| Age | Commit message (Collapse) | Author |
|---|
|
This allows shutting down dss domain when the screen blanks.
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
Add remaining PRM instances for the omap5 SoC. Additionally enable the
genpd support for them.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
We can now probe devices with ti-sysc interconnect driver and dts
data. Let's drop the related platform data and custom ti,hwmods
dts property.
Cc: Roger Quadros <rogerq@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
We can now probe devices with device tree only configuration using
ti-sysc interconnect target module driver.
Cc: Suman Anna <s-anna@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
Add remaining PRM instances for the dra7 SoC. Additionally enable the
genpd support for them.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
We can now probe devices with ti-sysc interconnect driver and dts
data. Let's drop the related platform data and custom ti,hwmods
dts property.
Cc: Roger Quadros <rogerq@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
We can now probe devices with ti-sysc interconnect driver and dts
data. Let's drop the related platform data and custom ti,hwmods
dts property.
Cc: Suman Anna <s-anna@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
This allows shutting down the dsp domain when not in use.
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
This allows shutting down dss domain when the screen blanks.
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
Add remaining PRM instances for the omap4 SoC. Additionally enable the
genpd support for them.
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
MAC0 was not functional in the AST2600A0 SoC. This has been resolved
with the A1, so allow use of this port on EVBs with the A1 and
subsequent revisions.
A0 EVBs will still boot with this change, but the first Ethernet device
will not be functional.
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Make them lowercase.
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20201002063414.275161-4-andrew@aj.id.au
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Some of the FSI-attached SPI controllers can't use the loop command
due to security requirements. Indicate this in the devicetree with
the restricted compatible string.
Signed-off-by: Eddie James <eajames@linux.ibm.com>
Link: https://lore.kernel.org/r/20201102194713.14812-1-eajames@linux.ibm.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Use IS_ERR_OR_NULL() to make the code cleaner.
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
clk_prepare_enable/clk_disable_unprepare
Because clk_prepare_enable() and clk_disable_unprepare() already checked
NULL clock parameter, so the additional checks are unnecessary, just
remove them.
Signed-off-by: Xu Wang <vulab@iscas.ac.cn>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
The variable ret has been initialized with '-ENOMEM'. The assignment
in the if branch is redundant. So remove it.
Signed-off-by: Jing Xiangfeng <jingxiangfeng@huawei.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
The returns pointer is NULL when kzalloc fails to apply for space, so fix
kfree NULL pointer.
Signed-off-by: Youling Tang <tangyouling@loongson.cn>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
If the clk_register fails, we should free hw before function returns to
prevent memleak.
Signed-off-by: Youling Tang <tangyouling@loongson.cn>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
According to the bosch,m_can.yaml bindings the first clock shall be the "hclk",
while the second clock "cclk".
This patch fixes the order accordingly.
Fixes: 0adbe832f21a ("ARM: dts: dra76x: Add MCAN node")
Cc: Faiz Abbas <faiz_abbas@ti.com>
Cc: Tony Lindgren <tony@atomide.com>
Cc: linux-omap@vger.kernel.org
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
Add device tree for the Kudo BMC. Kudo is an Ampere (Altra)
server platform manufactured by Fii and is based on a Nuvoton
NPCM730 SoC.
Signed-off-by: Mustatfa Shehabi <Mustafa.Shehabi@fii-na.com>
Signed-off-by: Mohaimen alsmarai <Mohaimen.alsamarai@fii-na.com>
Signed-off-by: Lancelot Kao <lancelot.kao@fii-usa.com>
Reviewed-by: Vivekanand Veeracholan <vveerach@google.com>
Reviewed-by: Benjamin Fair <benjaminfair@google.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20201116232127.7066-1-lancelot.kao@fii-usa.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Add Nuvoton NPCM7xx RunBMC Olympus Quanta board device tree.
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
The Quanta GSJ BMC uses the Nuvoton NPCM730 SoC.
