I've tested my OPi3 for best voltage numbers (for resolving stability problem after first patch), using StabilityTester and increasing slowly mV by 10 after every crash. Finaly, this values gives no error by running StabilityTester 100 times.
- change console welcome text
- changed display manager to lightdm by default and remove nodm completly
- desktop login manager enabled by default, toogle can be done with armbian-config
- added URL to the build script and commit hash to /etc/armbian-release file
- added debconf-utils,html2text,mmc-utils,sunxi-tools to the base and additional pkg list
- added synaptic package manager and onboard keyboard to the desktop base
- changed motd console welcome text to: Welcome to Debian Stretch with Armbian Linux 5.1.6-sunxi
- added "logout" to the panel/menu
- added normal users to additional groups: disk tty users games
* Adjust tabs and alter displaying build options at the end - hide empty ones
* Fix wrong MOTD message, adjust logic to autologin by default and request login for both notebooks
* Making fs cache naming more intuitive, rename script
* Move additional kernel drivers into separate file. Add AUFS for K5.1 this way, attach wireguard sources to last known working tag. master is broken atm
* Add two more extra wireless drivers 88x2BU, (improved) 8811EU and introduce switch EXTRAWIFI which is enabled by default
* Move 8812AU to the same version we have now. More recent has some performance issues in AP mode. Adjust configuration, add mvebut-dev 5.1.y kernel
* Enable Wireguard also on 3.10.y since it was fixed upstream. Attach bach to master branch since its working now
* Rockchip64 & RK3399 - upstream patches and remove deprecated patches, mvebu64 add DEV kernel attached to 5.1.y
* Enable BT on Tinkerboard CLI images as well
* cubox-i move to 5.1.y and enable missing HDMI
* Bugfixes for adding wireless driver to Kconfig. Adjusted RK3399 kernel due to upstream changes at FriendlyARM - does not boot yet. Changes has to be extract from their commits. Removed patches with RTL8188EU since we add more recent driver with a new precedure. Small fixes mvebu-default, udoo-next, s56818-next, ... remove deprecated patches.
* Boot-able RK3399-default config, add upstream patches for K4.4.y mvebu-default/udoo-next
* Adjust overclocking patch to meet upstream changes
* Odroid XU4 - remove deprecated kernel patches
new values are based on stabilityTester results
when using 930mV for 1.6GHz and 980mV for 1.8GHz, xhpl failed randomly.
So 930/980mV are the critical operation voltage.Then increase 10mV and
run 100 times test, result shows that xhpl (not always) failed 1~2 times.
It seems +10mV is still not enough to ensure dc-dc output voltage always
above operation voltage, that means the design of dc-dc converter is not
good enough that results in large ripple.
The testing script is "github.com/mzhboy/StabilityTester"
The following configuration have been tested.
1.08-1.32-1.48-1.64-1.8 GHz
900-910-920-930-990 *ok, 930 failed 2/100
900-900-910-940-990 *ok, 990 failed 1/100
890-900-910-940-990 *ok, 990 failed 1/100
880-880-910-940-1000 *ok, 940 failed 2/100
880-880-910-950-1000 *ok, 1000 failed 1/100
880-880-910-950-1010 *ok
use more powerfull 'cpuburn-a53' to do pressure test.
power meter shows that 'cpuburn-a53' consumes much more power than
'stress -c 4' in the same condition. 930mV is not enough when running
'cpuburn-a53' meanwhile 950mV is the minimum voltage.
When running 'cpuburn-a53' at 1.8G, h6 consumes lots of power and generate
such amount of heat that an active cooling is necessary otherwise h6
gets overheat and system crashes. Even with active cooling and heatsink
is in low temperature the junction temperature is still quit high that
it reaches about 88~90 ℃ . Maybe the high thermal resistance of plastic
package causes the problem.
* [sunxi-dev][h6]update orangepi lite2 dts file patch
* [sunxi-dev][h6]update patch for sun50i-h6.dtsi
* [sunxi-dev][h6] new operation voltage table of allwinner h6
The correct operation voltages are find out after many reboots and test (over 20+).
the original 1.16v@1.64/1.8GHz was too high since standard cell voltage
of 28nm@HPC is 0.9v, h6 got overheat very quickly over 110 Celsius and then system hangs,
a power-cycle off is needed to reset soc.
I use 'stress -c 4' to do pressure test, with a medium size passive heatsink on top of h6.
The pressure test proves that the soc has about 4 watt power consuming
that is much lesser than original 6~7.x watt when running at 1.8GHz(orangepi lite2 with
usb Ethernet rtl8153 and usb hub, wifi is connected but not used).
Soc runs much more stable and does not get overheat easily under heavy load
with new opp table.
* [sunxi-dev][h6] orangepi lite2 wifi fix
bcm43455-fmac is buggy, reset to bcm4329-fmac
Commit 683d9384e6 brought back the upstream A64 timer fixes from 5.1.y into to the 4.19.y Armbian NEXT tree but missed the activation in the device node.
This change is the same as commit efff71ba4f for the DEV tree and will re-introduce the activation for the fix.
Add patch to bring A64 pinctrl driver inline with other sunxi pintctrl drivers by disabling strict mode.
With kernel 4.15.y (and later):
* Strict mode was enabled by default via commit 1396007286b1e2fd5dd10ae6a5ccaaaed51ab762 which can/will cause breakage with existing implementations.
* The ability to configure strict mode was added via commit aae842a3ff3385f27f1df8a9ee1494a416ec032d to allow older drivers to maintain existing behavior and avoid breakage.
* Commit cd70387f892205bcd7b8093b0837269b0739cbe0 had then explicitly disabled strict mode for most other existing SoCs but did not include A64.
This change is to update the A64 pinctrl driver similar to the other pre-existing SoC pinctrl drivers.
See 63f8f3badf799c8b63ff33a489886bc138ce5d09 drm: bridge: Constify mode arguments to bridge .mode_set() operation
and fcd70cd36b9bf697122538c9e38e8cf954b2342b drm: Split out drm_probe_helper.h
arch/arm64/boot/dts/allwinner/overlay/sun50i-a64-spi-jedec-nor.dts:12.13-17.6:
Warning (spi_bus_reg): /fragment@0/__overlay__/spiflash: SPI bus unit
address format error, expected "0"
arch/arm64/boot/dts/allwinner/overlay/sun50i-a64-spi-jedec-nor.dts:26.13-31.6:
Warning (spi_bus_reg): /fragment@1/__overlay__/spiflash: SPI bus unit
address format error, expected "0"
Updated DT patch for allwinner unstable timer
Replace previous patch for DT timer node with new patch from upstream commit.
This matches the upstream change to the actual workaround in drivers/clocksource/arm_arch_timer.c
via commit c950ca8c35eeb32224a63adc47e12f9e226da241
seems like this patch headed upstream for rk3328-roc-cc:
https://patchwork.kernel.org/patch/10804863/
doesn't include the tx/rx delay changes in the original rock64 patch:
https://patchwork.kernel.org/patch/10178969/
My testing showed these changes help with a link down and reset that happens on heavy
exchanges, like streaming. Maybe we should sumbit upstream a patch that hadles this as well?
Interrupt pin of the Ethernet PHY is connected to GPIO. The PHY can be
configured to raise interrupt when it received WoL magic packet.
This patch use gpio-keys as wakeup source for Wake on LAN as currently
there is no way to use the GPIO directly under mvneta driver.
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
The SoC provides several waking options sourced by different peripherals
to take the system out of power save modes. One of the options is Wake
on GPIO.
This patch implements missing function to support gpio as wakeup source
and properly route it to upper interrupt controller (Arm GIC).
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
previous version of the patch has slightly higher performance impact
compared to disk activity trigger.
This updated patch should have similar impact as disk activity trigger.
- update to latest patchset from Baylibre
- Some bugs addressed
- Open items:
- HDMI does not want to work on boot
- If a "problematic" monitor is present at boot HDMI will not respond to replug
- changing monitors after boot results in 1/2 of the display appearing, second replug and it's ok.
- Changing HDMI resolution results in corrupted screen (aliasing, static, "lines" etc.)
