| Age | Commit message (Collapse) | Author |
|---|
|
into next
CI: https://source.denx.de/u-boot/custodians/u-boot-riscv/-/pipelines/26569
- SoC: add SPL support for licheepi4a
- RISC-V: raise SPL_SYS_MALLOC_SIZE to 8 MiB
|
|
Assign myself to develop U-Boot port of T-Head TH1520 SoC, and help
maintain related code and review patches.
Signed-off-by: Yao Zi <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
Proper U-Boot for Lichee Pi 4A now runs in S mode instead of M mode,
which means the extra firmware, OpenSBI, must be built and integrated
in the image, and the vendor U-Boot cannot chainload the result image
anymore as it runs in M mode.
Remove redundant information about chainloading and update build steps
to mention OpenSBI firmware.
Signed-off-by: Yao Zi <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
RISC-V software usually expects S mode when leaving the firmware, e.g.
UEFI applications could only run in S mode. Let's convert proper U-Boot
of Lichee Pi 4A port to run in S mode.
Signed-off-by: Yao Zi <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
Add an OpenSBI entry to the FIT image. As it expects an FDT to be
passed, corresponding FDT entry is generated with of-list as well.
As SPL now passes a full FDT for following stages, proper U-Boot image
is packed into u-boot-with-spl.bin without a devicetree copy included.
Signed-off-by: Yao Zi <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
These operations rely on a customized M-mode CSR, MHCR, which isn't
available when running in S mode.
Let's fallback to the generic weak stub when running in S mode to avoid
illegal accesses.
Signed-off-by: Yao Zi <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
Symbols in spl.c only function correctly in SPL stage. Build the file
for SPL only to avoid weak symbols in proper U-Boot being unexpectedly
reloaded.
Fixes: 5fe9ced3552 ("riscv: cpu: Add TH1520 CPU support")
Signed-off-by: Yao Zi <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
On several RISC-V boards we have seen that 1 MiB is a insufficient value
for CONFIG_SPL_SYS_MALLOC_SIZE.
For instance qemu-riscv32_spl_defconfig fails booting because u-boot.itb
exceeds 1 MiB.
8 MiB is a reasonable value that allows adding FPGA blobs or splash images
to main U-boot.
Reported-by: Simon Glass <[email protected]>
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Leo Yu-Chi Liang <[email protected]>
|
|
https://source.denx.de/u-boot/custodians/u-boot-rockchip
CI: https://source.denx.de/u-boot/custodians/u-boot-rockchip/-/pipelines/26117
- Allow to silent TPL/SPL debug console;
- enable exFAT support for Theobroma boards;
- Fix SD power initialization in SPL for rk3399-nanopi4
|
|
https://source.denx.de/u-boot/custodians/u-boot-efi
Pull request doc-2025-07-rc4
Documentation:
* fix typo in gcc.rst
* correct EFI_TCG2_PROTOCOL_MEASURE_DTB description
* Add missing reference to firmware for BB-AI64
* Tidy up the bootefi-command docs
|
|
Fix a typo error in gcc.rst, changing "out-out-tree" to
"out-of-tree" in the Out-of-tree section.
Signed-off-by: BehradElmi <[email protected]>
Reviewed-by: Heinrich Schuchardt <[email protected]>
|
|
There are backslashes in some of the tags which seems to be unnecessary.
Remove then.
Change the word 'either' to 'any' since there are three options.
Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Heinrich Schuchardt <[email protected]>
|
|
The details of the sysfw.itb from the R5 build that
also needs to be copied as part of the target images
is missing, but is included in the image formats a
little further down, so add it to the instructions.
Signed-off-by: Peter Robinson <[email protected]>
Reviewed-by: Dhruva Gole <[email protected]>
Acked-by: Nishanth Menon <[email protected]>
|
|
%s/data that change/data that changes/
%s/cannot be used has/cannot be used for/
%s/Otherwise/Otherwise,/
%s/allows better measurement/allows for better measurement/
Acked-by: Ilias Apalodimas <[email protected]>
Signed-off-by: Heinrich Schuchardt <[email protected]>
|
|
Tom Rini <[email protected]> says:
Taking inspiration from Heiko's patch[1] this series goes and cleans up
all of the arch/arm/dts/*.h files that can be easily removed. The big
challenge I ran in to here was that for some platforms that aren't using
OF_UPSTREAM were didn't have a sufficiently deep search path to find
files there rather than arch/arm/dts. This also showed that only ARM had
local header files to deal with.
