In case GRO is not as efficient as it should be or disabled,
we might have a user thread trapped in __release_sock() while
softirq handler flood packets up to the point we have to drop.

This patch balances work done from user thread and softirq,
to give more chances to __release_sock() to complete its work
before new packets are added the the backlog.

This also helps if we receive many ACK packets, since GRO
does not aggregate them.

This patch brings ~60% throughput increase on a receiver
without GRO, but the spectacular gain is really on
1000x release_sock() latency reduction I have measured.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Jean-Louis Dupond reported poor iscsi TCP receive performance
that we tracked to backlog drops.

Apparently we fail to send window updates reflecting the
fact that we are under stress.

Note that we might lack a proper window increase when
backlog is fully processed, since __release_sock() clears
sk->sk_backlog.len _after_ all skbs have been processed.

This should not matter in practice. If we had a significant
load through socket backlog, we are in a dangerous
situation.
Reported-by: Jean-Louis Dupond <jean-louis@dupond.be>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Tested-by: Jean-Louis Dupond<jean-louis@dupond.be>
Signed-off-by: David S. Miller <davem@davemloft.net>

Neal pointed out that non sack flows might suffer from ACK compression
added in the following patch ("tcp: implement coalescing on backlog queue")

Instead of tweaking tcp_add_backlog() we can take into
account how many ACK were coalesced, this information
will be available in skb_shinfo(skb)->gso_segs
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Tell the compiler that most TCP flows are using SACK these days.

There is no need to add the unlikely() clause in tcp_is_reno(),
the compiler is able to infer it.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Trace events are already present for the receive entry points, to indicate
how the reception entered the stack.

This patch adds the corresponding exit trace events that will bound the
reception such that all events occurring between the entry and the exit
can be considered as part of the reception context. This greatly helps
for dependency and root cause analyses.

Without this, it is not possible with tracepoint instrumentation to
determine whether a sched_wakeup event following a netif_receive_skb
event is the result of the packet reception or a simple coincidence after
further processing by the thread. It is possible using other mechanisms
like kretprobes, but considering the "entry" points are already present,
it would be good to add the matching exit events.

In addition to linking packets with wakeups, the entry/exit event pair
can also be used to perform network stack latency analyses.
Signed-off-by: Geneviève Bastien <gbastien@versatic.net>
CC: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
CC: Steven Rostedt <rostedt@goodmis.org>
CC: Ingo Molnar <mingo@redhat.com>
CC: David S. Miller <davem@davemloft.net>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org> (tracing side)
Signed-off-by: David S. Miller <davem@davemloft.net>

There are no such structs flow_dissector_key_flow_vlan or
 flow_dissector_key_flow_tags, the actual structs used are struct
 flow_dissector_key_vlan and struct flow_dissector_key_tags.  So correct the
 comments against FLOW_DISSECTOR_KEY_VLAN, FLOW_DISSECTOR_KEY_FLOW_LABEL and
 FLOW_DISSECTOR_KEY_CVLAN to refer to those.
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Total number of VFs supported by PF is used to determine the last
byte of VF's mac address. Number of VFs supported is not always
16, use the variable nvfs to get the number of VFs supported
rather than hard coding it to 16.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Daniel Borkmann says:

====================
bpf-next 2018-11-30

The following pull-request contains BPF updates for your *net-next* tree.

(Getting out bit earlier this time to pull in a dependency from bpf.)

The main changes are:

1) Add libbpf ABI versioning and document API naming conventions
   as well as ABI versioning process, from Andrey.

2) Add a new sk_msg_pop_data() helper for sk_msg based BPF
   programs that is used in conjunction with sk_msg_push_data()
   for adding / removing meta data to the msg data, from John.

3) Optimize convert_bpf_ld_abs() for 0 offset and fix various
   lib and testsuite build failures on 32 bit, from David.

4) Make BPF prog dump for !JIT identical to how we dump subprogs
   when JIT is in use, from Yonghong.

5) Rename btf_get_from_id() to make it more conform with libbpf
   API naming conventions, from Martin.