Note: by default Orange Pi Zero Plus2 H5 boards are missing part of their regulator
hardware circuit; as such the default regulator is limited to 1.1v.
Reworked assert-phy-reset-when-waking-up patch to fit into 4.19 kernel DWC2 USB driver
Adding usbphy reset lines patch for rk3288.dtsi in both rockchip next and dev kernels
Removed same reset lines from xt-q8l-v10 next and dev device trees
- Adds the specific shortname per ASUS to allow controlled enumeration of the onboard USB audio codec
- kernel 4.19 saw some rewriting of the naming code
* helios4: next branch use u-boot 2018
use new u-boot based on upstream 2018
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
* helios4: Tweak fancontrol configuration
Adjusted MINSTART and MINSTOP to suit both old and new fan.
Adjusted TEMP threshold so fan would stop or run in minimal speed when
the system idle.
Fixed wrong MINPWM value.
URL:https://wiki.kobol.io/pwm/#configuration-file
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
* kernel: mvebu-next: use upstream helios4 dts
Helios4 device tree has been merged on upstream since 4.19.
Rework the patch to produce same device tree.
URL:https://patchwork.kernel.org/patch/10449393/
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
* Helios4: Use boot-mvebu-next bootscript
Use boot-mvebu-next.cmd that make use of upstream's Generic Distro
Configuration.
Since loadaddr and fdt_addr varibales are no longer required, remove the
patch.
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
* Helios4: mvebu-next: Add workaround for SPI and SATA concurrent access issue
Concurrent access on SPI NOR and SATA drives can lead to unstable SATA.
Therefore as workaround, disable SATA controller when SPI flash access
is needed and make it as user configurable item in armbianEnv.txt
This workaround might applies to Clearfog too.
Refer to
Commit 59af84c07c ("Helios4: Add SPI bootloader install feature
(#1126)")
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
* Bootscripts: Fixed boot-mvebu-next.cmd
The environment setting (armbianEnv.txt) is loaded to RAM located on
${load_addr} but the env import use wrong address (${pxefile_addr_r}).
bootargs still expect ${boot_interface} which carried over from boot-
marvell.cmd. Upstream's distro configuration use ${devtype} instead.
Signed-off-by: Aditya Prayoga <aditya@kobol.io>
This merge request contains various files which add support for xt-q8l-v10 boards (TVBox) equipped with Rockchip RK3288 SoC, AP6330 WiSoC (BCM4330 WiFi + Bluetooth), 2 GB DRAM (LPDDR2 or DDR3), 8 Gb eMMC, Gigabit Ethernet, 3 USB (1 OTG), 1 microSD slot, SPDIF optical output, 1 HDMI.
Kernel patches:
This thouches all three linux-rockchip-* kernelconfigs, just adds brcmfmac and brcmutil modules and remote controller support. default flavor activates rockchip own remote controller driver, next and dev use the mainline GPIO CIR driver (dev has lirc userland support activated too).
About the remote controller, an additional kernel module is added to the existing keymaps which is activated via device tree.
About possibly clashing patches assert-phy-reset-when-waking-up-in-rk3288-platform.patch should be checked against other rk3288 boards because it addresses an errata in rk3288 which causes the USB Host ports to stop responding when exiting from autosleep. On my device if I connect the first USB device when the system is already running, the USB Host gets stuck without this patch. Probably to work correctly on other platforms the device tree should include the proper reset lines of the USB PHYs (for reference, check patch/kernel/rockchip-dev/xt-q8l-v10-add-device-tree.patch starting from line 869).
Patch 1-2-regulator-act8865-add-restart-handler-for-act8846.patch adds a restart handler which allows reboot using SIPC bit on act8846 power regulator. Possibly MiQi board is affected (is reboot working there?), others (tinkerboard) should not care.
Patch brcmfmac-add-ap6330-firmware.patch adds firmware file names for ap6330 , should be harmless in other cases.
