:17 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20111206230316.149318632@vyatta.com>
If user has configured a MAC address that is not one of the existing
ports of the bridge, then we need to add a special entry in the forwarding
table. This forwarding table entry has no outgoing port so it has to be
treated a little differently. The special entry is reported by the netlink
interface with ifindex of bridge, but ignored by the old interface since there
is no usable way to put it in the ABI.
Reported-by: Koki Sanagi <sanagi.koki@jp.fujitsu.com>
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
---
v2 - set ifindex in NEWNEIGH to bridge instead of zero
--- a/net/bridge/br_device.c 2011-12-07 10:54:21.616301619 -0800
+++ b/net/bridge/br_device.c 2011-12-07 11:28:14.490336392 -0800
@@ -170,8 +170,11 @@ static int br_set_mac_address(struct net
return -EINVAL;
spin_lock_bh(&br->lock);
- memcpy(dev->dev_addr, addr->sa_data, ETH_ALEN);
- br_stp_change_bridge_id(br, addr->sa_data);
+ if (compare_ether_addr(dev->dev_addr, addr->sa_data)) {
+ memcpy(dev->dev_addr, addr->sa_data, ETH_ALEN);
+ br_fdb_change_mac_address(br, addr->sa_data);
+ br_stp_change_bridge_id(br, addr->sa_data);
+ }
br->flags |= BR_SET_MAC_ADDR;
spin_unlock_bh(&br->lock);
--- a/net/bridge/br_fdb.c 2011-12-07 11:27:05.233784054 -0800
+++ b/net/bridge/br_fdb.c 2011-12-07 11:28:44.782581738 -0800
@@ -127,6 +127,18 @@ void br_fdb_changeaddr(struct net_bridge
spin_unlock_bh(&br->hash_lock);
}
+void br_fdb_change_mac_address(struct net_bridge *br, const u8 *newaddr)
+{
+ struct net_bridge_fdb_entry *f;
+
+ /* If old entry was unassociated with any port, then delete it. */
+ f = __br_fdb_get(br, br->dev->dev_addr);
+ if (f && f->is_local && !f->dst)
+ fdb_delete(br, f);
+
+ fdb_insert(br, NULL, newaddr);
+}
+
void br_fdb_cleanup(unsigned long _data)
{
struct net_bridge *br = (struct net_bridge *)_data;
@@ -250,7 +262,7 @@ int br_fdb_test_addr(struct net_device *
ret = 0;
else {
fdb = __br_fdb_get(port->br, addr);
- ret = fdb && fdb->dst->dev != dev &&
+ ret = fdb && fdb->dst && fdb->dst->dev != dev &&
fdb->dst->state == BR_STATE_FORWARDING;
}
rcu_read_unlock();
@@ -282,6 +294,10 @@ int br_fdb_fillbuf(struct net_bridge *br
if (has_expired(br, f))
continue;
+ /* ignore pseudo entry for local MAC address */
+ if (!f->dst)
+ continue;
+
if (skip) {
--skip;
continue;
@@ -468,7 +484,7 @@ static int fdb_fill_info(struct sk_buff
ndm->ndm_pad2 = 0;
ndm->ndm_flags = 0;
ndm->ndm_type = 0;
- ndm->ndm_ifindex = fdb->dst->dev->ifindex;
+ ndm->ndm_ifindex = fdb->dst ? fdb->dst->dev->ifindex : br->dev->ifindex;
ndm->ndm_state = fdb_to_nud(fdb);
NLA_PUT(skb, NDA_LLADDR, ETH_ALEN, &fdb->addr);
--- a/net/bridge/br_forward.c 2011-12-07 10:54:21.616301619 -0800
+++ b/net/bridge/br_forward.c 2011-12-07 11:28:14.490336392 -0800
@@ -98,7 +98,7 @@ static void __br_forward(const struct ne
/* called with rcu_read_lock */
void br_deliver(const struct net_bridge_port *to, struct sk_buff *skb)
{
- if (should_deliver(to, skb)) {
+ if (to && should_deliver(to, skb)) {
__br_deliver(to, skb);
return;
}
--- a/net/bridge/br_private.h 2011-12-07 10:54:21.616301619 -0800
+++ b/net/bridge/br_private.h 2011-12-07 11:28:14.490336392 -0800
@@ -348,6 +348,7 @@ extern void br_fdb_fini(void);
extern void br_fdb_flush(struct net_bridge *br);
extern void br_fdb_changeaddr(struct net_bridge_port *p,
const unsigned char *newaddr);
+extern void br_fdb_change_mac_address(struct net_bridge *br, const u8 *newaddr);
extern void br_fdb_cleanup(unsigned long arg);
extern void br_fdb_delete_by_port(struct net_bridge *br,
const struct net_bridge_port *p, int do_all);
^ permalink raw reply
* Re: [PATCH net-next 0/3] bridge forwarding database patches
From: Stephen Hemminger @ 2011-12-08 17:18 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20111207.135015.789061435692305493.davem@davemloft.net>
On Wed, 07 Dec 2011 13:50:15 -0500 (EST)
David Miller <davem@davemloft.net> wrote:
> From: Stephen Hemminger <shemminger@vyatta.com>
> Date: Tue, 06 Dec 2011 15:02:23 -0800
>
> > These three patches address issue of receiving packets when
> > bridge mac address is modified.
