Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
For Android use ICS variant of Fusion 7 and 10 multi-touch driver (See
Fusion 7 and 10 drivers for Linux.pdf and Linux Drivers Fusion 10.zip).
Tested to be working both with Android 4.0.4 for Colibri T20 as well as
Android 4.2.1 for Colibri T30.
|
|
Enable the back key as a wake-up source as well.
|
|
Use 32 rather than 16 bits per pixel as otherwise it looks so ugly.
|
|
This involved adding some callback methods, and pretending that this is a
YUV sensor.
Signed-off-by: Andrew Chew <achew@nvidia.com>
|
|
Add initial Android kernel configuration adopted from
tegra_android_defconfig.
|
|
Due to 720p not working well with most regular analogue screens and as
customers rather might want to evaluate using our 7 inch EDT
ET070080DH6 panel revert to 800x480 if not default VGA TFTLCD
resolution is chosen.
This reverts commit 0fc41ea361c79175f2076a349a0ba0dcc7e71131.
|
|
As customers might want to evaluate using a regular VGA screen or our
5.7 inch EDT ET057090DHU panel revert to default VGA TFTLCD resolution.
This is a partial revert of the following commit:
106ac0436290a98ecc5e3dfc75d0d7f104ef0ff4
|
|
The Android idc parser seems to have an issue with spaces in names:
D/EventHub( 159): No input device configuration file found for device
'wm97xx touchscreen'.
Rename the driver in order for the idc parser to recognise idc file
/system/usr/idc/wm97xx-ts.idc being a simple copy of the qwerty.idc
one.
|
|
Rotate touch for now due to missing resistive touch calibration
integration. This makes our 5.7 inch EDT ET057090DHU panel usable.
|
|
Due to a strong pull-down on the menu-key multiplexed ACC1_DETECT pin
Android unintentionally entered safe-mode. Fix this by making the
menu-key active-high as well.
|
|
Hack for Realtek HDA driver not to crash due to PCI structure not being
initialised as expected.
|
|
By default configure for LG LP156WF1 15.6 inch full HD dual channel
LVDS panel.
|
|
Integrate support for the Realtek ALC898 as found on the Apalis
Evaluation Board V1.0a.
Please note that the following HDA rewiring needs to be done:
remove jumpers X6-5 and 7 to 9
rewire as follows:
X5 X7
5 9
7 8
8 5
9 7
Please further note that currently only noise is heard during playback.
|
|
SODIMM-59, SODIMM-67 is no longer a GPIO and thus the PWM are usable again.
PWM<D>
|
|
Fix null pointer during USB OTG host to peripheral transition caused by
freeing the platform data twice. See the following Tegra OTG driver
commit for reference:
8acf989d5896b152cc88f8b29afd11d522f903e3
|
|
Integrate CAN support for the Colibri Evaluation Board V3.1a with built-in
MCP2515 SPI CAN controller resp. the Colibri Evaluation Board V2.1c with its
built-in SJA1000 CAN controller connected to the GMI bus.
The following kernel configuration needs to be enabled as well:
CONFIG_CAN
CONFIG_CAN_RAW
CONFIG_CAN_BCM
CONFIG_CAN_DEV
Plus depending on the Evaluation Board revision:
CONFIG_CAN_MCP251X
or
CONFIG_CAN_SJA1000
CONFIG_CAN_SJA1000_PLATFORM
Optional support for 32-bit GMI as well as xPOD CAN on MECS Tellurium for
Colibri T20 is provided via commented defines in
arch/arm/mach-tegra/board-colibri_t20.h.
|
|
Add initial Android kernel configuration adopted from
tegra3_android_defconfig.
|
|
Unless HDA driver is enabled device enters safe mode upon start-up.
|
|
NVIDIA's Android for their Cardhu developer tablet is looking for the
machine name in order to initialise certain things. For now just hack
it to be able to use their regular initrd.
|
|
Android uses 32 bits per pixel and fails otherwise.
While at it increase the default DVI-D aka HDMI resolution to full HD.
|
|
Integrate MAX9526 video decoder support as found on our analogous
camera module (ACM) connected to a Colibri Evaluation carrier board.