Patch 0010-GPU-Mali-Midgard-remove-rcu_read_lock-references.patch is from Miouyouyou. It should be harmless, it was suggested by him to do some tests with devfreq
Other patches just add the proper device trees, Kconfig and bits for supporting the board as a regular kernel supported board and should not interfere with anything else
U-Boot patches:
All the patches for u-boot are per-board, so nothing is added which may interfere with other existing boards here. They include the device tree and u-boot config and also a couple of patches to support the silergy power regulators driving current to CPU and GPU
* Initial commit to provide kernel and u-boot configuration and device trees for xt-q8-v10 as patches
Modification to rockchip config to add initialization bits for xt-q8-v10
* Committing correct path for rk3288_ddr_400Mhz... rockchip blob, moved assembling into another section to produce
immediately an u-boot working binary
* Enabled broadcom fmac driver in rockchip-next config
* Changed name definition of rk3288-xt-q8-v10 board to "TVBox"
Added bits to include support AP6330 and binary firmwares into the final image
* Fixed device tree file name in related patch, added patching of Makefile to produce the device tree binary accordingly
* Fixed xt-q8-v10 device tree patch
Added brcmfmac driver to rockchip dev and default kernel configs
* Syncing with upstream
* Splitted add-xt-q8... kernel patches into two separate patches
* Fixed bad extension while adding dtb in makefile for rockchip-default configuration
Updated device tree patches for all rockchip confs
* Enable mmc0 and usb in u-boot config
Fixed again makefile patch for kernel next
* Adding patches to reset the USB phy when kernel requires a reset, fixes autosuspend issue
* Changed xt-q8-v10 to proper xt-q8l-v10 in every string and every filename
Added power hold to u-boot, so now the device will boot and stay turned on without the need for the OTG cable anymore
* Changed names from 'Q8' to proper 'XT-Q8L-V10' in device tree patch files
* Legacy kernel device tree:
Fixed bluetooth gpio pin clashing
Fixed HDMI gpio pin clashing
Added support for PWM-based IR-Receiver, added driver in kernel default config too
Various other fixes to avoid some complaints from the kernel
* Added booting bluetooth systemd service for AP6330 (xt-q8l-v10) that loads patchram and invokes hciattach
Minor fixes to -next and -dev device trees for xt-q8l-v10
* Disabled OTG USB port in u-boot due to long timeout during initialization
Fixed warning during u-boot dts compilation
Added emmc as second boot device in dts
* Adding myself to licensing
* Committing modifications to device trees
* Fixed dmac_bus_s explicitly set to unused dmac, restored right dmac in xt-q8l-v10 dts only
Change PLL_CPLL frequency in device tree to 408 Mhz to avoid fractional divisor warnings
* Added proper xt-q8l-v10_rk3288 configuration to u-boot, now appearing in config menu and
correctly selectable as a real target
Fixed typo in device tree from rockchip
* Fixed missing semicolon in device tree for default configuration
Fixed patch files for u-boot appending themselves to files on each compilation
* Added bits to enable power to USB ports in u-boot, thus enabling booting from USB devices (only USB host port for now)
* Changed u-boot binary creation using the rockchip SPL properly
* Added boot order for xt-q8l-v10: sdcard, usb0, eMMC, network
* Added bionic:next in beta config for xt-q8l-v10 board
* Changed some minor bits in xt-q8l-v10 device tree files, added missing bits to dev flavour
Added patches to introduce fairchild fan53555/silergy82x regulators to u-boot and enabled in xt-q8l-v10 device tree
* Updated u-boot to version v2018.03 for xt-q8l-v10. Other rk3288 boards will gain v2018.05 from main armbian fork
Removed pre-reloc labels in u-boot device tree because they are not necessary since we don't use u-boot SPL for xt-q8l-v10
Removed vmmc-supply and vqmmc-supply in u-boot device tree to avoid hang on boot
* Tidied up a bit device trees, in particular some modifications are made to power regulator properties comparing them against the original q8l device tree
Removed unnecessary dummy regulator, removed unnecessary capacities to embedded eMMC
Disabled unused USB host
Removed vmmc-supply and vqmmc-supply from emmc section because it causes hang in u-boot v2018.03 and newer
* Restored previous regulator in u-boot dts
removed assert phy reset USB patch from rockchip-dev because of some upstream incompatible changes
