Olof Johansson [Thu, 24 Jun 2021 01:47:38 +0000 (18:47 -0700)]
Merge tag 'ixp4xx-arm-soc-v5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-nomadik into arm/soc
This is a major chunk of IXP4xx modernization:
- Fist we move some registers around to make room for
the predetermined PCI I/O space.
- Then we add some Kconfig options to make it possible
to use the old PCI driver in parallell with the new
shiny one.
- Then we add the new PCI driver and some bindings for
it.
- On top of this we add an (ages old) patch from Arnd
that centralize the CPU/SoC detection in drivers/soc
and make the header a standard Linux header to avoid
the <mach/*> business in drivers.
- Then we split out and modernize some platform data
headers for pata, and hwrandom, and top it up with
DT bindings and support for hwrandom.
* tag 'ixp4xx-arm-soc-v5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-nomadik:
ixp4xx: fix spelling mistake in Kconfig "Devce" -> "Device"
hw_random: ixp4xx: Add OF support
hw_random: ixp4xx: Add DT bindings
hw_random: ixp4xx: Turn into a module
hw_random: ixp4xx: Use SPDX license tag
hw_random: ixp4xx: enable compile-testing
pata: ixp4xx: split platform data to its own header
soc: ixp4xx: move cpu detection to linux/soc/ixp4xx/cpu.h
PCI: ixp4xx: Add a new driver for IXP4xx
PCI: ixp4xx: Add device tree bindings for IXP4xx
ARM/ixp4xx: Make NEED_MACH_IO_H optional
ARM/ixp4xx: Move the virtual IObases
Olof Johansson [Thu, 24 Jun 2021 01:45:21 +0000 (18:45 -0700)]
Merge tag 'omap-for-v5.14/fixes-not-urgent-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into arm/soc
Non-urgent fixes for omaps for v5.14 merge window
Warn and block suspend for am335x unless the PM related modules and
firmware are loaded and warn otherwise. Otherwise we easily end up
with a suspended system with nothing capable of waking it up.
We also drop a duplicated prototype for am33xx_init_early().
* tag 'omap-for-v5.14/fixes-not-urgent-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
ARM: OMAP2+: Block suspend for am3 and am4 if PM is not configured
ARM: OMAP2+: remove duplicated prototype
ARM: dts: dra7: Fix duplicate USB4 target module node
ARM: OMAP1: ams-delta: remove unused function ams_delta_camera_power
bus: ti-sysc: Fix flakey idling of uarts and stop using swsup_sidle_act
bus: ti-sysc: Fix am335x resume hang for usb otg module
ARM: OMAP2+: Fix build warning when mmc_omap is not built
ARM: OMAP1: isp1301-omap: Add missing gpiod_add_lookup_table function
ARM: OMAP1: Fix use of possibly uninitialized irq variable
bus: ti-sysc: Fix missing quirk flags for sata
Linus Walleij [Tue, 11 May 2021 12:15:42 +0000 (14:15 +0200)]
hw_random: ixp4xx: Turn into a module
Instead of just initializing always, which will invariably
create problems on multiplatform builds, turn this driver into
a module and pass a resource with the memory location.
The device only exist on the IXP46x so we only register
the device for the IXP46x SoC.
Arnd Bergmann [Sun, 25 Aug 2019 19:57:25 +0000 (21:57 +0200)]
soc: ixp4xx: move cpu detection to linux/soc/ixp4xx/cpu.h
Generic drivers are unable to use the feature macros from mach/cpu.h
or the feature bits from mach/hardware.h, so move these into a global
header file along with some dummy helpers that list these features as
disabled elsewhere.
Cc: David S. Miller <davem@davemloft.net> Cc: Jakub Kicinski <kuba@kernel.org> Cc: netdev@vger.kernel.org Cc: Zoltan HERPAI <wigyori@uid0.hu> Cc: Raylynn Knight <rayknight@me.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Linus Walleij [Sun, 2 May 2021 22:09:20 +0000 (00:09 +0200)]
PCI: ixp4xx: Add a new driver for IXP4xx
This adds a new PCI controller driver for the Intel IXP4xx
(IX425, IXP435 etc), based on the XScale microarchitecture.
This replaces the old driver in arch/arm/mach-ixp4xx/common-pci.c
which utilized the ARM-specific BIOS32 PCI framework,
and all parameterization for such things as memory and
IO space as well as interrupt swizzling is done from the
device tree.
The plan is to phase out and delete the old driver piecemal.
The __raw_writel() and __raw_readl() are used for accessing
the PCI controller for the same reason that these accessors
are used in the timer, IRQ and GPIO drivers: the platform
will alter its address bus pattern based on whether the
system is booted in big- or little-endian mode. For this
reason all register on IXP4xx must always be accessed in
native (CPU) endianness.
This driver supports 64MB of PCI memory space, but not the
indirect access of 1GB that is available in the old driver.
We can address that later if and only if there are users
that need all 1GB of PCI address space. Krzysztof reports
having to use indirect MMIO only once for a VGA card. There
is work ongoing for general indirect MMIO. (In practice
the indirect MMIO is performed by writing address and
writing and reading values into/from a controller
register.)
Tested by booting the NSLU2, attaching a USB stick, mounting
and browsing the drive.
In order to create a proper PCI driver for the IXP4xx
we need to make the old PCI driver and its reliance
on <mach/io.h> optional.
Create a new Kconfig symbol for the legacy PCI driver
IXP4XX_PCI_LEGACY and only activate NEED_MACH_IO_H
for this driver.
A few files need to be adjusted to explicitly include
the <mach/hardware.h> and <mach/cpu.h> headers that
they previously obtained implicitly using <linux/io.h>
that would include <mach/io.h> and in turn include
these two headers.
This breaks our reliance on the old PCI and indirect
PCI support so we can reimplement a proper purely
DT-based driver in the PCI subsystem.
UART1, UART2 and the expansion bus config registers
are the only registers mapped in a fixed location
when using device tree.
For device tree we also want to get rid of the custom
<mach/io.h> for IXP4xx. So we need to undefine
CONFIG_NEED_MACH_IO_H. Doing that activates the fixed
mapping of the PCI IO space to PCI_IO_VIRT_BASE which
is hardcoded to 0xFEE00000 and this would collide with
the old fixed mappings.