>
Resent 2 and 3, patch 1 is unchanged.
^ permalink raw reply
* Re: [PATCH net-next] sch_red: Adaptative RED AQM
From: Stephen Hemminger @ 2011-12-08 17:21 UTC (permalink / raw)
To: Eric Dumazet
Cc: Dave Taht, David Miller, linux-wireless,
netdev-u79uwXL29TY76Z2rM5mHXA, Hagen Paul Pfeifer, bloat-devel,
bloat
In-Reply-To: <1323360363.2521.30.camel@edumazet-HP-Compaq-6005-Pro-SFF-PC>
On Thu, 08 Dec 2011 17:06:03 +0100
Eric Dumazet <eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> Changes against our RED implementation are :
>
> max_p is no longer a negative power of two (1/(2^Plog)), but a Q0.32
> fixed point number, to allow full range described in Adatative paper.
>
> To deliver a random number, we now use a reciprocal divide (thats really
> a multiply), but this operation is done once per marked/droped packet
> when in RED_BETWEEN_TRESH window, so added cost (compared to previous
> AND operation) is near zero.
>
> dump operation gives current max_p value in a new TCA_RED_MAX_P
> attribute.
>
> Example on a 10Mbit link :
>
> tc qdisc add dev $DEV parent 1:1 handle 10: est 1sec 8sec red \
> limit 400000 min 30000 max 90000 avpkt 1000 \
> burst 55 ecn adaptative bandwidth 10Mbit
>
> # tc -s -d qdisc show dev eth3
> ...
> qdisc red 10: parent 1:1 limit 400000b min 30000b max 90000b ecn
> adaptative ewma 5 max_p=0.113335 Scell_log 15
> Sent 50414282 bytes 34504 pkt (dropped 35, overlimits 1392 requeues 0)
> rate 9749Kbit 831pps backlog 72056b 16p requeues 0
> marked 1357 early 35 pdrop 0 other 0
>
>
> Signed-off-by: Eric Dumazet <eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Is this backward compatible for users that don't specify
an adaptive parameter.
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: BCM43224 hanging [3.2.0-rc4-00248-gb835c0f]
From: Arend van Spriel @ 2011-12-08 17:34 UTC (permalink / raw)
To: Nico Schottelius, LKML, netdev, b43-dev, Greg KH
In-Reply-To: <20111207190821.GA1775@schottelius.org>
On 12/07/2011 08:08 PM, Nico Schottelius wrote:
> Update:
>
> The hang situation always happens, when the nic does *not* have
> a connection. And it still covers all network processes (i.e.
> including sudo, wpa_supplicant, postfix, etc.).
>
> Starting up the computer, having wpa_supplicant connect successfully,
> everything works fine.
>
> Until the connection is lost or I issue "select_network 4", which
> changes to another network.
>
> The problem also exists, if I startup wpa_supplicant and there
> is no network connection available.
>
> Verified that this bug exists in my pathced 3.1.0-rc6-g443452b
> as well as in 3.2.0-rc4-00248-gb835c0f.
>
> This is rather nasty, because it forces me to reboot as soon as
> I've started up wlan by accident.
>
> Is there a workaround available somewhere (besides using USB-LAN)?
>
> Attached are output of 3.1.0-rc6-g443452b, which is the same for
> all failing versions.
>
> Cheers,
>
> Nico
Could you try and see what happens when you kill wpa_supplicant.
Gr. AvS
^ permalink raw reply
* Re: 3.0.8 kernel : NULL ptr deref in skb_queue_purge()
From: Greg KH @ 2011-12-08 18:02 UTC (permalink / raw)
To: Grant Grundler
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, linux-usb-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <CANEJEGtJ3UmFNyui_SaZ6NF5FFVjZ+_UBg1RC2eif5Lu1YKDsQ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
On Wed, Dec 07, 2011 at 02:40:49PM -0800, Grant Grundler wrote:
> Hi,
> I'm testing asix (USB 100BT ethernet adapter with AX88772) driver
> initialization (and shut down) paths and reproduced a
> "skb_queue_purge" panic 3 times after a few hundred/thousand
> iterations of rmmod/modprobe. I'm inclined to believe
> skb_queue_purge() is a victim and not a culprit here.
>
> I don't know if all 3 "spontaneous reboots" I've seen have the same
> stack trace as the one I have a record for:
Have you tried this on 3.1, and especially, 3.2-rc? A number of asix
patches have gone into the 3.2-rc series, perhaps they might have
resolved this problem already?
thanks,
greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox