We're not guaranteed to see NETDEV_REGISTER/NETDEV_UNREGISTER notifications
only once per VF but we increase/decrease module refcount unconditionally.
Check vf_netdev to make sure we don't take/release it twice. We presume
that only one VF per netvsc device may exist.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We reset vf_inject on VF going down (netvsc_vf_down()) but we don't on
VF removal (netvsc_unregister_vf()) so vf_inject stays 'true' while
vf_netdev is already NULL and we're trying to inject packets into NULL
net device in netvsc_recv_callback() causing kernel to crash.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Here is a deadlock scenario:
- netvsc_vf_up() schedules netvsc_notify_peers() work and quits.
- netvsc_vf_down() runs before netvsc_notify_peers() gets executed. As it
  is being executed from netdev notifier chain we hold rtnl lock when we
  get here.
- we enter while (atomic_read(&net_device_ctx->vf_use_cnt) != 0) loop and
  wait till netvsc_notify_peers() drops vf_use_cnt.
- netvsc_notify_peers() starts on some other CPU but netdev_notify_peers()
  will hang on rtnl_lock().
- deadlock!

Instead of introducing additional synchronization I suggest we drop
gwrk.dwrk completely and call NETDEV_NOTIFY_PEERS directly. As we're
acting under rtnl lock this is legitimate.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

struct netvsc_device is not suitable for storing VF information as this
structure is being destroyed on MTU change / set channel operation (see
rndis_filter_device_remove()). Move all VF related stuff to struct
net_device_context which is persistent.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Added a condition to avoid bonding devices with same MAC registering
as VF.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use the new APIs for eliminating a copy on the receive path. These new APIs also
help in minimizing the number of memory barriers we end up issuing (in the
ringbuffer code) since we can better control when we want to expose the ring
state to the host.

The patch is being resent to address earlier email issues.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

I'm hitting 5 second timeout in rndis_filter_set_rss_param() while setting
RSS parameters for the device. When this happens we end up returning
-ETIMEDOUT from the function and rndis_filter_device_add() falls back to
setting

        net_device->max_chn = 1;
        net_device->num_chn = 1;
        net_device->num_sc_offered = 0;

but after a moment the rndis request succeeds and subchannels start to
appear. netvsc_sc_open() does unconditional nvscdev->num_sc_offered-- and
it becomes U32_MAX-1. Consequent rndis_filter_device_remove() will hang
while waiting for all U32_MAX-1 subchannels to appear and this is not
going to happen.

The immediate issue could be solved by adding num_sc_offered > 0 check to
netvsc_sc_open() but we're getting out of sync with the host and it's not
easy to adjust things later, e.g. in this particular case we'll be creating
queues without a user request for it and races are expected. Same applies
to other parts of the driver which have the same completion timeout.

Following the trend in drivers/hv/* code I suggest we remove all these
timeouts completely. As a guest we can always trust the host we're running
on and if the host screws things up there is no easy way to recover anyway.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The only caller rndis_filter_device_add() has 'struct net_device' pointer
already.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We unpack 'struct net_device' in netvsc_set_mac_addr() to get to
'struct hv_device' pointer which we use in rndis_filter_set_device_mac()
to get back to 'struct net_device'.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Both rndis_filter_open()/rndis_filter_close() use struct hv_device to
reach to struct netvsc_device only and all callers have it already.
While on it, rename net_device to nvdev in rndis_filter_open() as
net_device is misleading.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make it easier to get 'struct netvsc_device' from 'struct net_device' and
'struct hv_device' by introducing inline helpers.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

net_device_ctx is assigned in the very beginning of the function and 'net'
pointer doesn't change.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Added a condition to avoid vlan devices with same MAC registering
as VF.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Crash in netvsc_send() is observed when netvsc device is re-created on
mtu change/set channels. The crash is caused by dereferencing of NULL
channel pointer which comes from chn_table. The root cause is a mixture
of two facts:
- we set nvdev pointer in net_device_context in alloc_net_device()
  before we populate chn_table.
- we populate chn_table[0] only.