Move the fixed virtual IO base address from 0xFEF00000
to 0xFEC00000 in order to avoid the collision.
For the OF-only boot path let's even cut the reliance
on <mach/io.h> and just hardcode the one single virtbase
we need apart from the UART, which is hardcoded in
Kconfig.debug.
Olof Johansson [Tue, 15 Jun 2021 15:35:27 +0000 (08:35 -0700)]
Merge tag 'tegra-for-5.14-soc' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into arm/soc
soc/tegra: Changes for v5.14-rc1
These changes implement the core power domain for the PMC, and fix a
couple of minor issues as well as add stubs to help some drivers be
compile tested more easily.
* tag 'tegra-for-5.14-soc' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
soc/tegra: fuse: Fix Tegra234-only builds
soc/tegra: fuse: Don't return -ENOMEM when allocate lookups failed
soc/tegra: regulators: Support core domain state syncing
soc/tegra: pmc: Add driver state syncing
soc/tegra: pmc: Add core power domain
soc/tegra: fuse: Add stubs needed for compile-testing
soc/tegra: Add devm_tegra_core_dev_init_opp_table()
soc/tegra: Add stub for soc_is_tegra()
soc/tegra: regulators: Bump voltages on system reboot
regulator: core: Add regulator_sync_voltage_rdev()
Olof Johansson [Tue, 15 Jun 2021 15:13:30 +0000 (08:13 -0700)]
Merge tag 'qcom-dts-fixes-for-5.13' of https://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux into arm/fixes
Qualcomm DTS fixes for v5.13
Failure to properly define "secure-regions" as 64-bit ranges prevents
the two SDX55 boards from booting, so correct this.
* tag 'qcom-dts-fixes-for-5.13' of https://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux:
ARM: dts: qcom: sdx55-telit: Represent secure-regions as 64-bit elements
ARM: dts: qcom: sdx55-t55: Represent secure-regions as 64-bit elements
Olof Johansson [Tue, 15 Jun 2021 15:09:51 +0000 (08:09 -0700)]
Merge tag 'imx-soc-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into arm/soc
i.MX SoC changes for 5.14:
- Add SMP support for i.MX7D, as there are i.MX7D devices running in
field with U-Boot that doesn't support PSCI.
- Fix imx5_cpu_suspend_info reference in i.MX5 PM code comments.
- Add SoC ID initialization for i.MX50.
- Drop unnecessary of_platform_default_populate() call from i.MX6Q
.init_machine hook.
- A series from Oleksij Rempel to remove most of the i.MX6/7 board
specific PHY fixups.
* tag 'imx-soc-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: imx: add smp support for imx7d
ARM: imx6q: drop of_platform_default_populate() from init_machine
ARM: imx7d: remove Atheros AR8031 PHY fixup
ARM: imx6sx: remove Atheros AR8031 PHY fixup
ARM: imx6q: remove Atheros AR8035 SmartEEE fixup
ARM: imx6q: remove clk-out fixup for the Atheros AR8031 and AR8035 PHYs
ARM: imx6q: remove BMCR_PDOWN handler in ar8035_phy_fixup()
ARM: imx6q: remove part of ar8031_phy_fixup()
ARM: imx6q: remove PHY fixup for KSZ9031
ARM: imx: Initialize SoC ID on i.MX50
ARM: imx: pm-imx5: Fix references to imx5_cpu_suspend_info
Olof Johansson [Sat, 12 Jun 2021 15:55:15 +0000 (08:55 -0700)]
Merge tag 'v5.13-rockchip-dtsfixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip into arm/fixes
Fix PCIe address ranges that are affected by recent PCI changes.
There are 3 additional patches pending that handle the backward
compatiblity inside the PCI subsystem, but the address ranges
should be fixed anyway.
* tag 'v5.13-rockchip-dtsfixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip:
arm64: dts: rockchip: Update RK3399 PCI host bridge window to 32-bit address memory
Olof Johansson [Sat, 12 Jun 2021 15:39:06 +0000 (08:39 -0700)]
Merge tag 'samsung-soc-5.14' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux into arm/soc
Samsung mach/soc changes for v5.14
1. Fix: add missing of_node_put.
2. Extend Samsung maintainers entry to cover Samsung PWM driver files,
because they do not have a dedicated entry.
3. Minor cleanups.
* tag 'samsung-soc-5.14' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux:
MAINTAINERS: Include Samsung PWM in Samsung SoC entry
soc: samsung: pmu: drop EXYNOS_CENTRAL_SEQ_OPTION defines
ARM: exynos: add missing of_node_put for loop iteration
ARM: s3c: Remove unnecessary break in RX1950
Olof Johansson [Sat, 12 Jun 2021 15:32:34 +0000 (08:32 -0700)]
Merge tag 'omap-for-v5.14/cleanup-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into arm/soc
Clean up changes for omaps for v5.14 merge window
With the recent changes to drop legacy platform data for
am3/4, omap4/5, and dra7, we can drop the related unused
code. There is also a patch to remove leading spaces from
the Kconfig file.
* tag 'omap-for-v5.14/cleanup-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
ARM: OMAP2+: remove omap2_set_globals_control()
ARM: OMAP2+: CM: remove omap2_set_globals_cm()
ARM: OMAP1: Remove leading spaces in Kconfig
MAINTAINERS: remove obsolete OMAP HWMOD DATA FOR OMAP4-BASED DEVICES
Anson Huang [Wed, 26 May 2021 00:14:16 +0000 (21:14 -0300)]
ARM: imx: add smp support for imx7d
Add SMP support for i.MX7D, including CPU hotplug support, for
systems where TFA is not present.
The motivation for bringing up the second i.MX7D core inside the kernel
is that legacy vendor bootloaders usually do not implement PSCI support.
This is a significant blocker for systems in the field that are running old
bootloader versions to upgrade to a modern mainline kernel version, as only
one CPU of the i.MX7D would be brought up.
Bring up the second i.MX7D core inside the kernel to make the migration
path to mainline kernel easier for the existing iMX7D users.
