Commit graph

708,694 commits

Author SHA1 Message Date
Alexandre Belloni
ae3696c167 net: macb: fix phy interrupt parsing
Since 83a77e9ec4, the phydev irq is explicitly set to PHY_POLL when
there is no pdata. It doesn't work on DT enabled platforms because the
phydev irq is already set by libphy before.

Fixes: 83a77e9ec4 ("net: macb: Added PCI wrapper for Platform Driver.")
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-30 22:21:49 -04:00
Theodore Ts'o
72d622b422 ext4: replace BUG_ON with WARN_ONCE in ext4_end_bio()
Add fallback code and a WARN_ONCE() call instead of a BUG_ON() in
the ext4_end_bio() function.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-04-30 20:08:05 -04:00
Quentin Swain
b46c39a098 iio: ade7854: Convert symbolic permissions to octal
Convert symbolic S_IRUGO and S_IWUSR macros to octal to fix
warnings reported by checkpatch.pl

Signed-off-by: Quentin Swain <dudebrobro179@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-05-01 01:07:41 +01:00
Quentin Swain
78425b3290 iio: ade7758: Convert symbolic permissions to octal
Convert symbolic S_IRUGO and S_IWUSR macros to octal permissions to
fix warnings reported by checkpatch.pl

Signed-off-by: Quentin Swain <dudebrobro179@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-05-01 01:06:24 +01:00
Quentin Swain
37cbe596e2 iio: ade7754: Convert symbolic permissions to octal
Convert symbolic S_IRUGO and S_IWUSR macros to octal permissions
to resolve warnings reported by checkpatch.pl

Signed-off-by: Quentin Swain <dudebrobro179@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-05-01 01:05:17 +01:00
Quentin Swain
3e8eae9079 iio: ade7753 Convert: symbolic permissions to octal
Convert S_IRUGO and S_IWUSR macros to octal permissions to resolve
warnings reported by checkpatch.pl

Signed-off-by: Quentin Swain <dudebrobro179@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-05-01 01:01:44 +01:00
Quentin Swain
2451c5dda4 iio: ad9834 convert symbolic permissions to octal
Remove checkpatch warnings by converting symbolic S_IRUGO and S_IWUSR
permissions to octal

Signed-off-by: Quentin Swain <dudebrobro179@gmail.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-05-01 00:59:21 +01:00
Jan Kara
dddbd6ac8f ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.

Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:

274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.

After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:

0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s

[ Modified so we use an explicit do_map flag instead of relying on
  io_end not being allocated, the since io_end->inode is needed for I/O
  error handling. -- tytso ]

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-04-30 18:29:10 -04:00
Sebastian Reichel
09d9d82745 hwmon: (twl4030-madc) drop driver
This driver is no longer needed:

 * It has no mainline users
 * It has no DT support and OMAP is DT only
 * iio-hwmon can be used for madc, which also works with DT

Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
2017-04-30 11:45:31 -07:00
Linus Torvalds
97ce89f8a4 Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Thomas Gleixner:
 "The final fixes for 4.11:

   - prevent a triple fault with function graph tracing triggered via
     suspend to ram

   - prevent optimizing for size when function graph tracing is enabled
     and the compiler does not support -mfentry

   - prevent mwaitx() being called with a zero timeout as mwaitx() might
     never return. Observed on the new Ryzen CPUs"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  Prevent timer value 0 for MWAITX
  x86/build: convert function graph '-Os' error to warning
  ftrace/x86: Fix triple fault with graph tracing and suspend-to-ram
2017-04-30 11:44:18 -07:00
Linus Torvalds
14e07f023f Merge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fix from Thomas Gleixner:
 "A single fix for a cputime accounting regression which got introduced
  in the 4.11 cycle"

* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/cputime: Fix ksoftirqd cputime accounting regression
2017-04-30 11:35:00 -07:00
Lorenzo Bianconi
d3f770582a iio: imu: st_lsm6dsx: add system power management support
Add system sleep power management support to st_lsm6dsx driver.
In particular during suspend phase each sensor is disabled and
hw fifo is configured in bypass in order to avoid subsequent
I/O operations. The patch has been tested on HiKey board device

