cfg80211: Add support for aborting an ongoing scan
Implement new functionality for aborting an ongoing scan.
Add NL80211_CMD_ABORT_SCAN to the nl80211 interface. After
aborting the scan, driver shall provide the scan status by
calling cfg80211_scan_done().
Reviewed-by: Jouni Malinen <[email protected]> Signed-off-by: Vidyullatha Kanchanapally <[email protected]> Signed-off-by: Sunil Dutt <[email protected]>
[change command to take wdev instead of netdev so that it
can be used on p2p-device scans] Signed-off-by: Johannes Berg <[email protected]>
Johannes Berg [Thu, 26 Nov 2015 16:59:27 +0000 (17:59 +0100)]
mac80211_hwsim: do not actively scan DFS channels
We had another change to fix this in mac80211, but the hwsim
"hardware" scan should also be fixed. Obviously this one isn't
important since it's not real hardware, but we'd better be
consistent.
Eliad Peller [Tue, 17 Nov 2015 08:24:37 +0000 (10:24 +0200)]
mac80211: add atomic uploaded keys iterator
add ieee80211_iter_keys_rcu() to iterate over uploaded
keys in atomic context (when rcu is locked)
The station removal code removes the keys only after
calling synchronize_net(), so it's not safe to iterate
the keys at this point (and postponing the actual key
deletion with call_rcu() might result in some
badly-ordered ops calls).
Add a flag to indicate a station is being removed,
and skip the configured keys if it's set.
mac80211: allow the driver to send EOSP when needed
This can happen when the driver needs to send less frames
than expected and then needs to close the SP.
Mac80211 still needs to set the more_data properly based
on its buffer state (ps_tx_buffer and buffered frames on
other TIDs).
To that end, refactor the code that delivers frames upon
uAPSD trigger frames to be able to get only the more_data
bit without actually delivering those frames in case the
driver is just asking to set a NDP with EOSP and MORE_DATA
bit properly set.
Luca Coelho [Mon, 9 Nov 2015 10:17:37 +0000 (12:17 +0200)]
nl80211: clarify NL80211_ATTR_SCHED_SCAN_DELAY usage with net-detect
In this attribute's documentation, it was not clear whether the delay
started counting when WoWLAN net-detect was enabled or when the system
was suspended. The correct answer is that it starts when the system
suspends (which is when, in practice, the scan is scheduled). Clarify
that in the nl80211.h documentation.
Johannes Berg [Fri, 6 Nov 2015 11:34:24 +0000 (12:34 +0100)]
mac80211: remove string from unaligned packet warning
This really should never happen except very early in the process
of bringing up a new driver, at which point you'll have to add
more debugging in the driver and this string isn't useful. Remove
it and save some size (when it's even compiled in.)
Johannes Berg [Fri, 6 Nov 2015 11:02:31 +0000 (12:02 +0100)]
lib80211: ratelimit key index mismatch
This indicates a driver key selection issue, but even then there's
no point in printing it all the time, so ratelimit it. Also remove
the priv pointer from it -- people debugging will only have a single
device anyway and it's useless as anything but a cookie.
Johannes Berg [Fri, 23 Oct 2015 07:50:03 +0000 (09:50 +0200)]
cfg80211: replace ieee80211_ie_split() with an inline
The function is a very simple wrapper around another one,
just adds a few default parameters, so replace it with a
static inline instead of using EXPORT_SYMBOL, reducing
the module size slightly.
Johannes Berg [Thu, 22 Oct 2015 19:27:47 +0000 (22:27 +0300)]
cfg80211: add complete data to station add/change tracing
Complete the tracepoint with the missing data - it's not printed
by default (a lot of it is dynamic arrays) but will be recorded
and be available during post-processing.
Johannes Berg [Thu, 22 Oct 2015 15:35:19 +0000 (17:35 +0200)]
mac80211: allow driver to prevent two stations w/ same address
Some devices or drivers cannot deal with having the same station
address for different virtual interfaces, say as a client to two
virtual AP interfaces. Rather than requiring each driver with a
limitation like that to enforce it, add a hardware flag for it.
Linus Torvalds [Thu, 3 Dec 2015 23:45:16 +0000 (15:45 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block
Pull block fixes from Jens Axboe:
"A collection of fixes from this series. The most important here is a
regression fix for an issue that some folks would hit in blk-merge.c,
and the NVMe queue depth limit for the screwed up Apple "nvme"
controller.
In more detail, this pull request contains:
- a set of fixes for null_blk, including a fix for a few corner cases
where we could hang the device. From Arianna and Paolo.
- lightnvm:
- A build improvement from Keith.
