Age | Commit message (Collapse) | Author |
|
In the case of not using edma set the RXIDEN field. This sets the RDRF
flag even if there are received characters in the FIFO and the receiver
has been idle for the specified number of characters.
Related-to: ELB-3517
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
|
|
For low baudrates the uart baudrate divider can overflow, i.e. the
sbr register must stay within [1..8191].
With an input frequency of 80MHz and a baudrate of 1200 the following
register settings are calculated.
sbr: 16666, osr: 4
Fix this.
Related-to: ELB-3517
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
|
|
As the Fusion touchscreen is not belong to the specific board but a part
of peripheral, move it to the module level. Add the node name to use it
in an overlay.
Related-to: ELB-3398
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
This device is enabled in an overlay. Remains it disabled here.
Realetd-to: ELB-3395
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The parallel RGB interface (lcd) and all related stuff turn
on in an overlay. Keep them disabled in the main devicetree.
As these subsystems are provided by module, move their definitions
into the module-level devicetree.
Related-to: ELB-3395
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
Colibri Aster board supports two spidev devices. Add 2nd one.
Related-to: ELB-2532
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
All Toradex Colibri family boards support the ecspi interface.
Move the definition of spidev0 to the module devicetree and enable
it by default.
Colibri Eval board has a CAN SPI controller. Disable the spidev0
for this board to have the CAN controller been enabled by default.
Related-to: ELB-3401
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
All Toradex Colibri family boards support the ecspi interface.
Move the definition of spidev0 to the module devicetree and enable
it by default.
Colibri Eval board has a CAN SPI controller. Disable the spidev0
for this board to have the CAN controller been enabled by default.
Related-to: ELB-3401
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
All Toradex Colibri family boards support the ecspi interface.
Move the definition of spidev0 to the module devicetree and enable
it by default.
Colibri Eval board has a CAN SPI controller. Disable the spidev0
for this board to have the CAN controller been enabled by default.
Related-to: ELB-3401
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
All Toradex Colibri family boards support the ecspi interface.
Move the definition of spidev0 to the module devicetree and enable
it by default.
Colibri Eval board has a CAN SPI controller. Disable the spidev0
for this board to have the CAN controller been enabled by default.
Colibri Aster board supports 2 spidev interfaces. Enable both of them
and fix the CS configuration.
Related-to: ELB-3401
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The SODIMM_65 pin, which should be used as a 2nd CS signal, is
unavailable for customer on a Colibri Iris board.
Remove the 2nd CS signal configuration for all Iris board versions.
Related-to: ELB-3401
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The shutdown() callback of i2c client device calls for both scenarios,
power off and reboot the system.
In case of rebooting, the PMIC is configured to power off the power rails
when receiveing the PMIC_STBY_REQ request.
It leads to powering the module down and system hanging while suspending
the system after soft reboot.
Remove the shutdown() callback. Moreover, it is not needed now as the
power-down preparation of PMIC is done in the pfuze_power_off_prepare()
function.
Related-to: ELB-3473
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
There is only one spidev port is used on the Evaluation board.
Use the correct port address for the 1st device.
Related-to: ELB-3401
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
Add support of sub-functions disabling. It allows one to define
an stmpe sub-function device in devicetree, but keep it disabled.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
This solution is replaced with better upstream one.
This reverts commit 4daa584ef919a58b90e1957012ffb6eaad5dc6bd.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
Request and initialize the rest of GPIOx pins.
Related-to: ELB-3427
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
Detailing names for pins that serve GPIOx signals.
Related-to: ELB-3427
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The resume from a suspend state on SoC i.MX6Q leads to a
kernel hanging.
Adding a 10ms delay after clearing the bit TEST_POWERDOWN
of IOMUXC_GPR1 regiater fixes this issue.
There is not enough information about the testing "powerdown"
mode on i.MX6Q, but we can do some assumption.
According to "49.6.1.2 Power-On Reset" section of the
"i.MX6D/6Q APRM", the time between bringing up the PHY power
supplies and the reference clock is up and stable is
(< 10ms) + (> 10mks) ~= 10ms. So using this value seems reasonable.
Related-to: ELB-3012
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
In case of any error occurs after enabling the PM runtime framework
for the fsl-spdif-dai device, the next probing generates the
warning:
"fsl-spdif-dai 2004000.spdif: Unbalanced pm_runtime_enable!"
