When pKVM saves and restores the host floating point state on a SVE system,
it programs the vector length in ZCR_EL2.LEN to be whatever the maximum VL
for the PE is. But it uses a buffer allocated with kvm_host_sve_max_vl, the
maximum VL shared by all PEs in the system. This means that if we run on a
system where the maximum VLs are not consistent, we will overflow the buffer
on PEs which support larger VLs.
Since the host will not currently attempt to make use of non-shared VLs, fix
this by explicitly setting the EL2 VL to be the maximum shared VL when we
save and restore. This will enforce the limit on host VL usage. Should we
wish to support asymmetric VLs, this code will need to be updated along with
the required changes for the host:
https://lore.kernel.org/r/20240730-kvm-arm64-fix-pkvm-sve-vl-v6-0-cae8a2e0bd66@kernel.org
Fixes: b5b9955617 ("KVM: arm64: Eagerly restore host fpsimd/sve state in pKVM")
Signed-off-by: Mark Brown <broonie@kernel.org>
Tested-by: Fuad Tabba <tabba@google.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Link: https://lore.kernel.org/r/20240912-kvm-arm64-limit-guest-vl-v2-1-dd2c29cb2ac9@kernel.org
[maz: added punctuation to the commit message]
Signed-off-by: Marc Zyngier <maz@kernel.org>
On an error, hyp_vcpu will be accessed while this memory has already
been relinquished to the host and unmapped from the hypervisor. Protect
the CPTR assignment with an early return.
Fixes: b5b9955617 ("KVM: arm64: Eagerly restore host fpsimd/sve state in pKVM")
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Link: https://lore.kernel.org/r/20240919110500.2345927-1-vdonnefort@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
These drivers don't use the driver_data member of struct i2c_device_id,
so don't explicitly initialize this member.
This prepares putting driver_data in an anonymous union which requires
either no initialization or named designators. But it's also a nice
cleanup on its own.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
Link: https://lore.kernel.org/20240920153430.503212-13-u.kleine-koenig@baylibre.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Document the Top Level Mode Multiplexer on the QCS615 Platform.
It concisely explains the pin multiplexing and configuration in
the device tree, and includes simple examples of typical device
tree snippets, making it easier for designers to configure and
manage chip pins.
Signed-off-by: Lijuan Gao <quic_lijuang@quicinc.com>
Link: https://lore.kernel.org/20240920-add_qcs615_pinctrl_driver-v2-1-e03c42a9d055@quicinc.com
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Merge series from Igor Prusov <ivprusov@salutedevices.com>:
This series adds support for two NeoFidelity amplifiers. For both
amplifiers vendor provides software for equalizer and filters
configuration, which generates firmware files with registers values.
Since in both cases those files have same encoding, a common helper
module is added to get firmware via request_firmware() API and set
registers values.
Merge series from Javier Carrasco <javier.carrasco.cruz@gmail.com>:
Declare `snd_soc_component_driver` as const to move their declarations
to read-only sections for the drivers that do not modify the struct
after its declaration.
Apart from a single case under media/, the affected drivers are members
of the ASoC subsystem.
Merge series from Andrei Simion <andrei.simion@microchip.com>:
This patch set includes two updates for the MCHP SPDIF RX and TX drivers.
The patches remove the interface name from the stream_name, allowing the
interface name and index to be set in the Device Tree (DT) using the
sound-name-prefix string property.
Merge series from Andrei Simion <andrei.simion@microchip.com>:
This patch set includes two updates for the Atmel SSC DAI driver:
- Address the limitation with the S24_LE format.
- Add stream names for DPCM and future use-cases.
Merge series from Hongbo Li <lihongbo22@huawei.com>:
Many variable in macro are not used as we used macro_check
script to detect and mamually check, let us address these
issues.
Merge series from Vijendar Mukunda <Vijendar.Mukunda@amd.com>:
Currently different logics being used in the code for acp version
differentiation. This patch series refactors the code to use acp pci
revision id for handling acp version specific code.
Merge series from Shengjiu Wang <shengjiu.wang@nxp.com>:
Fix the usage of regmap_write_bits().
