Merge tag 'imx-soc-4.17' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into next/soc

Pull "i.MX SoC changes for 4.17" from Shawn Guo:

 - Add i.MX 6SoloLiteLite (i.MX6SLL) SoC support on top of the existing
   i.MX6SL platform code.
 - Improve the SoC revision mapping by utilizing the MAJOR field of
   ANATOP DIGPROG register.
 - Add CPUIDLE_FLAG_TIMER_STOP flag for cpuidle ARM power off state,
   so that we can use ARM generic timer for some i.MX6 SoC.
 - Set low-power interrupt mask for i.MX25 to support STOP mode.
 - Drop EPIT driver as there is no user of it.
 - Simplify the error path of imx6_pm_get_base() a bit.

* tag 'imx-soc-4.17' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
  ARM: imx: Add basic msl support for imx6sll
  ARM: imx: pm-imx6: Return the error directly
  ARM: imx: avic: set low-power interrupt mask for imx25
  ARM: imx: Improve the soc revision calculation flow
  ARM: imx: add timer stop flag to ARM power off state
  ARM: imx: Remove epit support

This patchs adds the minimal defconfig for the OXNAS ARMv6 SoCs
including the OX820 SoC and needed minimal configurations.
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Pull "Broadcom defconfig changes for 4.17" from Florian Fainelli:

This pull request contains multi_v7_defconfig changes for Broadcom
ARM-based SoC for 4.17, please pull the following:

- Tuomas enables the BCM2835 AUX driver which is necessary to provide
  console on the Raspberry Pi 3

- Stefan turns on support for the BCM43438 Bluetooth adapter connected
  via UART on the Raspberry Pi Zero Wireless and he also turns on
  support for the thermal driver on the Pi 2 and 3.

* tag 'arm-soc/for-4.17/defconfig' of https://github.com/Broadcom/stblinux:
  ARM: multi_v7_defconfig: Enable thermal driver for RPi 2/3
  ARM: bcm2385_defconfig: Enable BT support for BCM43438
  ARM: multi_v7_defconfig: Enable serial console on RPi 3

Pull "This is the pxa changes for v4.17 cycle" from Robert Jarzmik:

 - minor changes for property API
 - clock API fix for ULPI driver warning

It exceptionally contains a merge from the mtd tree from Boris
to prevent any merge conflicts in the PXA tree.

* tag 'pxa-for-4.17' of https://github.com/rjarzmik/linux:
  ARM: pxa/raumfeld: use PROPERTY_ENTRY_U32() directly
  ARM: pxa: ulpi: fix ulpi timeout and slowpath warn
  ARM: pxa: cm-x300: remove inline directive
  ARM: pxa: fix static checker warning in pxa3xx-ulpi
  MAINTAINERS: remove entry for deleted pxa3xx_nand driver
  arm: dts: pxa: use reworked NAND controller driver
  dt-bindings: mtd: remove pxa3xx NAND controller documentation
  mtd: nand: remove useless fields from pxa3xx NAND platform data
  mtd: nand: remove deprecated pxa3xx_nand driver
  mtd: nand: use Marvell reworked NAND controller driver with all platforms

Merge tag 'at91-ab-4.17-soc' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/abelloni/linux into next/soc

Pull "AT91 SoC for 4.17: from Alexandre Belloni:

 - Rename Atmel to Microhip in MAINTAINERS, Documentation and Kconfig

* tag 'at91-ab-4.17-soc' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:
  ARM: at91: Kconfig: Update company to Microchip
  Documentation: at91: Update Microchip SoC documentation
  MAINTAINERS: ARM: at91: update entry for ARM/Microchip

Pull "ARM64: hisilicon: defconfig updates for 4.17" from Wei Xu:

- Enable deadline IO scheduler to improve the performnace for some
usecaese without changing the default IO scheduler

* tag 'hisi-defconfig-for-4.17' of git://github.com/hisilicon/linux-hisi:
  arm64: defconfig: enable IOSCHED_DEADLINE

Add basic MSL support for i.MX6SLL.