* Added patch to enable IRQ for Midgard drivers which caused massive slowdown on dev kernel
Changed u-boot if-code for xt-q8l-v10 in rockchip.conf
Removed references to rk3288-linux.dtsi in xt-q8l-v10 device tree for default kernel
* Committing effective removal of USB reset assert for dev kernel
Committing changes to u-boot device tree
* Added patch to disable USB power down for rockchip devices broken on latest kernel
* Removed usb dwc2 patch to reinject it from specific branch
* Reverting some voltage changes for xt-q8l-v10 device in rockchip-dev
* Reverting some voltage changes for xt-q8l-v10 in u-boot section
* Added patch to make USB ports working again on rockchip devices with mainline
kernel >= 4.18
* Changed the 0 into false
* Moved xt-q8l-v10 u-boot patches into board_xt-q8l-v10 directory
* Changed some minor things in rockchip-dev dts for xt-q8l-v10, added mali midgard driver to dev kernel config
* Added devfreq support for Mali in rockchip-next flavour
* Remove manually applied patch (0007-drivers-drm...) because it has been
added to armbian main repo
* Removed duplicate patch which has added to main armbian repository
* Tidied up regulators for default/next/dev rockchip flavours for xt-q8l-v10, disabling those regulators which are not tied to anything
Enabled voltage regulator to make SPDIF connector work (thus not tested because I have no DAC)
Changed rockchip-dev and rockchip-next config files to enable gpio-ir-receiver module to enable bundled remote IR controller, including kernel patch for keymap
* Enabled back regulator REG7 to allow propert bluetooth functionaly
* Minor changes to u-boot device tree for xt-q8l-v10
Added patch to set act8846 SIPC to correctly reboot the device (thus require some power-hold at reboot to make reboot fully working)
* Fixed u-boot device tree
* Added configuration bits to support TPL in u-boot for xt-q8l-v10 (TPL is thrown away though) to allow faster reboot times and achieve a working reset feature activating power hold gpio pin as soon as possible. gpio pin is hardwired into spl_board_init() u-boot code because it is not possible to let it work via device tree
Fixed OTG USB port in u-boot, allowing devices detection and booting
Added proper vbus-supply properties for USB controllers in u-boot dts, so u-boot activates USB vbus itself
* Fixed dts makefile patching for next and dev rockchip kernel
* Fixed fdt_file renamed to fdtfile in armbianEnv.txt
* Changed xt-q8l-v10 board config as per recomendations
* Moved xt-q8l-v10 configuration to CSC
Restored linux-rockchip-* configurations, enabled brcmfmac driver, GPIO remote controller driver and lirc kernel compatibility interface
Polished a bit rockchip.conf
* Add patch to brcmfmac driver to search for ap6330 firmware
Removed copy-work from rockchip.conf about ap6330 firmware for xt-q8l-v10 and tidied up
Avoid using brcm_patchram_plus in ap6330-bluetooth-service putting proper firmware file in /etc/firmware for hciattach do firmware uploading itself
* Fixed bcm4330 bluetooth firmware linking for hciattach used by ap6330-bluetooth.service
* Removed foreign test patches from xt-q8l-v10 u-boot directory
the new cpu opp table in 4.19 now supports frequencies up to 1.368GHz, so this change
updates the 1.3GHz overclock overlay as appropriate. overlay name has been changed
to "cpu-clock-1.3GHz-1.3v" as this overlay is only needed for boards that provide a
fixed-voltage GPIO regulator/switcher that operate at 1.1v/1.3v.
Addresses were wrong, probably a mistake from the 64-bit DTS change for RK3288. This takes care of dmesg errors and reaching the I2C devices onboard.
@miouyouyou
* Initial commit BPi R2:
Currently working/not working:
- u-boot:
- builds with minor issues
- patching and cleanup happens on 'https://github.com/chwe17/u-boot-mt'
- next (https://github.com/frank-w/BPI-R2-4.14)
- boots with minor issues
- ETH doesn't work with nm
- sata works
- no wifi (needs driver from: https://github.com/frank-w/BPI-R2-4.4)
- USB not working (xhci-mtk 1a1c0000.usb: fail to get vbus) needs investigation
- appended device tree is needed due to u-boot doesn't work with fdt (kernelpacking needs adjustemts e.g. cat zImage dtb > zImage-dtb
- dev (kernel.org master)
- untested, needs adjustments in config (e.g. CONFIG_LOCALVERSION="")
- since no defconfig is available config is based on frank-w 4.16 Kernel
- Kernel builds without issues
- manual packing of kernel and dtb similar to next
The board boots, serial console is available but everything else must be expected as not working!
