imx_get_temp might be called before the sensor clock is prepared
thus resulting in a timeout of the first attempt to read temp:
thermal thermal_zone0: failed to read out thermal zone 0
Happened to me on a Utilite Standard with IMX6 Dual SoC.

Reason is that in imx_thermal_probe thermal_zone_device_register
is called before the sensor clock is prepared.
thermal_zone_device_register however calls
thermal_zone_device_update which eventually calls imx_get_temp.

Fix this by preparing the clock before calling
thermal_zone_device_register.
Signed-off-by: NHeiner Kallweit <heiner.kallweit@web.de>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

In exynos5260_tmu_registers tmu_ctrl entry is erroneously
assigned twice.  The second assignment (to EXYNOS_TMU_REG_CONTROL1
define which represents 0x24 value) overrides the first one
(to EXYNOS_TMU_REG_CONTROL define which represents 0x20 value)
which results in the wrong (according to the Exynos5260 SoC
documentation that I have) offset being used for TMU_CONTROL
register.  Fix it by removing the wrong assignment and then
remove no longer used EXYNOS_TMU_REG_CONTROL1 define.

Cc: Naveen Krishna Chatradhi <ch.naveen@samsung.com>
Cc: Amit Daniel Kachhap <amit.daniel@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

thermal driver should be regisetered after cpufreq driver has
been registered and probed. Doing so is to make sure that thermal
driver can get the max cpu cooling states correctly when calling
get_property.
Signed-off-by: NBai Ping <b51503@freescale.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Thermal hysteresis represents a temperature difference.
But the original code treats it as a temperature value,
Convert it from tenths of degree Kelvin to Milli-Celsius
by deducing 273200. This is not right.

Kelvin and Celsius have same degree size. From temperature
difference view, the conversion between tenths of degree
Kelvin unit and Milli-Celsius unit is just to multiply 100.
Signed-off-by: NLan Tianyu <tianyu.lan@intel.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

result is always zero when comes here.
Signed-off-by: NYao Dongdong <yaodongdong@huawei.com>
Acked-by: NEduardo Valentin <edubezval@gmail.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

* Factor out code for clearing raised IRQs from exynos_tmu_work() to
  exynos_tmu_clear_irqs().

* Add a comment about documentation bugs to exynos_tmu_clear_irqs().

  [ The documentation for Exynos3250, Exynos4412, Exynos5250 and
    Exynos5260 incorrectly states that INTCLEAR register has
    a different placing of bits responsible for FALL IRQs than
    INTSTAT register.  Exynos5420 and Exynos5440 documentation is
    correct (Exynos4210 doesn't support FALL IRQs at all). ]

* Use exynos_tmu_clear_irqs() in exynos_tmu_initialize() instead
  of open-coded code trying to clear IRQs according to predefined
  masks.  After this change exynos_tmu_initialize() just clears
  IRQs that are raised like it is already done in exynos_tmu_work().

  As a nice side-effect the code now uses the correct offset
  (16 instead of 12) for bits responsible for clearing FALL IRQs
  in INTCLEAR register on Exynos3250, Exynos4412 and Exynos5250.

* Remove no longer needed intclr_rise_[mask,shift] and
  intclr_fall_[mask,shift] fields from struct exynos_tmu_registers.

* Remove no longer needed defines.

This patch has been tested on Exynos4412 and Exynos5420 SoCs.

Cc: Amit Daniel Kachhap <amit.daniel@samsung.com>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Here on function return all temporarily used device nodes shall
decrement their usage counter. The problems are found with device
nodes allocated by for_each_child_of_node(), of_parse_phandle()
and of_find_node_by_name(), fix all problems at once.
Signed-off-by: NVladimir Zapolskiy <vz@mleia.com>
Cc: devicetree@vger.kernel.org
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

This patch add support for TRIM_RELOAD feature at Exynos3250. The TMu of
Exynos3250 has two TRIMINFO_CON register and must need to set RELOAD bit
before reading TRIMINFO register.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Amit Daniel Kachhap <amit.daniel@samsung.com>
Reviewed-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