Add the error handling path to keep the PM runtime usage counter balanced.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
Add drivers for e-con Systems AR0521 and AR1335 camera modules.
Related to: PMS-387
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
|
|
Mainly clocks need adjustment. Copy Paste from apalis-imx8 module
Related-to: ELB-3423
Signed-off-by: Philippe Schenker <philippe.schenker@toradex.com>
|
|
Related-to: ELB-3428, ELB-2790
Signed-off-by: Philippe Schenker <philippe.schenker@toradex.com>
|
|
They are used on MEK only and are blocking out GPIOs so disable them
Related-to: ELB-3428, ELB-2790
Signed-off-by: Philippe Schenker <philippe.schenker@toradex.com>
|
|
Reduce the length of pinmux name and mux the pin by default.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The USB_SS3_TC2 pin can be directly used as a USBO1_OC pin to
control over-current events. Use this mode.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The property "reset-on-resume" is useless without setting
"reset-gpio". It used as a workaround for earlier revisions
of Apalis iMX8QXP to fix Wi-Fi chip issue on resume.
Remove obsolete property.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
The cadence usb3 driver tries to turn down this regulator each time it
goes to idle. This signal enables on-board USB Hub and should be
always on.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
This reverts commit ce1e685adbd15203ca84c0be62cf752304e39df3.
That commit seems for Apalis iMX8QM and should have not been applied
for Apalis iMX8QXP.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
This keeps the numbering of the /sys/class/pwmchipX independent of
which overlays are loaded and allows the use of Verdin PWM_3_DSI from
user space if it is not used as a backlight PWM.
Related-to: ELB-3411
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
|
|
lpuart2 is the only UART where DMA request works and linux tries to use
a DMA. However this DMA does not work and output hangs, disbling the DMA
works for now.
This change was originally introduced in commit
1a8badaa585de ARM64: dts: apalis-imx8: disable dma for lpuart2
but lost when doing the devicetree-split.
Fixes: 2715972aadc8f arm64: dts: apalis-imx8: separate common module pa
Related-to: ELB-3178
Signed-off-by: Philippe Schenker <philippe.schenker@toradex.com>
|
|
Work around potential probe deferral, order and race condition issues.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
|
|
Both Verdin Developer Board V1.0A and V1.1A have their PHYs at address
7. Only V1.0B had it at the address 3!
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
|
|
Add SODIMM annotation to FEC pins.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
|
|
The following happens e.g. on the i.MX 8M Plus EVK which does not
specify any phy-supply:
[ 1.077851] imx-dwmac 30bf0000.ethernet: no regulator found
[ 1.083445] Unable to handle kernel NULL pointer dereference at
virtual address 0000000000000078
[ 1.092234] Mem abort info:
[ 1.095033] ESR = 0x96000004
[ 1.098087] EC = 0x25: DABT (current EL), IL = 32 bits
[ 1.103402] SET = 0, FnV = 0
[ 1.106456] EA = 0, S1PTW = 0
[ 1.109603] Data abort info:
[ 1.112489] ISV = 0, ISS = 0x00000004
[ 1.116329] CM = 0, WnR = 0
[ 1.119301] [0000000000000078] user address but active_mm is swapper
[ 1.125660] Internal error: Oops: 96000004 [#1] PREEMPT SMP
[ 1.131232] Modules linked in:
[ 1.134291] CPU: 2 PID: 1 Comm: swapper/0 Not tainted
5.4.77-05639-g581745444980 #10
[ 1.142032] Hardware name: NXP i.MX8MPlus EVK board (DT)
[ 1.147344] pstate: 60000005 (nZCv daif -PAN -UAO)
[ 1.152141] pc : regulator_enable+0x1c/0x78
[ 1.156326] lr : imx_dwmac_init+0x134/0x1a8
[ 1.160508] sp : ffff80001003bad0
Fixes: ca8811172bdc ("net: stmmac: dwmac-imx: add phy-supply")
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
|
|
Rework the read and write code paths in the driver to support operation
in atomic contexts. To achieve this, the driver must not rely on IRQs
and not call schedule(), e.g. via a sleep routine, in these cases.
With this patch the driver supports normal operation, DMA transfers and
now the polling mode or also called sleep-free or IRQ-less operation. It
makes the code not simpler or easier to read, but atomic I2C transfers
are needed on some hardware configurations, e.g. to trigger reboots on
an external PMIC chip.