Marco Felsch [Tue, 25 May 2021 10:01:08 +0000 (12:01 +0200)]
ARM: imx6q: drop of_platform_default_populate() from init_machine
Dropping the machine_init specific of_platform_default_populate() can be
done safely due to the following two facts:
1)
Since commit 44a7185c2ae6 ("of/platform: Add common method to populate
default bus") we have a common method to populate the devices.
2)
The machine_init code don't need bounded devices<>drivers since the code
uses syscon_regmap_lookup_by_compatible() which register a syscon
device on demand.
Signed-off-by: Marco Felsch <m.felsch@pengutronix.de> Signed-off-by: Shawn Guo <shawnguo@kernel.org>
The PCIe host bridge on RK3399 advertises a single 64-bit memory
address range even though it lies entirely below 4GB.
Previously the OF PCI range parser treated 64-bit ranges more
leniently (i.e., as 32-bit), but since commit 9d57e61bf723 ("of/pci:
Add IORESOURCE_MEM_64 to resource flags for 64-bit memory addresses")
the code takes a stricter view and treats the ranges as advertised in
the device tree (i.e, as 64-bit).
The change in behaviour causes failure when allocating bus addresses
to devices connected behind a PCI-to-PCI bridge that require
non-prefetchable memory ranges. The allocation failure was observed
for certain Samsung NVMe drives connected to RockPro64 boards.
Update the host bridge window attributes to treat it as 32-bit address
memory. This fixes the allocation failure observed since commit 9d57e61bf723.
Tony Lindgren [Tue, 25 May 2021 09:33:11 +0000 (12:33 +0300)]
ARM: OMAP2+: Block suspend for am3 and am4 if PM is not configured
If the PM related modules are not loaded and PM firmware not configured,
the system suspend fails to resume. Let's fix this by adding initial
platform_suspend_ops to block suspend and warn about missing modules.
When pm33xx and wkup_m3_ipc have been loaded and m3 coprocessor booted
with it's firmware, pm33xx sets up working platform_suspend_ops. Note
that we need to configure at least PM_SUSPEND_STANDBY to have
suspend_set_ops().
Cc: Dave Gerlach <d-gerlach@ti.com> Cc: Suman Anna <s-anna@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
With [1] USB4 target-module node got defined in dra74x.dtsi file.
However, the earlier definition in [2] was not removed, and this
duplication of the target module is causing boot failure on dra74
variant boards - dra7-evm, dra76-evm.
USB4 is only present in DRA74x variants, so keeping the entry in
dra74x.dtsi and removing it from the top level interconnect hierarchy
dra7-l4.dtsi file. This change makes the USB4 target module no longer
visible to AM5718, DRA71x and DRA72x so removing references to it in
their respective dts files.
[1]: commit c7b72abca61ec ("ARM: OMAP2+: Drop legacy platform data for
dra7 dwc3")
[2]: commit 549fce068a311 ("ARM: dts: dra7: Add l4 interconnect
hierarchy and ti-sysc data")
Fixes: c7b72abca61ec ("ARM: OMAP2+: Drop legacy platform data for dra7 dwc3") Signed-off-by: Gowtham Tammana <g-tammana@ti.com> Reviewed-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Icenowy Zheng [Wed, 9 Jun 2021 08:38:43 +0000 (16:38 +0800)]
arm64: dts: allwinner: a64-sopine-baseboard: change RGMII mode to TXID
Although the schematics of Pine A64-LTS and SoPine Baseboard shows both
the RX and TX internal delay are enabled, they're using the same broken
RTL8211E chip batch with Pine A64+, so they should use TXID instead, not
ID.
In addition, by checking the real components soldered on both a SoPine
Baseboard and a Pine A64-LTS, RX delay is not enabled (GR69 soldered and
GR70 NC) despite the schematics says it's enabled. It's a common
situation for Pine64 boards that the NC information on schematics is not
the same with the board.
So the RGMII delay mode should be TXID on these boards.
Olof Johansson [Sat, 5 Jun 2021 22:43:46 +0000 (15:43 -0700)]
Merge tag 'ti-k3-dt-fixes-for-v5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/nmenon/linux into arm/fixes
Devicetree fixes for TI K3 platforms for v5.13 merge window:
These minor fixes include:
* Fixups for device tree discovered during yaml conversion
* Fixups for missing dma-coherent property in j7200
* Removal of camera sensor node from am65 evm dts to overlay
as camera sensor boards are variable.
* tag 'ti-k3-dt-fixes-for-v5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/nmenon/linux:
arm64: dts: ti: k3*: Introduce reg definition for interrupt routers
arm64: dts: ti: k3-am65|j721e|am64: Map the dma / navigator subsystem via explicit ranges
arm64: dts: ti: k3-*: Rename the TI-SCI node
arm64: dts: ti: k3-am65-wakeup: Drop un-necessary properties from dmsc node
arm64: dts: ti: k3-am65-wakeup: Add debug region to TI-SCI node
arm64: dts: ti: k3-*: Rename the TI-SCI clocks node name
arm64: dts: ti: j7200-main: Mark Main NAVSS as dma-coherent
arm64: dts: ti: k3-am654-base-board: remove ov5640
Olof Johansson [Sat, 5 Jun 2021 22:41:40 +0000 (15:41 -0700)]
Merge tag 'omap-for-v5.13/fixes-pm' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into arm/fixes
PM and build warning fixes for omaps
While chasing system suspend related regressions, I noticed few other
issues related to PM would be good to have fixed:
- UART idling does not always work for hardware autoidle features
- am335x resume works only the first time unless musb module is loaded
Then there are three patches for omap1 related warnings caused by the gpio
changes, and one build warning fix for legacy mmc platform code when mmc
is built as a loadable module.
These can all be merged whenever suitable naturally. I've sent the more
urgent SATA regression fix separately although it appears in this pull
request too because of the branches merged.