This patch support many TRIMINFO_CTRL registers if specific Exynos SoC
has one more TRIMINFO_CTRL registers. Also this patch uses proper 'RELOAD'
shift/mask bit operation to set RELOAD feature instead of static value.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Amit Daniel Kachhap <amit.daniel@samsung.com>
Reviewed-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Currently these SoCs claim TRIM_RELOAD support but don't have
triminfo_ctrl register address defined in their struct
exynos_tmu_registers entries.  This causes incorrect write of
value "1" to data->base + 0x00 address (which happens to be
TRIMINFO register).  Additionally according to the documentation
that I have neither Exynos5260 nor Exynos5420 support/require
TRIM_RELOAD feature.  Thus fix the aforementioned issue by
removing TMU_SUPPORT_TRIM_RELOAD flag for both Exynos5260 and
Exynos5420.

Cc: Naveen Krishna Chatradhi <ch.naveen@samsung.com>
Cc: Amit Daniel Kachhap <amit.daniel@samsung.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

There is no need for abstracting configuration for registers that
are identical on all SoC types.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

pdata->reference_voltage and pdata->gain are always defined
to non-zero values so remove the redundant checks from
exynos_tmu_control().

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Cache number of non-hardware trigger levels in a new pdata field
(non_hw_trigger_levels) and convert code in exynos_tmu_initialize()
accordingly.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Reviewed-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

* Remove dead temp check from temp_to_code() (this function users
  in exynos_tmu_initialize() always pass correct temperatures and
  exynos_tmu_set_emulation() returns early for EXYNOS4210 because
  TMU_SUPPORT_EMULATION flag is not set on this SoC).

* Move temp_code check from code_to_temp() to exynos_tmu_read()
  (code_to_temp() only user).

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Reviewed-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Remove runtime checks for negative return values of temp_to_code()
from exynos_tmu_initialize().

The current level temperature data hardcoded in pdata will never
cause a negative temp_to_code() return values and checking itself
is not proper.  The checks in question are done at runtime in
a production code for data that is hardcoded inside driver during
development time and later it doesn't change.  Such data should
be verified during development and review time (i.e. by a script
parsing relevant data from exynos_tmu_data.c, one can also argue
that verification to be done is so simple that the review by
a maintainer should be enough).

Theres should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Remove runtime checks for pdata sanity from exynos_tmu_initialize().

The current values hardcoded in pdata will never trigger the checks
and checking itself is not proper.  The checks in question are done
at runtime in a production code for data that is hardcoded inside
driver during development time and later it doesn't change.  Such
data should be verified during development and review time (i.e. by
a script parsing relevant data from exynos_tmu_data.c, one can also
argue that verification to be done is so simple that the review by
a maintainer should be enough).

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

The commit 1928457e ("thermal: exynos: Add hardware mode thermal
calibration support") has added HW_MODE feature but it has never
been enabled.  As such it has been a dead code for over a year
now and should be removed from the kernel.

We don't keep the unused/untested features in the kernel just
in case that some future hardware might need it.  Such code has
a real maintainance cost (all other code changes have to take
the dead code into account) and usually makes future changes
more difficult, not easier (i.e. recent additions of Exynos5420
SoC and Exynos5260 SoC thermal support has not made use of any
of the driver's currently unused/untested features, moreover
the recently added code is more complex than needed because of
the existing dead code).  Also all removed dead code is still
accessible in the kernel git repository and can be easily
brought back if/when needed.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Remove unused / write-only entries from struct exynos_tmu_registers.
Then remove unused defines while at it.