- Update the qemu pci id detection from Matias.
- Error handling fixes for leaks and other little fixes from
Sudip and Wenwei.
- fix from Eric where BLKRRPART would not return EBUSY for whole
device mounts, only when partitions were mounted.
- fix from Jan Kara, where EOF O_DIRECT reads would return
negatively.
- remove check for rq_mergeable() when checking limits for cloned
requests. The check doesn't make any sense. It's assuming that
since NOMERGE is set on the request that we don't have to
recalculate limits since the request didn't change, but that's not
true if the request has been redirected. From Hannes.
- correctly get the bio front segment value set for single segment
bio's, fixing a BUG() in blk-merge. From Ming"
* 'for-linus' of git://git.kernel.dk/linux-block:
nvme: temporary fix for Apple controller reset
null_blk: change type of completion_nsec to unsigned long
null_blk: guarantee device restart in all irq modes
null_blk: set a separate timer for each command
blk-merge: fix computing bio->bi_seg_front_size in case of single segment
direct-io: Fix negative return from dio read beyond eof
block: Always check queue limits for cloned requests
lightnvm: missing nvm_lock acquire
lightnvm: unconverted ppa returned in get_bb_tbl
lightnvm: refactor and change vendor id for qemu
lightnvm: do device max sectors boundary check first
lightnvm: fix ioctl memory leaks
lightnvm: free memory when gennvm register fails
lightnvm: Simplify config when disabled
Return EBUSY from BLKRRPART for mounted whole-dev fs
Linus Torvalds [Thu, 3 Dec 2015 23:23:17 +0000 (15:23 -0800)]
Merge tag 'trace-v4.4-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fix from Steven Rostedt:
"During the merge window I added a new file that is used to filter
trace events on pids. It filters all events where only tasks with
their pid in that file exists. It also handles the sched_switch and
sched_wakeup trace events where the current task does not have its pid
in the file, but the task either being switched to or awaken does.
Unfortunately, I forgot about sched_wakeup_new and sched_waking. Both
of these tracepoints use the same class as the sched_wakeup
tracepoint, and they too should be included in what gets filtered by
the set_event_pid file"
* tag 'trace-v4.4-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Add sched_wakeup_new and sched_waking tracepoints for pid filter
David S. Miller [Thu, 3 Dec 2015 20:56:22 +0000 (15:56 -0500)]
Merge tag 'mac80211-for-davem-2015-12-02' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
Johannes Berg says:
====================
A small set of fixes for 4.4:
* fix scanning in mac80211 to not actively scan radar
channels (from Antonio)
* fix uninitialized variable in remain-on-channel that
could lead to treating frame TX as remain-on-channel
and not sending the frame at all
* remove NL80211_FEATURE_FULL_AP_CLIENT_STATE again, it
was broken and needs more work, we'll enable it later
* fix call_rcu() induced use-after-reset/free in mesh
(that was suddenly causing issues in certain tests)
* always request block-ack window size 64 as we found
some APs will otherwise crash (really ...)
* fix P2P-Device teardown sequence to avoid restarting
with uninitialized data
====================
Jérôme Pouiller [Thu, 3 Dec 2015 09:02:35 +0000 (10:02 +0100)]
net: phy: reset only targeted phy
It is possible to address another chip on same MDIO bus. The case is
correctly handled for media advertising. It is taken into account
only if mii_data->phy_id == phydev->addr. However, this condition
was missing for reset case.
Stephen Boyd [Thu, 3 Dec 2015 07:55:15 +0000 (23:55 -0800)]
stmmac: ipq806x: Return error values instead of pointers
Typically we return error pointers when we want to use those
pointers in the non-error case, but this function is just
returning error pointers or NULL for success. Change the style to
plain int to follow normal kernel coding styles.
Jon Paul Maloy [Wed, 2 Dec 2015 20:19:37 +0000 (15:19 -0500)]
tipc: fix node reference count bug
Commit 5405ff6e15f40f2f ("tipc: convert node lock to rwlock")
introduced a bug to the node reference counter handling. When a
message is successfully sent in the function tipc_node_xmit(),
we return directly after releasing the node lock, instead of
continuing and decrementing the node reference counter as we
should do.
David S. Miller [Thu, 3 Dec 2015 20:18:10 +0000 (15:18 -0500)]
Merge branch 'mvneta-ethtool-autoneg'
Stas Sergeev says:
====================
mvneta: implement ethtool autonegotiation control
These 2 patches add an ability to control the
autonegotiation via ethtool. For example:
ethtool -s eth0 autoneg off
ethtool -s eth0 autoneg on
This is needed if you want to connect the mvneta's MII
to different switches or PHYs: the ones the do support
the in-band status, and the ones that do not.