* tag 'omap-for-v5.13/fixes-pm' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
ARM: OMAP1: ams-delta: remove unused function ams_delta_camera_power
bus: ti-sysc: Fix flakey idling of uarts and stop using swsup_sidle_act
bus: ti-sysc: Fix am335x resume hang for usb otg module
ARM: OMAP2+: Fix build warning when mmc_omap is not built
ARM: OMAP1: isp1301-omap: Add missing gpiod_add_lookup_table function
ARM: OMAP1: Fix use of possibly uninitialized irq variable
Olof Johansson [Sat, 5 Jun 2021 22:39:22 +0000 (15:39 -0700)]
Merge tag 'amlogic-fixes-v5.13-rc1' of https://git.kernel.org/pub/scm/linux/kernel/git/amlogic/linux into arm/fixes
Amlogic fixes for v5.13-rc1
- arm64: meson: select COMMON_CLK to select a proper implementation of the clock API
- soc: amlogic: meson-clk-measure: remove redundant dev_err call in meson_msr_probe()
* tag 'amlogic-fixes-v5.13-rc1' of https://git.kernel.org/pub/scm/linux/kernel/git/amlogic/linux:
arm64: meson: select COMMON_CLK
soc: amlogic: meson-clk-measure: remove redundant dev_err call in meson_msr_probe()
Olof Johansson [Sat, 5 Jun 2021 22:24:10 +0000 (15:24 -0700)]
Merge tag 'imx-fixes-5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into arm/fixes
i.MX fixes for 5.13:
- Fix missing-prototypes warning of 'imx27_pm_init' in i.MX27 platform
pm code.
- A couple of patches from Fabio Estevam to fix 'tuning-step' property
in imx7d-meerkat96 and imx7d-pico DT.
- Fix '#gpio-cells' of nxp,pca8574 device in imx6qdl-emcon-avari DT.
- A couple of patches from Lucas Stach to fix regulator and voltage for
imx8mq-zii-ultra board.
- Add missing regulators for imx6q-dhcom to avoid possible instability
issues.
- Fix memory-controller settings for fsl-ls1028a DT.
- Fix RGMII clock and voltage for a couple of fsl-ls1028a-kontron-sl28
boards.
- Fix RGMII connection to QCA8334 switch for imx6dl-yapp4 board.
* tag 'imx-fixes-5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: dts: imx: emcon-avari: Fix nxp,pca8574 #gpio-cells
ARM: dts: imx7d-pico: Fix the 'tuning-step' property
ARM: dts: imx7d-meerkat96: Fix the 'tuning-step' property
arm64: dts: freescale: sl28: var1: fix RGMII clock and voltage
arm64: dts: freescale: sl28: var4: fix RGMII clock and voltage
ARM: imx: pm-imx27: Include "common.h"
arm64: dts: zii-ultra: fix 12V_MAIN voltage
arm64: dts: zii-ultra: remove second GEN_3V3 regulator instance
arm64: dts: ls1028a: fix memory node
ARM: dts: imx6q-dhcom: Add PU,VDD1P1,VDD2P5 regulators
ARM: dts: imx6dl-yapp4: Fix RGMII connection to QCA8334 switch
Yang Yingliang [Mon, 12 Apr 2021 14:05:27 +0000 (22:05 +0800)]
soc/tegra: fuse: Don't return -ENOMEM when allocate lookups failed
fuse->base can not be unmapped if allocate lookups failed in
tegra_init_fuse(), because it is an early_initcall, the driver
will be loaded anyway and fuse->base will be accessed by other
functions later, so remove the return -ENOMEM after allocating
lookups failed to make less confusing.
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
[treding@nvidia.com: drop error message, out-of-memory is noisy anyway] Signed-off-by: Thierry Reding <treding@nvidia.com>
Dmitry Osipenko [Tue, 1 Jun 2021 02:31:19 +0000 (05:31 +0300)]
soc/tegra: regulators: Support core domain state syncing
The core voltage shall not drop until state of core domain is synced,
i.e. all device drivers that use core domain are loaded and ready.
Support core domain state syncing. The core domain driver invokes the
core-regulator voltage syncing once the state of domain is synced, at
this point the core voltage is allowed to go lower than the level left
after bootloader.
Tested-by: Peter Geis <pgwipeout@gmail.com> # Ouya T30 Tested-by: Paul Fertser <fercerpav@gmail.com> # PAZ00 T20 Tested-by: Nicolas Chauvet <kwizart@gmail.com> # PAZ00 T20 and TK1 T124 Tested-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30 Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Dmitry Osipenko <digetx@gmail.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
Dmitry Osipenko [Tue, 1 Jun 2021 02:31:18 +0000 (05:31 +0300)]
soc/tegra: pmc: Add driver state syncing
Add driver state syncing that is invoked once all PMC consumers are
attached and ready. The consumers are the power domain clients.
The synchronization callback is invoked once all client drivers are
probed, the driver core handles this for us. This callback informs
PMC driver that all voltage votes are initialized by each PD client
and it's safe to begin voltage scaling of the core power domain.
Dmitry Osipenko [Tue, 1 Jun 2021 02:31:17 +0000 (05:31 +0300)]
soc/tegra: pmc: Add core power domain
NVIDIA Tegra SoCs have multiple power domains, each domain corresponds
to an external SoC power rail. Core power domain covers vast majority of
hardware blocks within a Tegra SoC. The voltage of a power domain should
be set to a level which satisfies all devices within the power domain.
Add support for the core power domain which controls voltage state of the
domain. This allows us to support system-wide DVFS on Tegra20-210 SoCs.
The PMC powergate domains now are sub-domains of the core domain, this
requires device-tree updating, older DTBs are unaffected and will continue
to work as before.
Tested-by: Peter Geis <pgwipeout@gmail.com> # Ouya T30 Tested-by: Paul Fertser <fercerpav@gmail.com> # PAZ00 T20 Tested-by: Nicolas Chauvet <kwizart@gmail.com> # PAZ00 T20 and TK1 T124 Tested-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30 Signed-off-by: Dmitry Osipenko <digetx@gmail.com> Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
[treding@nvidia.com: squash lockdep class removal patch] Signed-off-by: Thierry Reding <treding@nvidia.com>
Dmitry Osipenko [Thu, 27 May 2021 23:54:01 +0000 (02:54 +0300)]
soc/tegra: regulators: Bump voltages on system reboot
Ensure that SoC voltages are at a level suitable for a system reboot.
This is important for some devices that use CPU reset method for the
rebooting. SoC CPU and core voltages now are be restored to a level
that is suitable for rebooting. This patch fixes hang on reboot on
Asus Transformer TF101, it was also reported as fixing some of reboot
issues on Toshiba AC100.