Included features:
1. Image partitions
2. PWM fan controller
3. USB device
4. Serial port
5. FIU
6. LEDs and GPIOs
Signed-off-by: Fran Hsu <Fran.Hsu@quantatw.com>
Reviewed-by: Benjamin Fair <benjaminfair@google.com>
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Add pinctrl definition for the Quanta GSJ BMC machine.
Signed-off-by: Fran Hsu <Fran.Hsu@quantatw.com>
Reviewed-by: Benjamin Fair <benjaminfair@google.com>
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
The Nuvoton NPCN730 SoC is a part of the Nuvoton NPCM7xx SoCs family.
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Reserve a 1.5MiB region of memory to record kmsg dumps, console and
userspace message state into 16kiB ring-buffer slots. The sizing allows
for up to 16 dumps to be captured and read out.
Set max-reason to KMSG_DUMP_EMERG to capture bad-path reboots.
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20201022014731.2035438-7-andrew@aj.id.au
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Reserve a 1.5MiB region of memory to record kmsg dumps, console and
userspace message state into 16kiB ring-buffer slots. The sizing allows
for up to 16 dumps to be captured and read out.
Set max-reason to KMSG_DUMP_EMERG to capture bad-path reboots.
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20201022014731.2035438-6-andrew@aj.id.au
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
The mismatch lead to a miscalculation of regions in another patch, and
shouldn't be mismatched anyway, so make them consistent.
Fixes: 575640201e66 ("ARM: dts: aspeed: tacoma: Use 64MB for firmware memory")
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20201022014731.2035438-2-andrew@aj.id.au
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Drop the dma_direct_set_offset export and move the declaration to
dma-map-ops.h now that the Allwinner drivers have stopped calling it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
|
|
This driver uses GPIO descriptors to drive the touchscreen RESET line. In
the existing device trees this has in conflict with intution been flagged
as GPIO_ACTIVE_HIGH and the driver then applies the reverse action by
driving the line low (setting to 0) to enter reset state and driving the
line high (setting to 1) to get out of reset state.
The correct way to handle active low GPIO lines is to provide the
GPIO_ACTIVE_LOW in the device tree (thus properly describing the hardware)
and letting the GPIO framework invert the assertion (driving high) to a
low level and vice versa.
This is considered a bug since the device trees are incorrectly
mis-specifying the line as active high.
Fix the driver and all device trees specifying a reset line.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Philippe Schenker <philippe.schenker@toradex.com>
Acked-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20201104153032.1387747-1-linus.walleij@linaro.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
|
|
The DLCI driver (dlci.c) implements the Frame Relay protocol. However,
we already have another newer and better implementation of Frame Relay
provided by the HDLC_FR driver (hdlc_fr.c).
The DLCI driver's implementation of Frame Relay is used by only one
hardware driver in the kernel - the SDLA driver (sdla.c).
The SDLA driver provides Frame Relay support for the Sangoma S50x devices.
However, the vendor provides their own driver (along with their own
multi-WAN-protocol implementations including Frame Relay), called WANPIPE.
I believe most users of the hardware would use the vendor-provided WANPIPE
driver instead.
(The WANPIPE driver was even once in the kernel, but was deleted in
commit 8db60bcf3021 ("[WAN]: Remove broken and unmaintained Sangoma
drivers.") because the vendor no longer updated the in-kernel WANPIPE
driver.)
Cc: Mike McLagan <mike.mclagan@linux.org>
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Link: https://lore.kernel.org/r/20201114150921.685594-1-xie.he.0141@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This property is only useful if a node has children as it allows them
to then use io-channel properties in the parent. Here there are no
children.
This is harmless, but we are planning to shortly drop this property
as it is rarely used correctly and there is little reason it would
ever be needed as we can just provide the io-channels property to
any child nodes that need it.
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/20201115192951.1073632-8-jic23@kernel.org
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
This property is relevant to consumers of io-channels. Here it is
used by a provider.
dt-schema will now report and error as this property must be in the
same node as io-channels and it is not here.
