In DC world, GSO packets initially cooked by tcp_sendmsg() are usually
big, as sk_pacing_rate is high.

When network is congested, cwnd can be smaller than the GSO packets
found in socket write queue. tcp_write_xmit() splits GSO packets
using the available cwnd, and we end up sending a single GSO packet,
consuming all available cwnd.

With GRO aggregation on the receiver, we might handle a single GRO
packet, sending back a single ACK.

1) This single ACK might be lost
   TLP or RTO are forced to attempt a retransmit.
2) This ACK releases a full cwnd, sender sends another big GSO packet,
   in a ping pong mode.

This behavior does not fill the pipes in the best way, because of
scheduling artifacts.

Make sure we always have at least two GSO packets in flight.

This allows us to safely increase GRO efficiency without risking
spurious retransmits.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Reallocation is only required for shrinking and expanding and both rely
on a mutex for synchronization and callers of rhashtable_init() are in
non atomic context. Therefore, no reason to continue passing allocation
hints through the API.

Instead, use GFP_KERNEL and add __GFP_NOWARN | __GFP_NORETRY to allow
for silent fall back to vzalloc() without the OOM killer jumping in as
pointed out by Eric Dumazet and Eric W. Biederman.
Signed-off-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Or Gerlitz says:

====================
mlx4: Flexible (asymmetric) allocation of EQs and MSI-X vectors

This series from Matan Barak is built as follows:

The 1st two patches fix small bugs w.r.t firmware spec. Next
are two patches which do more re-factoring of the init/fini flow
and a patch that adds support for the QUERY_FUNC firmware command,
these are all pre-steps for the major patch of the series. In this
patch (#6) we change the order of talking/querying the firmware
and enabling SRIOV. This allows to remote worst-case assumption
w.r.t the number of available MSI-X vectors and EQs per function.

The last patch easily enjoys this ordering change, to enable
supports > 64 VFs over a firmware that allows for that.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We now allow up to 126 VFs. Note though that certain firmware
versions only allow up to 80 VFs. Moreover, old HCAs only support 64 VFs.
In these cases, we limit the maximum number of VFs to 64.
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Previously, the driver queried the firmware in order to get the number
of supported EQs. Under SRIOV, since this was done before the driver
notified the firmware how many VFs it actually needs, the firmware had
to take into account a worst case scenario and always allocated four EQs
per VF, where one was used for events while the others were used for completions.

Now, when the firmware supports the asymmetric allocation scheme, denoted
by exposing num_sys_eqs > 0 (--> MLX4_DEV_CAP_FLAG2_SYS_EQS), we use the
QUERY_FUNC command to query the firmware before enabling SRIOV. Thus we
can get more EQs and MSI-X vectors per function.

Moreover, when running in the new firmware/driver mode, the limitation
that the number of EQs should be a power of two is lifted.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

QUERY_FUNC firmware command could be used in order to query the
number of EQs, reserved EQs, etc for a specific function.
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Refactor mlx4_load_one, as a preparation step for a new and
more complicated load function. The goal is to support both
newer firmware that required init_hca to be done before
enable_sriov and legacy firmwares that requires things to
be done the other way around.
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Refactoring mlx4_cmd_init and mlx4_cmd_cleanup such that partial init
and cleanup are possible. After this refactoring, calling mlx4_cmd_init
several times is safe.

This is necessary in the VF init flow when mlx4_init_hca returns -EACCESS,
we need to issue cleanup and re-attempt to call it with the slave flag.
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We've used an incorrect type for the loop counter and the
mlx4_QUERY_FUNC_CAP function. The current input modifier
is either a port or a boolean.
Since the number of ports is always a positive value < 255,
we should use u8 instead of an integer with casting.
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We mistakenly read the reserved_eqs field as a standard
numeric value rather than a log2 value.
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Herbert Xu says:

====================
rhashtable: Allow local locks to be used and tested

This series moves mutex_is_held entirely under PROVE_LOCKING so
there is zero foot print when we're not debugging.  More importantly
it adds a parrent argument to mutex_is_held so that we can test
local locks rather than global ones (e.g., per-namespace locks).
====================
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently mutex_is_held can only test locks in the that are global
since it takes no arguments.  This prevents rhashtable from being
used in places where locks are lock, e.g., per-namespace locks.