- to do:
- u-boot needs a cleanup
- binary blobs needs investigation
- eMMC is **not tested** and needs for sure adjustements! (don't try nand-sata-install!!!)
- bootscript isn't tested (only manual boot over u-boot console at the moment)
- both kernelconfigs aren't 'armbian standard' modules need to be adjusted
Only use it when you know're familiar with u-boot commands!
* Small fixes, add bootscript
- add bootscript for patched u-boot (needs patching of u-boot, currently under investigation and not working properly)
- revert cat zImage dtb > zImage-dtb (cause not working)
- first cleanup of dev kernelconfig (remove CONFIG_LOCALVERSION="")
* Minor fixup
-kernelconfig for next was adjustet so that USB is recognized
-firt bootscripts was written to boot with source command (doesn't work 100% reliable)
-slightly adjustd boardconfig (e.g. bootscript)
must still be considered as early wip!
* small update (FDT works now)
-working bootscript
-adjusted boardconfig
-earlyprintk activated in kernelconfig
* First attempt to get ETH working
-all interfaces are bridged together
-performance sucks at the moment (not clear if it's related to insane network configuration or I miss something obvious... :P )
* Major update (see below for changes):
- 4.14 kernel was dropped due to https://forum.armbian.com/topic/7296-bananapi-r2-csc-mt7623-as-new-boardfamily/?do=findComment&comment=55194 and following
- boot.cmd was adjusted to 'clean' mainline behaviour
- next build opition and its defaultconfg was removed (at the moment dev only)
- rework of the network default configuration
- default configuration will be over network.d **not** NetworkManager
- per default all wired interfaces are bridged together to br0 (still wip)
- old 'interfaces' configuration was removed
- boardconfig is adjusted (no desktop until I've prove that HDMI works)
- what works/ not works:
- board boots up without manual u-boot hacking
- SATA, USB3 (massstorage) is tested and works without issues
- due to rework of network configuration this is still wip and must be considered as 'not working'
* moved to network.d for configuration of wired network
- renamed bsp packages
- blacklist wired interfaces for NetworkManager
- defined all wired interfaces as br0 in systemd/network
- defined networkd als default renderer for bionic (not tested yet!)
* The houskeeping commit:
- BOARDFAMILY was renamed to mt7623 instead of mt7623n (including patchfolders etc.)
- Network.d has no fully control over wired networkes (block NM from controll, start networkd on firstrun etc.)
- Further cleanup kernelconfig
- CPU temp is visible from userspace
- missed switch driver loaded
- cryptodrivers are there (not tested)
- still a bunch of work
* Add DMA mem alloc patch
-under testing!
* Add next option (4.17.y) stick dev to master (4.18-rc1 untested)
* minor fixes:
- solve kconfig issues between 4.17 and 4.18
- stick next branch to 4.18
- first attempts to bring up gmac2 (doesn't work currently)
* switch to upstream u-boot
- fix ext4 dependency in patch series
- kernelpacking currently broken
* add bootz & cmd_ext4
* resolved merge conflict
* apply ugly u-boot patch, fix bootscript
* remove unneeded stuff, apply forgotten stuff
* update config for dev & clean up boardconf (remove xenial)
* - drop dev and next, move into default. There is no intention to provide stock kernel
- fixed board description, renamed to WIP, where it can be merged later
- packing boot firmware to the u-boot package
- docker dependencies (aufs will be added later. not essential)
- attached to 4.19.y
- loading armbianEnv.txt
- UUID support (tested)
- enabled ZRAM (tested)
- enabled eMMC install (not quite working properly yet)
- Bionic has some issues with systemd networking.
* gov to ondemand (thermals are conservative anyway), remove unneeded kernelconfigs
* Added AUFS, remove debug from kernel boot parameters, add ath10 mPCI support (which works)
* add power-off-key and rtc
* Added onboard wireless, but it's enabling is disabled by default. Too fragile.