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/20201115192951.1073632-6-jic23@kernel.org
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
This property is only relevant to consumers of io-channels, not providers.
All these dtsi files have it alongside #io-channel-cells which indicates
they are providers of io-channels, not consumers.
Note that dt-schema will now flag this up due to a dependency between
this property and io-channels.
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/20201115192951.1073632-5-jic23@kernel.org
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
Add a device-tree for the Rainier 4U system. Change the model name
on the existing Rainier device-tree to 2U, and remove the two
extra power supplies that are only present on the 4U system. Also
add labels to the fan nodes for use in the 4U device-tree.
Signed-off-by: Eddie James <eajames@linux.ibm.com>
Reviewed-by: Andrew Jeffery <andrew@aj.id.au>
Link: https://lore.kernel.org/r/20201028185647.14565-1-eajames@linux.ibm.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Add bindings for the KS8851 ethernet present on the STM32MP1 DHCOM SoM.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Cc: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Cc: linux-stm32@st-md-mailman.stormreply.com
To: linux-arm-kernel@lists.infradead.org
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>
|
|
Add another mux option for FMC EBI bus, this is used on DHCOM SoM
for the second ethernet and on the PDK2 devkit for SRAM.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Cc: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Cc: linux-stm32@st-md-mailman.stormreply.com
To: linux-arm-kernel@lists.infradead.org
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>
|
|
On the production DHCOM STM32MP15xx SoM, the PHY IRQ line is connected
to the PI11 pin. Describe it in the DT as well, so the PHY IRQ can be
used e.g. to detect cable insertion and removal.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Cc: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Cc: linux-stm32@st-md-mailman.stormreply.com
To: linux-arm-kernel@lists.infradead.org
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>
|
|
Enable FIFO mode with half-full threshold.
Signed-off-by: Hugues Fruchet <hugues.fruchet@st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@st.com>
|
|
Add the following new device nodes to NPCM750 evolution board device
tree:
- NPCM7xx Pin controller and GPIO
- NPCM7xx PWM and FAN
- NPCM7xx EHCI USB
- NPCM7xx KCS
- NPCM Reset
- NPCM Peripheral SPI
- NPCM FIU SPI
- NPCM HWRNG
- NPCM I2C
- STMicro STMMAC
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20200929131807.15378-6-tmaimon77@gmail.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Add the following new device nodes to NPCM7XX:
- NPCM7xx PWM and FAN.
- NPCM7xx EHCI USB.
- NPCM7xx KCS.
- NPCM Reset.
- NPCM Peripheral SPI.
- NPCM FIU SPI.
- NPCM HWRNG.
- NPCM I2C.
- STMicro STMMAC.
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20200929131807.15378-5-tmaimon77@gmail.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Add pin controller and GPIO node to NPCM7XX device tree.
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20200929131807.15378-4-tmaimon77@gmail.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Modify NPCM7xx device tree timer register size
from 0x50 to 0x1C to control only the timer registers
and not other hw modules.
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20200929131807.15378-3-tmaimon77@gmail.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
Modify NPCM7xx device tree clock parameter to clock constants that
define at include/dt-bindings/clock/nuvoton,npcm7xx-clock.h file.
Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20200929131807.15378-2-tmaimon77@gmail.com
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc
Pull ARM SoC fixes from Arnd Bergmann:
"Around one third of the fixes this time are for dts files that list
their ethernet controller as using 'phy-mode="rgmii"' but are changed
to 'phy-mode="rgmii-id"' now, because the PHY drivers (realtek,
ksz9031, dp83867, ...) now configure the internal delay based on that
when they used to stay on the hardware default.
The long story is archived at
https://lore.kernel.org/netdev/CAMj1kXEEF_Un-4NTaD5iUN0NoZYaJQn-rPediX0S6oRiuVuW-A@mail.gmail.com/
I was trying to hold off on the bugfixes until there was a solution
that would avoid breaking all boards, but that does not seem to be
happening any time soon, so I am now sending the correct version of
the dts files to ensure that at least these machines can use their
network devices again.