Reported-by: Nikola Milosavljević <mnidza@outlook.com> Tested-by: Nikola Milosavljević <mnidza@outlook.com> # TF101 Signed-off-by: Dmitry Osipenko <digetx@gmail.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
Some NVIDIA Tegra devices use a CPU soft-reset method for the reboot and
in this case we need to restore the coupled voltages to the state that is
suitable for hardware during boot. Add new regulator_sync_voltage_rdev()
helper which is needed by regulator drivers in order to sync voltage of
a coupled regulators.
Acked-by: Mark Brown <broonie@kernel.org> Signed-off-by: Dmitry Osipenko <digetx@gmail.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
This fix the recent removal of clock drivers selection.
While it is not necessary to select the clock drivers themselves, we need
to select a proper implementation of the clock API, which for the meson, is
CCF
Fixes: ba66a25536dd ("arm64: meson: ship only the necessary clock controllers") Reviewed-by: Neil Armstrong <narmstrong@baylibre.com> Signed-off-by: Jerome Brunet <jbrunet@baylibre.com> Reviewed-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com> Signed-off-by: Kevin Hilman <khilman@baylibre.com> Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Link: https://lore.kernel.org/r/20210429083823.59546-1-jbrunet@baylibre.com
Qiheng Lin [Fri, 9 Apr 2021 11:02:43 +0000 (19:02 +0800)]
soc: amlogic: meson-clk-measure: remove redundant dev_err call in meson_msr_probe()
There is a error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.
Reported-by: Hulk Robot <hulkci@huawei.com> Signed-off-by: Qiheng Lin <linqiheng@huawei.com> Signed-off-by: Kevin Hilman <khilman@baylibre.com> Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Link: https://lore.kernel.org/r/20210409110243.41-1-linqiheng@huawei.com
Manivannan Sadhasivam [Wed, 12 May 2021 05:01:41 +0000 (10:31 +0530)]
ARM: dts: qcom: sdx55-telit: Represent secure-regions as 64-bit elements
The corresponding MTD code expects the regions to be of 64-bit elements.
Hence, prefix "/bits/ 64", otherwise the regions will not be parsed
correctly.
Manivannan Sadhasivam [Wed, 12 May 2021 05:01:40 +0000 (10:31 +0530)]
ARM: dts: qcom: sdx55-t55: Represent secure-regions as 64-bit elements
The corresponding MTD code expects the regions to be of 64-bit elements.
Hence, prefix "/bits/ 64", otherwise the regions will not be parsed
correctly.
Krzysztof Kozlowski [Thu, 6 May 2021 20:27:29 +0000 (16:27 -0400)]
MAINTAINERS: Include Samsung PWM in Samsung SoC entry
The Samsung PWM and PWM-based timer/clocksource drivers lacked dedicated
maintainers entry. They are used on all Samsung SoC designs (although
timer/clocksource driver only on older platforms), so include them in
Samsung SoC entry maintained by Krzysztof Kozlowski.
Tony Lindgren [Tue, 25 May 2021 06:08:23 +0000 (09:08 +0300)]
bus: ti-sysc: Fix flakey idling of uarts and stop using swsup_sidle_act
Looks like the swsup_sidle_act quirk handling is unreliable for serial
ports. The serial ports just eventually stop idling until woken up and
re-idled again. As the serial port not idling blocks any deeper SoC idle
states, it's adds an annoying random flakeyness for power management.
Let's just switch to swsup_sidle quirk instead like we already do for
omap3 uarts. This means we manually idle the port instead of trying to
use the hardware autoidle features when not in use.
For more details on why the serial ports have been using swsup_idle_act,
see commit 66dde54e978a ("ARM: OMAP2+: hwmod-data: UART IP needs software
control to manage sidle modes"). It seems that the swsup_idle_act quirk
handling is not enough though, and for example the TI Android kernel
changed to using swsup_sidle with commit 77c34c84e1e0 ("OMAP4: HWMOD:
UART1: disable smart-idle.").
Fixes: b4a9a7a38917 ("bus: ti-sysc: Handle swsup idle mode quirks") Cc: Carl Philipp Klemm <philipp@uvos.xyz> Cc: Ivan Jelincic <parazyd@dyne.org> Cc: Merlijn Wajer <merlijn@wizzup.org> Cc: Pavel Machek <pavel@ucw.cz> Cc: Sebastian Reichel <sre@kernel.org> Cc: Sicelo A. Mhlongo <absicsz@gmail.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Commit bbc4d71d6354 ("net: phy: realtek: fix rtl8211e rx/tx delay
config") sets the RX/TX delay according to the phy-mode property in the
device tree. For the Orange Pi Plus board this is "rgmii", which is the
wrong setting.
Following the example of a900cac3750b ("ARM: dts: sun7i: a20: bananapro:
Fix ethernet phy-mode") the phy-mode is changed to "rgmii-id" which gets
the Ethernet working again on this board.
Linus Torvalds [Sun, 23 May 2021 16:32:40 +0000 (06:32 -1000)]
Merge tag 'perf-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf fixes from Thomas Gleixner:
"Two perf fixes:
- Do not check the LBR_TOS MSR when setting up unrelated LBR MSRs as
this can cause malfunction when TOS is not supported
- Allocate the LBR XSAVE buffers along with the DS buffers upfront
because allocating them when adding an event can deadlock"
* tag 'perf-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/lbr: Remove cpuc->lbr_xsave allocation from atomic context
perf/x86: Avoid touching LBR_TOS MSR for Arch LBR
Linus Torvalds [Sun, 23 May 2021 16:30:08 +0000 (06:30 -1000)]
Merge tag 'locking-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull locking fixes from Thomas Gleixner:
"Two locking fixes:
- Invoke the lockdep tracepoints in the correct place so the ordering
is correct again
- Don't leave the mutex WAITER bit stale when the last waiter is
dropping out early due to a signal as that forces all subsequent
lock operations needlessly into the slowpath until it's cleaned up
again"
* tag 'locking-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
locking/mutex: clear MUTEX_FLAGS if wait_list is empty due to signal
locking/lockdep: Correct calling tracepoints
Linus Torvalds [Sun, 23 May 2021 16:28:20 +0000 (06:28 -1000)]
Merge tag 'irq-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull irq fixes from Thomas Gleixner:
"A few fixes for irqchip drivers:
- Allocate interrupt descriptors correctly on Mainstone PXA when
SPARSE_IRQ is enabled; otherwise the interrupt association fails
- Make the APPLE AIC chip driver depend on APPLE
- Remove redundant error output on devm_ioremap_resource() failure"
* tag 'irq-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip: Remove redundant error printing
irqchip/apple-aic: APPLE_AIC should depend on ARCH_APPLE
ARM: PXA: Fix cplds irqdesc allocation when using legacy mode
Linus Torvalds [Sun, 23 May 2021 16:12:25 +0000 (06:12 -1000)]
Merge tag 'x86_urgent_for_v5.13_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Borislav Petkov:
- Fix how SEV handles MMIO accesses by forwarding potential page faults
instead of killing the machine and by using the accessors with the
exact functionality needed when accessing memory.