The i.MX 6SoloLiteLite application processors are NXP's latest
additions to a growing family of multimedia-focused products
offering high-performance processing optimized for lowest power
consumption. The i.MX 6SoloLiteLite processors feature NXP's advanced
implementation of the ARM Cortex-A9 core, which can be interfaced
with LPDDR3 and LPDDR2 DRAM memory devices.
Signed-off-by: Bai Ping <ping.bai@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>

Merge tag 'samsung-soc-4.17' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/krzk/linux into next/soc

Pull "Samsung mach/soc changes for v4.17" from Krzysztof Kozłowski:

1. Add few remaining SPDX license identifiers.
2. Add cpuidle support to all Midas-based boards (including new
   GT-I9300/GT-I9305/GT-N7100/GT-N7105).

* tag 'samsung-soc-4.17' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/krzk/linux:
  ARM: EXYNOS: Extend cpuidle support to Midas boards
  ARM: EXYNOS: Add SPDX license identifiers

Merge tag 'omap-for-v4.17/ti-sysc-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into next/soc

Pull "Driver changes for ti-sysc for v4.17" from Tony Lindgren:

This series of changes enables the use device tree based sysconfig
data for ti-sysc driver. As we already have SmartReflex data configured,
we use that as the first driver to enable. To do that in a way where
SmartReflex is not probed twice, we need to prepare the SmartReflex
driver before flipping dts data on for it in the last patch of the
series.

To avoid regressions, we are checking the passed dts data against
existing platform data since we still have it available. Then after the
dts files are converted, we can simply drop the related platform data
at some point in the future.

* tag 'omap-for-v4.17/ti-sysc-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
  ARM: OMAP2+: Enable ti-sysc to use device tree data for smartreflex
  PM / AVS: SmartReflex: Prepare to use device tree based probing
  ARM: OMAP2+: Try to parse earlycon from parent too
  ARM: OMAP2+: Add checks for device tree based sysconfig data
  ARM: OMAP2+: Add functions to allocate module data from device tree
  bus: ti-sysc: Handle some devices in omap_device compatible way
  bus: ti-sysc: Add support for platform data callbacks
  bus: ti-sysc: Remove unnecessary debugging statements
  bus: ti-sysc: Improve handling for no-reset-on-init and no-idle-on-init
  bus: ti-sysc: Handle stdout-path for debug console
  bus: ti-sysc: Add suspend and resume handling
  bus: ti-sysc: Add fck clock alias for children with notifier_block
  ARM: OMAP2+: Prepare to pass auxdata for smartreflex

Merge tag 'omap-for-v4.17/am-pm-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into next/soc

Pull "Add am335x and am437x PM code for v4.17" from Tony Lindgren:

This series of changes from Dave Gerlach adds the PM related
code to allow low-power suspend states. The code consists of
the SoC specific assembly code and a related PM driver.

* tag 'omap-for-v4.17/am-pm-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
  soc: ti: Add pm33xx driver for basic suspend support
  ARM: OMAP2+: pm33xx-core: Add platform code needed for PM
  ARM: OMAP2+: Introduce low-level suspend code for AM43XX
  ARM: OMAP2+: Introduce low-level suspend code for AM33XX

Merge tag 'omap-for-v4.17/timer-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into next/soc

Pull "Move omap timer to drivers for 4.17" from Tony Lindgren:

This series from Keerthy and Ladislav Michl move omap dmtimer code
to drivers. As we don't want to export custom timer functions to
random drivers, we also need to update the related PWM driver to
pass the timer specific functions in platform data.

Note that this series is based on a merge of omap1 specific timer
fix and omap2+ platform data clean-up to keep things working and
make the move a bit simpler.