The following kernel configuration options are required to be enabled:
CONFIG_VIDEO_V4L2=y
CONFIG_SOC_CAMERA=y
CONFIG_SOC_CAMERA_MAX9526=y
CONFIG_VIDEO_TEGRA=y
|
|
|
|
The VI client within drivers/video/tegra is mutually exclusive with the
Tegra V4L2 framework, since they both want to own the camera hardware.
Signed-off-by: Andrew Chew <achew@nvidia.com>
|
|
set some platform_data.
Signed-off-by: Andrew Chew <achew@nvidia.com>
|
|
arch/arm/mach-tegra/common.c
drivers/ata/ahci-tegra.c
|
|
Move GPIO defines to header file.
While at it get rid of spurious AVDD_DSI_CSI_ENB_GPIO define.
|
|
Re-arrange defines in header file.
|
|
Get rid of spurious AC_PRESENT_GPIO/INT defines.
|
|
Explicitly configure MMC/SD card as 4-bit only by default.
|
|
Rename UARTs to be consistent with Colibri FF/BT/STDUART naming.
|
|
Rename PWM LEDs to be consistent with Colibri PWM<X> naming.
|
|
Migrate from static TEGRA_GPIO_TO_IRQ macro to run-time gpio_to_irq
function.
While at it clean-up some comments and fix some long lines in order to
adhere to Linux kernel coding style.
|
|
Clean-up clocks. HDA clocks are of course required on Apalis T30.
|
|
Fix naming of fixed regulator initialisation function.
|
|
Fix naming of CAN initialisation function.
|
|
Move Apalis T30 specific entries for alphabetical consistency. Put
Colibri T30 items in alphabetical order.
|
|
Enable back key wake-up source.
|
|
Due to missing pull-ups on the gpio-key pins Android unintentionally
entered safe-mode.
|
|
Rotate touch for now due to missing resistive touch calibration
integration. This makes our 7 inch EDT ET070080DH6 panel usable.
|
|
Activate 24 bpp RGB output.
Please not that this is just the backend configuration. For true 24 bpp
output the framebuffer as well as driver have to be configured for 24
resp. 32 bpp as well.
|
|
The audio_port_id in the audio platform data really has to be the index
of the registered tegra_i2s_device rather than the I2S interface
instance.
|
|
Enable building b43 driver as a module to demo Broadcom AirForce
BCM4311 based mini-PCIe Wi-Fi 802.11a/b/g module.
|
|
Reset PLX PEX 8605 PCIe Switch on Apalis Evaluation Board to enable
X43, X44 and X45.
|
|
Add defines for the generic Apalis GPIOs 1 to 8 and the special use
thereof on the Apalis Evaluation Board:
FAN_EN on GPIO8: enable FAN supply on X35
PEX_PERST_N on GPIO7: reset PCIe switch for X43, X44 and X45
TOUCH_WIPER on GPIO6: optional touch wiper GPIO
|
|
Fix the following build error using our OpenEmbedded gcc-4.7:
arch/arm/mach-tegra/pcie.c:335:22: error: 'reg_pmc_base' defined but
not used [-Werror=unused-variable]
cc1: all warnings being treated as errors
|
|
Add new PCIe driver fixing multi-port operation. Pulled from
Linux_Kernel-src_CARMA_DevKit-ver3.1.10_l4t_Rev545.tgz.
|
|
This patch integrates multi-touch for the TouchRevolution Fusion 7 and
10 panels connected to Iris as follows:
Iris X16-5, SODIMM pin 194, I2C SDA
Iris X16-6, SODIMM pin 196, I2C SCL
Iris X16-15, SODIMM pin 103, pen down interrupt
Iris X16-16, SODIMM pin 101, reset
The following kernel configuration needs to be enabled:
Device Drivers -> Input device support -> Touchscreens
-> TouchRevolution Fusion F0710A Touchscreens
Please note that level shifters are required.
Please further note that our current OpenEmbedded BSP does not yet
integrate multi-touch support in the XInput as well as X server layers.
|