- Fix a confusion with Clang LTO compiler switches passed to the it
- Handle the case gracefully when VMGEXIT has been executed in
userspace
* tag 'x86_urgent_for_v5.13_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/sev-es: Use __put_user()/__get_user() for data accesses
x86/sev-es: Forward page-faults which happen during emulation
x86/sev-es: Don't return NULL from sev_es_get_ghcb()
x86/build: Fix location of '-plugin-opt=' flags
x86/sev-es: Invalidate the GHCB after completing VMGEXIT
x86/sev-es: Move sev_es_put_ghcb() in prep for follow on patch
Linus Torvalds [Sun, 23 May 2021 16:07:33 +0000 (06:07 -1000)]
Merge tag 'powerpc-5.13-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fixes from Michael Ellerman:
- Fix breakage of strace (and other ptracers etc.) when using the new
scv ABI (Power9 or later with glibc >= 2.33).
- Fix early_ioremap() on 64-bit, which broke booting on some machines.
Thanks to Dmitry V. Levin, Nicholas Piggin, Alexey Kardashevskiy, and
Christophe Leroy.
* tag 'powerpc-5.13-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/64s/syscall: Fix ptrace syscall info with scv syscalls
powerpc/64s/syscall: Use pt_regs.trap to distinguish syscall ABI difference between sc and scv syscalls
powerpc: Fix early setup to make early_ioremap() work
Michael Walle [Fri, 14 May 2021 18:55:53 +0000 (20:55 +0200)]
arm64: dts: freescale: sl28: var1: fix RGMII clock and voltage
During hardware validation it was noticed that the clock isn't
continuously enabled when there is no link. This is because the 125MHz
clock is derived from the internal PLL which seems to go into some kind
of power-down mode every once in a while. The LS1028A expects a contiuous
clock. Thus enable the PLL all the time.
Also, the RGMII pad voltage is wrong, it was configured to 2.5V (that is
the VDDH regulator). The correct voltage is 1.8V, i.e. the VDDIO
regulator.
This fix is for the freescale/fsl-ls1028a-kontron-sl28-var1.dts.
Michael Walle [Fri, 14 May 2021 18:55:52 +0000 (20:55 +0200)]
arm64: dts: freescale: sl28: var4: fix RGMII clock and voltage
During hardware validation it was noticed that the clock isn't
continuously enabled when there is no link. This is because the 125MHz
clock is derived from the internal PLL which seems to go into some kind
of power-down mode every once in a while. The LS1028A expects a contiuous
clock. Thus enable the PLL all the time.
Also, the RGMII pad voltage is wrong. It was configured to 2.5V (that is
the VDDH regulator). The correct voltage is 1.8V, i.e. the VDDIO
regulator.
This fix is for the freescale/fsl-ls1028a-kontron-sl28-var4.dts.
Oleksij Rempel [Tue, 11 May 2021 04:37:29 +0000 (06:37 +0200)]
ARM: imx6q: remove PHY fixup for KSZ9031
Starting with:
bcf3440c6dd7 ("net: phy: micrel: add phy-mode support for the KSZ9031 PHY")
the micrel phy driver started respecting phy-mode for the KSZ9031 PHY.
At least with kernel v5.8 configuration provided by this fixup was
overwritten by the micrel driver.
This configuration is outside of the recommended RGMII clock skew delays
and about in the middle of: rgmii-idrx and rgmii-id
Since most embedded systems do not have enough place to introduce
significant clock skew, rgmii-id is the way to go.
In case this patch breaks network functionality on your system, build
kernel with enabled MICREL_PHY. If it is still not working then try
following device tree options:
1. Set (or change) phy-mode in DT to:
phy-mode = "rgmii-id";
This actives internal delay for both RX and TX.
1. Set (or change) phy-mode in DT to:
phy-mode = "rgmii-idrx";
This actives internal delay for RX only.
3. Use following DT properties:
phy-mode = "rgmii";
txen-skew-psec = <0>;
rxdv-skew-psec = <0>;
rxd0-skew-psec = <0>;
rxd1-skew-psec = <0>;
rxd2-skew-psec = <0>;
rxd3-skew-psec = <0>;
rxc-skew-psec = <1860>;
txc-skew-psec = <1860>;
This activates the internal delays for RX and TX, with the value as
the fixup that is removed in this patch.
Lucas Stach [Fri, 7 May 2021 19:44:39 +0000 (21:44 +0200)]
arm64: dts: zii-ultra: remove second GEN_3V3 regulator instance
When adding the sound support a second instance of the GEN_3V3 regulator was
added by accident. Remove it and point the consumers to the first instance.
Linus Torvalds [Sun, 23 May 2021 01:20:20 +0000 (15:20 -1000)]
Merge branch 'akpm' (patches from Andrew)
Merge misc fixes from Andrew Morton:
"10 patches.
Subsystems affected by this patch series: mm (pagealloc, gup, kasan,
and userfaultfd), ipc, selftests, watchdog, bitmap, procfs, and lib"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
userfaultfd: hugetlbfs: fix new flag usage in error path
lib: kunit: suppress a compilation warning of frame size
proc: remove Alexey from MAINTAINERS
linux/bits.h: fix compilation error with GENMASK
watchdog: reliable handling of timestamps
kasan: slab: always reset the tag in get_freepointer_safe()
tools/testing/selftests/exec: fix link error
ipc/mqueue, msg, sem: avoid relying on a stack reference past its expiry
Revert "mm/gup: check page posion status for coredump."
mm/shuffle: fix section mismatch warning
Mike Kravetz [Sun, 23 May 2021 00:42:11 +0000 (17:42 -0700)]
userfaultfd: hugetlbfs: fix new flag usage in error path
In commit d6995da31122 ("hugetlb: use page.private for hugetlb specific
page flags") the use of PagePrivate to indicate a reservation count
should be restored at free time was changed to the hugetlb specific flag
HPageRestoreReserve. Changes to a userfaultfd error path as well as a
VM_BUG_ON() in remove_inode_hugepages() were overlooked.