Move mclk clock enablement to late stage.
Enable the micfil error interrupt.
Originally, we have an active flag to record whether we have set PHY once.
After impending MLO support, there will be dual-PHY and they can be set
individually on Wi-Fi 7 chips. So, we now have active flag per PHY and
handle them individually.
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20240925020119.13170-3-pkshih@realtek.com
To support MLO, we initialize things on dual HW bands of Wi-Fi 7 chip.
And, each link will indicate which HW band it's bound to. So, in link
control flow, we control major things based on target link's HW band
instead of hardcode like RTW89_PHY_X or RTW89_MAC_X.
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20240925020119.13170-2-pkshih@realtek.com
Read calibration value of PCI RX offset, and set to manual mode as the
value at PCI probe to prevent abnormal calibration results at runtime.
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20240925013901.9835-5-pkshih@realtek.com
To use manual mode to set value of filter out EQ, read the source value of
filter out EQ coded in gray code, and write to the target register in
binary number. The function to convert from gray code to binary number is
wrong originally, causing unexpected EQ, so fix it.
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20240925013901.9835-4-pkshih@realtek.com
For adaption EQ circuit, this HW design and affected by EIEOS (Electrical
Idle Exit Order Set) amplitude from platform and process from IC, so
disable EQ to improve that.
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20240925013901.9835-3-pkshih@realtek.com
This patch doesn't change runtime at all, it's just for kernel hardening.
The "count" here comes from the user and on 32bit systems, it leads to
integer wrapping when we pass it to compute_user_elem_size():
alloc_size = compute_user_elem_size(private_size, count);
However, the integer over is harmless because later "count" is checked
when we pass it to snd_ctl_new():
err = snd_ctl_new(&kctl, count, access, file);
These days as part of kernel hardening we're trying to avoid integer
overflows when they affect size_t type. So to avoid the integer overflow
copy the check from snd_ctl_new() and do it at the start of the
snd_ctl_elem_add() function as well.
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Jaroslav Kysela <perex@perex.cz>
Link: https://patch.msgid.link/5457e8c1-01ff-4dd9-b49c-15b817f65ee7@stanley.mountain
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The PCI settings aren't always persistent after chip suspends, so
reconfigure the settings after chip resumes. Since most of these settings
are the same, consolidate them into a function to avoid missing somewhere.
Fix the missing case of 8922AE resume flow.
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20240925013901.9835-2-pkshih@realtek.com
Configuration of the K230 is similar to that of the K210. However, in
K210, the 256 functions for each pin are shared, whereas in K230,
multiplex functions are different for every pin.
`drv_data` of `pinctrl_pin_desc` is pointing to currently activated
group, which is used to print the name of current function of pin in
`pin_dbg_show` and will be updated in `set_mux`, so they are not set
const.
Signed-off-by: Ze Huang <18771902331@163.com>
Link: https://lore.kernel.org/20240926-k230-pinctrl-v2-2-a9a36fba4b34@163.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Use the more concise interrupts-extended property to fully describe the
interrupt.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20240927095414.10241-10-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Use the more concise interrupts-extended property to fully describe the
interrupt.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20240927095414.10241-9-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Using the inherited interrupt-parent is discouraged. Use
interrupts-extended to fully describe the interrupt.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20240927095414.10241-7-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
The driver does not use this property and all upstream SoCs use the same
value anyhow. Remove it and get rid of a lot of dtbs_check warnings.
Tested with a Lager (R-Car H2) board and with the soon to be added
Genmai (RZA1) MMCIF support.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20240925150904.3582-2-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Meanwhile, bindings for the audio codec exist. Add #sound-dai-cells
property and limit max frequency accordingly.
codec@0: '#sound-dai-cells' is a required property
codec@0: spi-max-frequency: 5000000 is greater than the maximum of 526000
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20240925090812.14497-2-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
The IPQ5424 SoC comes with a TLMM block, like all other Qualcomm
platforms, so add a driver for it.
