Call to board_debug_uart_init() is useless, as mainline U-Boot can
not build TPL for rk3308, and proprietary ddr.bin to be used as TPL
is responsible to init debug uart. Moreover current implementation
of board_debug_uart_init() is not compatible with ROCK Pi S, as it
sets pins for UART2 channel 1 breaking access to sdmmc due to pinmux
conflict. Debug uart for ROCK Pi S is UART0.
Thus, avoid ROCKCHIP_RK3308 to select DEBUG_UART_BOARD_INIT and allow
to deselct it in rock-pi-s-rk3308_defconfig. The DEBUG_UART_BOARD_INIT
is already implied by ARCH_ROCKCHIP, therefore other boards based on
rk3308 chip are not affected by change.
Signed-off-by: Massimo Pegorer <massimo.pegorer@vimar.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Definition of function board_debug_uart_init() must be under
CONFIG_DEBUG_UART_BOARD_INIT and not under CONFIG_DEBUG_UART,
as it was: see debug_uart.h. In this way the debug uart can
be used but its board-specific initialization skipped by
configuration, if useless.
Signed-off-by: Massimo Pegorer <massimo.pegorer@vimar.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
- Resync some of the K3 DTS files with the kernel, and pull in some
required related updates to keep drivers in sync with the dts files
now. Bring in some incremental fixes on top of one of the series I
applied recently as well as updating the iot2050 platform. Also do a
few small updates to the K2 platforms.
- Enable pcie support for rk3568;
- Add boards:
rk3399: Radxa ROCK 4SE;
rk3328: Orange Pi R1 Plus, Orange Pi R1 Plus LTS
rk3568: FriendlyARM NanoPi R5S/R5C, Hardkernel ODROID-M1
rk3588: Edgeble Neu6B
- support OP-TEE with binman;
- support Winbond SPI flash;
- rk3588 usbdp phy support;
- dts and config updates for different boards;
Move to using .env file for setting up environment variables for K2x_evm.
Signed-off-by: Andrew Davis <afd@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
This is more complex than it needs to be and makes converting these
boards over to plain text env files more difficult. Remove setting
mtdparts as the DTS already contain the partitions. While here also
drop the conflicting definitions from the K2 defconfigs.
Signed-off-by: Andrew Davis <afd@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
The following checks are more reasonable as the previous logs were a bit
misleading as we could still get the logs that the authetication is
being skipped but still authenticate. Move the debug prints and checks
to proper locations.
Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
Similar to get_fdt_mmc make get_overlays_mmc look at /boot/dtb/* path
for overlay files.
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
Reviewed-by: Nikhil M Jain <n-jain1@ti.com>
All K3 SoCs use same set of args to load kernel for MMC. So move this to
common place to avoid duplication.
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
Reviewed-by: Nikhil M Jain <n-jain1@ti.com>
This was regressed by the following commit and is required to build with
additional configs enabled.
Fixes: 14439cd71c ("configs: k3: make consistent bootcmd across all k3 socs")
Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
Tested-by: Nikhil M Jain <n-jain1@ti.com>
Fix missing declaration of fdt_del_node_path() while compiling am625_fdt.c and
missing common_fdt.h include in common_fdt.c
Fixes: 70aa5a94d4 ("arm: mach-k3: am62: Fixup CPU core, gpu and pru nodes in fdt")
Signed-off-by: Emanuele Ghidoli <emanuele.ghidoli@toradex.com>
Only accept SPACE to stop autobooting. This is safer to avoid accidental
interruptions on unattended devices.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
ATF is now called BL31, and OP-TEE since 3.21 suggests to use
tee-raw.bin instead of (the still identical) tee-pager_v2.bin.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
This avoids having to maintain to defconfigs that are 99% equivalent.
The approach is to use binman to generate two flash images,
flash-pg1.bin and flash-pg2.bin. With the help of a template dtsi, we
can avoid duplicating the common binman image definitions.
Suggested-by: Andrew Davis <afd@ti.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
The underlying issue was fixed in the meantime. Also signing the U-Boot
proper fit image now works. Just supporting custom cert templates
remains a todo.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
The common env bits now come via ti_armv7_common.env, include it.
Furthermore restore the board-specific boot targets and their ordering
that is now enforced k3-wide differently. Finally, enable
CONFIG_LEGACY_IMAGE_FORMAT explicitly which got lost while turning
FIT_SIGNATURE on by default for k3 devices.