====================
Stas Sergeev [Wed, 2 Dec 2015 17:35:11 +0000 (20:35 +0300)]
mvneta: implement ethtool autonegotiation control
This patch allows to do
ethtool -s eth0 autoneg off
ethtool -s eth0 autoneg on
to disable or enable autonegotiation at run-time.
Without that functionality, the only way to control the autonegotiation
is to modify the device tree.
This is needed if you plan to use the same kernel with
different ethernet switches, the ones that support the in-band
status and the ones that not.
Thierry Reding [Wed, 2 Dec 2015 16:30:29 +0000 (17:30 +0100)]
net: mv643xx: Use platform_register/unregister_drivers()
These new helpers simplify implementing multi-driver modules and
properly handle failure to register one driver by unregistering all
previously registered drivers.
Thierry Reding [Wed, 2 Dec 2015 16:30:28 +0000 (17:30 +0100)]
net: mpc52xx: Use platform_register/unregister_drivers()
These new helpers simplify implementing multi-driver modules and
properly handle failure to register one driver by unregistering all
previously registered drivers.
Thierry Reding [Wed, 2 Dec 2015 16:30:27 +0000 (17:30 +0100)]
net: bcm63xx: Use platform_register/unregister_drivers()
These new helpers simplify implementing multi-driver modules and
properly handle failure to register one driver by unregistering all
previously registered drivers.
Thierry Reding [Wed, 2 Dec 2015 16:30:26 +0000 (17:30 +0100)]
net: bfin_mac: Use platform_register/unregister_drivers()
These new helpers simplify implementing multi-driver modules and
properly handle failure to register one driver by unregistering all
previously registered drivers.
Michael Chan [Wed, 2 Dec 2015 06:54:08 +0000 (01:54 -0500)]
bnxt_en: Setup uc_list mac filters after resetting the chip.
Call bnxt_cfg_rx_mode() in bnxt_init_chip() to setup uc_list and
mc_list mac address filters. Before the patch, uc_list is not
setup again after chip reset (such as ethtool ring size change)
and macvlans don't work any more after that.
Modify bnxt_cfg_rx_mode() to return error codes appropriately so
that the init chip sequence can detect any failures.
Jeffrey Huang [Wed, 2 Dec 2015 06:54:07 +0000 (01:54 -0500)]
bnxt_en: enforce proper storing of MAC address
For PF, the bp->pf.mac_addr always holds the permanent MAC
addr assigned by the HW. For VF, the bp->vf.mac_addr always
holds the administrator assigned VF MAC addr. The random
generated VF MAC addr should never get stored to bp->vf.mac_addr.
This way, when the VF wants to change the MAC address, we can tell
if the adminstrator has already set it and disallow the VF from
changing it.
v2: Fix compile error if CONFIG_BNXT_SRIOV is not set.
Jeffrey Huang [Wed, 2 Dec 2015 06:54:06 +0000 (01:54 -0500)]
bnxt_en: Fixed incorrect implementation of ndo_set_mac_address
The existing ndo_set_mac_address only copies the new MAC addr
and didn't set the new MAC addr to the HW. The correct way is
to delete the existing default MAC filter from HW and add
the new one. Because of RFS filters are also dependent on the
default mac filter l2 context, the driver must go thru
close_nic() to delete the default MAC and RFS filters, then
open_nic() to set the default MAC address to HW.
David S. Miller [Thu, 3 Dec 2015 20:05:56 +0000 (15:05 -0500)]
Merge branch 'vsock-virtio'
Stefan Hajnoczi says:
====================
Add virtio transport for AF_VSOCK
v2:
* Rebased onto Linux v4.4-rc2
* vhost: Refuse to assign reserved CIDs
* vhost: Refuse guest CID if already in use
* vhost: Only accept correctly addressed packets (no spoofing!)
* vhost: Support flexible rx/tx descriptor layout
* vhost: Add missing total_tx_buf decrement
* virtio_transport: Fix total_tx_buf accounting
* virtio_transport: Add virtio_transport global mutex to prevent races
* common: Notify other side of SOCK_STREAM disconnect (fixes shutdown
semantics)
* common: Avoid recursive mutex_lock(tx_lock) for write_space (fixes deadlock)
* common: Define VIRTIO_VSOCK_TYPE_STREAM/DGRAM hardware interface constants
* common: Define VIRTIO_VSOCK_SHUTDOWN_RCV/SEND hardware interface constants
* common: Fix peer_buf_alloc inheritance on child socket
This patch series adds a virtio transport for AF_VSOCK (net/vmw_vsock/).
