Judy Hsiao [Fri, 26 Aug 2022 06:56:21 +0000 (06:56 +0000)]
arm64: dts: qcom: sc7280: Fix Dmic no sound on villager-r1
Fix the DMIC no sound issue of villager-r1 by using "PP1800_L2C" as the
DMIC power source to match the hardware schematic.
This patch:
1. set vdd-micb-supply to PP1800_L2C as the MIC Bias voltage regulator.
2. In audio-routing, set VA DMIC01~VA DMIC03 to use the vdd-micb-supply
setting.
Co-developed-by: Srinivasa Rao Mandadapu <quic_srivasam@quicinc.com> Signed-off-by: Srinivasa Rao Mandadapu <quic_srivasam@quicinc.com> Signed-off-by: Judy Hsiao <judyhsiao@chromium.org> Reviewed-by: Stephen Boyd <swboyd@chromium.org> Reviewed-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/20220826065621.2255795-1-judyhsiao@chromium.org
arm64: dts: qcom: sc7280: Add sound node for CRD 3.0/3.1
Add dt nodes for sound card support on rev5+ (aka CRD 3.0/3.1) boards,
which is using WCD9385 headset playback, capture, I2S speaker playback
and DMICs via VA macro.
arm64: dts: qcom: sc7280: Add sound node for CRD 1.0/2.0 and IDP boards
Add dt nodes for sound card support on revision 3, 4
(aka CRD 1.0 and 2.0) and IDP boards, which is using WCD9385 headset
playback, capture, I2S speaker playback and DMICs via VA macro.
arm64: dts: qcom: sc7280: Add wcd9385 codec node for CRD 1.0/2.0 and IDP boards
Add wcd9385 codec node for audio use case on sc7280 based platforms
of revision 3, 4 (aka CRD 1.0 and 2.0) and IDP boards.
Add tlmm gpio property for switching CTIA/OMTP Headset.
arm64: dts: qcom: sc7280: Add nodes for soundwire and va tx rx digital macro codecs
SC7280 has VA, TX and RX macros with SoundWire Controllers to attach with
external codecs using soundwire masters. Add these nodes for sc7280 based
platforms audio use case.
When this property is set, the remoteproc is used to boot the
LPASS and therefore lpass_q6ss_ahbm_clk and lpass_q6ss_ahbs_clk
clocks would be used to bring LPASS out of reset and the rest of
the lpass clocks would be controlled directly by the remoteproc.
This is a cleanup done to handle overlap of regmap of
lpasscc and lpass_aon blocks.
Signed-off-by: Taniya Das <quic_tdas@quicinc.com> Signed-off-by: Satya Priya <quic_c_skakit@quicinc.com> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: Stephen Boyd <sboyd@kernel.org> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/1662005846-4838-2-git-send-email-quic_c_skakit@quicinc.com
This adds sc7280-herobrine-villager-r1.dts for villager device tree files.
Herobrine-r1 is exactly the same as -r0 except that it uses a
different audio solution (it uses the same one as the CRD).
arm64: dts: qcom: sa8155p-adp: Remove unused properties from eth node
The 'snps,ptp-ref-clk-rate' and 'snps,ptp-req-clk-rate' properties
are not supported by the stmmac driver currently, so remove
them from the sa8155p-adp ethernet node as well.
Cc: Bjorn Andersson <andersson@kernel.org> Cc: Rob Herring <robh@kernel.org> Cc: Vinod Koul <vkoul@kernel.org> Cc: David Miller <davem@davemloft.net> Signed-off-by: Bhupesh Sharma <bhupesh.sharma@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/20220907204153.2039776-1-bhupesh.sharma@linaro.org
Add magnetometer Yamaha YAS537 to the DeviceTree of samsung-serranove.
The YAS537 variant was recently added to the Yamaha YAS magnetometers
driver [1].
In the DeviceTree of samsung-serranove for the Android kernel, there is
unfortunately no information on interrupts or pinctrl [2].
In the Android kernel driver for magnetometer Yamaha YAS537, there is a
device-specific matrix to correct an ellipsoid shape of the measure values
into a sphere shape [3]. This could be converted and applied to a mount-matrix.
However, the current state of the mainline Yamaha YAS537 driver needs
post-process calibration in userspace anyway, as it lacks a formula to center
the measure values around zero. The correction of the ellipsoid into a sphere
can be done in the post-process calibration as well.
A mount-matrix is needed nonetheless. When putting samsung-serranove flat on
a table in portrait orientation heading north, the Yamaha YAS537 magnetometer
axes natively point X+ to north, Y+ to east and Z+ into the ground, which
corresponds to a common way to define the Earth's magnetic field coordinate
system [4]. According to the IIO definition, it should be Y+ to north, X+ to
east and Z+ upwards [5], which corresponds to a common device coordinate system
and eases sensor fusing.
arm64: dts: qcom: sc8280xp-pmics: Remove reg entry & use correct node name for pmc8280c_lpg node
Commit eeca7d46217c ("arm64: dts: qcom: pm8350c: Drop PWM reg declaration")
dropped PWM reg declaration for pm8350c pwm(s), but there is a leftover
'reg' entry inside the lpg/pwm node in sc8280xp dts file. Remove the same.
While at it, also remove the unused unit address in the node
label.
Also, since dt-bindings expect LPG/PWM node name to be "pwm",
use correct node name as well, to fix the following
error reported by 'make dtbs_check':
Johan Hovold [Mon, 5 Sep 2022 09:16:02 +0000 (11:16 +0200)]
arm64: dts: qcom: fix syscon node names
Some recent changes that added new syscon nodes used misspelled node names.
Fixes: 86d7c9460e2c arm64: dts: qcom: sm8150: split TCSR halt regs out of mutex Fixes: 0da603387225 arm64: dts: qcom: sdm630: split TCSR halt regs out of mutex Fixes: 8a8531e69b2d arm64: dts: qcom: sdm845: split TCSR halt regs out of mutex Fixes: d9a2214d6ba5 arm64: dts: qcom: sc7280: split TCSR halt regs out of mutex Fixes: ce1ac53c7faa arm64: dts: qcom: sc7180: split TCSR halt regs out of mutex Fixes: fc10cfa38580 arm64: dts: qcom: msm8998: split TCSR halt regs out of mutex Fixes: 100ce2205924 arm64: dts: qcom: msm8996: split TCSR halt regs out of mutex Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/20220905091602.20364-1-johan+linaro@kernel.org
arm64: dts: qcom: pmk8350: drop interrupt-names from ADC
The SPMI PMIC VADC and Thermal Monitoring ADC have only one interrupt
line and their bindings do not allow interrupt-names. None of other
variants use them, so drop it from DTSI.
arm64: dts: qcom: pmk8350: drop incorrect io-channel-ranges
Since commit 044b32fa5229 ("dt-bindings:iio:qcom-spmi-vadc drop
incorrect io-channel-ranges from example") the io-channel-ranges are not
allowed in the Qualcomm SPMI PMIC ADC and anyway they are not correct
for IIO provider.
Akhil P Oommen [Sun, 28 Aug 2022 19:40:55 +0000 (01:10 +0530)]
arm64: dts: qcom: sc7280: Update gpu opp table
On the lite sku where GPU Fmax is 550Mhz, voting for a slightly higher
bandwidth at the highest gpu opp helps to improve "Manhattan offscreen"
score by 10%. Update the gpu opp table such that this is applicable only
on SKUs which has 550Mhz as GPU Fmax.
Add the alias 'wifi0' for the WiFi interface on the Qcard. The alias
is needed by the BIOS which patches the WiFi MAC address read from
the VPD (Vital Product Data) into the device tree.
Add the registers and clock for the Inline Crypto Engine (ICE) to the
device tree node for the UFS host controller on sm8450. This makes
ufs_qcom support inline encryption when CONFIG_SCSI_UFS_CRYPTO=y.