Users could see incorrect hugetlb reserve counts if they experience an
error with a UFFDIO_COPY operation. Specifically, this would be the
result of an unlikely copy_huge_page_from_user error. There is not an
increased chance of hitting the VM_BUG_ON.
Link: https://lkml.kernel.org/r/20210521233952.236434-1-mike.kravetz@oracle.com Fixes: d6995da31122 ("hugetlb: use page.private for hugetlb specific page flags") Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com> Reviewed-by: Mina Almasry <almasry.mina@google.com> Cc: Oscar Salvador <osalvador@suse.de> Cc: Michal Hocko <mhocko@suse.com> Cc: Muchun Song <songmuchun@bytedance.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: David Hildenbrand <david@redhat.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Miaohe Lin <linmiaohe@huawei.com> Cc: Mina Almasry <almasrymina@google.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Zhen Lei [Sun, 23 May 2021 00:42:08 +0000 (17:42 -0700)]
lib: kunit: suppress a compilation warning of frame size
lib/bitfield_kunit.c: In function `test_bitfields_constants':
lib/bitfield_kunit.c:93:1: warning: the frame size of 7456 bytes is larger than 2048 bytes [-Wframe-larger-than=]
}
^
As the description of BITFIELD_KUNIT in lib/Kconfig.debug, it "Only useful
for kernel devs running the KUnit test harness, and not intended for
inclusion into a production build". Therefore, it is not worth modifying
variable 'test_bitfields_constants' to clear this warning. Just suppress
it.
Rikard Falkeborn [Sun, 23 May 2021 00:42:02 +0000 (17:42 -0700)]
linux/bits.h: fix compilation error with GENMASK
GENMASK() has an input check which uses __builtin_choose_expr() to
enable a compile time sanity check of its inputs if they are known at
compile time.
However, it turns out that __builtin_constant_p() does not always return
a compile time constant [0]. It was thought this problem was fixed with
gcc 4.9 [1], but apparently this is not the case [2].
Switch to use __is_constexpr() instead which always returns a compile time
constant, regardless of its inputs.
Petr Mladek [Sun, 23 May 2021 00:41:59 +0000 (17:41 -0700)]
watchdog: reliable handling of timestamps
Commit 9bf3bc949f8a ("watchdog: cleanup handling of false positives")
tried to handle a virtual host stopped by the host a more
straightforward and cleaner way.
But it introduced a risk of false softlockup reports. The virtual host
might be stopped at any time, for example between
kvm_check_and_clear_guest_paused() and is_softlockup(). As a result,
is_softlockup() might read the updated jiffies and detects a softlockup.
A solution might be to put back kvm_check_and_clear_guest_paused() after
is_softlockup() and detect it. But it would put back the cycle that
complicates the logic.
In fact, the handling of all the timestamps is not reliable. The code
does not guarantee when and how many times the timestamps are read. For
example, "period_ts" might be touched anytime also from NMI and re-read in
is_softlockup(). It works just by chance.
Fix all the problems by making the code even more explicit.
1. Make sure that "now" and "period_ts" timestamps are read only once.
They might be changed at anytime by NMI or when the virtual guest is
stopped by the host. Note that "now" timestamp does this implicitly
because "jiffies" is marked volatile.
2. "now" time must be read first. The state of "period_ts" will
decide whether it will be used or the period will get restarted.
3. kvm_check_and_clear_guest_paused() must be called before reading
"period_ts". It touches the variable when the guest was stopped.
As a result, "now" timestamp is used only when the watchdog was not
touched and the guest not stopped in the meantime. "period_ts" is
restarted in all other situations.
Link: https://lkml.kernel.org/r/YKT55gw+RZfyoFf7@alley Fixes: 9bf3bc949f8aeefeacea4b ("watchdog: cleanup handling of false positives") Signed-off-by: Petr Mladek <pmladek@suse.com> Reported-by: Sergey Senozhatsky <senozhatsky@chromium.org> Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Yang Yingliang [Sun, 23 May 2021 00:41:53 +0000 (17:41 -0700)]
tools/testing/selftests/exec: fix link error
Fix the link error by adding '-static':
gcc -Wall -Wl,-z,max-page-size=0x1000 -pie load_address.c -o /home/yang/linux/tools/testing/selftests/exec/load_address_4096
/usr/bin/ld: /tmp/ccopEGun.o: relocation R_AARCH64_ADR_PREL_PG_HI21 against symbol `stderr@@GLIBC_2.17' which may bind externally can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: /tmp/ccopEGun.o(.text+0x158): unresolvable R_AARCH64_ADR_PREL_PG_HI21 relocation against symbol `stderr@@GLIBC_2.17'
/usr/bin/ld: final link failed: bad value
collect2: error: ld returned 1 exit status
make: *** [Makefile:25: tools/testing/selftests/exec/load_address_4096] Error 1
Link: https://lkml.kernel.org/r/20210514092422.2367367-1-yangyingliang@huawei.com Fixes: 206e22f01941 ("tools/testing/selftests: add self-test for verifying load alignment") Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Cc: Chris Kennelly <ckennelly@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Varad Gautam [Sun, 23 May 2021 00:41:49 +0000 (17:41 -0700)]
ipc/mqueue, msg, sem: avoid relying on a stack reference past its expiry
do_mq_timedreceive calls wq_sleep with a stack local address. The
sender (do_mq_timedsend) uses this address to later call pipelined_send.
This leads to a very hard to trigger race where a do_mq_timedreceive
call might return and leave do_mq_timedsend to rely on an invalid
address, causing the following crash:
1. do_mq_timedreceive calls wq_sleep with the address of `struct
ext_wait_queue` on function stack (aliased as `ewq_addr` here) - it
holds a valid `struct ext_wait_queue *` as long as the stack has not
been overwritten.