[1]: https://lore.kernel.org/u-boot/[email protected]/
Link: https://lore.kernel.org/r/[email protected]
|
|
We have this file in both arch/arm/dts and
dts/upstream/src/arm64/freescale. This file is identical save for
changes which have been made upstream. Remove our local copy to get in
sync with upstream now.
Acked-by: Peng Fan <[email protected]>
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
We have these files in both arch/arm/dts and
dts/upstream/src/arm/microchip. These files are identical save for
changes which have been made upstream. Remove our local copy to get in
sync with upstream now.
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream DT related
header files we have a header that is under the arch/arm/dts directory
and differ in being a strict subset of what is found upstream. We can
remove this now to prevent future conflicts.
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream DT related
header files we have a number of these headers that are under the
arch/arm/dts directory and differ only in combinations of spacing
changes and/or switching to SPDX license tags. We can remove these now
to prevent future conflicts.
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream DT related
header files we have a number of these headers that are under the
arch/arm/dts directory and are currently identical to the versions in
dts/upstream. We can remove these now to prevent future conflicts.
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream DT related
header files we have a header that is under the arch/arm/dts directory
and now unused. We can remove this now to prevent any future conflicts.
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
A problem with the logic in upstream_dtsi_include currently is that it
does not list directories such as dts/upstream/src/arm/nxp/imx and so
will not findi "imx6ul-pinfunc.h" for example as it is normally and
correctly included without vendor sub-paths. Expand the current wildcard
glob to catch these directories too.
Reviewed-by: Peter Robinson <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream dt-bindings
headers we have a number of these headers that are in our include
directory and while they are not a strict subset of the upstream version
of the headers, all platforms build with the new headers as well. We can
remove the copies under include/dt-bindings now to prevent future
conflicts.
Signed-off-by: Tom Rini <[email protected]>
|
|
Aside from SPDX tags, the only difference between our version of this
header and upstream is that BCM6318_RST_HOSTMIPS was defined to 11 (the
same as BCM6318_RST_PHYMIPS) and is now defined to 12.
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream dt-bindings
headers we have a number of these headers that are in our include
directory and differ in being a strict subset of what is found upstream.
We can safely remove the copies under include/dt-bindings now to prevent
future conflicts.
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream dt-bindings
headers we have a number of these headers that are in our include
directory and differ only in combinations of spacing changes and/or
switching to SPDX license tags. We can safely remove the copies under
include/dt-bindings now to prevent future conflicts.
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream dt-bindings
headers we have a number of these headers that are in our include
directory and are currently identical to the versions in dts/upstream.
We can remove these now to prevent future conflicts.
Signed-off-by: Tom Rini <[email protected]>
|
|
As part of moving to using OF_UPSTREAM and so the upstream dt-bindings
headers we have a number of these headers that are in our include
directory and not referenced by any code outside of dts/upstream. We can
remove these now to prevent future conflicts.
Signed-off-by: Tom Rini <[email protected]>
|
|
The val (the bits to set) is the second member of the reg_data structure
and mask the third one. We obviously want to clear bits 6 and 7 in order
to only set bit 7 in there instead of only clearing bit 7 in order to
write bits 6 and 7 (which makes no sense).
Fortunately, according to the datasheet, bit 6 value doesn't matter when
bit 7 is set so this is essentially just a cosmetic change, no intended
change in behavior.
Fixes: f172575d92cd ("power: rk8xx: add support for RK806")
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The NanoPi RK3399 boards support UHS-I (up to SDR104) SD cards, however
using any of these 1.8v modes results in a boot failure in SPL upon soft
reboot.
The issue is that the "vcc_sdio" regulator is left at 1.8v on reboot
and the corresponding GPIO defaults to 3.3v. This prevents the SD card
from being reinitialized and read successfully.