6) Add a missing BPF kselftest config item, from Naresh.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

During porting libbpf to bcc, I got some warnings like below:
  ...
  [  2%] Building C object src/cc/CMakeFiles/bpf-shared.dir/libbpf/src/libbpf.c.o
  /home/yhs/work/bcc2/src/cc/libbpf/src/libbpf.c:12:0:
  warning: "_GNU_SOURCE" redefined [enabled by default]
   #define _GNU_SOURCE
  ...
  [  3%] Building C object src/cc/CMakeFiles/bpf-shared.dir/libbpf/src/libbpf_errno.c.o
  /home/yhs/work/bcc2/src/cc/libbpf/src/libbpf_errno.c: In function ‘libbpf_strerror’:
  /home/yhs/work/bcc2/src/cc/libbpf/src/libbpf_errno.c:45:7:
  warning: assignment makes integer from pointer without a cast [enabled by default]
     ret = strerror_r(err, buf, size);
  ...

bcc is built with _GNU_SOURCE defined and this caused the above warning.
This patch intends to make libpf _GNU_SOURCE friendly by
  . define _GNU_SOURCE in libbpf.c unless it is not defined
  . undefine _GNU_SOURCE as non-gnu version of strerror_r is expected.
Signed-off-by: Yonghong Song <yhs@fb.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

Reported-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We can remove the loop and conditional branches
and compute wscale efficiently thanks to ilog2()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

skb_linearization can fail due to memory allocation failure.
In such a case, the driver will drop the packet. In such a case
The driver used to print an error message.
This patch replaces this error message by a dedicated statistic.
Signed-off-by: Michael Shteinbok <michael.shteinbok@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add comments in the switch statement for XDP action to indicate
fallthrough is intended.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Kunihiko Hayashi says:

====================
Add suspend/resume support for AVE ethernet driver

This series adds support for suspend/resume to AVE ethernet driver.

And to avoid the error that wol state of phy hardware is enabled by default,
this sets initial wol state to disabled and add preservation the state in
suspend/resume sequence.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Since the wol state forces to be initialized after reset, the state should
be preserved in suspend/resume sequence.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If wol state of phy hardware is enabled after reset, phy_ethtool_get_wol()
returns that wol.wolopts is true.

However, since net_device.wol_enabled is zero and this doesn't apply wol
state until calling ethtool_set_wol(), so mdio_bus_phy_may_suspend()
returns true, that is, it's in a state where phy can suspend even though
wol state is enabled.

In this inconsistency, phy_suspend() returns -EBUSY, and at last,
suspend sequence fails with the following message:

    dpm_run_callback(): mdio_bus_phy_suspend+0x0/0x58 returns -16
    PM: Device 65000000.ethernet-ffffffff:01 failed to suspend: error -16
    PM: Some devices failed to suspend, or early wake event detected

In order to fix the above issue, this patch forces to set initial wol state
to disabled as default.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch introduces suspend and resume functions to ave driver.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge tag 'linux-can-next-for-4.21-20181128' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
This is a pull request for net-next/master consisting of 18 patches.

The first patch is by Colin Ian King and fixes the spelling in the ucan
driver.

The next three patches target the xilinx driver. YueHaibing's patch
fixes the return type of ndo_start_xmit function. Two patches by
Shubhrajyoti Datta add support for the CAN FD 2.0 controllers.

Flavio Suligoi's patch for the sja1000 driver add support for the ASEM
CAN raw hardware.

Wolfram Sang's and Kuninori Morimoto's patches switch the rcar driver to
use SPDX license identifiers.

The remaining 111 patches improve the flexcan driver. Pankaj Bansal's
patch enables the driver in Kconfig on all architectures with IOMEM
support. The next four patches by me fix indention, add missing
parentheses and comments. Aisheng Dong's patches add self wake support
and document it in the DT bindings. The remaining patches by Pankaj
Bansal first fix the loopback support and prepare the driver for the
CAN-FD support needed for the LX2160A SoC. The actual CAN-FD support
will be added in a later patch series.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Trivial conflict in net/core/filter.c, a locally computed
'sdif' is now an argument to the function.
Signed-off-by: David S. Miller <davem@davemloft.net>

Cannot cast a u64 to a pointer on 32-bit without an intervening (long)
cast otherwise GCC warns.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

CONFIG_FTRACE_SYSCALLS=y is required for get_cgroup_id_user test case
this test reads a file from debug trace path
/sys/kernel/debug/tracing/events/syscalls/sys_enter_nanosleep/id
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

John Fastabend says:

====================
After being able to add metadata to messages with sk_msg_push_data we
have also found it useful to be able to "pop" this metadata off before
sending it to applications in some cases. This series adds a new helper
sk_msg_pop_data() and the associated patches to add tests and tools/lib
support.

