Commit graph

661,541 commits

Author SHA1 Message Date
Jonathan Cameron
d1bf1a865f staging:iio:accel:sca3000 drop sca3000_register_ring_funcs
This was needed when the buffer support was optional. Pointless wrapper
now so drop it.

Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Reviewed-by: Lars-Peter Clausen <lars@metafoo.de>
2016-10-23 19:33:58 +01:00
Jonathan Cameron
ced5c03d36 staging:iio:accel:sca3000 merge files into one.
In the early days of IIO we were much more inclined to keep the impact
of the core IIO elements to the minimum.  As time has moved on it has
become clear that hardly any builds are done without buffer support
and that it adds considerable complexity to the drivers.

Hence merge down the buffer and non buffer elements of the sca3000 driver
also allowing us to drop the header file used for the interfaces between
the two.

Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Reviewed-by: Lars-Peter Clausen <lars@metafoo.de>
2016-10-23 19:33:57 +01:00
Jonathan Cameron
e3ef7532a3 staging:iio:accel:sca3000 Fix a use before setting of the indio_dev->buffer pointer.
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Reviewed-by: Lars-Peter Clausen <lars@metafoo.de>
2016-10-23 19:33:57 +01:00
Brian Masney
a57504144c staging: iio: isl29018: move out of staging
Move ISL29018/ISL29023/ISL29035 driver out of staging into mainline.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:56 +01:00
Brian Masney
af06437dae staging: iio: isl29018: add ABI documentation for infrared suppression
Add ABI documentation from the ISL29018 Data Sheet (FN6619.4, Oct 8,
2012) about the infrared suppression that can be controlled
with the proximity_on_chip_ambient_infrared_suppression sysfs attribute.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:55 +01:00
Alison Schofield
ac13980584 iio: pressure: mpl3115: claim direct mode during raw reads
Driver was checking for direct mode but not locking it.  Use
claim/release helper functions to guarantee the device stays
in direct mode during raw reads.

Signed-off-by: Alison Schofield <amsfield22@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:55 +01:00
Jonathan Cameron
79de2ee469 iio: accel: mma8452: claim direct mode during write raw
Driver was checking for direct mode but not locking it.  Use
claim/release helper functions to guarantee the device stays
in direct mode during all write raw operations.

Signed-off-by: Alison Schofield <amsfield22@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:54 +01:00
Alison Schofield
4d9b0413e2 iio: accel: mma8452: claim direct mode during raw reads
Driver was checking for direct mode but not locking it.  Use
claim/release helper functions to guarantee the device stays
in direct mode during raw reads.

Signed-off-by: Alison Schofield <amsfield22@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:54 +01:00
Eva Rachel Retuya
4e2cf0e21f tools: iio: iio_generic_buffer: add -A to force-enable all channels
If attribute/s is/are already enabled (by default or via scripts or
manual interaction), issuing -a will fail to enable the channels thereby
one has to manually disable the said attribute/s before proceeding with
auto-enabling.

Add a command-line option -A to force-activate all channels regardless
of their current state.

Suggested-by: Alison Schofield <amsfield22@gmail.com>
Signed-off-by: Eva Rachel Retuya <eraretuya@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:53 +01:00
Tomas Novotny
b440f1d90e iio: dac: mcp4725: use regulator framework
Use a standard framework to get the reference voltage. It is done that way
in the iio subsystem and it will simplify extending of the driver.

Structure mcp4725_platform_data is left undeleted because it used in the
next patch.

This change breaks the current users of the driver, but there is no
mainline user of struct mcp4725_platform_data.

