Commit graph

661,541 commits

Author SHA1 Message Date
Hans de Goede
7508f44129 power: supply: axp288_charger: Actually get and use the USB_HOST extcon device
Nothing was setting info->otg.cable, so the extcon_get_cable_state_
calls on it would always return -EINVAL.

This commit fixes this by actually setting info->otg.cable using the new
extcon_get_extcon_dev_by_cable_id function.

This commit also makes failing to register the extcon notifier for the
USB_HOST cable an error rather then a warning, because we MUST have this
notfier to properly disable the VBUS path when in host mode so that we're
not drawing current from the 5V boost converter which is supplying power
to the otg port when in host mode.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:02:52 +01:00
Hans de Goede
d96e07350a power: supply: axp288_charger: Move init_hw_regs call before supply registration
Move the charger_init_hw_regs() above the power_supply_register call,
the axp288_charger_usb_set_property() uses axp288_chrg_info.max_cv and
.max_cc which get set by charger_init_hw_regs().

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:02:47 +01:00
Hans de Goede
42e2008a66 power: supply: axp288_charger: Register extcon notifers after power_supply
The extcon notifier work calls power_supply_changed on the power_supply
we register, so the extcon notifiers should be registered after we
register the power_supply.

While touching this code anyways, refactor the code for the 3 cable types
into a loop to avoid code repetition.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:02:39 +01:00
Hans de Goede
e3668e37f9 power: supply: axp288_charger: Use devm_power_supply_register
Use devm_power_supply_register instead of power_supply_register,
this avoids the need to do manual cleanup and results in quite
a nice code cleanup.

Note it may seem excessive to add a "struct device *dev" helper local
variable for the 1 time it is used in this patch, but future patches
in this series also use it.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:02:31 +01:00
Sebastian Reichel
8bb17b6c83 Merge branch 'psy-mfd-axp288-immutable' into psy-next 2017-01-04 22:01:42 +01:00
Nicolas Saenz Julienne
feb583e37f power: supply: add sbs-charger driver
This adds support for sbs-charger compilant chips as defined here:
http://sbs-forum.org/specs/sbc110.pdf

This was tested on a arm board connected to an LTC4100 battery charger
chip.

Signed-off-by: Nicolas Saenz Julienne <nicolas.saenz@prodys.net>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:01:00 +01:00
Nicolas Saenz Julienne
63219a2fe4 dt-bindings: power: supply: sbs-charger bindings
Adds device tree documentation for SBS charger compilant devices as defined
here: http://sbs-forum.org/specs/sbc110.pdf

Signed-off-by: Nicolas Saenz Julienne <nicolas.saenz@prodys.net>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:55 +01:00
Peter Rosin
a7d143d42b power: supply: bq24735: always check for AC adapter presence in probe
So what if there is a status_gpio specified? bq24735_charger_is_present()
do have a working fallback for the case of no status_gpio.

Simplify this by not special casing setups w/o status_gpio, folding
two consecutive if-blocks in the process.

Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:52 +01:00
Peter Rosin
a07bea32c6 power: supply: bq24735: configure the charger as part of enabling it
During probe, it makes no sense to take care to first not issue any
i2c commands to verify if the connected part really is a bq24735, to
later simply fail the probe in the next step when trying to configure
the charger. So, delay configuration of the charging parameters until
the charger is accessible (i.e. when the AC adapter is present) as
part of enabling the charging.

This also fixes the rather serious issue that the charging parameters
are lost when the AC adapter is disconnected.

Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:48 +01:00
Peter Rosin
2e66585ca2 power: supply: bq24735: move down bq24735_{en,dis}able_charging
bq24735_enable_charging() needs to call bq24735_config_charging(),
which is something to change later, this is just a preparatory patch.

Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:41 +01:00
Milo Kim
5b903a1555 power: supply: tps65217: Use generic charger name
"tps65217-charger" is more appropriate name because the driver supports
not only AC but also USB charger.