Thanks!

v2: Daniel caught that we missed adding sk_msg_pop_data to the changes
    data helper so that the verifier ensures BPF programs revalidate
    data after using this helper. Also improve documentation adding a
    return description and using RST syntax per Quentin's comment. And
    delta calculations for DROP with pop'd data (albeit a strange set
    of operations for a program to be doing) had potential to be
    incorrect possibly confusing user space applications, so fix it.
====================
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

Similar to msg_pull_data and msg_push_data add a set of options to
have msg_pop_data() exercised.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

Add the necessary header definitions to tools for new
msg_pop_data_helper.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

This adds a BPF SK_MSG program helper so that we can pop data from a
msg. We use this to pop metadata from a previous push data call.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

Pull networking fixes from David Miller:

 1) ARM64 JIT fixes for subprog handling from Daniel Borkmann.

 2) Various sparc64 JIT bug fixes (fused branch convergance, frame
    pointer usage detection logic, PSEODU call argument handling).

 3) Fix to use BH locking in nf_conncount, from Taehee Yoo.

 4) Fix race of TX skb freeing in ipheth driver, from Bernd Eckstein.

 5) Handle return value of TX NAPI completion properly in lan743x
    driver, from Bryan Whitehead.

 6) MAC filter deletion in i40e driver clears wrong state bit, from
    Lihong Yang.

 7) Fix use after free in rionet driver, from Pan Bian.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (53 commits)
  s390/qeth: fix length check in SNMP processing
  net: hisilicon: remove unexpected free_netdev
  rapidio/rionet: do not free skb before reading its length
  i40e: fix kerneldoc for xsk methods
  ixgbe: recognize 1000BaseLX SFP modules as 1Gbps
  i40e: Fix deletion of MAC filters
  igb: fix uninitialized variables
  netfilter: nf_tables: deactivate expressions in rule replecement routine
  lan743x: Enable driver to work with LAN7431
  tipc: fix lockdep warning during node delete
  lan743x: fix return value for lan743x_tx_napi_poll
  net: via: via-velocity: fix spelling mistake "alignement" -> "alignment"
  qed: fix spelling mistake "attnetion" -> "attention"
  net: thunderx: fix NULL pointer dereference in nic_remove
  sctp: increase sk_wmem_alloc when head->truesize is increased
  firestream: fix spelling mistake: "Inititing" -> "Initializing"
  net: phy: add workaround for issue where PHY driver doesn't bind to the device
  usbnet: ipheth: fix potential recvmsg bug and recvmsg bug 2
  sparc: Adjust bpf JIT prologue for PSEUDO calls.
  bpf, doc: add entries of who looks over which jits
  ...

Pull Xtensa fixes from Max Filippov:

 - fix kernel exception on userspace access to a currently disabled
   coprocessor

 - fix coprocessor data saving/restoring in configurations with multiple
   coprocessors

 - fix ptrace access to coprocessor data on configurations with multiple
   coprocessors with high alignment requirements

* tag 'xtensa-20181128' of git://github.com/jcmvbkbc/linux-xtensa:
  xtensa: fix coprocessor part of ptrace_{get,set}xregs
  xtensa: fix coprocessor context offset definitions
  xtensa: enable coprocessors that are being flushed

Jeff Kirsher says:

====================
Intel Wired LAN Driver Fixes 2018-11-28

This series contains fixes to igb, ixgbe and i40e.

Yunjian Wang from Huawei resolves a variable that could potentially be
NULL before it is used.

Lihong fixes an i40e issue which goes back to 4.17 kernels, where
deleting any of the MAC filters was causing the incorrect syncing for
the PF.

Josh Elsasser caught that there were missing enum values in the link
capabilities for x550 devices, which was preventing link for 1000BaseLX
SFP modules.

Jan fixes the function header comments for XSK methods.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The response for a SNMP request can consist of multiple parts, which
the cmd callback stages into a kernel buffer until all parts have been
received. If the callback detects that the staging buffer provides
insufficient space, it bails out with error.
This processing is buggy for the first part of the response - while it
initially checks for a length of 'data_len', it later copies an
additional amount of 'offsetof(struct qeth_snmp_cmd, data)' bytes.