Fixes: 53873974 ("include: armv7: Enable distroboot across all configs")
Fixes: 4ae1a247 ("env: Make common bootcmd across all k3 devices")
Fixes: 86fab110 ("Kconfig: Enable FIT_SIGNATURE if ARM64")
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Update the am62 and am625 device-trees from linux v6.5-rc1. This needed
the following tweaks to the u-boot specific dtsi as well:
- Switch tick-timer to the main_timer as it's now defined in the main dtsi
- Secure proxies are defined in SoC dtsi
- Drop duplicate nodes - u-boot.dtsi is includes in r5-sk, no need for
either the definitions from main.dtsi OR duplication from u-boot.dtsi
Reviewed-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Tested-by: Maxime Ripard <mripard@kernel.org>
Cc: Francesco Dolcini <francesco@dolcini.it>
Cc: Sjoerd Simons <sjoerd@collabora.com>
Cc: Wadim Egorov <w.egorov@phytec.de>
Signed-off-by: Nishanth Menon <nm@ti.com>
Timer0 is used by u-boot as the tick timer; Add it to the soc devices
list so it can be enabled via the k3 power controller.
Signed-off-by: Sjoerd Simons <sjoerd@collabora.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Roger Quadros <rogerq@kernel.org>
Tested-by: Maxime Ripard <mripard@kernel.org>
Tested-by: Ravi Gunasekaran <r-gunasekaran@ti.com>
Tested-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Cc: Francesco Dolcini <francesco@dolcini.it>
Cc: Wadim Egorov <w.egorov@phytec.de>
Signed-off-by: Nishanth Menon <nm@ti.com>
The TI AM654 timer is compatible with the omap-timer implementation,
so add it to the compatible id list.
Signed-off-by: Sjoerd Simons <sjoerd@collabora.com>
Reviewed-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Tested-by: Maxime Ripard <mripard@kernel.org>
Tested-by: Ravi Gunasekaran <r-gunasekaran@ti.com>
Tested-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Cc: Francesco Dolcini <francesco@dolcini.it>
Cc: Wadim Egorov <w.egorov@phytec.de>
Signed-off-by: Nishanth Menon <nm@ti.com>
Hardkernel ODROID-M1 is a single board computer with a RK3568B2 SoC,
a slightly modified version of the RK3568 SoC.
Features tested on a ODROID-M1 8GB v1.0 2022-06-13:
- SD-card boot
- eMMC boot
- SPI Flash boot
- PCIe/NVMe/AHCI
- SATA port
- USB host
Device tree is imported from linux v6.4.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Stefan Agner <stefan@agner.ch>
Tested-by: Stefan Agner <stefan@agner.ch>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Building U-Boot with CMD_INI=y result in following build warning:
cmd/ini.c: In function 'memgets':
include/linux/kernel.h:184:24: warning: comparison of distinct pointer types lacks a cast
184 | (void) (&_min1 == &_min2); \
| ^~
cmd/ini.c:92:15: note: in expansion of macro 'min'
92 | len = min((end - *mem) + newline, num);
| ^~~
Fix this by adding an int cast to the pointer arithmetic result.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
The dwc_ahci driver use platform specific defines, place the platform
specific code behind a ifdef CONFIG_ARCH_OMAP2PLUS to allow build and
use of the driver on Rockchip platform.
Fixes: 02a4b42979 ("drivers: block: dwc_ahci: Implement a driver for Synopsys DWC sata device")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
The standard boot path expects the kernel_comp_addr_r and kernel_comp_size
variables for booting compressed kernel images. Define them using the previous
kernel_addr_c value (likely initially meant for this purpose) and usual size.
This was tested on the PX30 EVB to successfully boot compressed Linux kernel
images.
Signed-off-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Update config, IO and memory regions used based on [1] with pcie3x2
config reg address and reg size corrected.
Before this change:
PCI Autoconfig: Bus Memory region: [0-3eefffff],
PCI Autoconfig: Bus I/O region: [3ef00000-3effffff],
After this change:
PCI Autoconfig: Bus Memory region: [40000000-7fffffff],
PCI Autoconfig: Bus I/O region: [f0100000-f01fffff],
[1] https://lore.kernel.org/lkml/20221112114125.1637543-2-aholmes@omnom.net/
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add missing pinctrl and defconfig options to enable PCIe and NVMe
support on Radxa ROCK 3 Model A.
Use of pcie20m1_pins and pcie30x2m1_pins ensure IO mux selection M1.
The following pcie_reset_h and pcie3x2_reset_h ensure GPIO func is
restored to the perstn pin, a workaround to avoid having to define
a new rockchip,pins.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
The commit 12df2c182ccb ("regulator: dt-bindings: fixed-regulator: allow
gpios property") in linux v6.3-rc1 added support for use of either a
gpios or gpio prop with a fixed-regulator.
This adds support for the new gpios prop to the fixed-regulator driver.
gpios prop is used by vcc3v3-pcie-regulator on Radxa ROCK 3 Model A.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
The Root Complex BARs default to claim the full 1 GiB memory region on
RK3568, leaving no space for any attached device.