* tag 'omap-for-v4.17/timer-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
  clocksource: timer-ti-dm: Check prescaler value
  clocksource: timer-ti-dm: Consolidate set source
  clocksource: timer-ti-dm: Make unexported functions static
  ARM: OMAP: pdata-quirks: Remove unused timer pdata
  pwm: pwm-omap-dmtimer: Adapt driver to utilize dmtimer pdata ops
  clocksource: timer-ti-dm: Hook device platform data if not already assigned
  clocksource: timer-ti-dm: Populate the timer ops to the pdata
  clocksource: timer-ti-dm: Add timer ops to the platform data structure
  ARM: OMAP: Move dmtimer driver out of plat-omap to drivers under clocksource
  clocksource: timer-ti-dm: Replace architecture
  ARM: OMAP: Move dmtimer.h out of plat-omap
  ARM: OMAP: timer: Wrap the inline functions under OMAP2PLUS define
  ARM: OMAP: dmtimer: Remove all the exports
  ARM: OMAP: Fix dmtimer init for omap1

Merge tag 'omap-for-v4.17/soc-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into next/soc

Pull "SoC changes for omaps for v4.17" from Tony Lindgren:

This series is mostly a series from Suman Anna to remove now unused
platform data that is now coming from device tree. This also make it
a bit simpler to move the timer code to live under drivers in a
separate series of patches. There are also few minor clean-ups for
omap4 PM code.

* tag 'omap-for-v4.17/soc-signed' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
  ARM: OMAP2+: Fix typo for wakeup_ns_pa_addr
  ARM: OMAP2+: Use v7_invalidate_l1 in omap4_finish_suspend
  ARM: OMAP2+: Remove unused gpio header file references
  ARM: OMAP: Move plat/i2c.h into mach-omap1 folder
  ARM: OMAP2+: Cleanup omap_mcbsp_dev_attr and other legacy data
  ARM: OMAP2+: Cleanup omap2_spi_dev_attr and other legacy data
  ARM: OMAP2+: Cleanup omap_timer_capability_dev_attr usage
  ARM: OMAP2+: Cleanup omap_i2c_dev_attr usage
  ARM: OMAP2+: Cleanup omap_gpio_dev_attr usage
  ARM: OMAP2+: Include types.h directly for hwmod data

Merge tag 'stm32-defconfig-for-v4.17-1' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/atorgue/stm32 into next/soc

Pull "STM32 defconfig updates for v4.17, round 1" from Alexandre Torgue:

Highlights:
----------
 -Add STM32 support (for MPU) to multi-v7-defconfig
 -Enable MMCI and EXT3_FS support

Merge tag 'stm32-soc-for-v4.17-1' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/atorgue/stm32 into next/soc

Pull "STM32 SOC updates for v4.17, round 1." from Alexandre Torgue:

Highlights:
----------
 -Add STM32F769 MCU support
 -Add STM32MP157 MPU support
 -Add AMBA PL180 MMCI support for F4 and F7 MCUs
 -Move STM32 documentation to rst format

* tag 'stm32-soc-for-v4.17-1' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/atorgue/stm32:
  ARM: stm32: restore reboot capabilities
  ARM: stm32: Add AMBA support for STM32F4 and STM32F7 SoCs
  ARM: stm32: add initial support for STM32MP157
  ARM: stm32: prepare stm32 family to welcome armv7 architecture
  Documentation: arm: stm32: move to rst format
  ARM: stm32: Fix copyright
  ARM: stm32: add new STM32F769 MCU
  ARM: stm32: Kconfig: introduce MACH_STM32F769 flag

Add Spreadtrum's maintainer entry and Orson, Baolin and me
as maintainers.

Cc: Orson Zhai <orsonzhai@gmail.com>
Cc: Baolin Wang <baolin.wang@linaro.org>
Signed-off-by: Chunyan Zhang <zhang.lyra@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

As SoCFPGA uses the standard suspend_valid_only_mem() for its
platform_suspend_ops.valid() callback, its platform_suspend_ops.enter()
callback will never be called with state equal to PM_SUSPEND_STANDBY.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Tony Lindgren <tony@atomide.com>
Acked-by: Dinh Nguyen <dinguyen@kernel.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

nspire_map_io() just calls debug_ll_io_init().

If machine_desc.map_io is not set, devicemaps_init() in the
common ARM code will call debug_ll_io_init(), so nspire_map_io() and the
initialization of .map_io can be removed.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Reorder the mach-* inclusions so the comment about them being ordered
alphabetically ends up being true.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'davinci-for-v4.17/defconfig' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/nsekhar/linux-davinci into next/soc

Pull "DaVinci defconfig updates for v4.17" from Sekhar Nori:

Enable SYSCON_REBOOT_MODE needed for Lego EV3 and build the DaVinci
watchdog driver in so reboot works even when modules are not loaded.