AF_VSOCK is designed for communication between virtual machines and
hypervisors. It is currently only implemented for VMware's VMCI transport.
This series implements the proposed virtio-vsock device specification from
here:
http://comments.gmane.org/gmane.comp.emulators.virtio.devel/855
Most of the work was done by Asias He and Gerd Hoffmann a while back. I have
picked up the series again.
The QEMU userspace changes are here:
https://github.com/stefanha/qemu/commits/vsock
Why virtio-vsock?
-----------------
Guest<->host communication is currently done over the virtio-serial device.
This makes it hard to port sockets API-based applications and is limited to
static ports.
virtio-vsock uses the sockets API so that applications can rely on familiar
SOCK_STREAM and SOCK_DGRAM semantics. Applications on the host can easily
connect to guest agents because the sockets API allows multiple connections to
a listen socket (unlike virtio-serial). This simplifies the guest<->host
communication and eliminates the need for extra processes on the host to
arbitrate virtio-serial ports.
Overview
--------
This series adds 3 pieces:
1. virtio_transport_common.ko - core virtio vsock code that uses vsock.ko
2. virtio_transport.ko - guest driver
3. drivers/vhost/vsock.ko - host driver
Howto
-----
The following kernel options are needed:
CONFIG_VSOCKETS=y
CONFIG_VIRTIO_VSOCKETS=y
CONFIG_VIRTIO_VSOCKETS_COMMON=y
CONFIG_VHOST_VSOCK=m
Launch QEMU as follows:
# qemu ... -device vhost-vsock-pci,id=vhost-vsock-pci0,guest-cid=3
Guest and host can communicate via AF_VSOCK sockets. The host's CID (address)
is 2 and the guest is automatically assigned a CID (use VMADDR_CID_ANY (-1) to
bind to it).
Status
------
There are a few design changes I'd like to make to the virtio-vsock device:
1. The 3-way handshake isn't necessary over a reliable transport (virtqueue).
Spoofing packets is also impossible so the security aspects of the 3-way
handshake (including syn cookie) add nothing. The next version will have a
single operation to establish a connection.
2. Credit-based flow control doesn't work for SOCK_DGRAM since multiple clients
can transmit to the same listen socket. There is no way for the clients to
coordinate buffer space with each other fairly. The next version will drop
credit-based flow control for SOCK_DGRAM and only rely on best-effort
delivery. SOCK_STREAM still has guaranteed delivery.
3. In the next version only the host will be able to establish connections
(i.e. to connect to a guest agent). This is for security reasons since
there is currently no ability to provide host services only to certain
guests. This also matches how AF_VSOCK works on modern VMware hypervisors.
====================
Roopa Prabhu [Wed, 2 Dec 2015 06:18:11 +0000 (22:18 -0800)]
mpls: support for dead routes
Adds support for RTNH_F_DEAD and RTNH_F_LINKDOWN flags on mpls
routes due to link events. Also adds code to ignore dead
routes during route selection.
Unlike ip routes, mpls routes are not deleted when the route goes
dead. This is current mpls behaviour and this patch does not change
that. With this patch however, routes will be marked dead.
dead routes are not notified to userspace (this is consistent with ipv4
routes).
dead routes:
-----------
$ip -f mpls route show
100
nexthop as to 200 via inet 10.1.1.2 dev swp1
nexthop as to 700 via inet 10.1.1.6 dev swp2
$ip link set dev swp1 down
$ip link show dev swp1
4: swp1: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast state DOWN mode
DEFAULT group default qlen 1000
link/ether 00:02:00:00:00:01 brd ff:ff:ff:ff:ff:ff
$ip -f mpls route show
100
nexthop as to 200 via inet 10.1.1.2 dev swp1 dead linkdown
nexthop as to 700 via inet 10.1.1.6 dev swp2
linkdown routes:
----------------
$ip -f mpls route show
100
nexthop as to 200 via inet 10.1.1.2 dev swp1
nexthop as to 700 via inet 10.1.1.6 dev swp2
$ip link show dev swp1
4: swp1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
state UP mode DEFAULT group default qlen 1000
link/ether 00:02:00:00:00:01 brd ff:ff:ff:ff:ff:ff
/* carrier goes down */
$ip link show dev swp1
4: swp1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast
state DOWN mode DEFAULT group default qlen 1000
link/ether 00:02:00:00:00:01 brd ff:ff:ff:ff:ff:ff
$ip -f mpls route show
100
nexthop as to 200 via inet 10.1.1.2 dev swp1 linkdown
nexthop as to 700 via inet 10.1.1.6 dev swp2
net: lpc_eth: remove irq > NR_IRQS check from probe()
If the driver is used on an ARM platform with SPARSE_IRQ defined,
semantics of NR_IRQS is different (minimal value of virtual irqs) and
by default it is set to 16, see arch/arm/include/asm/irq.h.