Signed-off-by: Tomas Novotny <tomas@novotny.cz>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:51 +01:00
Crt Mori
3f5def652f iio: ms65611_core: Fixes dereferencing regulator pointer
Change in handling of the regulator description means that static
checkers correctly assume we could be using dereferenced pointer to the
regulator. In reality we will never get the -ENODEV error, as current
behavior flow does not predict it, because:
If the device tree or board file does not define suitable regulators for
the component, it will be substituted by a dummy regulator, or, if
regulators are disabled altogether, by stubs.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Suggested-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Crt Mori <cmo@melexis.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:50 +01:00
Dan Carpenter
d5d4602e04 Staging: iio: fix a MAINTAINERS entry
The "drivers/" part of the path name was missing.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:44 +01:00
sayli karnik
dd4ba3fb22 iio: bmi160_core: Fix sparse warning due to incorrect type in assignment
There is a type mismatch between the buffer which is of type s16 and the
samples stored, which are declared as __le16.

Fix the following sparse warning:
drivers/iio/imu/bmi160/bmi160_core.c:411:26: warning: incorrect type
in assignment (different base types)

drivers/iio/imu/bmi160/bmi160_core.c:411:26: expected signed short
[signed] [short] [explicitly-signed] <noident>
drivers/iio/imu/bmi160/bmi160_core.c:411:26: got restricted __le16
[addressable] [usertype] sample

This is a cosmetic-type patch since it does not alter code behaviour.
The le16 is going into a 16bit buf element, and is labelled as IIO_LE in the
channel buffer definition.

Signed-off-by: sayli karnik <karniksayli1995@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:44 +01:00
Brian Masney
bac80fd0cf staging: iio: isl29018: rename description in Kconfig for consistency
Rename description from "ISL 29018" to "Intersil 29018" in Kconfig for
consistency with other drivers in mainline.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:43 +01:00
Brian Masney
b2f85b08c9 staging: iio: isl29018: remove blank line for consistency
Remove blank line between MODULE_DEVICE_TABLE() and its corresponding
structure for consistency with the other device table entries.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:43 +01:00
Brian Masney
aba1a8d77f staging: iio: isl29018: combine two return statements into one
Use the return value from isl29018_set_integration_time() as the return
value for isl29018_chip_init() since this is the last piece of work
inside that function.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:42 +01:00
Brian Masney
19a00a9d4c staging: iio: isl29018: fix multiline comment syntax
Change multiline comments from:

/* line1
 * line2
 * ...
 */

to

/*
 * line1
 * line2
 * ...
 */

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:41 +01:00
Brian Masney
8281101c2d staging: iio: isl29018: fix poorly named function
isl29035_detect() did not just do chip detection. Move functionality directly
into isl29018_chip_init() to avoid naming confusion.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:41 +01:00
Brian Masney
744ef62c2e staging: iio: isl29018: add newlines to improve readability
Add newlines to improve code readability in preparation for moving the
driver out of staging.

Signed-off-by: Brian Masney <masneyb@onstation.org>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:40 +01:00
Jonathan Cameron
92da01a08d staging:iio:TODO drop outdated entries in this todo.
There is still stuff to be done in the remaining drivers but pretty much
nothing was left from this original TODO.

Sorry Greg, should have been keeping this up to date.

Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-10-23 19:33:40 +01:00
sayli karnik
f08a150847 staging: iio: cdc: ad7152: Replace mlock with a local mutex lock
mlock is intended to protect only switches between modes.
Given this driver doesn't support more than one mode (sysfs polled reads
only), replace mlock with a local mutex lock.

Signed-off-by: sayli karnik <karniksayli1995@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:33:37 +01:00
Eva Rachel Retuya
2296c0623e staging: iio: cdc: ad7746: implement IIO_CHAN_INFO_SAMP_FREQ
This driver predates the availability of IIO_CHAN_INFO_SAMP_FREQ
attribute wherein usage has some advantages like it can be accessed by
in-kernel consumers as well as reduces the code size.

Therefore, use IIO_CHAN_INFO_SAMP_FREQ to implement the
sampling_frequency attribute instead of using IIO_DEVICE_ATTR() macro.

Move code from the functions associated with IIO_DEVICE_ATTR() into
respective read and write hooks with the mask set to
IIO_CHAN_INFO_SAMP_FREQ.