The address and size of the register range, and the minimum and maximum
frequency of the ICE core clock, all match the values used downstream.
I've validated this on an SM8450 HDK using the 'encrypt' group of
xfstests on ext4 with MOUNT_OPTIONS="-o inlinecrypt".
Stephen Boyd [Sat, 27 Aug 2022 00:49:00 +0000 (17:49 -0700)]
arm64: dts: qcom: sc7180-trogdor: Keep pm6150_adc enabled for TZ
There's still a thermal zone using pm6150_adc in the pm6150.dtsi file,
pm6150_thermal. It's not super obvious because it indirectly uses the
adc through an iio channel in pm6150_temp. Let's keep this enabled on
lazor and coachz so that reading the temperature of the pm6150_thermal
zone continues to work. Otherwise we get -EINVAL when reading the zone,
and I suspect the PMIC temperature trip doesn't work properly so we
don't shutdown when the PMIC overheats.
Bryan O'Donoghue [Sun, 28 Aug 2022 13:26:48 +0000 (14:26 +0100)]
arm64: dts: qcom: pm8350c: Drop PWM reg declaration
The PWM is a part of the SPMI PMIC block and maps several different
addresses within the SPMI block. It is not accurate to describe as pwm@reg
as a result.
Bob Moragues [Thu, 28 Jul 2022 21:33:21 +0000 (14:33 -0700)]
arm64: dts: qcom: sc7280: Add support for zoglin
Zoglin is a Hoglin Chromebook with SPI Flash reduced from 64MB to 8MB.
Zoglin is identical to Hoglin except for the SPI Flash.
The actual SPI Flash is dynamically probed at and not specified in DTS.
Bob Moragues [Thu, 28 Jul 2022 21:33:20 +0000 (14:33 -0700)]
dt-bindings: arm: qcom: document zoglin board
Zoglin is a Hoglin Chromebook with SPI Flash reduced from 64MB to 8MB.
Zoglin is identical to Hoglin except for the SPI Flash.
The actual SPI Flash is dynamically probed at and not specified in DTS.
Add LEDs found on the Xiaomi MSM8996 devices. The devices share
a status RGB LED mounted on the front, as well as a PWM-driven
IR LED for remote control (sometimes known as an IR blaster).
The Mi Note 2 has an additional pair of white LEDs used as backlights
for the touchkeys driven by the PM8994 LPG block.
arm64: dts: qcom: sdm845-db845c: drop power-domains from CCI I2C sensors
The Camera Control Interface I2C controller device node belongs to
TITAN_TOP_GDSC power domain, so its children do not need to specify it
again. The OV7251 and OV8856 camera sensor bindings do not allow
power-domains.
Do not allow the RPMh regulators to switch to low-power mode with an
exception for the UFS regulators (l3c, l6c, l10c and l17c) as UFS
supports an idle mode.
This specifically avoids having regulators be but in low-power mode when
only some consumers specify loads while the actual total load really
warrants high-power mode.
Do not allow the RPMh regulators to switch to low-power mode.
This specifically avoids having regulators be but in low-power mode when
only some consumers specify loads while the actual total load really
warrants high-power mode.
Do not allow the RPMh regulators to switch to low-power mode with an
exception for the UFS regulators (l7c and l3d) as UFS supports an idle
mode.
This specifically avoids having regulators be but in low-power mode when
only some consumers specify loads while the actual total load really
warrants high-power mode.
Abel Vesa [Fri, 12 Aug 2022 10:12:39 +0000 (13:12 +0300)]
arm64: dts: qcom: sdm845: Add the RPMh stats node
SDM845 is a special case compared to the other platforms that use RPMh
stats, since it only has 2 stats (aosd and cxsd), while the others have
a 3rd one (ddr).
So lets add the node but with a SDM845 dedicated compatible to make
the driver aware of the different stats config.