Related-to: ELB-3308
Related-to: ELB-1403
Signed-off-by: Stefan Lengfeld <contact@stefanchrist.eu>
[m.felsch@pengutronix.de: integrate https://patchwork.ozlabs.org/patch/1085943/ review feedback]
[m.felsch@pengutronix.de: adapt commit message]
Signed-off-by: Marco Felsch <m.felsch@pengutronix.de>
Acked-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Stefan Agner <stefan@agner.ch>
Tested-by: Stefan Lengfeld <contact@stefanchrist.eu>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
(cherry picked from commit 3a5ee18d2a32bda6b9a1260136f6805848e3839d)
|
|
tempmon driver requires the platform driver imx-cpufreq-dt that is enabled
for imx7 with the option ARM_IMX_CPUFREQ_DT. Adding this option fixes
tempmon driver working.
Related-to: ELB-3339
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|
|
Prepare the nodes for the Verdin DSI to LVDS Adapter overlay.
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
|
|
Prepare the nodes for the DSI to HDMI Adapter overlay.
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
|
|
The OV5640 driver crashes on NXP's i.MX 8M Plus if loaded as a module.
Related to: ELB-3209
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
|
|
From the SD-Card specification:
> The host shall supply power to the card so that the voltage is
> reached to V DD _min within 250ms and start to supply at least
> 74 SD clocks to the SD card with keeping CMD line to high. In case of
> SPI mode, CS shall be held to high during 74 clock cycles.
According to this information I think we should pull the CMD data
line like on other boards to 3.3V. Also on other boards we
do have the CLK signal pulled.
This looks like this got mixed that up.
Signed-off-by: Philippe Schenker <philippe.schenker@toradex.com>
|
|
The pcie_phy uses ext_osc info in addition to the pcie node.
While at it add the bus-range property to prevent:
No bus range found for /pcie@33800000, using [bus 00-ff]
Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
|
|
Correct the clock mode of PCIe PHY.
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit 2c23d9de9fc41d5436a7d3f8510fc95f3bf1c56b)
|
|
- Verify the both internal PLL_SYS and external OSC reference clock
modes on iMX8MP EVK board, and pass the PCIe compliance tests.
- Remove the no-needed bypass setting.
- PHY configration should be completed before CMN_RSTN is set to 1b1
- To manually initiate the speed change to make sure GEN2 is linked up:
- Write to LINK_CONTROL2_LINK_STATUS2_REG.PCIE_CAP_TARGET_LINK_SPEED
in the local device
- De-assert GEN2_CTRL_OFF.DIRECT_SPEED_CHANGE in the local device
- Assert GEN2_CTRL_OFF.DIRECT_SPEED_CHANGE in the local device
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit 12589ba98b1145e2d2a9f439b20064197aea8ca9)
|
|
Add the PERST for iMX8MP PCIe.
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit 0a91df355e20aa53a4bcaa9d352952e5fa9e4ef6)
|
|
Verify the PCIe PLL_SYS reference clock source on EVK board.
The external OSC clock is used as PCIe REF clock source in default.
- sequence should be the following one.
phy configuration--> CMN_RSTN--> wait for pll lock
- add the calibrate callback to fit the correct init sequence of phy
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit 9afaf7a465858970aee8858fb52067f2ef152c7f)
|
|
Update the clock modes of iMX8MP PCIe PHY in binding DOC.
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit eade3750bf840948b29db26f0beabf10dea5f981)
|
|
Verify the PCIe PLL_SYS reference clock source on EVK board.
The external OSC clock is used as PCIe REF clock source in default.
NOTE: Change the ext_osc of pcie/pcie_phy to '0' when enable SYS_PLL
clock mode.
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit 1bda33273eccae3c0d878d34660eca9da1765db0)
|
|
Configure the L1 latency of iMX8M's RC to less than 64us, otherwise,
the L1/L1SS wouldn't be enabled by ASPM.
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Fugang Duan <fugang.duan@nxp.com>
(cherry picked from commit eb9fdb1d10c10dc4d4bbe9bc6fc060afe1485989)
|
|
Signed-off-by: Jaskaranjeet Singh <jaskaranjeet.singh@toradex.com>
|
|
Add support of sub-functions disabling. It allows one to
define and turn on sub-function devices in different
devicetree files.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
|