* Change to CSC target
Adds preliminary patches for 4.19
Known Issues:
- Does not like a lot of HDMI adapters
Needs tested by anyone with a Le Potato, Odroid C2, or Nanopi K2
- update patchset to reflect @superna9999 's meta-meson patches
- notable update adds video decoder and canvas drivers
As always many thanks to BayLibre and Neil Armstrong
This change brings in the Rockchip version of the FUSB302 USB controller driver into
the kernel (with appropriate changes to enable it to build in kernel 4.19.y), and
configures the module build for the rk3399 kernel. This provides proper operation
of all <&fusb0> "extcon" references for the rk3399-nanopi4 boards, thereby enabling
all board USB host ports (e.g., now including <&u2phy0>) to now function properly.
- Fixes horrible screen performance/long boot/etc
- Thank you @Miouyouyou
Current issues:
- No Pulse Audio. Probably related to the I2S issues, needs checked
- General config review needed
- Seems to be topping out at 1.6 GHz
- only known bug so far: Somehow the entire screen has been rotated 180 degrees, making the touch screen not align with the graphics (can rotate in settings)
- probably an easy fix, but finished for tonight.
* Bump to 4.18, removing the obvious, fixing build problems, put some on waiting.
* Pin 4.18 to DEV, rollback 4.14 to NEXT, adjust configs, remove one deprecated patch from NEXT and add board-h3-address-some-stability-issues.patch
* Adjust few boards in development to new reality, removing it from NEXT for now
* Adjust few board configs
* Board config adjustement
* Adjust few boards configs
* Port NeoCore2 and Neo21.1 to 4.14.y
* Adjust board config
* Adjust board config
-- Thanks again to @ntemis for bringing these together and letting me know about it
- Tinker reboot now seemingly solved by a proper method, should no longer require OF workarounds
- various fixes from various sources, see patches for complete information.
- boots, still needs debugging of BT/touchscreen/sound
- some patch warnings exist, again WIP update for those willing to take a look. (@chwe17 perhaps?)
Based on work of David Huang https://github.com/hjc4869 Changed in the transition:
- firefly family was renamed to rk3399, which we actually never started to use. It was made for Odroid N1, which will never be sold
- adjusted compiler toolchain2 parameter
- added standard wireless drivers
- kernel config with the following changes: Docker dependencies, ZRAM, CPUfreq info, ...
- added upstream patches
- made test Ubuntu Bionic desktop and CLI Stretch build, bootlog: http://ix.io/1jVu
TBD: wifi and BT support, mainline kernel, ...
There are two versions of the NanoPi NEO2 board; this change adds the default DT for the
v1.1 board. It includes the correct LED configuration and the MP2143DJ VDD regulator
definition.
This change introduces an optional overlay that can be used to modify
the default CPU clock operating table to support a maximum core clock
of 1.4GHz. These higher clock rates will only be allowed if the
board's CPU VDD regulator can support operation at 1.4v (or greater),
e.g., SY8106A.
This change adds a board configuration for the NanoPi NEO Core2 LTS device ("nanopineocore2").
This board configuration provides support for the Core2's SY8106A I2C-based voltage regulator.
This regulator provides higher CPU voltage operation, allowing higher CPU clock rates to be
utilized.
* [Early WIP] Update sunxi-next to kernel 4.17
* Switch Allwinner 32 and 64bit to U-boot 2018.05
* Adjust patched for 4.17.y / sunxi-next
- adjust both configurations
- removing FAT support from u-boot (breaks if you try to save)
Tested those boards:
Cubietruck: wlan fails http://ix.io/1fYS USB OK, HDMI yes
Bananapi R40: http://ix.io/1fZm USB OK, HDMI yes
Lime A64: USB no, HDMI no, wireless buggy, eMMC yes
Orangepi prime H5: OK http://ix.io/1fZJ DVFS no
Orangepi2e: DVFS OK, HDMI OK, net OK, wifi OK, eMMC ok, http://ix.io/1fZT
* Kernel config update, enabling HDMI on CT+
* Trying to fix A64 HDMI but failed. Fixed M64 ethernet instead
* Update orangepioneplus.wip
* Update orangepioneplus.wip
* Fix H6 build process
* Add regulator bits for Orangepizero+, thanks to @5kft
* add H5 support for optional 1.3v regulator and 1.3GHz operation
This patch adds two optional overlays that can be used to:
1) enable the 1.1v/1.3v regulator on boards that provide the necessary compatible H/W support
2) modify the default CPU clock operating table to add new 1.2GHz and 1.3GHz clocks
Note that the generated regulator overlay will only support boards whose 1.1v/1.3v regulator
is controlled by GPIO PL6.