The issue could be papered over by checking channel != NULL in
netvsc_send() but populating the whole chn_table and writing the
nvdev pointer afterwards seems more appropriate.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When netvsc device is removed during mtu change or channels setup we get
into troubles as both paths are trying to remove the device. Synchronize
them with start_remove flag and rtnl lock.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Simplify netvsvc pointer graph by getting rid of the redundant ndev
pointer. We can always get a pointer to struct net_device from somewhere
else.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We have the following structures keeping netvsc adapter state:
- struct net_device
- struct net_device_context
- struct netvsc_device
- struct rndis_device
- struct hv_device
and there are pointers/dependencies between them:
- struct net_device_context is contained in struct net_device
- struct hv_device has driver_data pointer which points to
  'struct net_device' OR 'struct netvsc_device' depending on driver's
  state (!).
- struct net_device_context has a pointer to 'struct hv_device'.
- struct netvsc_device has pointers to 'struct hv_device' and
  'struct net_device_context'.
- struct rndis_device has a pointer to 'struct netvsc_device'.

Different functions get different structures as parameters and use these
pointers for traveling. The problem is (in addition to keeping in mind
this complex graph) that some of these structures (struct netvsc_device
and struct rndis_device) are being removed and re-created on mtu change
(as we implement it as re-creation of hyper-v device) so our travel using
these pointers is dangerous.

Simplify this to a the following:
- add struct netvsc_device pointer to struct net_device_context (which is
  a part of struct net_device and thus never disappears)
- remove struct hv_device and struct net_device_context pointers from
  struct netvsc_device
- replace pointer to 'struct netvsc_device' with pointer to
  'struct net_device'.
- always keep 'struct net_device' in hv_device driver_data.

We'll end up with the following 'circular' structure:

net_device:
 [net_device_context] -> netvsc_device -> rndis_device -> net_device
                      -> hv_device -> net_device

On MTU change we'll be removing the 'netvsc_device -> rndis_device'
branch and re-creating it making the synchronization easier.

There is one additional redundant pointer left, it is struct net_device
link in struct netvsc_device, it is going to be removed in a separate
commit.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

netvsc_link_change() can race with netvsc_change_mtu() or
netvsc_set_channels() as these functions destroy struct netvsc_device and
rndis filter. Use start_remove flag for syncronization. As
netvsc_change_mtu()/netvsc_set_channels() are called with rtnl lock held
we need to take it before checking start_remove value in
netvsc_link_change().
Reported-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

struct netvsc_device is destroyed on mtu change so keeping the
protection flag there is not a good idea. Move it to struct
net_device_context which is preserved.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

RNDIS_STATUS_NETWORK_CHANGE event is handled as two "half events" --
media disconnect & connect. The second half should be added to the list
head, not to the tail. So all events are processed in normal order.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Support VF drivers on Hyper-V. On Hyper-V, each VF instance presented to
the guest has an associated synthetic interface that shares the MAC address
with the VF instance. Typically these are bonded together to support
live migration. By default, the host delivers all the incoming packets
on the synthetic interface. Once the VF is up, we need to explicitly switch
the data path on the host to divert traffic onto the VF interface. Even after
switching the data path, broadcast and multicast packets are always delivered
on the synthetic interface and these will have to be injected back onto the
VF interface (if VF is up).
This patch implements the necessary support in netvsc to support Linux
VF drivers.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Reorder the code in netvsc_sc_open(), so num_sc_offered is only decremented
after vmbus_open() is called. This avoid pontential race of removing device
before all channels are setup.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The VRSS_CHANNEL_MAX is the max number of channels supported by Hyper-V
hosts. We use it for the related array sizes instead of using NR_CPUS,
which may be set to several thousands.
This patch reduces possible memory allocation failures.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

struct netvsc_device is freed in rndis_filter_device_remove(). So we save
the nvdev->num_chn into a temp variable for later usage.