Signed-off-by: Abel Vesa <abel.vesa@linaro.org> Tested-by: Caleb Connolly <caleb.connolly@linaro.org> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/20220812101240.1869605-3-abel.vesa@linaro.org
Abel Vesa [Fri, 12 Aug 2022 10:12:37 +0000 (13:12 +0300)]
arm64: dts: qcom: sdm845: Reduce reg size for aoss_qmp
Like on the other platforms that provide RPMh stats, on SDM845, the
aoss_qmp reg size needs to be reduced to its actual size of 0x400,
otherwise it will overlap with the RPMh stats reg base, node that will
be added later on.
arm64: dts: qcom: sm8150: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/sm8150-mtp.dtb: hwlock: 'reg' is a required property
qcom/sm8150-mtp.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: sm8150: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
arm64: dts: qcom: sdm630: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/sdm636-sony-xperia-ganges-mermaid.dtb: hwlock: 'reg' is a required property
qcom/sdm636-sony-xperia-ganges-mermaid.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: sdm630: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
arm64: dts: qcom: qcs404: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/qcs404-evb-4000.dtb: hwlock: 'reg' is a required property
qcom/qcs404-evb-4000.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: sdm845: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/sdm845-shift-axolotl.dtb: hwlock: 'reg' is a required property
qcom/sdm845-shift-axolotl.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: sdm845: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
arm64: dts: qcom: sc7280: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
This also describes more accurately the devices and their IO address
space, and allows to remove incorrect syscon compatible from TCSR mutex:
qcom/sc7280-herobrine-crd.dtb: hwlock@1f40000: compatible: ['qcom,tcsr-mutex', 'syscon'] is too long
arm64: dts: qcom: sc7180: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/sc7180-trogdor-wormdingler-rev1-inx.dtb: hwlock: 'reg' is a required property
qcom/sc7180-trogdor-wormdingler-rev1-inx.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: sc7180: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
arm64: dts: qcom: msm8998: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/msm8998-asus-novago-tp370ql.dtb: hwlock: 'reg' is a required property
qcom/msm8998-asus-novago-tp370ql.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: msm8998: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
arm64: dts: qcom: msm8996: switch TCSR mutex to MMIO
The TCSR mutex bindings allow device to be described only with address
space (so it uses MMIO, not syscon regmap). This seems reasonable as
TCSR mutex is actually a dedicated IO address space and it also fixes DT
schema checks:
qcom/msm8996-xiaomi-natrium.dtb: hwlock: 'reg' is a required property
qcom/msm8996-xiaomi-natrium.dtb: hwlock: 'syscon' does not match any of the regexes: 'pinctrl-[0-9]+'
arm64: dts: qcom: msm8996: split TCSR halt regs out of mutex
The TCSR halt regs are next to TCSR mutex (in one address block called
TCSR_MUTEX), so before converting the TCSR mutex into device with
address space, we need to split the halt regs to its own syscon device.
arm64: dts: qcom: msm8916-samsung-a2015: Rename touchscreen analog regulator
reg_vdd_tsp: regulator-vdd-tsp is actually used as an analog regulator for
touchscreen on all of a2015 and e2015 devices.
Rename it into reg_vdd_tsp_a: regulator-vdd-tsp-a to reduce confusion.
On the Samsung Galaxy E5 and E7 the touch key is supplied by a single
fixed regulator (enabled via GPIO 97) that supplies both MCU and LED.
Add it to the device tree.
ARM: dts: qcom-msm8916-samsung-e2015: Include dts from arm64
After adding all necessary support for MSM8916 SMP/cpuidle without PSCI
on ARM32, build the Samsung Galaxy E5/E7/Grand Max device trees from the
arm64 tree together with the ARM32 include to allow booting this device on
ARM32.
The approach to include device tree files from other architectures is
inspired from e.g. the Raspberry Pi (bcm2711-rpi-4-b.dts) where this is
used to build the device tree for both ARM32 and ARM64.
arm64: dts: qcom: msm8916-samsung-e2015: Add initial common dtsi
Samsung Galaxy E5, E7 and Grand Max are smartphones using the MSM8916 SoC
released in 2015.
e2015 and a2015 are similar, with some differences in accelerometer,
MUIC and Vibrator. The common parts are shared in
msm8916-samsung-a2015-common.dtsi to reduce duplication.