The other changes this time are:
- Updating the MAINTAINER lists for Allwinner and Samsung SoCs
- Multiple i.MX8MN machines get updates for their CPU operating
points to match the data sheet
- A revert for a dts patch that caused a regression in USB support on
Odroid U3
- Two fixes for the AMD Tee driver, addressing a memory leak and
missing locking
- Mark the network subsystem on qoriq-fman3 as cache coherent for
correctness as better performance.
- Minor dts fixes elsewhere, addressing dtc warnings and similar
problems"
* tag 'arm-soc-fixes-v5.10-2' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc: (48 commits)
ARM: dts: exynos: revert "add input clock to CMU in Exynos4412 Odroid"
ARM: dts: imx50-evk: Fix the chip select 1 IOMUX
arm64: dts: imx8mm: fix voltage for 1.6GHz CPU operating point
ARM: dts: stm32: Keep VDDA LDO1 always on on DHCOM
ARM: dts: stm32: Enable thermal sensor support on stm32mp15xx-dhcor
ARM: dts: stm32: Define VIO regulator supply on DHCOM
ARM: dts: stm32: Fix LED5 on STM32MP1 DHCOM PDK2
ARM: dts: stm32: Fix TA3-GPIO-C key on STM32MP1 DHCOM PDK2
arm64: dts: renesas: r8a774e1: Add missing audio_clk_b
tee: amdtee: synchronize access to shm list
tee: amdtee: fix memory leak due to reset of global shm list
arm64: dts: agilex/stratix10: Fix qspi node compatible
ARM: dts: imx6q-prti6q: fix PHY address
ARM: dts: vf610-zii-dev-rev-b: Fix MDIO over clocking
arm: dts: imx6qdl-udoo: fix rgmii phy-mode for ksz9031 phy
arm64: dts imx8mn: Remove non-existent USB OTG2
arm64: dts: imx8mm-beacon-som: Fix Choppy BT audio
arm64: dts: fsl: DPAA FMan DMA operations are coherent
arm64: dts: fsl: fix endianness issue of rcpm
arm64: dts: imx8mn-evk: fix missing PMIC's interrupt line pull-up
...
|
|
The P4 Note family contains a couple of variants of the Galaxy Note 10.1
tablet with mainly different modems. The GT-N8010/GT-N8013 is the WiFi
only version.
Signed-off-by: Martin Jücker <martin.juecker@gmail.com>
Link: https://lore.kernel.org/r/20201113212525.13455-3-martin.juecker@gmail.com
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
The Atmel MXT touchscreen can load firmware and settings from the
/lib/firmware directory, it makes sense to have it as a module to have
more control over the loading process.
Signed-off-by: Martin Jücker <martin.juecker@gmail.com>
Link: https://lore.kernel.org/r/20201113212525.13455-5-martin.juecker@gmail.com
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
Enable the STMPE ADC driver for the P4 Note device family in exynos and
multi_v7 defconfigs.
Signed-off-by: Martin Jücker <martin.juecker@gmail.com>
Link: https://lore.kernel.org/r/20201113212525.13455-4-martin.juecker@gmail.com
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
The P4 Note family supports cpuidle, so allow it to make use of this
feature.
Signed-off-by: Martin Jücker <martin.juecker@gmail.com>
Link: https://lore.kernel.org/r/20201113212525.13455-6-martin.juecker@gmail.com
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into arm/fixes
i.MX fixes for 5.10, round 4:
- Fix MDIO over clocking on vf610-zii-dev-rev-b board to get switch
device work reliably.
- Fix imx50-evk IOMUX for the chip select 1 to use GPIO4_13 instead of
the native CSPI_SSI function.
- Fix voltage for 1.6GHz CPU operating point on i.MX8MM to match
hardware datasheet.
- Fix phy-mode for KSZ9031 PHY on imx6qdl-udoo board.