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:37 +01:00
Milo Kim
757620c4e4 power: supply: tps65217: Use generic name for get_property()
Rename it as tps65217_charger_get_property().

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:33 +01:00
Milo Kim
da50b3a584 power: supply: tps65217: Use generic name for power supply property
Replace 'ac_props' with 'charger_props'.

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:29 +01:00
Milo Kim
3967d1f91c power: supply: tps65217: Use generic name for power supply structure
Replace 'ac' of tps65217_charger structure with 'psy'.

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:25 +01:00
Milo Kim
3c2e58a6d7 power: supply: tps65217: Use generic name for charger online
This driver supports AC and USB chargers. Generic name is preferred.
Replace 'ac_online' with 'online'.

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:21 +01:00
Milo Kim
20a7e173c2 power: supply: tps65217: Support USB charger interrupt
TPS65217 has two charger interrupts - AC and USB power status change.

Interrupt number in the TPS65217 driver data:
  IRQ number is only used on requesting the interrupt, so no need to keep
  it inside the driver data.

Interrupt handler:
  Check not only AC but also USB charger status.
  In both cases, enable charging operation.

Interrupt request:
  If an interrupt number is invalid, then use legacy polling thread.
  Otherwise, create IRQ threads to handle AC and USB charger event.

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:18 +01:00
Milo Kim
9ef0bf1184 power: supply: tps65217: Use 'poll_task' on unloading the module
Use the task_struct variable for running polling thread. If polling task
is activated, then use it to stop running thread.
This is a preceding step of supporting two interrupts of TPS65217 charger,
so checking single IRQ number is not appropriate when the module is removed.

Signed-off-by: Milo Kim <woogyom.kim@gmail.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:14 +01:00
Chanwoo Choi
bdca5d9e6b power: supply: qcom_smbb: Replace the deprecated extcon API
This patch replaces the deprecated extcon API as following:
- extcon_set_cable_state_() -> extcon_set_state_sync()

Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:07 +01:00
Chanwoo Choi
c31480348f power: supply: axp288_charger: Replace the extcon API
This patch uses the resource-managed extcon API for extcon_register_notifier()
and replaces the deprecated extcon API as following:
- extcon_get_cable_state_() -> extcon_get_state()

Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 22:00:02 +01:00
Srikant Ritolia
a50b0dbbaf power: supply: max8997_charger: Using device managed API and remove OOM print
Use managed resource function devm_power_supply_register instead of
power_supply_register to simplify the error path by allowing unregistering
to happen automatically on error and remove.
Removing max8997_battery_remove function also as it is now redundant.

Also removing out of memory printk message after kzalloc as there is
already enough information on failure.

Signed-off-by: Srikant Ritolia <s.ritolia@samsung.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:56 +01:00
Quentin Schulz
50111d3f88 power: supply: axp20x_usb_power: add 100mA max current limit for AXP223
The X-Powers AXP223 shares most of its behaviour with the AXP221 PMIC
but allows the VBUS power supply max current to be set to 100mA (like
the AXP209 PMIC).

This basically adds a new compatible to the VBUS power supply driver and
adds a check on the compatible when setting current max limit.

Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>
Acked-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:46 +01:00
Quentin Schulz
ac88bebeda dt-bindings: power: supply: axp20x_usb_power: add axp223 compatible
This adds the "x-powers,axp223-usb-power-supply" to the list of
compatibles for AXP20X VBUS power supply driver.

Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>
Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Acked-by: Rob Herring <robh@kernel.org>
Acked-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:40 +01:00
Quentin Schulz
5c3ff59b4e power: supply: axp20x_usb_power: set min voltage and max current from sysfs
AXP20X and AXP22X PMICs allow setting the min voltage and max current of
VBUS power supply. This adds entries in sysfs to allow to do so.

Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>
Acked-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:34 +01:00
Quentin Schulz
0dcc70ca86 power: supply: axp20x_usb_power: use of_device_id data field instead of device_is_compatible
This replaces calls to of_device_is_compatible to check data field of
of_device_id matched when probing the driver.

Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>
Acked-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:30 +01:00
Peter Rosin
bf383fea1f power: supply: bq24735-charger: optionally poll the ac-detect gpio
If the ac-detect gpio does not support interrupts, provide a fallback
to poll the gpio at a configurable interval.

Signed-off-by: Peter Rosin <peda@axentia.se>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:23 +01:00
Hans de Goede
248efcf006 power: supply: axp288_fuel_gauge: Read 12 bit values 2 registers at a time
In order for the MSB -> LSB latching to work correctly we must read the
2 8 bit registers of a 12 bit value in one consecutive read.

This fixes voltage_ocv reporting inconsistent values on my tablet.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:17 +01:00
Hans de Goede
4949fc5e07 power: supply: axp288_fuel_gauge: Read 15 bit values 2 registers at a time
In order for the MSB -> LSB latching to work correctly we must read the
2 8 bit registers of a 15 bit value in one consecutive read.

This fixes charge_full reporting 3498768 on some reads and 3354624 one
other reads on my tablet (for the 3354624 value the raw LSB is 0x00).

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:13 +01:00
Hans de Goede
6f074bc878 power: supply: axp288_fuel_gauge: Fix fuel_gauge_reg_readb return on error
If reading the register fails, return the actual error code, instead
of the uninitialized val variable;

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:59:08 +01:00
Peter Rosin
94c7073fe3 dt-bindings: power: supply: bq24735: reverse the polarity of ac-detect
The ACOK pin on the bq24735 is active-high, of course meaning that when
AC is OK the pin is high. However, all Tegra dts files have incorrectly
specified active-high even though the signal is inverted on the Tegra
boards. This has worked since the Linux driver has also inverted the
meaning of the GPIO. Fix this situation by simply specifying in the
bindings what everybody else agrees on; that the ti,ac-detect-gpios is
active on AC adapter absence.

Signed-off-by: Peter Rosin <peda@axentia.se>
Acked-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:58:53 +01:00
Peter Rosin
d0ddcba9e9 power: supply: bq24735-charger: simplify register update to stop charging
Providing value bits outside of the mask is pointless.

Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:58:42 +01:00
Hans de Goede
888f97435a power: supply: axp288_fuel_gauge: Drop platform_data dependency
When the axp288_faul_gauge driver was originally merged, it was
merged with a dependency on some other driver providing platform
data for it.

However the battery-data-framework which should provide that data
never got merged, resulting in x86 tablets / laptops with an axp288
having no working battery monitor, as before this commit the driver
would simply return -ENODEV if there is no platform data.

This commit removes the dependency on the platform_data instead
checking that the firmware has initialized the fuel-gauge and
reading the info back from the pmic.

What is missing from the read-back info is the table to map raw adc
values to temperature, so this commit drops the temperature and
temperature limits properties. The min voltage, charge design and
model name info is also missing. Note that none of these are really
important for userspace to have.

All other functionality is preserved and actually made available
by this commit.

BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=88471
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:56:10 +01:00
Hans de Goede
eac53b3664 power: supply: axp288_charger: Drop platform_data dependency
When the axp288_charger driver was originally merged, it was merged with
a dependency on some other driver providing platform data for it.

However the battery-data-framework which should provide that data never
got merged, so the axp288_charger as merged upstream has never worked,
its probe method simply always returns -ENODEV.

This commit removes the dependency on the platform_data instead reading
back the charging current and charging voltage that the firmware has set
and using those values as the maximum values the user may set.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:54:58 +01:00
Hans de Goede
d556f21cb0 power: supply: axp288_charger: Make charger_init_hw_regs propagate i2c errors
Make charger_init_hw_regs propagate i2c errors, instead of only warning
about them and then ignoring them.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
2017-01-04 21:54:53 +01:00
Takashi Iwai
c7efff9284 ALSA: hda - Apply asus-mode8 fixup to ASUS X71SL
Although the old quirk table showed ASUS X71SL with ALC663 codec being
compatible with asus-mode3 fixup, the bugzilla reporter explained that
asus-model8 fits better for the dual headphone controls.  So be it.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=191781
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2017-01-04 21:38:16 +01:00
Jonathan Cameron
49b2fd6ea6 docs: IIO documentation sphinx conversion
This is a manual conversion of the existing DocBook documentation
for IIO.  The intent is not to substantially change any of the
content in this patch, but to give a base to build upon.

Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
2017-01-04 13:24:41 -07:00
Paul Gortmaker
c6ef7fd40e vfio-mdev: fix non-standard ioctl return val causing i386 build fail
What appears to be a copy and paste error from the line above gets
the ioctl a ssize_t return value instead of the traditional "int".

The associated sample code used "long" which meant it would compile
for x86-64 but not i386, with the latter failing as follows:

  CC [M]  samples/vfio-mdev/mtty.o
samples/vfio-mdev/mtty.c:1418:20: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types]
  .ioctl          = mtty_ioctl,
                    ^
samples/vfio-mdev/mtty.c:1418:20: note: (near initialization for ‘mdev_fops.ioctl’)
cc1: some warnings being treated as errors

Since in this case, vfio is working with struct file_operations; as such:

    long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long);
    long (*compat_ioctl) (struct file *, unsigned int, unsigned long);

...and so here we just standardize on long vs. the normal int that user
space typically sees and documents as per "man ioctl" and similar.

Fixes: 9d1a546c53 ("docs: Sample driver to demonstrate how to use Mediated device framework.")
Cc: Kirti Wankhede <kwankhede@nvidia.com>
Cc: Neo Jia <cjia@nvidia.com>
Cc: kvm@vger.kernel.org
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
2017-01-04 13:22:38 -07:00
Geliang Tang
30965eeab5 of: drop duplicate headers
Drop duplicate headers string.h and of_platform.h.

Signed-off-by: Geliang Tang <geliangtang@gmail.com>
Signed-off-by: Rob Herring <robh@kernel.org>
2017-01-04 14:01:28 -06:00
Smitha T Murthy
a3b398e6f2 of: reserved_mem: set dma_ops for devices using reserved mem
For some IPs, there may be virtual child devices created and for them its
necessary to set the dma_ops if it's using reserved memory else it will call
the dummy dma_ops during buffer operations for the child devices which will
lead to memory mapping failure.

Signed-off-by: Smitha T Murthy <smitha.t@samsung.com>
Signed-off-by: Pankaj Dubey <pankaj.dubey@samsung.com>
Acked-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Rob Herring <robh@kernel.org>
2017-01-04 14:00:45 -06:00
Vladimir Zapolskiy
cf9e1672a6 mtd: nand: lpc32xx: fix invalid error handling of a requested irq
Semantics of NR_IRQS is different on machines with SPARSE_IRQ option
disabled or enabled, in the latter case IRQs are allocated starting
at least from the value specified by NR_IRQS and going upwards, so
the check of (irq >= NR_IRQ) to decide about an error code returned by
platform_get_irq() is completely invalid, don't attempt to overrule
irq subsystem in the driver.

The change fixes LPC32xx NAND MLC driver initialization on boot.

Fixes: 8cb17b5ed0 ("irqchip: Add LPC32xx interrupt controller driver")
Cc: stable@kernel.org # v4.7+
Signed-off-by: Vladimir Zapolskiy <vz@mleia.com>
Acked-by: Sylvain Lemieux <slemieux.tyco@gmail.com>
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
2017-01-04 20:50:18 +01:00
Edward Cree
4fdda95893 sfc: don't report RX hash keys to ethtool when RSS wasn't enabled
If we failed to set up RSS on EF10 (e.g. because firmware declared
 RX_RSS_LIMITED), ethtool --show-nfc $dev rx-flow-hash ... should report
 no fields, rather than confusingly reporting what fields we _would_ be
 hashing on if RSS was working.