Fix this by disable the unused BAR 0 and BAR 1 of the RC.
Signed-off-by: Jon Lin <jon.lin@rock-chips.com>
[jonas@kwiboo.se: Move to rk_pcie_configure and use PCI_BASE_ADDRESS_0/1 const]
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Use a similar pattern and delay values as the linux mainline driver to
speed up failing when nothing is connected.
Reduce fail speed from around 5+ seconds down to around one second on a
Radxa ROCK 3 Model A, where pcie2x1 is probed before pcie3x2 M2 slot.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
The vpcie3v3 regulator is typically a fixed regulator controlled using
gpio. Change to use enable and disable calls on the regulator instead
of trying to set a voltage value.
Also remove the delay to match linux driver, for a fixed regulator the
startup-delay-us prop can be used in case a startup delay is needed.
Limited testing on ROCK 3A, ROCK 5B, Quartz64, Odroid-M1 has shown that
this delay was not needed.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Get the config region to use from the reg prop. Also update the
referenced region index used in comment.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add dev_read_addr_size_index_ptr function with the same functionality as
dev_read_addr_size_index, but instead a return pointer is given.
Use map_sysmem() function as cast for the return.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Some veyron boards seem to have Winbond SPI flash chips instead of
GigaDevice ones. At the very least, coreboot builds for veyron boards
have them enabled [1]. Enable support for them here as well.
[1] https://review.coreboot.org/c/coreboot/+/9719
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add board-specific devicetree/config for the RK3399T-based Radxa ROCK 4SE
board. This board offers similar peripherals in a similar form-factor to
the existing ROCK Pi 4B but uses the cost-optimised RK3399T processor
(which has different OPP table than the RK3399) and other minimal hardware
changes.
Kernel tag: next-20230719
Kernel commits:
- 86a0e14a82ea ("arm64: dts: rockchip: Add Radxa ROCK 4SE")
Signed-off-by: Christopher Obbard <chris.obbard@collabora.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
To prepare for ROCK 4 SE support, changes are needed to the common ROCK
Pi 4 devicetree to move the OPP from the common devicetree to individual
board devicetrees. Sync the Rockchip RK3399 ROCK Pi 4-related DTs from
Linux to gain from these changes.
Kernel tag: next-20230719
Kernel commits:
cfa12c32b96f ("arm64: dts: rockchip: correct wifi interrupt flag in Rock \
Pi 4B")
cee572756aa2 ("arm64: dts: rockchip: Disable HS400 for eMMC on ROCK Pi 4")
2bd1d2dd808c ("arm64: dts: rockchip: Disable HS400 for eMMC on ROCK 4C+")
fd2762a62646 ("arm64: dts: rockchip: Move OPP table from ROCK Pi 4 dtsi")
Signed-off-by: Christopher Obbard <chris.obbard@collabora.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
This updates the evb_rk3229's README on howto create / use the FIT image
created by binman.
Also fix some wrong paths and update filenames which have changed in recent
upstream optee-os versions.
Signed-off-by: Alex Bee <knaerzche@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
For RK322x series ARM SoCs the OP-TEE is non-optional, as besides the TEE
it also provides the PSCI implementation, which is expected to be available
by upstream linux.
Select CONFIG_SPL_OPTEE_IMAGE if an FIT image is built.
Signed-off-by: Alex Bee <knaerzche@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
An OP-TEE FIT image will fail to extract in SPL because the malloc stack
size is currently limited to 0x2000 for evb-rk3229 board.
In SPL we do not have to care about size limitations, since we are no
longer bound to SRAM limits after DRAM initialization has been done in TPL.
Use the default value for CONFIG_SPL_STACK_R_MALLOC_SIMPLE_LEN in order
successfully unpack the FIT image.
Signed-off-by: Alex Bee <knaerzche@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
CONFIG_SPL_OPTEE_IMAGE option is used during DRAM size detection for
Rockchip ARM platform to indicate that an OP-TEE binary was already loaded
and a Trusted Execution Environment (TEE) is available in order to
block/reserve a memory-region for it.
This adds a bunch of new `#if's` to u-boot-rockchip.dtsi to include the
OP-TEE binary in the FIT image for ARM SOCs if CONFIG_SPL_OPTEE_IMAGE is
selected.
That makes it a little harder to read, but I opted for that, because all
the duplicates in an extra ARM-OP-TEE-specfic .dtsi would be the greater
evil, IMHO. Besides it's more likley being "forgotten" to sync when changes
in u-boot-rockchip.dtsi are made.
The no longer required rockchip-optee.dtsi and it's inclusions are dropped.
The hardcoded load address is common across all OP-TEE implemenations for
Rockchip (vendor and upstream).