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@st.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-04-30 17:55:15 +01:00
Lorenzo Bianconi
535de397b2 iio: imu: st_lsm6dsx: modify st_lsm6dsx_flush_fifo and st_lsm6dsx_set_fifo_mode scope
Remove static qualifier from st_lsm6dsx_flush_fifo() and
st_lsm6dsx_set_fifo_mode() in order to use them in system sleep pm support

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@st.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-04-30 17:55:12 +01:00
Sebastian Reichel
e530cc8408 iio: isl29028: add isl29030 support
isl29030 is basically the same chip. The only difference
is the chip's first pin. For isl29028 its named ADDR0 and
can be used to change the chip's i2c address. For isl29030
on the other hand that pin is named Ials and is an analog
current output proportional to ALS/IR. This change is
irrelevant for the Linux driver.

This has been tested on Motorola Droid 4.

Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2017-04-30 17:26:04 +01:00
David S. Miller
8dd5b698c2 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 2017-04-30

This series contains updates to i40e and i40evf only.

Jake provides majority of the changes in this series, starting with the
renaming of a flag to avoid confusion.  Then renamed a variable to a
more meaningful name to clarify what is actually being done and to
reduce confusion.  Amortizes the wait time when initializing or disabling
lots of VFs by using i40e_reset_all_vfs() and
i40e_vsi_stop_rings_no_wait().  Cleaned up a unnecessary delay since
pci_disable_sriov() already has its own delay, so need to add a additional
delay when removing VFs.  Avoid using the same name flags for both
vsi->state and pf->state, to make code review easier and assist future
work to use the correct state field when checking bits.  Use
DECLARE_BITMAP() to ensure that we always allocate enough space for flags.
Replace hw_disabled_flags with the new _AUTO_DISABLED flags, which are
more readable because we are not setting an *_ENABLED flag to
disable the feature.

Alex corrects a oversight where we were not reprogramming the ports
after a reset, which was causing us to lose all of the receive tunnel
offloads.

Arnd Bergmann moves the declaration of a local variable to avoid a
warning seen on architectures with larger pages about an unused variable.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-30 11:33:37 -04:00
David S. Miller
5b36d8f5e5 Merge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue
Jeff Kirsher says:

====================
1GbE Intel Wired LAN Driver Updates 2017-04-30

This series contains updates to e1000e only.

Jarod Wilson fixes an issue where the workaround for 82574 & 82583
is needed for i218 as well, so set the appropriate flags.

Sasha adds support for the upcoming new i219 devices for the client
platform (CannonLake), which includes the support for 38.4MHz frequency
to support PTP on CannonLake.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-30 11:30:13 -04:00
Marcel Holtmann
71653eb64b Bluetooth: Add selftest for ECDH key generation
Since the ECDH key generation takes a different path, it needs to be
tested as well. For this generate the public debug key from the private
debug key and compare both.

This also moves the seeding of the private key into the SMP calling code
to allow for easier re-use of the ECDH key generation helper.

Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2017-04-30 16:52:43 +03:00
Marcel Holtmann
f958315358 Bluetooth: zero kpp input for key generation
When generating new ECDH keys with kpp, the shared secret input needs to
be set to NULL. Fix this by including kpp_request_set_input call.