This value may be less than the actual number of virtual irqs, which
may break the driver initialization. The check removal allows to use
the driver on such a platform, and, if irq controller driver works
correctly, the check is not needed on legacy platforms.
Fixes a runtime problem:
lpc-eth 31060000.ethernet: error getting resources.
lpc_eth: lpc-eth: not found (-6).
Simon Horman [Wed, 2 Dec 2015 05:58:33 +0000 (14:58 +0900)]
ravb: add device tree support for r8a779[123]
Simply document new compatibility strings.
As a previous patch adds a generic R-Car Gen2 compatibility string
there appears to be no need for a driver updates.
Simon Horman [Wed, 2 Dec 2015 05:58:32 +0000 (14:58 +0900)]
ravb: add fallback compatibility strings
Add fallback compatibility strings for R-Car Gen 2 & 3 SoC Families.
This is in keeping with the fallback scheme being adopted wherever appropriate
for drivers for Renesas SoCs.
mq/mqprio have their own ways to report qlen/drops by folding stats on
all their queues, with appropriate locking.
A second problem is that qdisc_tree_decrease_qlen() calls qdisc_lookup()
without proper locking : concurrent qdisc updates could corrupt the list
that qdisc_match_from_root() parses to find a qdisc given its handle.
Fix first problem adding a TCQ_F_NOPARENT qdisc flag that
qdisc_tree_decrease_qlen() can use to abort its tree traversal,
as soon as it meets a mq/mqprio qdisc children.
Second problem can be fixed by RCU protection.
Qdisc are already freed after RCU grace period, so qdisc_list_add() and
qdisc_list_del() simply have to use appropriate rcu list variants.
A future patch will add a per struct netdev_queue list anchor, so that
qdisc_tree_decrease_qlen() can have more efficient lookups.
Phil Sutter [Tue, 1 Dec 2015 21:45:15 +0000 (22:45 +0100)]
net: ipv6: restrict hop_limit sysctl setting to range [1; 255]
Setting a value bigger than 255 resulted in using only the lower eight
bits of that value as it is assigned to the u8 header field. To avoid
this unexpected result, reject such values.
Setting a value of zero is technically possible, but hosts receiving
such a packet have to treat it like hop_limit was set to one, according
to RFC2460. Therefore I don't see a use-case for that.
Setting a route's hop_limit to zero in iproute2 means to use the sysctl
default, which is not the case here: Setting e.g.
net.conf.eth0.hop_limit=0 will not make the kernel use
net.conf.all.hop_limit for outgoing packets on eth0. To avoid these
kinds of confusion, reject zero.
Paolo Abeni [Tue, 1 Dec 2015 17:33:36 +0000 (18:33 +0100)]
openvswitch: fix hangup on vxlan/gre/geneve device deletion
Each openvswitch tunnel vport (vxlan,gre,geneve) holds a reference
to the underlying tunnel device, but never released it when such
device is deleted.
Deleting the underlying device via the ip tool cause the kernel to
hangup in the netdev_wait_allrefs() loop.
This commit ensure that on device unregistration dp_detach_port_notify()
is called for all vports that hold the device reference, properly
releasing it.
Fixes: 614732eaa12d ("openvswitch: Use regular VXLAN net_device device") Fixes: b2acd1dc3949 ("openvswitch: Use regular GRE net_device instead of vport") Fixes: 6b001e682e90 ("openvswitch: Use Geneve device.") Signed-off-by: Paolo Abeni <[email protected]> Acked-by: Flavio Leitner <[email protected]> Acked-by: Pravin B Shelar <[email protected]> Signed-off-by: David S. Miller <[email protected]>
David S. Miller [Thu, 3 Dec 2015 19:17:51 +0000 (14:17 -0500)]
Merge branch 'netronome-NFP4000-NFP6000'
Jakub Kicinski says:
====================
Netronome NFP4000/NFP6000 NIC VF driver
This patchset adds support for VFs of Netronome's NFP-4000 and NFP-6000
based NICs. We are currently also preparing the submission for the PF
driver, but it is not quite ready yet. The PF driver can be found on
GitHub:
David S. Miller [Thu, 3 Dec 2015 17:11:00 +0000 (12:11 -0500)]
Merge branch '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue
Jeff Kirsher says:
====================
40GbE Intel Wired LAN Driver Updates 2015-12-03
This series contains updates to i40e and i40evf only.