Fix the calculation of 'data_len' for the first part of the response.
This also nicely cleans up the memcpy code.

Fixes: 1da177e4 ("Linux-2.6.12-rc2")
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/qualcomm/rmnet/rmnet_map_command.c:26:6:
 warning: variable 'ip_family' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qualcomm/rmnet/rmnet_map_command.c:27:6:
 warning: variable 'fc_seq' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qualcomm/rmnet/rmnet_map_command.c:28:6:
 warning: variable 'qos_id' set but not used [-Wunused-but-set-variable]

It never used since introduction in commit
ceed73a2 ("drivers: net: ethernet: qualcomm: rmnet: Initial implementation")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c:4011:5:
 warning: variable 'max_hw_rings' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c:4013:6:
 warning: variable 'cur_rings' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c:2996:25:
 warning: variable 'tx_desc_info' set but not used [-Wunused-but-set-variable]

'cur_rings, max_hw_rings' never used since introduction in commit
34e8c406 ("qlcnic: refactor Tx/SDS ring calculation and validation in driver.")
'tx_desc_info' never used since commit
95b3890a ("qlcnic: Enhance Tx timeout debugging.")
Also 'queue_type' only can be QLCNIC_RX_QUEUE/QLCNIC_TX_QUEUE,
so make a trival cleanup on if statement.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Shahed Shaikh <shahed.shaikh@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/neterion/vxge/vxge-traffic.c:1698:35:
 warning: variable 'txdl_priv' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/neterion/vxge/vxge-traffic.c:1699:6:
 warning: variable 'max_frags' set but not used [-Wunused-but-set-variable]

It never used since introduction in
commit 11324132 ("Neterion: New driver: Traffic & alarm handler")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

The following patchset contains Netfilter fixes for net:

1) Disable BH while holding list spinlock in nf_conncount, from
   Taehee Yoo.

2) List corruption in nf_conncount, also from Taehee.

3) Fix race that results in leaving around an empty list node in
   nf_conncount, from Taehee Yoo.

4) Proper chain handling for inactive chains from the commit path,
   from Florian Westphal. This includes a selftest for this.

5) Do duplicate rule handles when replacing rules, also from Florian.

6) Remove net_exit path in xt_RATEEST that results in splat, from Taehee.

7) Possible use-after-free in nft_compat when releasing extensions.
   From Florian.

8) Memory leak in xt_hashlimit, from Taehee.

9) Call ip_vs_dst_notifier after ipv6_dev_notf, from Xin Long.

10) Fix cttimeout with udplite and gre, from Florian.

11) Preserve oif for IPv6 link-local generated traffic from mangle
    table, from Alin Nastac.

12) Missing error handling in masquerade notifiers, from Taehee Yoo.

13) Use mutex to protect registration/unregistration of masquerade
    extensions in order to prevent a race, from Taehee.

14) Incorrect condition check in tree_nodes_free(), also from Taehee.

15) Fix chain counter leak in rule replacement path, from Taehee.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Ioana Ciocoi Radulescu says:

====================
dpaa2-eth: Introduce XDP support

Add support for XDP programs. Only XDP_PASS, XDP_DROP and XDP_TX
actions are supported for now. Frame header changes are also
allowed.

v2: - count the XDP packets in the rx/tx inteface stats
    - add message with the maximum supported MTU value for XDP
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add counters for xdp processed frames to the channel statistics.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove unused counter. Reorder fields in channel stats structure
to match the ethtool strings order and make it easier to print them
with ethtool -S.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Send frames back on the same port for XDP_TX action.
Since the frame buffers have been allocated by us, we can recycle
them directly into the Rx buffer pool instead of requesting a
confirmation frame upon transmission complete.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to support enqueueing Rx FDs back to hardware, we need to
DMA map Rx buffers as bidirectional.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Instead of freeing the RX buffers, release them back into the pool.
We wait for the maximum number of buffers supported by a single
release command to accumulate before issuing the command.

Also, don't unmap the Rx buffers at the beginning of the Rx routine
anymore, since that would require remapping them before release.
Instead, just do a DMA sync at first and only unmap if the frame is
meant for the stack.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>