Fixes: 58771c1c ("Bluetooth: convert smp and selftest to crypto kpp
API")
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2017-04-30 16:52:39 +03:00
Mark Brown
ae53b5dbaf Merge remote-tracking branch 'regulator/topic/vctrl' into regulator-next 2017-04-30 22:17:44 +09:00
Mark Brown
81bc8e386f Merge remote-tracking branches 'regulator/topic/notifier', 'regulator/topic/pfuze100', 'regulator/topic/settle', 'regulator/topic/tps65132' and 'regulator/topic/twl6030' into regulator-next 2017-04-30 22:17:36 +09:00
Mark Brown
0603b37e1e Merge remote-tracking branches 'regulator/topic/helpers', 'regulator/topic/hi655x', 'regulator/topic/lm363x', 'regulator/topic/ltc3589' and 'regulator/topic/ltc3676' into regulator-next 2017-04-30 22:17:31 +09:00
Mark Brown
59e4c636df Merge remote-tracking branches 'regulator/topic/anatop', 'regulator/topic/arizona', 'regulator/topic/bd9571mvw-m' and 'regulator/topic/const' into regulator-next 2017-04-30 22:17:25 +09:00
Mark Brown
bde1e61b8e Merge remote-tracking branch 'regulator/topic/core' into regulator-next 2017-04-30 22:17:23 +09:00
Mark Brown
6a8007c83a Merge remote-tracking branches 'regulator/fix/isl9305', 'regulator/fix/rk808' and 'regulator/fix/tps65023' into regulator-linus 2017-04-30 22:17:18 +09:00
Mark Brown
20d5c84bef Merge remote-tracking branches 'asoc/topic/wm8960', 'asoc/topic/wm8978' and 'asoc/topic/zte-tdm' into asoc-next 2017-04-30 22:16:48 +09:00
Mark Brown
258ced9d08 Merge remote-tracking branches 'asoc/topic/uda1380', 'asoc/topic/unregister', 'asoc/topic/ux500', 'asoc/topic/wm5100' and 'asoc/topic/wm8903' into asoc-next 2017-04-30 22:16:41 +09:00
Mark Brown
7b7b384957 Merge remote-tracking branches 'asoc/topic/tegra', 'asoc/topic/tlv320aic23', 'asoc/topic/topology', 'asoc/topic/twl6040' and 'asoc/topic/txx9' into asoc-next 2017-04-30 22:16:37 +09:00
Mark Brown
65fd5252b4 Merge remote-tracking branches 'asoc/topic/sta529', 'asoc/topic/sti', 'asoc/topic/stm', 'asoc/topic/sunxi' and 'asoc/topic/tas2552' into asoc-next 2017-04-30 22:16:31 +09:00
Mark Brown
ae17a14f14 Merge remote-tracking branches 'asoc/topic/samsung', 'asoc/topic/sgtl5000', 'asoc/topic/simple', 'asoc/topic/sirf' and 'asoc/topic/ssm4567' into asoc-next 2017-04-30 22:16:26 +09:00
Mark Brown
fc180c0eaf Merge remote-tracking branches 'asoc/topic/rockchip', 'asoc/topic/rt5514', 'asoc/topic/rt5645' and 'asoc/topic/rt5677' into asoc-next 2017-04-30 22:16:21 +09:00
Mark Brown
757fc30a1a Merge remote-tracking branches 'asoc/topic/nau8824', 'asoc/topic/omap', 'asoc/topic/pxa' and 'asoc/topic/qcom' into asoc-next 2017-04-30 22:16:18 +09:00
Mark Brown
ccf9fc86d6 Merge remote-tracking branches 'asoc/topic/max9878', 'asoc/topic/max98927', 'asoc/topic/mtk' and 'asoc/topic/nau8540' into asoc-next 2017-04-30 22:16:04 +09:00
Mark Brown
50946b2aec Merge remote-tracking branches 'asoc/topic/fsl-ssi', 'asoc/topic/hi6220' and 'asoc/topic/imx' into asoc-next 2017-04-30 22:16:01 +09:00
Mark Brown
0f57c12ab4 Merge remote-tracking branches 'asoc/topic/es7134', 'asoc/topic/es8328', 'asoc/topic/fsl', 'asoc/topic/fsl-asrc' and 'asoc/topic/fsl-esai' into asoc-next 2017-04-30 22:15:55 +09:00
Mark Brown
99dd3c53f6 Merge remote-tracking branches 'asoc/topic/cs35l35', 'asoc/topic/cs53l30', 'asoc/topic/da7213', 'asoc/topic/dio2125' and 'asoc/topic/dwc' into asoc-next 2017-04-30 22:15:50 +09:00
Mark Brown
2f42a77661 Merge remote-tracking branches 'asoc/topic/adsp', 'asoc/topic/ak4613', 'asoc/topic/au1761' and 'asoc/topic/blackfin' into asoc-next 2017-04-30 22:15:45 +09:00
Mark Brown
cce9b271e5 Merge remote-tracking branch 'asoc/topic/rt5665' into asoc-next 2017-04-30 22:15:43 +09:00
Mark Brown
245e302a4d Merge remote-tracking branch 'asoc/topic/rcar' into asoc-next 2017-04-30 22:15:42 +09:00
Mark Brown
0c2964cb38 Merge remote-tracking branch 'asoc/topic/intel' into asoc-next 2017-04-30 22:15:41 +09:00
Mark Brown
d872f04606 Merge remote-tracking branch 'asoc/topic/core' into asoc-next 2017-04-30 22:15:40 +09:00
Mark Brown
9dfc53942d Merge remote-tracking branch 'asoc/fix/cs4271' into asoc-linus 2017-04-30 22:15:37 +09:00
Mark Brown
d0a906d3a1 Merge remote-tracking branch 'asoc/fix/rcar' into asoc-linus 2017-04-30 22:15:36 +09:00
Mark Brown
ec7f9844ef ASoC: Fixes for v4.11
A few last minute fixes for v4.11, the STI fix is relatively large but
 driver specific and has been cooking in -next for a little while now:
 