Mitch updates the i40evf driver by increasing the maximum number of queues,
since future devices will allow for more queue pairs. Cleans up a
duplicate printing of the driver info string done in init, since it is
already done in probe. Cleaned up the several allocations which did
not need to be at atomic level, where GFP_KERNEL would work just fine.
Then makes i40e_sync_vsi_filters() a more mature function, make having
a common exit point so it will properly release the busy lock on the VSI
and propagate errors to the callers. Then does some whitespace
housekeeping in i40evf.
Kiran moves and updates the detection/recovery of transmit queue hang code
to service_task from tx_timeout function. Also fixed memory leak when
users program flow-director filter using ethtool (sideband filter
programming), the cause being the check of 'tx_buffer->skb' was preventing
'raw_buf' from being freed as part of the cleanup.
Jesse enabled the ability to turn off/on packet split using ethtool priv
flags. Then does some housekeeping for both the i40e and i40evf drivers
which includes: remove unused/useless code, correct whitespace, remove
duplicate #include, fix incorrect comment, etc...
Neerav cleans up functions to gather Flow Control Rx XOFF stats, since
the recent change in the driver logic for checking transmit hang has been
moved, so these functions do not do anything meaningful any longer.
====================
This patchset introduces the support of Ethernet SRIOV in ConnectX-4
family of 100G Ethernet NICs.
Some features are still missing, but all the basic SRIOV functionalities
are there already.
Basic Introduction:
ConnectX-4 HW architecture provides two kinds of underlying HW switches.
MPFS (Multi Physical Function Switch) or L2 Table in Software terms:
The HCA has one MPFS switch per physical port, this switch is responsible
of forwarding Unicast traffic to the various overlying Physical Functions (PFs).
Multicast traffic is flooded amongst all the PFs, Each PF can request to
forward a unicast MAC to its E-Switch Uplink vport (which we will cover later)
through SET_L2_TABLE_ENTRY HW command.
MPFS has five ports, four are connected to PFs (one for each) and one is connected
directly to the Physical Port (Physical Link).
E-Switch (Ethernet Switch):
The HCA has one per physical function. The main responsibility of this component is
to forward Unicast/Multicast and vlan tagged/untagged traffic to the various
Virtual Functions (VFs) allocated by the PF. Unlike MPFS, the PF needs to explicitly
create the E-Switch FDB table, Which is a HW flow table managed by the PF driver
whenever vport_group_manager capability bit is set for this PF.
E-Switch has Virtual Ports (vports) entities as its ports, vport0 and uplink vport
are special kind of vports that represents PF vport (vport0) and uplink vport which
is connected to the MPFS switch (if exists) as the PF external link.
vport1..vportN represent VF0..VF(N-1) egress/ingress ports.
E-Switch FDB contains forwarding rules such as:
UC MAC0 -> vport0(PF).
UC MAC1 -> vport1.
UC MAC2 -> vport2.
MC MACX -> vport0, vport2, Uplink.
MC MACY -> vport1, Uplink.
For unmatched traffic FDB has the following default rules:
Unmatched Traffic (src vport != Uplink) -> Uplink.
Unmatched Traffic (src vport == Uplink) -> vport0(PF).
NIC VPort context:
Each NIC (VF/PF) has its own vport context which will be used to store the current
NIC vport context (UC/MC and vlan lists) and other NIC properties such as MTU, promisc
mode, etc.. NIC (VF/PF) driver is responsible of constantly updating this context.
FDB rules population:
Each NIC vport (VF/PF) will notify E-Switch manager of its UC/MC vport
context changes via modify vport context command, which will be
translated to an event that will be handled by E-Switch manager (PF)
which will update FDB table accordingly.
Both PF and VF use the same driver and submit commands directly to the firmware.
The PF sees the vport_group_manager capability bit and as such runs the code
to populate the embedded switches as explained above.
The patch goes as follows:
Patches 1-2 introduces the basic PCI SRIOV functionalities and the support of
Connectx4 to enable specific VFs via enable/disable HCA commands. These two
patches will be also in use later for the IB SRIOV flow.
Patches 3-8 Introduces the basic E-Switch capabilities and commands to be used later by
VF to modify and update its NIC vport context, and by PF (E-Switch Manager) driver to
Query the VF NIC context and acts accordingly.
Patches 9-10 Provide the needed functionality of a NIC driver VF/PF to support SRIOV,
mainly vport context update support.
Patch 11 ("net/mlx5: Introducing E-Switch and l2 table"), Introduces the basic
E-Switch support and infrastructure to read vport context events and to update
MPFS L2 Table of the UC mac addresses request by the PF.