This patch adds a parent field to mutex_is_held and rhashtable_params
so that local locks can be used (and tested).
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The rhashtable function mutex_is_held is only used when PROVE_LOCKING
is enabled.  This patch makes the mutex_is_held field in rhashtable
optional depending on PROVE_LOCKING.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The rhashtable function mutex_is_held is only used when PROVE_LOCKING
is enabled.  This patch modifies netfilter so that we can rhashtable.h
itself can later make mutex_is_held optional depending on PROVE_LOCKING.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The rhashtable function mutex_is_held is only used when PROVE_LOCKING
is enabled.  This patch modifies netlink so that we can rhashtable.h
itself can later make mutex_is_held optional depending on PROVE_LOCKING.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The device tree probing for R-Car M2N (r8a7793) is added.
Signed-off-by: NHisashi Nakamura <hisashi.nakamura.ak@renesas.com>
Signed-off-by: NYoshihiro Kaneko <ykaneko0929@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When using RMMI mode, it is necessary to change in probe.
Signed-off-by: NHisashi Nakamura <hisashi.nakamura.ak@renesas.com>
Signed-off-by: NYoshihiro Kaneko <ykaneko0929@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Large receive offloading is known to cause problems if received packets
are passed to other host. Therefore the kernel disables it by calling
dev_disable_lro() whenever a network device is enslaved in a bridge or
forwarding is enabled for it (or globally). For virtual devices we need
to disable LRO on the underlying physical device (which is actually
receiving the packets).

Current dev_disable_lro() code handles this  propagation for a vlan
(including 802.1ad nested vlan), macvlan or a vlan on top of a macvlan.
It doesn't handle other stacked devices and their combinations, in
particular propagation from a bond to its slaves which often causes
problems in virtualization setups.

As we now have generic data structures describing the upper-lower device
relationship, dev_disable_lro() can be generalized to disable LRO also
for all lower devices (if any) once it is disabled for the device
itself.

For bonding and teaming devices, it is necessary to disable LRO not only
on current slaves at the moment when dev_disable_lro() is called but
also on any slave (port) added later.

v2: use lower device links for all devices (including vlan and macvlan)
Signed-off-by: NMichal Kubecek <mkubecek@suse.cz>
Acked-by: NVeaceslav Falico <vfalico@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There was a missing break statement so we set everything to
PKT_HASH_TYPE_L3 even when we intended to use PKT_HASH_TYPE_L4.

Fixes: 5b9dfe29 ('amd-xgbe: Provide support for receive side scaling')
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Refactored all macros used in cxgb4i as part of previously started cxgb4 macro
names cleanup. Makes them more uniform and avoids namespace collision.
Minor changes in other drivers where required as some of these macros are used
 by multiple drivers, affected drivers are iw_cxgb4, cxgb4(vf) & csiostor
Signed-off-by: NAnish Bhatt <anish@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes two issues after using iovec iterators:
- vlan_offset should be initialized to zero, otherwise unexpected offset
  will be used in skb_copy_datagram_iter()
- advance iovec iterator when vnet_hdr_sz is greater than sizeof(gso), this
  is the case when mergeable rx buffer were enabled for a virt guest.

Fixes e0b46d0e ("tun: Use iovec iterators")
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NJason Wang <jasowang@redhat.com>
Acked-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

net/ipv4/fou.c: In function ‘ip_tunnel_encap_del_fou_ops’:
net/ipv4/fou.c:861:1: warning: no return statement in function returning non-void [-Wreturn-type]

Fixes: a8c5f90f ("ip_tunnel: Ops registration for secondary encap (fou, gue)")
Signed-off-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With commit d75b1ade ("net: less interrupt masking in NAPI") napi
repoll is done only when work_done == budget. bcm_sysport_tx_poll()
always returns 0 whether or not we completed the poll quantum.