  - A fix from Takashi for some suspend/resume related crashes in the
    Intel drivers.
  - A fix from Mousumi Jana for issues with incorrectly created
    enumeration controls generated from topology files which could cause
    problems for userspace.
  - Fixes from Arnaud Pouliquen for some crashes due to races with the
    interrupt handler in the STI driver.
 -----BEGIN PGP SIGNATURE-----
 
 iQFHBAABCAAxFiEEreZoqmdXGLWf4p/qJNaLcl1Uh9AFAlj/auUTHGJyb29uaWVA
 a2VybmVsLm9yZwAKCRAk1otyXVSH0JEJB/9fZoEPm2Flfp9d4tXDmNihyv/STARr
 UbIIBxJUsDG2U6nMxv1kTKRk/9puuzDfNjt4giiLyMNDMbh4G5QRbgMF53ryeTuM
 8Nqq6PEvhZ4/YftOnh4uD4PmsVKTV8OauQDzZquLPB9/lRF6wLvSskSfZXTsYbv9
 Oem4Sop1XzXt68cUIcMHQlwXrhxIaQUHiOS8QfhQXNw1y5xgGB5Rj0ak5fWdn+KY
 6j10ldQgrwYSCx4zbqLsS6Fjlo2zdUj/Rn7fQrarzsjQtrZq+k6iYvAWAblkMhWi
 7hL/RK4U0konSeo12ftIESDpZFjW2Yi9Q74+4i6soEYbOVTzMam+XMo+
 =qmpO
 -----END PGP SIGNATURE-----

Merge tag 'asoc-fix-v4.11-rc7' into asoc-linus

ASoC: Fixes for v4.11

A few last minute fixes for v4.11, the STI fix is relatively large but
driver specific and has been cooking in -next for a little while now:

 - A fix from Takashi for some suspend/resume related crashes in the
   Intel drivers.
 - A fix from Mousumi Jana for issues with incorrectly created
   enumeration controls generated from topology files which could cause
   problems for userspace.
 - Fixes from Arnaud Pouliquen for some crashes due to races with the
   interrupt handler in the STI driver.

# gpg: Signature made Wed 26 Apr 2017 00:27:33 KST
# gpg:                using RSA key ADE668AA675718B59FE29FEA24D68B725D5487D0
# gpg:                issuer "broonie@kernel.org"
# gpg: key 0D9EACE2CD7BEEBC: no public key for trusted key - skipped
# gpg: key 0D9EACE2CD7BEEBC marked as ultimately trusted
# gpg: key CCB0A420AF88CD16: no public key for trusted key - skipped
# gpg: key CCB0A420AF88CD16 marked as ultimately trusted
# gpg: key 162614E316005C11: no public key for trusted key - skipped
# gpg: key 162614E316005C11 marked as ultimately trusted
# gpg: key A730C53A5621E907: no public key for trusted key - skipped
# gpg: key A730C53A5621E907 marked as ultimately trusted
# gpg: key 276568D75C6153AD: no public key for trusted key - skipped
# gpg: key 276568D75C6153AD marked as ultimately trusted
# gpg: Good signature from "Mark Brown <broonie@sirena.org.uk>" [ultimate]
# gpg:                 aka "Mark Brown <broonie@debian.org>" [ultimate]
# gpg:                 aka "Mark Brown <broonie@kernel.org>" [ultimate]
# gpg:                 aka "Mark Brown <broonie@tardis.ed.ac.uk>" [ultimate]
# gpg:                 aka "Mark Brown <broonie@linaro.org>" [ultimate]
# gpg:                 aka "Mark Brown <Mark.Brown@linaro.org>" [ultimate]
2017-04-30 22:15:35 +09:00
Daniel Baluta
82bab88910 ASoC: codec: wm8960: Relax bit clock computation when using PLL
Bitclk is derived from sysclk using bclk_divs.
Sysclk can be derived in two ways:
	(1) directly from MLCK
	(2) MCLK via PLL

Commit 3c01b9ee2a ("ASoC: codec: wm8960: Relax bit clock
computation")
relaxed bitclk computation when sysclk is directly derived from MCLK.