Patches 12-18 Introduces SRIOV enablemenet and E-Switch FDB table management
It adds the Basic E-Swtich public API to set and get sriov properties to be used
in PF netdev sriov ndos.
Patchset was applied ontop of commit 3f8c0f7 "gianfar: use of_property_read_bool()"
Saeed, Eli and Or.
changes from V0, addressed feedback from Alex Duyck:
- patch 09, remove the loop to seek the device address
- patch 09, avoid using array as returned value from helper function
- patch 10, fix possible buffer over-run
changes from V1, addressed feedback from and Julia Lawall and kbuild test robot
- patch 11 check the right variable for allocation failure
- patch 18 eliminated unneeded semicolon
====================
Saeed Mahameed [Tue, 1 Dec 2015 16:03:20 +0000 (18:03 +0200)]
net/mlx5: E-Switch, Add SR-IOV (FDB) support
Enabling E-Switch SRIOV for nvfs+1 vports.
Create E-Switch FDB for L2 UC/MC mac steering between VFs/PF and
external vport (Uplink).
FDB contains forwarding rules such as:
UC MAC0 -> vport0(PF).
UC MAC1 -> vport1.
UC MAC2 -> vport2.
MC MACX -> vport0, vport2, Uplink.
MC MACY -> vport1, Uplink.
For unmatched traffic FDB has the following default rules:
Unmached Traffic (src vport != Uplink) -> Uplink.
Unmached Traffic (src vport == Uplink) -> vport0(PF).
FDB rules population:
Each NIC vport (VF) will notify E-Switch manager of its UC/MC vport
context changes via modify vport context command, which will be
translated to an event that will be handled by E-Switch manager (PF)
which will update FDB table accordingly.
Saeed Mahameed [Tue, 1 Dec 2015 16:03:18 +0000 (18:03 +0200)]
net/mlx5: Introducing E-Switch and l2 table
E-Switch is the software entity that represents and manages ConnectX4
inter-HCA ethernet l2 switching.
E-Switch has its own Virtual Ports, each Vport/vNIC/VF can be
connected to the device through a vport of an e-switch.
Each e-switch is managed by one vNIC identified by
HCA_CAP.vport_group_manager (usually it is the PF/vport[0]),
and its main responsibility is to forward each packet to the
right vport.
e-Switch needs to manage its own l2-table and FDB tables.
L2 table is a flow table that is managed by FW, it is needed for
Multi-host (Multi PF) configuration for inter HCA switching between
PFs.
FDB table is a flow table that is totally managed by e-Switch driver,
its main responsibility is to switch packets between e-Swtich internal
vports and uplink vport that belong to the same.
This patch introduces only e-Swtich l2 table management, FDB managemnt
will come later when ethernet SRIOV/VFs will be enabled.
preperation for ethernet sriov and l2 table management.
Saeed Mahameed [Tue, 1 Dec 2015 16:03:11 +0000 (18:03 +0200)]
net/mlx5: Update access functions to Query/Modify vport MAC address
In preparation for SR-IOV we add here an API to enable each e-switch
client (PF/VF) to configure its L2 MAC addresses and for the e-switch
manager (usually the PF) to access them in order to be able to
configure them into the e-switch.
Therefore we now pass vport num parameter to
mlx5_query_nic_vport_context, so PF can access other vports contexts.
preperation for ethernet sriov and l2 table management.
Eli Cohen [Tue, 1 Dec 2015 16:03:09 +0000 (18:03 +0200)]
net/mlx5_core: Add base sriov support
This patch adds SRIOV base support for mlx5 supported devices. The same
driver is used for both PFs and VFs; VFs are identified by the driver
through the flag MLX5_PCI_DEV_IS_VF added to the pci table entries.
Virtual functions are created as usual through writing a value to the
sriov_numvs sysfs file of the PF device. Upon instantiating VFs, they will
all be probed by the driver on the hypervisor. One can gracefully unbind
them through /sys/bus/pci/drivers/mlx5_core/unbind.
mlx5_wait_for_vf_pages() was added to ensure that when a VF dies without
executing proper teardown, the hypervisor driver waits till all of the
pages that were allocated at the hypervisor to maintain its operation
are returned.
In order for the VF to be operational, the PF needs to call enable_hca
for it. This can be done before the VFs are created through a call to
pci_enable_sriov.
If the there are VFs assigned to a VMs when the driver of the PF is
unloaded, all the VF will experience system error and PF driver unloads
cleanly; in this case pci_disable_sriov is not called and the devices
will show when running lspci. Once the PF driver is reloaded, it will
sync its data structures which maintain state on its VFs.
Modify these functions to have func_id argument to state which device we
are referring to. This is done as a preparation for SRIOV support where
a PF driver needs to control its virtual functions.