Add a common device tree for with initial support for:
- GPIO keys
- GPIO LEDs for Grand Max
- Regulator haptic
- Hall sensor (except Grand Max)
- SDHCI (internal and external storage)
- USB Device Mode
- UART (on USB connector via the SM5504 MUIC)
- WCNSS (WiFi/BT)
- Regulators
- S3FWRN5 NFC (except Grand Max)
The three devices (and all other variants of E5/E7/Grand Max released in
2015) are very similar, with some differences in display, touchscreen,
sensors and NFC. The common parts are shared in
msm8916-samsung-e2015-common.dtsi to reduce duplication.
Unfortunately, some E5/E7/Grand Max were released with outdated 32-bit
only firmware and never received any update from Samsung. Since the 32-bit
TrustZone firmware is signed there seems to be no way currently to
actually boot this device tree on arm64 Linux on those variants at the
moment.
However, it is possible to use this device tree by compiling an ARM32
kernel instead. The device tree can be easily built on ARM32 with
an #include and it works really well there. To avoid confusion for others
it is still better to add this device tree on arm64. Otherwise it's easy
to forget to update this one when making some changes that affect all
MSM8916 devices.
Maybe someone finds a way to boot ARM64 Linux on those device at some
point. In this case I expect that this device tree can be simply used
as-is.
dt-bindings: arm: qcom: Document additional sku6 for sc7180 pazquel
The difference between sku6 and sku4 is that there is no esim
The different SKUs are:
LTE with physical SIM _and_ eSIM
LTE with only a physical SIM
WiFi only
Both sku4 and sku6 are LTE SKUs.
One has the eSIM stuffed and one doesn't.
There is a single shared device tree for the two.
Konrad Dybcio [Thu, 14 Jul 2022 12:34:06 +0000 (14:34 +0200)]
arm64: dts: qcom: Add device tree for Sony Xperia 1 IV
Add support for Sony Xperia 1 IV, a.k.a PDX223. This device is a part
of the SoMC SM8450 Nagara platform and currently it is the only
device based on that board, so no -common DTSI is created until (if?)
other Nagara devices appear.
This commit brings support for:
* SD Card
* USB (*including SuperSpeed*)
* ADSP/CDSP/SLPI (modem remains untested for now)
* Most regulators (some GPIO-enabled ones require PMIC GPIOs but
trying to access any SPMI device crashes the device..)
* Part of I2C-connected peripherals (notably no touch due to a
driver bug)
* PCIe0 (PCIe1 is unused)
Do note display via simplefb is not supported, as the display is blanked
upon exiting XBL.
To create a working boot image, you need to run:
cat arch/arm64/boot/Image.gz arch/arm64/boot/dts/qcom/sm8450-sony-xperia-\
nagara-pdx223.dtb > .Image.gz-dtb
Then, you need to flash it on the device and get rid of all the
vendor_boot/dtbo mess:
// You have to either pull vbmeta{"","_system"} from
// /dev/block/bootdevice/by-name/ or build one as a part of AOSP build process
fastboot --disable-verity --disable-verification flash vbmeta vbmeta.img
fastboot --disable-verity --disable-verification flash vbmeta_system \
vbmeta_system.img
Where emptydtbo.img is a tiny file that consists of 2 bytes (all zeroes), doing
a "fastboot erase" won't cut it, the bootloader will go crazy and things will
fall apart when it tries to overlay random bytes from an empty partition onto a
perfectly good appended DTB.
Konrad Dybcio [Thu, 14 Jul 2022 12:34:04 +0000 (14:34 +0200)]
arm64: dts: qcom: sm8450: Adjust memory map
Fix up the camera region (cross-referenced different vendors' msm-5.10
drops, 9f500000 is the default location for SM8450) and reserve the
second chunk occupied by xbl_sc.
projects / thirdparty / kernel / linux.git / log