We don't keep the unused/untested features in the kernel just
in case that some future hardware might need it.  Such code has
a real maintainance cost (all other code changes have to take
the dead code into account) and usually makes future changes
more difficult, not easier (i.e. recent additions of Exynos5420
SoC and Exynos5260 SoC thermal support has not made use of any
of the driver's currently unused/untested features, moreover
the recently added code is more complex than needed because of
the existing dead code).  Also all removed dead code is still
accessible in the kernel git repository and can be easily
brought back if/when needed.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Tested-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

we share the same driver for both ACPI predefined Fan device
and INT3404 Fan device, thus we should select the ACPI Fan
driver when int340x thermal drivers are enabeld.
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

The kernel used to contain two functions for length-delimited,
case-insensitive string comparison, strnicmp with correct semantics and
a slightly buggy strncasecmp.  The latter is the POSIX name, so strnicmp
was renamed to strncasecmp, and strnicmp made into a wrapper for the new
strncasecmp to avoid breaking existing users.

To allow the compat wrapper strnicmp to be removed at some point in the
future, and to avoid the extra indirection cost, do
s/strnicmp/strncasecmp/g.
Signed-off-by: NRasmus Villemoes <linux@rasmusvillemoes.dk>
Acked-by: NZhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

ACPI _TRT and _ART parsing code has been moved to acpi_thermal_rel such
that it can be used by other devices in the future. Use the parsing APIs
in acpi_thermal_rel.c instead.
Signed-off-by: NJacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

ACPI 4.0 introduced two thermal relationship tables via _ART
(active cooling) and  _TRT (passive cooling) objects. These
tables contain many to many relationships among thermal sensors
and cooling devices.

This patch parses _ART and _TRT and makes the result available to
the userspace via an misc device interface. At the same time,
kernel drivers can also request parsing results from internal
kernel APIs.

The results include source and target devices, influence, and
sampling rate in case of _TRT. For _ART, the result shows source
device, target device, and weight percentage.
Signed-off-by: NJacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

ACPI INT3403 device object can be used to retrieve temperature date
from temperature sensors present in the system, and to expose
device' performance control.

The previous INT3403 thermal driver supports temperature reporting only,
thus remove it and introduce this new & enhanced one.
Signed-off-by: NLan Tianyu <tianyu.lan@intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

ACPI INT3402 device object could report temperature for the memory module.
To expose such information to user space, a thermal zone device is registered
for it so that the thermal sysfs interface can expose such information for
userspace to use.
Signed-off-by: NAaron Lu <aaron.lu@intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

Signed-off-by: NZhang Rui <rui.zhang@intel.com>

Signed-off-by: NZhang Rui <rui.zhang@intel.com>

Introduce int3400 thermal driver. And make INT3400 driver
enumerate the other int340x thermal components shown in _ART/_TRT.
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

It turns out that some boards can have instance->lower greater than 0 and
when thermal trend is dropping it results with next_target equal to -1.

Since the next_target is defined as unsigned long it is interpreted as
0xFFFFFFFF and larger than instance->upper.
As a result the next_target is set to instance->upper which ramps up to
maximal cooling device target when the temperature is steadily decreasing.
Signed-off-by: NLukasz Majewski <l.majewski@samsung.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

Newer laptops and tablets that use ACPI may have thermal sensors and
other devices with thermal control capabilities outside the core CPU/SOC,
for thermal safety reasons.
They are exposed for the OS to use via
1) INT3400 ACPI device object as the master.
2) INT3401 ~ INT340B ACPI device objects as the slaves.

This patch introduces a scan handler to enumerate the INT3400
ACPI device object to platform bus, and prevent its slaves
from being enumerated before the controller driver being probed.
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

mach-kirkwood has been removed, now that kirkwood lives in mach-mvebu.
Depend on MACH_KIRKWOOD, which will be set when kirkwood is built as
part of ARCH_MVEBU.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: linux-pm@vger.kernel.org
Acked-by: NEduardo Valentin <edubezval@gmail.com>
Link: https://lkml.kernel.org/r/1409417172-6846-4-git-send-email-andrew@lunn.chSigned-off-by: NJason Cooper <jason@lakedaemon.net>

The bang-bang thermal governor uses a hysteresis to switch abruptly on
or off a cooling device.  It is intended to control fans, which can
not be throttled but just switched on or off.
Bang-bang cannot be set as default governor as it is intended for
special devices only.  For those special devices the driver needs to
explicitely request it.

Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Andreas Mohr <andi@lisas.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Javi Merino <javi.merino@arm.com>
Cc: linux-pm@vger.kernel.org
Signed-off-by: NPeter Feuerer <peter@piie.net>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

i.MX6SX has some new features of thermal interrupt function,
there are LOW, HIGH and PANIC irq for thermal sensor, so add
platform data to separate different thermal version;

The reset value of LOW ALARM is 0 which means the highest
temp, so the LOW ALARM will be triggered once irq is enabled,
so we need to correct it before enabling thermal irq;

Enable PANIC ALARM as critical trip point, it will trigger
system reset via SRC module once PANIC IRQ is triggered, it
is pure hardware function, so use it instead of software
reset by cooling device.
Signed-off-by: NAnson Huang <b20788@freescale.com>
Tested-by: NShawn Guo <shawn.guo@linaro.org>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Add support to check status of the thermal zone before registering the
zone. This will help on disabling some non-existing thermal zone from
the top level DTS file out of common dtsi thermalzone file.

For example,
we have 3 platforms almost same but thermal zones on this platform are
little bit different. Platform 1 and 2 have three thermal zones and
platform 3 has two thermal zones. To avoid duplication of the thermal
zones entries on each DTS file of platforms,we created one common
dtsi file for thermal zone and included this dtsi file from these
3 platform's top level dts file.

On common thermal zone com dtsi file, all thermal zone are enabled and
need to disable one of thermal zone on platform 3 dts file. For this, we
just added entry status = "disabled" for that thermal zone on platform 3
dts file and along with this change to make it work.

This way, we reuse the common file and control the enable/disable of the
thermal zone from top level dts file.
Signed-off-by: NLaxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Create a new event to trace when the temperature is above a trip
point. Use the trace-point when handling non-critical and critical
trip pionts.

Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: NPunit Agrawal <punit.agrawal@arm.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Introduce and use an event to trace when a cooling device's state is
updated. This is useful to follow the effect of governor decisions on
cooling devices.

Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: NPunit Agrawal <punit.agrawal@arm.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Create a new event to trace the temperature of a thermal zone. Using
this event trace the temperature changes of the thermal zone every-time
it is updated.

Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: NPunit Agrawal <punit.agrawal@arm.com>
Signed-off-by: NEduardo Valentin <edubezval@gmail.com>

Cc: Eduardo Valentin <eduardo.valentin@ti.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: NJavi Merino <javi.merino@arm.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

This patch add registers, bit fields and compatible strings for Exynos3250 TMU
(Thermal Management Unit). Exynos3250 uses the Cortex-A7 dual cores and has
a target speed of 1.0 GHz.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
[Add MUX address setting bits by Jonghwa Lee]
Signed-off-by: NJonghwa Lee <jonghwa3.lee@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Reviewed-by: Amit Daniel Kachhap<amit.daniel@samsung.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

It might not be a problem currently but unregister/uninitialize things
in the reverse order that they are registered/initialized.
Signed-off-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Reviewed-by: NAmit Daniel Kachhap <amit.daniel@samsung.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>

DT-enabled Dove has moved from ARCH_DOVE in mach-dove to MACH_DOVE
in mach-mvebu. As non-DT ARCH_DOVE will stay to rot for a while, add a new
DT-only MACH_DOVE to thermal Kconfig.

This was originally supposed to go in via "ARM: dove: prepare new Dove DT Kconfig"
patch from Sebastian Hesselbarth for 3.15, but slipped through the cracks.

I've tested on CuBox that without this patch you can't compile
dove_thermal into a mach-mvebu based kernel, and with this patch I can
build the driver and it works as expected run-time.

v2: non-ascii char creeped in somehow
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Cc: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Cc: linux-pm@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Acked-by: NSebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>