Andrew Lunn [Tue, 1 Dec 2015 15:31:08 +0000 (16:31 +0100)]
ipv4: igmp: Allow removing groups from a removed interface
When a multicast group is joined on a socket, a struct ip_mc_socklist
is appended to the sockets mc_list containing information about the
joined group.
If the interface is hot unplugged, this entry becomes stale. Prior to
commit 52ad353a5344f ("igmp: fix the problem when mc leave group") it
was possible to remove the stale entry by performing a
IP_DROP_MEMBERSHIP, passing either the old ifindex or ip address on
the interface. However, this fix enforces that the interface must
still exist. Thus with time, the number of stale entries grows, until
sysctl_igmp_max_memberships is reached and then it is not possible to
join and more groups.
The previous patch fixes an issue where a IP_DROP_MEMBERSHIP is
performed without specifying the interface, either by ifindex or ip
address. However here we do supply one of these. So loosen the
restriction on device existence to only apply when the interface has
not been specified. This then restores the ability to clean up the
stale entries.
Signed-off-by: Andrew Lunn <[email protected]> Fixes: 52ad353a5344f "(igmp: fix the problem when mc leave group") Signed-off-by: David S. Miller <[email protected]>
This patchset introduces needed infrastructure for link aggregation
offload - for both team and bonding. It also implements the offload
in mlxsw driver.
Particulary, this patchset introduces possibility for upper driver
(bond/team/bridge/..) to pass type-specific info down to notifier listeners.
Info is passed along with NETDEV_CHANGEUPPER/NETDEV_PRECHANGEUPPER
notifiers. Listeners (drivers of netdevs being enslaved) can react
accordingly.
Other extension is for run-time use. This patchset introduces
new netdev notifier type - NETDEV_CHANGELOWERSTATE. Along with this
notification, the upper driver (bond/team/bridge/..) can pass some
information about lower device change, particulary link-up and
TX-enabled states. Listeners (drivers of netdevs being enslaved)
can react accordingly.
The last part of the patchset is implementation of LAG offload in mlxsw,
using both previously introduced infrastructre extensions.
Note that bond-speficic (and ugly) NETDEV_BONDING_INFO used by mlx4
can be removed and mlx4 can use the extensions this patchset adds.
I plan to convert it and get rid of NETDEV_BONDING_INFO in
a follow-up patchset.
v2->v3:
- one small fix in patch 1
v1->v2:
- added patch 1 and 2 per Andy's request
- couple of more or less cosmetic changes described in couple other patches
====================
Jiri Pirko [Thu, 3 Dec 2015 11:12:29 +0000 (12:12 +0100)]
mlxsw: spectrum: Implement FDB add/remove/dump for LAG
Implement FDB offloading for lagged ports, including learning LAG FDB
entries, adding/removing static FDB entries and dumping existing LAG FDB
entries.
Jiri Pirko [Thu, 3 Dec 2015 11:12:23 +0000 (12:12 +0100)]
mlxsw: core: Add support for packets received from LAG port
Lower layer (pci) has information if the packet is received via LAG port.
If that is the case, it fills up rx_info accordingly. However upper
layer does not care about lag_id/port_index for received packets so
convert it to local_port before passing it up. For that conversion, lag
mapping array is introduced. Upper layer is responsible for setting up
the mapping according to what is set in HW.
Jiri Pirko [Thu, 3 Dec 2015 11:12:21 +0000 (12:12 +0100)]
bonding: set inactive flags on release
Be correct and symmetric to enslave and set inactive flags during release.
That gives LAG offload drivers - lower state change listeners - possibility
to do proper cleanup.
Jiri Pirko [Thu, 3 Dec 2015 11:12:19 +0000 (12:12 +0100)]
bonding: allow notifications for bond_set_slave_link_state
Similar to state notifications.
We allow caller to indicate if the notification should happen now or later,
depending on if he holds rtnl mutex or not. Introduce bond_slave_link_notify
function (similar to bond_slave_state_notify) which is later on called
with rtnl mutex and goes over slaves and executes delayed notification.
Jiri Pirko [Thu, 3 Dec 2015 11:12:15 +0000 (12:12 +0100)]
net: introduce change lower state notifier
When lower device like bonding slave, team/bridge port, etc changes its
state, it is useful for others to notice this change. Currently this is
implemented specificly for bonding as NETDEV_BONDING_INFO notifier. This
patch aims to replace this specific usage and make this more generic to
be used for all upper-lower devices.
Introduce NETDEV_CHANGELOWERSTATE netdev notifier type and
netdev_lower_state_changed() helper.
projects / linux.git / log