Co-developed-by: Varadarajan Narayanan <quic_varada@quicinc.com>
Signed-off-by: Varadarajan Narayanan <quic_varada@quicinc.com>
Signed-off-by: Sricharan Ramabadhran <quic_srichara@quicinc.com>
Link: https://lore.kernel.org/20240927065244.3024604-6-quic_srichara@quicinc.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Today we are proxying the `console=` command line args to the
`param_setup_earlycon()` handler. This is done because the following are
equivalent:
console=uart[8250],mmio,<addr>[,options]
earlycon=uart[8250],mmio,<addr>[,options]
Both invocations enable an early `bootconsole`. `console=uartXXXX` is
just an alias for `earlycon=uartXXXX`.
In addition, when `earlycon=` (empty value) or just `earlycon`
(no value) is specified on the command line, we enable the earlycon
`bootconsole` specified by the SPCR table or the DT.
The problem arises when `console=` (empty value) is specified on the
command line. It's intention is to disable the `console`, but what
happens instead is that the SPRC/DT console gets enabled.
This happens because we are proxying the `console=` (empty value)
parameter to the `earlycon` handler. The `earlycon` handler then sees
that the parameter value is empty, so it enables the SPCR/DT
`bootconsole`.
This change makes it so that the `console` or `console=` parameters no
longer enable the SPCR/DT `bootconsole`. I also cleans up the hack in
`main.c` that would forward the `console` parameter to the `earlycon`
handler.
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Tested-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20240911123507.v2.1.Id08823b2f848237ae90ce5c5fa7e027e97c33ad3@changeid
Signed-off-by: Petr Mladek <pmladek@suse.com>
Add bindings for the pin controllers on the T-Head TH1520 RISC-V SoC.
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Tested-by: Thomas Bonnefille <thomas.bonnefille@bootlin.com>
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com>
[dfustini: add thead,pad-group to select the pin controller instance]
Signed-off-by: Drew Fustini <dfustini@tenstorrent.com>
Link: https://lore.kernel.org/20240930-th1520-pinctrl-v3-1-32cea2bdbecb@tenstorrent.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This adds support for lan969x SoC pinctrl, reusing the existing ocelot
driver. There are 66 General Purpose I/O pins that are individually
configurable to multiple interfaces.
Signed-off-by: Daniel Machon <daniel.machon@microchip.com>
Link: https://lore.kernel.org/20240917-lan969x-pinctrl-v2-2-ea02cbc56831@microchip.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Lan969x is going to reuse the existing Ocelot pinctrl driver - document
that by adding compatible strings for the different SKU's that we
support.
Signed-off-by: Daniel Machon <daniel.machon@microchip.com>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/20240917-lan969x-pinctrl-v2-1-ea02cbc56831@microchip.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This aligns with the use of SOC_IMX* as dependencies for all the other
mach-imx pinctrl drivers. Enabling i.MX RT pinctrl drivers for a kernel
with out i.MX RT SOC support is pointless.
Signed-off-by: Esben Haabendal <esben@geanix.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/20240506-imx-pinctrl-optional-v2-2-bdff75085156@geanix.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Making pinctrl drivers and subsequently the pinctrl framework
user-controllable, allows building a kernel without this.
While in many (most) cases, this could make the system unbootable, it
does allow building smaller kernels for those situations where picntrl
is not needed.
One such situation is when building a kernel for NXP LS1021A systems,
which does not have run-time controllable pinctrl, so pinctrl framework
and drivers are 100% dead-weight.
Signed-off-by: Esben Haabendal <esben@geanix.com>
Acked-by: Shawn Guo <shawnguo@kernel.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/20240506-imx-pinctrl-optional-v2-1-bdff75085156@geanix.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Add compatible for TLMM block representing support on SA8255p.
SA8255p uses the same TLMM block as SA8775p however the ownership
of pins are split between Firmware VM and Linux VM on SA8255p. For
example, pins used by UART are owned and configured by Firmware VM
while pins used by ethernet are owned and configured by Linux VM.
Therefore, adding a sa8255p specific compatible to mark the difference.
Signed-off-by: Nikunj Kela <quic_nkela@quicinc.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/20240905200218.3810712-1-quic_nkela@quicinc.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