* tag 'davinci-for-v4.17/defconfig' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/nsekhar/linux-davinci:
  ARM: davinci_all_defconfig: set CONFIG_DAVINCI_WATCHDOG=y
  ARM: davinci_all_defconfig: enable SYSCON_REBOOT_MODE

Merge tag 'davinci-for-v4.17/soc' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/nsekhar/linux-davinci into next/soc

Pull "Miscellaneous DaVinci SoC support improvements for v4.17" from Sekhar Nori:

* rationalization of con_id names for phy clocks to make DT conversion easy

* A patch to move away from syscon as platform device. This is needed for
  common clock framework conversion as well as helps get rid of
  syscon_regmap_lookup_by_pdevname() by removing the last known user.

* convert mach-davinci to use reset support available in watchdog driver

* a non-critical warning fix. It has been around since beginning so not sending
  as a standalone fix for -rc cycle.

* moving mach-davinci clock init to .init_time() for legacy boot. This is
  again in preparation for CCF conversion.

* tag 'davinci-for-v4.17/soc' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/nsekhar/linux-davinci:
  ARM: davinci: move davinci_clk_init() to init_time
  ARM: davinci: board-da830-evm: fix unused const variable warning
  ARM: davinci: remove watchdog reset
  ARM: da8xx: use platform data for CFGCHIP syscon regmap
  phy: da8xx-usb: rename clock con_ids

Remove the pxa3xx_nand driver (replaced by marvell_nand).

Instead of using PROPERTY_ENTRY_INTEGER() with explicitly supplied type,
use PROPERTY_ENTRY_U32() dedicated macro.

It will help modify internals of built-in device properties API.

No functional change intended.
Acked-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>

Both cm-x300 and pxa3xx-ulpi use the plain clk_{en,dis}able() API.
With the new clocking framework this results in warnings of type:
------------[ cut here ]------------
WARNING: CPU: 0 PID: 1 at drivers/clk/clk.c:714 clk_core_enable+0x90/0x9c
Modules linked in:
CPU: 0 PID: 1 Comm: swapper Not tainted 4.15.0-rc5-cm-x300+ #15
Hardware name: CM-X300 module
[<c001007c>] (unwind_backtrace) from [<c000df94>] (show_stack+0x10/0x14)
[<c000df94>] (show_stack) from [<c00199a8>] (__warn+0xd8/0x100)
[<c00199a8>] (__warn) from [<c0019a0c>] (warn_slowpath_null+0x3c/0x48)
[<c0019a0c>] (warn_slowpath_null) from [<c024e8c0>] (clk_core_enable+0x90/0x9c)
[<c024e8c0>] (clk_core_enable) from [<c024ea54>] (clk_core_enable_lock+0x18/0x2c)
[<c024ea54>] (clk_core_enable_lock) from [<c0016994>] (cm_x300_u2d_init+0x4c/0xe8)
[<c0016994>] (cm_x300_u2d_init) from [<c00163e0>] (pxa3xx_u2d_probe+0xe0/0x244)
[<c00163e0>] (pxa3xx_u2d_probe) from [<c0283de0>] (platform_drv_probe+0x38/0x88)
...
------------[ cut here ]------------
and alike...

And finally, it results in:
------------[ cut here ]------------
pxa310_ulpi_poll: ULPI access timed out!
OTG transceiver init failed
------------[ cut here ]------------

It might be that disabling the warning in kernel config would also do
the job, but IMO a better solution would be to switch to
clk_prepare_enable() and clk_disable_unprepare() APIs.
Signed-off-by: Igor Grinberg <grinberg@compulab.co.il>
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>

cm_x300_u2d_init() function is only used through its function pointer
that is passed through platform_data structure to the driver.
Therefore it can never be inlined.
Remove the inline directive.
Signed-off-by: Igor Grinberg <grinberg@compulab.co.il>
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>

Static checker reports the following warning:

arch/arm/mach-pxa/pxa3xx-ulpi.c:336 pxa3xx_u2d_probe()
warn: did you mean to pass the address of 'u2d'