Fix this by returning either 0 when we did complete the TX ring reclaim,
or budget to trigger a repoll.

Fixes: d75b1ade ("net: less interrupt masking in NAPI")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With the commit d75b1ade ("net: less interrupt masking in NAPI") napi repoll
is done only when work_done == budget. In tx napi poll we always return 0.
So tx napi is not called again and we do not clean up the tx ring.
Signed-off-by: NGovindarajulu Varadarajan <_govind@gmx.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix the build failures that result from the use of pr_debug
without the referenced char * arrays being defined.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the host uses packet encapsulation feature, the MTU may be reduced by the
host due to headroom reservation for encapsulation. This patch handles this
new MTU value.
Signed-off-by: NHaiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change the types of the descriptor entries in the xgbe_ring_desc struct
from u32 to __le32 to fix endian warnings issued by sparse.
Signed-off-by: NTom Lendacky <thomas.lendacky@amd.com>
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is to ensure the net_device's dev.parent is set before we used it
in dma_zalloc_coherent() from init_hash_table().
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Acked-by: NAntoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix static checker warning that got introduced in commit e2ac9628
("cxgb4: Cleanup macros so they follow the same style and look consistent, part
2") due to accidental checkin of bogus line.
Signed-off-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of calling fou and gue functions directly from ip_tunnel
use ops for these that were previously registered. This patch adds the
logic to add and remove encapsulation operations for ip_tunnel,
and modified fou (and gue) to register with ip_tunnels.

This patch also addresses a circular dependency between ip_tunnel
and fou that was causing link errors when CONFIG_NET_IP_TUNNEL=y
and CONFIG_NET_FOU=m. References to fou an gue have been removed from
ip_tunnel.c
Reported-by: NRandy Dunlap <rdunlap@infradead.org>
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Standardize function pointer uses.

Convert calling style from:
	(*foo)(args...);
to:
	foo(args...);

Other miscellanea:

o Add braces around loops with single ifs on multiple lines
o Realign arguments around these functions
o Invert logic in if to return immediately.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Hayes Wang says:

====================
Code adjustment

v3:
 Remove the test_bit for patch #2.

v2:
 Correct the spelling error for the comment of patch #3.

v1:
Adjust some codes to make them more reasonable.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the device is unplugged or !netif_running(), the workqueue
doesn't need to wake the device, and could return directly.
Signed-off-by: NHayes Wang <hayeswang@realtek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Clear the flag of SCHEDULE_TASKLET in bottom_half() to avoid
re-schedule the tasklet again by workqueue.
Signed-off-by: NHayes Wang <hayeswang@realtek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The INIT_LIST_HEAD(&tp->rx_done) would be done in rtl_start_rx(),
so remove the unnecessary one in alloc_all_mem().
Signed-off-by: NHayes Wang <hayeswang@realtek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Florian Fainelli says:

====================
net: phy: bcm7xxx: workaround updates

This patch series contains some updates to the Broadcom BCM7xxx internal
PHY driver, including:

- removing an annonying print that would appear during interface up/down and
  suspend/resume cycles
- drop a workaround sequence for a non-production PHY revision
- add new workarounds for the latest and greatest PHY devices found out there
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PHY revisions E0 and F0 share the same shorter workaround initialization
sequence. Dedicate a special function for these two PHY revisions to
perform the needed workaround sequence.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PHY revision D0 requires a specific workaround sequence which needs to
be applied to get the HW to behave properly in all corner cases
conditions. Do this based on the revision we just read out of the HW
using a specific function.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This function performs a R/RC calibration reset and will start being
used by more than one function in the next patches, create a helper
function to factor code.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

bcm7445_config_init() was working around non-production version of the
PHY HW block, so just remove it entirely.

bcm7xxx_28nm_afe_config_init() was running for all PHY revisions greater
than B0, but this workaround sequence is really specific to the B0 PHY
revision, so rename the function accordingly and update the GPHY macro
to use the generic config_init callback.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>