Signed-off-by: Eva Rachel Retuya <eraretuya@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:28:04 +01:00
sayli karnik
6572389bcc staging: iio: cdc: ad7152: Implement IIO_CHAN_INFO_SAMP_FREQ attribute
Attributes that were once privately defined become standard with time
and hence a special global define is used. Hence update driver ad7152 to use
IIO_CHAN_INFO_SAMP_FREQ which is a global define instead of
IIO_DEV_ATTR_SAMP_FREQ.
Move functionality from IIO_DEV_ATTR_SAMP_FREQ attribute into
IIO_CHAN_INFO_SAMP_FREQ to implement the sampling_frequency attribute.
Modify ad7152_read_raw() and ad7152_write_raw() to allow reading and
writing the element as well. Also add a lock in the driver's private
data.

Signed-off-by: sayli karnik <karniksayli1995@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2016-10-23 19:28:01 +01:00
Michael Turquette
7ae74aaf38 Merge branch 'clk-next-oxnas' into clk-next 2016-10-23 10:19:53 -07:00
Neil Armstrong
d783f36fce dt-bindings: clk: oxnas,stdclk: Add OX820 bindings
Add OX820 bindings and remove clock indices from bindings since they are present
in the dt-bindings headers files.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Link: lkml.kernel.org/r/20161005150752.22618-7-narmstrong@baylibre.com
2016-10-23 10:19:03 -07:00
Neil Armstrong
6df4393daf clk: oxnas: Add OX820 Gate clocks
Add support for the Oxford Semiconductor OX820 SoC gate clocks
along the OX810SE SoC support.
This rework on concerns the gate clocks since they are different.
Future PLL handling code will be added for OX820.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Link: lkml.kernel.org/r/20161005150752.22618-6-narmstrong@baylibre.com
2016-10-23 10:18:45 -07:00
Neil Armstrong
5a9e54a81b clk: oxnas: Refactor to make use of devm_clk_hw_register()
Make usage of static tables identified by the OF match table to
feed devm_clk_hw_register() and use of_clk_add_hw_provider().

This structure is cleaner and simplifies adding new SoC support while
having common probe and gate ops code.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Link: lkml.kernel.org/r/20161005150752.22618-5-narmstrong@baylibre.com
2016-10-23 10:18:42 -07:00
Neil Armstrong
1a2cfd0070 clk: oxnas: Rename to clk_oxnas_gate
Rename clock ops to clk_oxnas_gate in ops and structures.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Link: lkml.kernel.org/r/20161005150752.22618-4-narmstrong@baylibre.com
2016-10-23 10:18:39 -07:00
Neil Armstrong
5881456295 clk: oxnas: Add dt-bindings include file for OX820
In order to support the Oxford Semiconductor Gate clocks, add a
dedicated dt-binding include file for gate indexes.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Link: lkml.kernel.org/r/20161005150752.22618-3-narmstrong@baylibre.com
2016-10-23 10:18:36 -07:00
Neil Armstrong
6fc8ec8bfd clk: oxnas: Add dt-bindings include file for OX810SE
In order to prepare support for the Oxford Semiconductor OX820, add
a dt-bindings include file used by the ox810se dtsi.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Link: lkml.kernel.org/r/20161005150752.22618-2-narmstrong@baylibre.com
2016-10-23 10:18:33 -07:00
Vadim Pasternak
304887041d platform/x86: Introduce support for Mellanox hotplug driver
Enable system support for the Mellanox Technologies hotplug platform
driver, which provides support for the next Mellanox basic systems:
"msx6710", "msx6720", "msb7700", "msn2700", "msx1410", "msn2410",
"msb7800", "msn2740", "msn2100" and also various number of derivative
systems from the above basic types.
This driver handles hot-plug events for the power suppliers, power
cables and fans for the above systems.