Fix it by passing the correct pointer.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Igor Grinberg <grinberg@compulab.co.il>
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>

Adds basic support for the Nuvoton NPCM750 BMC.
Signed-off-by: Brendan Higgins <brendanhiggins@google.com>
Reviewed-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Avi Fishman <avifishman70@gmail.com>
Tested-by: Tomer Maimon <tmaimon77@gmail.com>
Tested-by: Avi Fishman <avifishman70@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Add maintainers and reviewers for the Nuvoton NPCM architecture.
Signed-off-by: Brendan Higgins <brendanhiggins@google.com>
Reviewed-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Avi Fishman <avifishman70@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Enable support in arm64's defconfig for Socionext SynQuacer based
platforms, by enabling the arch Kconfig symbol, and enabling builtin
support for the ethernet, GPIO and SDHCI controllers
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

New crypto drivers have been introduced in v4.16 that implement the
SHA-512, SHA3 and SM3 secure hash algorithms using ARMv8.2 optional
instructions. Add these drivers to arm64's defconfig as modules.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

To be able to reboot the platform we need to use armv7m_restart for STM32
SoCs using ARMv7 cores.

Fixes: e0644101bd2f ("ARM: mach-stm32: prepare stm32 family to welcome armv7 architecture)
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

As both STM32F4 and STM32F7 SoCs embeds an AMBA PL180 mmci IP,
we need to enable AMBA support in mach-stm32.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

This patch adds initial support of STM32MP157 microprocessor (MPU)
based on Arm Cortex-A7. New Cortex-A infrastructure (gic, timer,...)
are selected if ARCH_MULTI_V7 is defined.
Signed-off-by: Ludovic Barre <ludovic.barre@st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

This patch prepares the STM32 machine for the integration of Cortex-A
based microprocessor (MPU), on top of the existing Cortex-M
microcontroller family (MCU). Since both MCUs and MPUs are sharing
common hardware blocks we can keep using ARCH_STM32 flag for most of
them. If a hardware block is specific to one family we can use either
ARM_SINGLE_ARMV7M or ARCH_MULTI_V7 flag.
Signed-off-by: Ludovic Barre <ludovic.barre@st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

This patch rewrites stm32 documentation to rst
(ReStructuredText) format.
Signed-off-by: Ludovic Barre <ludovic.barre@st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

Uniformize STMicroelectronics copyrights header
Add SPDX identifier
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@st.com>
CC: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Acked-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

Add new st,stm32f769 compatible machine name for STM32F769 MCU and update
documentation.
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

This patch introduces the MACH_STM32F769 to make possible to only select
STM32F769 pinctrl driver.

By default, all the MACH_STM32Fxxx flags will be set with STM32 defconfig.
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>

Pull x86 fixes from Thomas Gleixner:
 "A small set of fixes for x86:

   - Add missing instruction suffixes to assembly code so it can be
     compiled by newer GAS versions without warnings.

   - Switch refcount WARN exceptions to UD2 as we did in general

   - Make the reboot on Intel Edison platforms work

   - A small documentation update so text and sample command match"

* 'x86/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  Documentation, x86, resctrl: Make text and sample command match
  x86/platform/intel-mid: Handle Intel Edison reboot correctly
  x86/asm: Add instruction suffixes to bitops
  x86/entry/64: Add instruction suffix
  x86/refcounts: Switch to UD2 for exceptions

Pull x86/pti fixes from Thomas Gleixner:
 "Three fixes related to melted spectrum:

   - Sync the cpu_entry_area page table to initial_page_table on 32 bit.

     Otherwise suspend/resume fails because resume uses
     initial_page_table and triggers a triple fault when accessing the
     cpu entry area.

   - Zero the SPEC_CTL MRS on XEN before suspend to address a
     shortcoming in the hypervisor.

   - Fix another switch table detection issue in objtool"

* 'x86-pti-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/cpu_entry_area: Sync cpu_entry_area to initial_page_table
  objtool: Fix another switch table detection issue
  x86/xen: Zero MSR_IA32_SPEC_CTRL before suspend