Fixes: dcb4123cbe ("sfc: disable RSS when unsupported")
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 14:18:19 -05:00
Arjun V
8eb9f2f9e4 cxgb4: Support compressed error vector for T6
t6fw-1.15.15.0 enabled compressed error vector in cpl_rx_pkt for T6.
Updating driver to take care of these changes.

Signed-off-by: Santosh Rastapur <santosh@chelsio.com>
Signed-off-by: Arjun V <arjun@chelsio.com>
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 14:01:53 -05:00
David S. Miller
cee3548db1 Merge branch 'sh_eth-intrs-cleanup'
Sergei Shtylyov says:

====================
sh_eth: E-MAC interrupt handler cleanups

   Here's a set of 3 patches against DaveM's 'net-next.git' repo. I'm cleaning
up the E-MAC interrupt handling with the main goal of factoring out the E-MAC
interrupt handler into a separate function.

[1/3] sh_eth: handle only enabled E-MAC interrupts
[2/3] sh_eth: no need for *else* after *goto*
[3/3] sh_eth: factor out sh_eth_emac_interrupt()
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:47:55 -05:00
Sergei Shtylyov
9b39f05ce8 sh_eth: factor out sh_eth_emac_interrupt()
The E-MAC interrupt (EESR.ECI) is not always caused  by an error condition,
so  it really shouldn't be handled by sh_eth_error(). Factor out the E-MAC
interrupt handler, sh_eth_emac_interrupt(),  removing the ECI bit from the
EESR's values throughout the driver...

Update Cogent Embedded's copyright and clean up the whitespace in Renesas'
copyright, while at it...

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:47:55 -05:00
Sergei Shtylyov
1940f24076 sh_eth: no need for *else* after *goto*
Well, checkpatch.pl complains about *else* after *return* and *break* but
not after *goto*... and it probably should have complained about the code
in sh_eth_error().  Win couple LoCs by removing that *else*. :-)

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:47:54 -05:00
Sergei Shtylyov
4063469971 sh_eth: handle only enabled E-MAC interrupts
The driver should only handle the enabled E-MAC interrupts, like it does
for the E-DMAC interrupts since commit 3893b27345 ("sh_eth: workaround
for spurious ECI interrupt"),  so mask ECSR with  ECSIPR when reading it
in sh_eth_error().

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:47:54 -05:00
David S. Miller
aa9773be2a Merge branch 'dpaa_eth-fixes'
Madalin Bucur says:

====================
dpaa_eth: a couple of fixes

Add cleanup on PHY initialization failure path, avoid using
uninitialized memory at CGR init.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:45:09 -05:00
Roy Pledge
0fbb0f24dd dpaa_eth: Initialize CGR structure before init
The QBMan CGR options needs to be zeroed before calling the init
function

Signed-off-by: Roy Pledge <roy.pledge@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:45:09 -05:00
Madalin Bucur
3fe61f0940 dpaa_eth: cleanup after init_phy() failure
Signed-off-by: Madalin Bucur <madalin.bucur@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-04 13:45:08 -05:00
Chris Wilson
b1ed35d917 drm/i915: Revoke fenced GTT mmapings across GPU reset
The fence registers are clobbered by a GPU reset. If there is concurrent
user access to a fenced region via a GTT mmaping, the access will not be
fenced during the reset (until we restore the fences afterwards). In order
to prevent invalid access during the reset, before we clobber the fences
first we must invalidate the GTT mmapings. Access to the mmap will then
be forced to fault in the page, and in handling the fault, i915_gem_fault()
will take the struct_mutex and wait upon the reset to complete.

v2: Fix up commentary.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99274
Testcase: igt/gem_mmap_gtt/hang
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170104145110.1486-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
2017-01-04 18:44:30 +00:00
Andy Shevchenko
2501452160 spi: pxa2xx-pci: Enable DMA for Intel Merrifield
SPI controller on Intel Merrifield is backed by DMA engine. Add necessary bits
to support it.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
2017-01-04 18:36:41 +00:00