2. `ewq_addr` gets added to info->e_wait_q[RECV].list in wq_add, and
do_mq_timedsend receives it via wq_get_first_waiter(info, RECV) to call
__pipelined_op.
3. Sender calls __pipelined_op::smp_store_release(&this->state,
STATE_READY). Here is where the race window begins. (`this` is
`ewq_addr`.)
4. If the receiver wakes up now in do_mq_timedreceive::wq_sleep, it
will see `state == STATE_READY` and break.
5. do_mq_timedreceive returns, and `ewq_addr` is no longer guaranteed
to be a `struct ext_wait_queue *` since it was on do_mq_timedreceive's
stack. (Although the address may not get overwritten until another
function happens to touch it, which means it can persist around for an
indefinite time.)
6. do_mq_timedsend::__pipelined_op() still believes `ewq_addr` is a
`struct ext_wait_queue *`, and uses it to find a task_struct to pass to
the wake_q_add_safe call. In the lucky case where nothing has
overwritten `ewq_addr` yet, `ewq_addr->task` is the right task_struct.
In the unlucky case, __pipelined_op::wake_q_add_safe gets handed a
bogus address as the receiver's task_struct causing the crash.
do_mq_timedsend::__pipelined_op() should not dereference `this` after
setting STATE_READY, as the receiver counterpart is now free to return.
Change __pipelined_op to call wake_q_add_safe on the receiver's
task_struct returned by get_task_struct, instead of dereferencing `this`
which sits on the receiver's stack.
As Manfred pointed out, the race potentially also exists in
ipc/msg.c::expunge_all and ipc/sem.c::wake_up_sem_queue_prepare. Fix
those in the same way.
Link: https://lkml.kernel.org/r/20210510102950.12551-1-varad.gautam@suse.com Fixes: c5b2cbdbdac563 ("ipc/mqueue.c: update/document memory barriers") Fixes: 8116b54e7e23ef ("ipc/sem.c: document and update memory barriers") Fixes: 0d97a82ba830d8 ("ipc/msg.c: update and document memory barriers") Signed-off-by: Varad Gautam <varad.gautam@suse.com> Reported-by: Matthias von Faber <matthias.vonfaber@aox-tech.de> Acked-by: Davidlohr Bueso <dbueso@suse.de> Acked-by: Manfred Spraul <manfred@colorfullife.com> Cc: Christian Brauner <christian.brauner@ubuntu.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Michal Hocko [Sun, 23 May 2021 00:41:46 +0000 (17:41 -0700)]
Revert "mm/gup: check page posion status for coredump."
While reviewing [1] I came across commit d3378e86d182 ("mm/gup: check
page posion status for coredump.") and noticed that this patch is broken
in two ways. First it doesn't really prevent hwpoison pages from being
dumped because hwpoison pages can be marked asynchornously at any time
after the check. Secondly, and more importantly, the patch introduces a
ref count leak because get_dump_page takes a reference on the page which
is not released.
It also seems that the patch was merged incorrectly because there were
follow up changes not included as well as discussions on how to address
the underlying problem [2]
Arnd Bergmann [Sun, 23 May 2021 00:41:43 +0000 (17:41 -0700)]
mm/shuffle: fix section mismatch warning
clang sometimes decides not to inline shuffle_zone(), but it calls a
__meminit function. Without the extra __meminit annotation we get this
warning:
WARNING: modpost: vmlinux.o(.text+0x2a86d4): Section mismatch in reference from the function shuffle_zone() to the function .meminit.text:__shuffle_zone()
The function shuffle_zone() references
the function __meminit __shuffle_zone().
This is often because shuffle_zone lacks a __meminit
annotation or the annotation of __shuffle_zone is wrong.
shuffle_free_memory() did not show the same problem in my tests, but it
could happen in theory as well, so mark both as __meminit.
Link: https://lkml.kernel.org/r/20210514135952.2928094-1-arnd@kernel.org Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: David Hildenbrand <david@redhat.com> Reviewed-by: Nathan Chancellor <nathan@kernel.org> Cc: Nick Desaulniers <ndesaulniers@google.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Wei Yang <richard.weiyang@linux.alibaba.com> Cc: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* tag 'block-5.13-2021-05-22' of git://git.kernel.dk/linux-block:
block: fix a race between del_gendisk and BLKRRPART
block: prevent block device lookups at the beginning of del_gendisk
nvme-fc: clear q_live at beginning of association teardown
nvme-tcp: rerun io_work if req_list is not empty
nvme-tcp: fix possible use-after-completion
nvme-loop: fix memory leak in nvme_loop_create_ctrl()
nvmet: fix memory leak in nvmet_alloc_ctrl()
Linus Torvalds [Sat, 22 May 2021 17:33:09 +0000 (07:33 -1000)]
Merge tag 'for-linus-5.13b-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen fixes from Juergen Gross:
- a fix for a boot regression when running as PV guest on hardware
without NX support
- a small series fixing a bug in the Xen pciback driver when
configuring a PCI card with multiple virtual functions
* tag 'for-linus-5.13b-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen-pciback: reconfigure also from backend watch handler
xen-pciback: redo VF placement in the virtual topology
x86/Xen: swap NX determination and GDT setup on BSP
Michael Walle [Thu, 8 Apr 2021 11:02:18 +0000 (13:02 +0200)]
arm64: dts: ls1028a: fix memory node
While enabling EDAC support for the LS1028A it was discovered that the
memory node has a wrong endianness setting as well as a wrong interrupt
assignment. Fix both.
This was tested on a sl28 board. To force ECC errors, you can use the
error injection supported by the controller in hardware (with
CONFIG_EDAC_DEBUG enabled):
# enable error injection
$ echo 0x100 > /sys/devices/system/edac/mc/mc0/inject_ctrl
# flip lowest bit of the data
$ echo 0x1 > /sys/devices/system/edac/mc/mc0/inject_data_lo
Fixes: 8897f3255c9c ("arm64: dts: Add support for NXP LS1028A SoC") Signed-off-by: Michael Walle <michael@walle.cc> Signed-off-by: Shawn Guo <shawnguo@kernel.org>
projects / users / hch / dma-mapping.git / log