Lets do the same thing when sysclk is derived via PLL.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
2017-04-30 22:14:32 +09:00
Daniel Baluta
66772eda0e ASoC: codec: wm9860: avoid maybe-uninitialized warning
The new PLL configuration code triggers a harmless warning:

sound/soc/codecs/wm8960.c: In function 'wm8960_configure_clocking':
sound/soc/codecs/wm8960.c:735:3: error: 'best_freq_out' may be used
uninitialized in this function [-Werror=maybe-uninitialized]
   wm8960_set_pll(codec, freq_in, best_freq_out);
   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
sound/soc/codecs/wm8960.c:699:12: note: 'best_freq_out' was declared
here

Fix this by reworking the code such that:

1) When there is no PLL freq available return -EINVAL and make
sure *bclk_idx, *dac_idx, *sysclk_idx are initialized with
invalid values.

2) When there is a PLL freq available initialize *bclk_idx,
*dac_idx and *sysclk_idx with correct values and immediately
return the freq available.

Fixes: 84fdc00d51 ("ASoC: codec: wm9860: Refactor PLL out freq search")
Fixes: 303e8954af ("ASoC: codec: wm8960: Stop when a matching PLL freq is found")
Suggested-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
2017-04-30 22:14:27 +09:00
Eli Cohen
0a0ab1d2cc net/mlx5: E-Switch, Avoid redundant memory allocation
struct esw_mc_addr is a small struct that can be part of struct
mlx5_eswitch. Define it as a field and not as a pointer and save the
kzalloc call and then error flow handling.

Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2017-04-30 16:03:21 +03:00
Eran Ben Elisha
0f6e4cf674 net/mlx5e: Disable HW LRO when PCI is slower than link on striding RQ
We will activate the HW LRO only on servers with PCI BW > MAX LINK BW,
or when PCI BW > 16Gbps. On other cases we do not want LRO by default as
LRO sessions might get timeout and add redundant software overhead.

Tested:
	ethtool -k <ifs-name> | grep large-receive-offload
	On systems with and without the limitations.

Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Cc: kernel-team@fb.com
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2017-04-30 16:03:20 +03:00
Tariq Toukan
b1b03bded1 net/mlx5e: Use u8 as ownership type in mlx5e_get_cqe()
CQE ownership indication is as small as a single bit.
Use u8 to speedup the comparison.

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Cc: kernel-team@fb.com
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2017-04-30 16:03:19 +03:00
Tariq Toukan
ad78af9b88 net/mlx5e: Use prefetchw when a write is to follow
"prefetchw()" prefetches the cacheline for write. Use it for
skb->data, as soon we'll be copying the packet header there.

Performance:
Single-stream packet-rate tested with pktgen.
Packets are dropped in tc level to zoom into driver data-path.
Larger gain is expected for smaller packets, as less time
is spent on handling SKB fragments, making the path shorter
and the improvement more significant.

---------------------------------------------
packet size | before    | after     | gain  |
64B         | 4,113,306 | 4,778,720 |  16%  |
1024B       | 3,633,819 | 3,950,593 | 8.7%  |

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Cc: kernel-team@fb.com
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2017-04-30 16:03:17 +03:00
Tariq Toukan
1f5b1e47ee net/mlx5e: Optimize poll ICOSQ completion queue
UMR operations are more frequent and important.
Check them first, and add a compiler branch predictor hint.

According to current design, ICOSQ CQ can contain at most one
pending CQE per napi. Poll function is optimized accordingly.

Performance:
Single-stream packet-rate tested with pktgen.
Packets are dropped in tc level to zoom into driver data-path.
Larger gain is expected for larger packet sizes, as BW is higher
and UMR posts are more frequent.

---------------------------------------------
packet size | before    | after     | gain  |
64B         | 4,092,370 | 4,113,306 |  0.5% |
1024B       | 3,421,435 | 3,633,819 |  6.2% |

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Cc: kernel-team@fb.com
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2017-04-30 16:03:16 +03:00