(Please also include this patch into stable branch.)
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

During hot add, vmbus_device_register() is called from vmbus_onoffer(), on
the same workqueue as the subchannel offer message work-queue, so
subchannel offer won't be processed until the vmbus_device_register()/...
/netvsc_probe() is done.
Also, vmbus_device_register() is called with channel_mutex locked, which
prevents subchannel processing too. So the "waiting for sub-channel
processing" will not success in hot add case. But, in usual module loading,
the netvsc_probe() is called from different code path, and doesn't fail.

This patch resolves the deadlock during NIC hot-add, and speeds up NIC
loading time.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch allows the user to set and retrieve speed and duplex of the
hv_netvsc device via ethtool.

Example:
$ ethtool eth0
Settings for eth0:
...
    Speed: Unknown!
    Duplex: Unknown! (255)
...
$ ethtool -s eth0 speed 1000 duplex full
$ ethtool eth0
Settings for eth0:
...
    Speed: 1000Mb/s
    Duplex: Full
...

This is based on patches by Roopa Prabhu and Nikolay Aleksandrov.
Signed-off-by: NSimon Xiao <sixiao@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Enable skb_tx_timestamp in hyperv netvsc.
Signed-off-by: NSimon Xiao <sixiao@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit c0eb4540 ("hv_netvsc: Don't ask for additional head room in the
skb") got rid of needed_headroom setting for the driver. With the change I
hit the following issue trying to use ptkgen module:

[   57.522021] kernel BUG at net/core/skbuff.c:1128!
[   57.522021] invalid opcode: 0000 [#1] SMP DEBUG_PAGEALLOC
...
[   58.721068] Call Trace:
[   58.721068]  [<ffffffffa0144e86>] netvsc_start_xmit+0x4c6/0x8e0 [hv_netvsc]
...
[   58.721068]  [<ffffffffa02f87fc>] ? pktgen_finalize_skb+0x25c/0x2a0 [pktgen]
[   58.721068]  [<ffffffff814f5760>] ? __netdev_alloc_skb+0xc0/0x100
[   58.721068]  [<ffffffffa02f9907>] pktgen_thread_worker+0x257/0x1920 [pktgen]

Basically, we're calling skb_cow_head(skb, RNDIS_AND_PPI_SIZE) and crash on
    if (skb_shared(skb))
        BUG();

We probably need to restore needed_headroom setting (but shrunk to
RNDIS_AND_PPI_SIZE as we don't need more) to request the required headroom
space. In theory, it should not give us performance penalty.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

1. Adding NETIF_F_TSO6 feature flag;
2. Adding NETIF_F_HW_CSUM. NETIF_F_IPV6_CSUM and NETIF_F_IP_CSUM are
being deprecated;
3. Cleanup the coding style of flag assignment by using macro.
Signed-off-by: NSimon Xiao <sixiao@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since eliminating send_completion_tid from struct hv_netvsc_packet, we
haven't add proper book keeping for the skb of the batched packet. This
patch fixes this issue and allows the previous skb is properly freed.
Otherwise, a panic may happen.
Thanks to Simon Xiao <sixiao@microsoft.com> for bisecting and analysis.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Recent changes to 'struct flow_keys' (e.g commit d34af823 ("net: Add
VLAN ID to flow_keys")) introduced a performance regression in netvsc
driver. Is problem is, however, not the above mentioned commit but the
fact that netvsc_set_hash() function did some assumptions on the struct
flow_keys data layout and this is wrong.

Get rid of netvsc_set_hash() by switching to skb_get_hash(). This change
will also imply switching to Jenkins hash from the currently used Toeplitz
but it seems there is no good excuse for Toeplitz to stay.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In commit 2a04ae8a ("hv_netvsc: remove locking in netvsc_send()"), the
locking for MSD (Multi-Send Data) field was removed. This could cause a
race condition between RNDIS control messages and data packets processing,
because these two types of traffic are not synchronized.
This patch fixes this issue by sending control messages out directly
without reading MSD field.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate vlan_tci from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate status from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate xmit_more from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate completion_func from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate is_data_pkt from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate send_completion_tid from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eliminate page_buf from struct hv_netvsc_packet.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Packet scheduler guarantees there won't be multiple senders for the same
queue and as we use q_idx for multi_send_data the spinlock is redundant.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>