* updates for the NanoPi NEO Plus2
This change introduces a patch that provides two changes for the NanoPi NEO Plus2:
* Configure the "cpu0" to use the "vdd_cpux" regulator; this enables the ability to use higher CPU clocks
* Correct the configurations of the on-board power and status LEDs
* Adjust nightly building and few boards config cleanup
* implemented improved support for the FriendlyArm NanoPi Fire3
* introduced new WIP board configuration "nanopifire3"
* added and enabled the FA SPU1705 regulator
* enabled CPU-based thermal support
* implemented DT-based DVFS table (400MHz-1.4GHz)
* fixed and updated the thermal table and cooling maps
* added support for the onboard LED
* minor patch cleanup
the I2S clock is apparently problematic right now, or the dts definition is off. Either way it is causing problems at the moment. This simply doesn't add the device tree entries for the cards.
* Initial commit, cleanup of Rockchip default
* clean up patches:
- 100_DTS_Tinker_opp_LED was splitted into 3 patches
- 100_DTS_opp for opp (adjusted to RKs 3 point description of opp Voltage)
- 110_DTS_8723bs for the wifi driver
-rename of 270-dts-tinker-bluetooth.patch -> 111_DTS_disable_AP6212_bluetooth.patch to keep wifi-chip stuff together
- 120_DTS_led1_heartbeat
- 190_dts_tinker_s_emmc was reworked and splitted
- gpiomem node was removed and has its own patch (260_DTS_gpiomem_node)
- i2c1 node was removed and has its own patch 270_DTS_activate_i2c1
- spi2 part was removed (defined in rk3288.dtsi), default CS will now be CS0 not 1 (no reason to change this)
- vdd_log regulator is removed (the counterpart is missing anyway, and actually disabled in ASUS and Rockchips branch) --> own patch in case it gets activated
- 120_workaround_tinker_board_reboot -> 130_workaround_tinker_board_reboot to get some free numbers... ;)
* Renamed 190_DTS_tinker_s_emmc and splitted into DTS-Part and driver part (191_driver_mmc_adjustment)
* Boots, Still no hdmi console
* Rockchip Kernel config updates
Thanks to @Ntemis for the config options, they fixed the framebuffer
Still have runtime GPU errors related to power model
- merging C2 and Meson family on the source level, C2 is still making own kernels, but sources, patches and configs are the same
- merged boot scripts for meson family. support for non existing legacy kernel was dropped from boot scripts, added UUID support, armbianEnv.txt
- Meson default remain on 4.14.y, NEXT moved to 4.17.y and DEV remains on master
- remove deprecated patches, remove wrong permissions in patches
- tested on C2, K2 and Le Potate. USB troubles are still present and common, while K2 and Lepotato might have boot/reboot troubles on 4.16+
- boards configuration cleanup
- moving K2 to the supported list. It's the same level of support as Le Potato
renamed board-47-02-enable-dvfs-opi-one from safe-need-more-inspection/47-02-enable-dvfs-opi-one
renamed board-47-05-enable-dvfs-opi-lite from safe-need-more-inspection/47-05-enable-dvfs-opi-lite
renamed board-47-01-enable-dvfs-opi-zero from safe-need-more-inspection/47-01-enable-dvfs-opi-zero
File: unresolved/12-dw-hdmi-add-H3-glue.patch
now included upstream in drivers/gpu/drm/sun4i/Kconfig b/drivers/gpu/drm/sun4i/Kconfig
and in drivers/gpu/drm/sun4i/Makefile
and provided upstream in drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