The Kconfig currently controlling compilation of this code is:
driver/platform/x86:config MLX_CPLD_PLATFORM
                       tristate "Mellanox platform hotplug driver support"

Signed-off-by: Vadim Pasternak <vadimp@mellanox.com>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
2016-10-23 07:52:57 -07:00
Zhengyu Shen
e76bdfd740 ARM: imx: Added perf functionality to mmdc driver
MMDC is a multi-mode DDR controller that supports DDR3/DDR3L x16/x32/x64
and LPDDR2 two channel x16/x32 memory types. MMDC is configurable, high
performance, and optimized. MMDC is present on i.MX6 Quad and i.MX6
QuadPlus devices, but this driver only supports i.MX6 Quad at the moment.
MMDC provides registers for performance counters which read via this
driver to help debug memory throughput and similar issues.

$ perf stat -a -e mmdc/busy-cycles/,mmdc/read-accesses/,mmdc/read-bytes/,mmdc/total-cycles/,mmdc/write-accesses/,mmdc/write-bytes/ dd if=/dev/zero of=/dev/null bs=1M count=5000
Performance counter stats for 'dd if=/dev/zero of=/dev/null bs=1M count=5000':

         898021787      mmdc/busy-cycles/
          14819600      mmdc/read-accesses/
            471.30 MB   mmdc/read-bytes/
        2815419216      mmdc/total-cycles/
          13367354      mmdc/write-accesses/
            427.76 MB   mmdc/write-bytes/

       5.334757334 seconds time elapsed

Signed-off-by: Zhengyu Shen <zhengyu.shen@nxp.com>
Signed-off-by: Frank Li <frank.li@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
2016-10-23 20:39:01 +08:00
Jaret Cantu
371a8dd6fa ARM: dts: imx: b650v3: Calibrate USB PHY to pass eye diagram test
Calibrate the USB PHY TX settings to pass the eye diagram signal
integrity test.  The settings are taken from the i.MX6 reference
manual's recommended configuration for USB certification (66.2.6).

Signed-off-by: Jaret Cantu <jaret.cantu@timesys.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
2016-10-23 20:13:21 +08:00
David S. Miller
f18a37ff24 Merge branch 'qed-fixes'
Sudarsana Reddy Kalluru says:

====================
qed*: fix series.

The patch series contains several minor bug fixes for qed/qede modules.
Please consider applying this to 'net' branch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:07 -04:00
Manish Chopra
fabd545c6d qede: Fix incorrrect usage of APIs for un-mapping DMA memory
Driver uses incorrect APIs to unmap DMA memory which were
mapped using dma_map_single(). This patch fixes it to use
appropriate APIs for un-mapping DMA memory.

Signed-off-by: Manish Chopra <manish.chopra@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:07 -04:00
Sudarsana Reddy Kalluru
15c6de2c65 qed: Zero-out the buffer paased to dcbx_query() API
qed_dcbx_query_params() implementation populate the values to input
buffer based on the dcbx mode and, the current negotiated state/params,
the caller of this API need to memset the buffer to zero.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:07 -04:00
Sudarsana Reddy Kalluru
ed0dd91515 qede: Reconfigure rss indirection direction table when rss count is updated
Rx indirection table entries are in the range [0, (rss_count - 1)]. If
user reduces the rss count, the table entries may not be in the ccorrect
range. Need to reconfigure the table with new rss_count as a basis.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:07 -04:00
Sudarsana Reddy Kalluru
0e19182738 qed*: Reduce the memory footprint for Rx path
With the current default values for Rx path i.e., 8 queues of 8Kb entries
each with 4Kb size, interface will consume 256Mb for Rx. The default values
causing the driver probe to fail when the system memory is low. Based on
the perforamnce results, rx-ring count value of 1Kb gives the comparable
performance with Rx coalesce timeout of 12 seconds. Updating the default
values.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:07 -04:00
Sudarsana Reddy Kalluru
837d4eb6ed qede: Loopback implementation should ignore the normal traffic
During the execution of loopback test, driver may receive the packets which
are not originated by this test, loopback implementation need to skip those
packets.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:07 -04:00
Sudarsana Reddy Kalluru
ba300ce351 qede: Do not allow RSS config for 100G devices
RSS configuration is not supported for 100G adapters.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:06 -04:00
Sudarsana Reddy Kalluru
bdc8cbd34d qede: get_channels() need to populate max tx/rx coalesce values
Recent changes in kernel ethtool implementation requires the driver
callback for get_channels() has to populate the values for max tx/rx
coalesce fields.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:08:06 -04:00
David S. Miller
67dc159673 Merge branch 'bpf-numa-id'
Daniel Borkmann says:

====================
Add BPF numa id helper

This patch set adds a helper for retrieving current numa node
id and a test case for SO_REUSEPORT.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:06:09 -04:00
Daniel Borkmann
3c2c3c16aa reuseport, bpf: add test case for bpf_get_numa_node_id
The test case is very similar to reuseport_bpf_cpu, only that here
we select socket members based on current numa node id.

  # numactl -H
  available: 2 nodes (0-1)
  node 0 cpus: 0 1 2 3 4 5 12 13 14 15 16 17
  node 0 size: 128867 MB
  node 0 free: 120080 MB
  node 1 cpus: 6 7 8 9 10 11 18 19 20 21 22 23
  node 1 size: 96765 MB
  node 1 free: 87504 MB
  node distances:
  node   0   1
    0:  10  20
    1:  20  10

  # ./reuseport_bpf_numa
  ---- IPv4 UDP ----
  send node 0, receive socket 0
  send node 1, receive socket 1
  send node 1, receive socket 1
  send node 0, receive socket 0
  ---- IPv6 UDP ----
  send node 0, receive socket 0
  send node 1, receive socket 1
  send node 1, receive socket 1
  send node 0, receive socket 0
  ---- IPv4 TCP ----
  send node 0, receive socket 0
  send node 1, receive socket 1
  send node 1, receive socket 1
  send node 0, receive socket 0
  ---- IPv6 TCP ----
  send node 0, receive socket 0
  send node 1, receive socket 1
  send node 1, receive socket 1
  send node 0, receive socket 0
  SUCCESS

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:05:52 -04:00
Daniel Borkmann
2d0e30c30f bpf: add helper for retrieving current numa node id
Use case is mainly for soreuseport to select sockets for the local
numa node, but since generic, lets also add this for other networking
and tracing program types.

Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:05:52 -04:00
David S. Miller
a10b91b8b8 Merge branch 'udpmem'
Paolo Abeni says:

====================
udp: refactor memory accounting

This patch series refactor the udp memory accounting, replacing the
generic implementation with a custom one, in order to remove the needs for
locking the socket on the enqueue and dequeue operations. The socket backlog
usage is dropped, as well.

The first patch factor out pieces of some queue and memory management
socket helpers, so that they can later be used by the udp memory accounting
functions.
The second patch adds the memory account helpers, without using them.
The third patch replacse the old rx memory accounting path for udp over ipv4 and
udp over ipv6. In kernel UDP users are updated, as well.

The memory accounting schema is described in detail in the individual patch
commit message.

The performance gain depends on the specific scenario; with few flows (and
little contention in the original code) the differences are in the noise range,
while with several flows contending the same socket, the measured speed-up
is relevant (e.g. even over 100% in case of extreme contention)

Many thanks to Eric Dumazet for the reiterated reviews and suggestions.

v5 -> v6:
 - do not orphan the skb on enqueue, skb_steal_sock() already did
   the work for us

v4 -> v5:
 - use the receive queue spin lock to protect the memory accounting
 - several minor clean-up

v3 -> v4:
 - simplified the locking schema, always use a plain spinlock

v2 -> v3:
 - do not set the now unsed backlog_rcv callback

v1 -> v2:
 - changed slighly the memory accounting schema, we now perform lazy reclaim
 - fixed forward_alloc updating issue
 - fixed memory counter integer overflows
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:05:13 -04:00
Paolo Abeni
850cbaddb5 udp: use it's own memory accounting schema
Completely avoid default sock memory accounting and replace it
with udp-specific accounting.