* tag 'imx-fixes-5.10-4' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: dts: imx50-evk: Fix the chip select 1 IOMUX
arm64: dts: imx8mm: fix voltage for 1.6GHz CPU operating point
ARM: dts: vf610-zii-dev-rev-b: Fix MDIO over clocking
arm: dts: imx6qdl-udoo: fix rgmii phy-mode for ksz9031 phy
Link: https://lore.kernel.org/r/20201116090702.GM5849@dragon
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
|
|
Stefan Agner reported a bug when using zsram on 32-bit Arm machines
with RAM above the 4GB address boundary:
Unable to handle kernel NULL pointer dereference at virtual address 00000000
pgd = a27bd01c
[00000000] *pgd=236a0003, *pmd=1ffa64003
Internal error: Oops: 207 [#1] SMP ARM
Modules linked in: mdio_bcm_unimac(+) brcmfmac cfg80211 brcmutil raspberrypi_hwmon hci_uart crc32_arm_ce bcm2711_thermal phy_generic genet
CPU: 0 PID: 123 Comm: mkfs.ext4 Not tainted 5.9.6 #1
Hardware name: BCM2711
PC is at zs_map_object+0x94/0x338
LR is at zram_bvec_rw.constprop.0+0x330/0xa64
pc : [<c0602b38>] lr : [<c0bda6a0>] psr: 60000013
sp : e376bbe0 ip : 00000000 fp : c1e2921c
r10: 00000002 r9 : c1dda730 r8 : 00000000
r7 : e8ff7a00 r6 : 00000000 r5 : 02f9ffa0 r4 : e3710000
r3 : 000fdffe r2 : c1e0ce80 r1 : ebf979a0 r0 : 00000000
Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
Control: 30c5383d Table: 235c2a80 DAC: fffffffd
Process mkfs.ext4 (pid: 123, stack limit = 0x495a22e6)
Stack: (0xe376bbe0 to 0xe376c000)
As it turns out, zsram needs to know the maximum memory size, which
is defined in MAX_PHYSMEM_BITS when CONFIG_SPARSEMEM is set, or in
MAX_POSSIBLE_PHYSMEM_BITS on the x86 architecture.
The same problem will be hit on all 32-bit architectures that have a
physical address space larger than 4GB and happen to not enable sparsemem
and include asm/sparsemem.h from asm/pgtable.h.
After the initial discussion, I suggested just always defining
MAX_POSSIBLE_PHYSMEM_BITS whenever CONFIG_PHYS_ADDR_T_64BIT is
set, or provoking a build error otherwise. This addresses all
configurations that can currently have this runtime bug, but
leaves all other configurations unchanged.
I looked up the possible number of bits in source code and
datasheets, here is what I found:
- on ARC, CONFIG_ARC_HAS_PAE40 controls whether 32 or 40 bits are used
- on ARM, CONFIG_LPAE enables 40 bit addressing, without it we never
support more than 32 bits, even though supersections in theory allow
up to 40 bits as well.
- on MIPS, some MIPS32r1 or later chips support 36 bits, and MIPS32r5
XPA supports up to 60 bits in theory, but 40 bits are more than
anyone will ever ship
- On PowerPC, there are three different implementations of 36 bit
addressing, but 32-bit is used without CONFIG_PTE_64BIT
- On RISC-V, the normal page table format can support 34 bit
addressing. There is no highmem support on RISC-V, so anything
above 2GB is unused, but it might be useful to eventually support
CONFIG_ZRAM for high pages.
Fixes: 61989a80fb3a ("staging: zsmalloc: zsmalloc memory allocation library")
Fixes: 02390b87a945 ("mm/zsmalloc: Prepare to variable MAX_PHYSMEM_BITS")
Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Reviewed-by: Stefan Agner <stefan@agner.ch>
Tested-by: Stefan Agner <stefan@agner.ch>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Link: https://lore.kernel.org/linux-mm/bdfa44bf1c570b05d6c70898e2bbb0acf234ecdf.1604762181.git.stefan@agner.ch/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
|