The OP-TEE-binary is non-optional if CONFIG_SPL_OPTEE_IMAGE is selected and
there will be an error if the file does not exist and/or `TEE=` build
option is missing.
Signed-off-by: Alex Bee <knaerzche@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Enable CONFIG_PCI_INIT_R for rock5b pci enumeration during boot in order
to autodetect the PCI ethernet NIC during the boot process.
Signed-off-by: Christopher Obbard <chris.obbard@collabora.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Current SYS_MALLOC_F_LEN of 0x2000 (8 KB) used in SPL is too small for
some RK3568 boards. SPL will print following during boot:
alloc space exhausted
Increase the default SYS_MALLOC_F_LEN to 0x20000 (128 KB) to mitigate.
Fixes: 2a950e3ba5 ("rockchip: Add rk3568 architecture core")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Rockchip boot blocks are written per 4 x 512 byte sectors per page.
Each page must have a page address (PA) pointer in OOB to the next page.
Pages are written in a pattern depending on the NAND chip ID.
This logic used to build a page pattern table is not fully disclosed and
is not easy to fit in the MTD framework.
The formula in rk_nfc_write_page_hwecc() function is not correct.
Make hwecc and raw behavior identical.
Generate boot block page address and pattern for hwecc in user space
and copy PA data to/from the already reserved last 4 bytes before EEC
in the chip->oob_poi data layout.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Puma supports upstream TF-A and is configured to output serial on UART0
instead of the default UART2. Since U-Boot is properly configured to
output on UART0, let's pass the DT to TF-A so there is no need for a
custom TF-A to make the latter output to UART0 too.
Cc: Quentin Schulz <foss+uboot@0leil.net>
Signed-off-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Long are gone the times TF-A couldn't handle the FDT passed by U-Boot.
Specifically, since commit e7b586987c0a ("rockchip: don't crash if we
get an FDT we can't parse") in TF-A, failure to parse the FDT will use
the fallback mechanism. This patch was merged in TF-A v2.4-rc0 from two
years ago.
New boards should likely have this option disabled or explicitly enable
it in their respective defconfig.
Because existing boards might depend on a TF-A version that predates
v2.4, let's just enable this option in all RK3399 defconfigs.
Maintainers of each board can decide for themselves if they would prefer
to disable this option and allow U-Boot to pass the DT to TF-A.
Cc: Quentin Schulz <foss+uboot@0leil.net>
Signed-off-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Neural Compute Module 6B(Neu6B) is a 96boards SoM-CB compute module
based on Rockchip RK3588J from Edgeble AI.
Add support for this SoM and IO board.
Signed-off-by: Jagan Teki <jagan@edgeble.ai>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Neural Compute Module 6B(Neu6B) is a 96boards SoM-CB compute module
based on Rockchip RK3588J from Edgeble AI.
General features:
- Rockchip RK3588J
- up to 32GB LPDDR4x
- up to 128GB eMMC
- 2x MIPI CSI2 FPC
- On module WiFi6/BT
Neural Compute Module 6B(Neu6B) IO board is an industrial form factor
ready-to-use IO board from Edgeble AI.
General features:
- microSD slot
- 1x HDMI Out
- 1x HDMI In
- 2x DP
- 1x eDP
- 2x MIPI DSI connector
- 4x MIPI CSI2 connector
- 2x USB Host
- 2x USB 3.0 OTG/Host
- 1x SATA
- 1x 2.5Gbps Ethernet
- 1x M.2 B-Key for 4G/5G cards
- 1x M.2 M-Key slot
- 1x Onboard PoE
- 1x RS485, RS232, CAN
- 1x Audio, MIC port
- RTC battery slot
- 40-pin GPIO expansion
Neu6B needs to mount on top of this IO board in order to create a
complete Edgeble Neural Compute Module 6B(Neu6B) IO platform.
Kernel commits:
commit <5f06c3f508f7> ("arm64: dts: rockchip: Add rk3588 Edgeble Neu6
Model B SoM")
commit <3a9181a43b94> ("arm64: dts: rockchip: Add rk3588 Edgeble Neu6
Model B IO")
Signed-off-by: Jagan Teki <jagan@edgeble.ai>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add rk3588j-u-boot.dtsi for adding U-Boot specific nodes and
properties for Rockchip RK3588J SoC.
Signed-off-by: Jagan Teki <jagan@edgeble.ai>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Rockchip RK3588J is the industrial-grade version of RK3588 SoC and
is operated with -40 °C to +85 °C temparature.
Add rk3588j specific dtsi for adding rk3588j specific operating points
and other changes to be add in future.
Kernel commit:
commit <8274a04ff1dc> ("arm64: dts: rockchip: Add Rockchip RK3588J")
Signed-off-by: Jagan Teki <jagan@edgeble.ai>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