Since the new memory accounting model encapsulates completely
the required locking, remove the socket lock on both enqueue and
dequeue, and avoid using the backlog on enqueue.

Be sure to clean-up rx queue memory on socket destruction, using
udp its own sk_destruct.

Tested using pktgen with random src port, 64 bytes packet,
wire-speed on a 10G link as sender and udp_sink as the receiver,
using an l4 tuple rxhash to stress the contention, and one or more
udp_sink instances with reuseport.

nr readers      Kpps (vanilla)  Kpps (patched)
1               170             440
3               1250            2150
6               3000            3650
9               4200            4450
12              5700            6250

v4 -> v5:
  - avoid unneeded test in first_packet_length

v3 -> v4:
  - remove useless sk_rcvqueues_full() call

v2 -> v3:
  - do not set the now unsed backlog_rcv callback

v1 -> v2:
  - add memory pressure support
  - fixed dropwatch accounting for ipv6

Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:05:05 -04:00
Paolo Abeni
f970bd9e3a udp: implement memory accounting helpers
Avoid using the generic helpers.
Use the receive queue spin lock to protect the memory
accounting operation, both on enqueue and on dequeue.

On dequeue perform partial memory reclaiming, trying to
leave a quantum of forward allocated memory.

On enqueue use a custom helper, to allow some optimizations:
- use a plain spin_lock() variant instead of the slightly
  costly spin_lock_irqsave(),
- avoid dst_force check, since the calling code has already
  dropped the skb dst
- avoid orphaning the skb, since skb_steal_sock() already did
  the work for us

The above needs custom memory reclaiming on shutdown, provided
by the udp_destruct_sock().

v5 -> v6:
  - don't orphan the skb on enqueue

v4 -> v5:
  - replace the mem_lock with the receive queue spin lock
  - ensure that the bh is always allowed to enqueue at least
    a skb, even if sk_rcvbuf is exceeded

v3 -> v4:
  - reworked memory accunting, simplifying the schema
  - provide an helper for both memory scheduling and enqueuing

v1 -> v2:
  - use a udp specific destrctor to perform memory reclaiming
  - remove a couple of helpers, unneeded after the above cleanup
  - do not reclaim memory on dequeue if not under memory
    pressure
  - reworked the fwd accounting schema to avoid potential
    integer overflow

Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:05:05 -04:00
Paolo Abeni
f8c3bf00d4 net/socket: factor out helpers for memory and queue manipulation
Basic sock operations that udp code can use with its own
memory accounting schema. No functional change is introduced
in the existing APIs.

v4 -> v5:
  - avoid whitespace changes

v2 -> v4:
  - avoid exporting __sock_enqueue_skb

v1 -> v2:
  - avoid export sock_rmem_free

Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 17:05:05 -04:00
WANG Cong
396a30cce1 ipv4: use the right lock for ping_group_range
This reverts commit a681574c99
("ipv4: disable BH in set_ping_group_range()") because we never
read ping_group_range in BH context (unlike local_port_range).

Then, since we already have a lock for ping_group_range, those
using ip_local_ports.lock for ping_group_range are clearly typos.

We might consider to share a same lock for both ping_group_range
and local_port_range w.r.t. space saving, but that should be for
net-next.

Fixes: a681574c99 ("ipv4: disable BH in set_ping_group_range()")
Fixes: ba6b918ab2 ("ping: move ping_group_range out of CONFIG_SYSCTL")
Cc: Eric Dumazet <edumazet@google.com>
Cc: Eric Salo <salo@google.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 16:23:12 -04:00
David S. Miller
860e0217fc Merge branch 'dsa-bcm_sf2-do-not-rely-on-kexec_in_progress'
Florian Fainelli says:

====================
net: dsa: bcm_sf2: Do not rely on kexec_in_progress

These are the two patches following the discussing we had on kexec_in_progress.

Feel free to apply or discard them, thanks!
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2016-10-22 16:17:54 -04:00