This change enables the RK8XX regulators and Rockchip IO-domain drivers
in SPL, which initializes "vcc_sdio" regulator to 3.0v and configures
the GPIO for the correct level on boot.
Signed-off-by: Justin Klaassen <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
This change removes the "regulator-always-on" property from the
"vcc3v0_sd" (vmmc-supply) and "vcc_sdio" (vqmmc-supply) regulators,
which otherwise prevents the MMC driver from being able to power cycle
the SD card as part of the initialization procedure.
It also removes the "regulator-boot-on" from the "vcc_sdio" regulator,
which could theoretically damage a SD card that is already initialized
in a low voltage mode.
Signed-off-by: Justin Klaassen <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Allows use of the regulator functions of the RK8XX PMIC in SPL, which is
necessary to support the functionality of the Rockchip IO-domain driver
on relevant platforms.
Signed-off-by: Justin Klaassen <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Allows use of the Rockchip IO-domain driver in SPL to configure
the GPIO to match the voltage supplied by specific regulators
(e.g. "vcc_sdio").
Signed-off-by: Justin Klaassen <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Log the value of the regulators during initialization of the IO-domain
driver to aid in debugging GPIO voltage configuration problems.
Signed-off-by: Justin Klaassen <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Our upcoming Mass Flasher solution will be storing boot artifacts on an
exFAT partition so enable its support.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Our upcoming Mass Flasher solution will be storing boot artifacts on an
exFAT partition so enable its support.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Our upcoming Mass Flasher solution will be storing boot artifacts on an
exFAT partition so enable its support.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Our upcoming Mass Flasher solution will be storing boot artifacts on an
exFAT partition so enable its support.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Because DEBUG_UART_BOARD_INIT depends on DEBUG_UART, hard dependency
to DEBUG_UART_BOARD_INIT in ROCKCHIP_PX30 can cause warnings if
DEBUG_UART is disabled.
The DEBUG_UART_BOARD_INIT is already implied by ARCH_ROCKCHIP entry.
Remove hard dependency from ROCKCHIP_PX30, so that it will be
consistent with other rockchip boards.
Signed-off-by: Lukasz Czechowski <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Allow to disable serial console in TPL and SPL. Weak dependency
to SPL_SERIAL and TPL_SERIAL is also used in other Rockchip boards.
Signed-off-by: Lukasz Czechowski <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
In case DEBUG UART is not used, define dummy macros replacing
the actual function implementations that will not be available.
This allows to compile code and avoid linker errors.
Redefine the DEBUG_UART_FUNCS macro if DEBUG UART is not available,
to avoid compilation errors.
Signed-off-by: Lukasz Czechowski <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The RAM_ROCKCHIP_DEBUG can be used only if DEBUG_UART is
available.
The next commit introduces changes in definition of debug
uart functions, so that DEBUG_UART is required to be defined
in order to initialize uart and use print functions.
Signed-off-by: Lukasz Czechowski <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
|
|
Update definition of _debug_uart_putc to static inline.
This will allow to avoid compilation warnings about unused code
after introduction of patch changing debug uart functions to
dummies if CONFIG_DEBUG_UART is not set.
This also matches the instructions in include/debug_uart.h and
provides consistency with implementations for other platforms.
Signed-off-by: Lukasz Czechowski <[email protected]>
Reviewed-by: Tom Rini <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Update the definition of _debug_uart_putc to static inline.
This matches the instructions in include/debug_uart.h and
provides consistency with implementations for other platforms.
Signed-off-by: Lukasz Czechowski <[email protected]>
Reviewed-by: Tom Rini <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
This file is a duplicate and also comes with the sync of the linux
mainline dts files. By removing this the one from the dts folder should
be taken that is more up-to-date.
Signed-off-by: Heiko Thiery <[email protected]>
|
|
Clang is stricter than GCC when it comes to inline assembly and expects the
register to be written with explicitly same type of variable.
Fixes: c0e1775a867c ("armv8: Add arch-specific sysinfo platform driver")
Signed-off-by: Raymond Mao <[email protected]>
Reviewed-by: Ilias Apalodimas <[email protected]>
|
|
The changes upstream since the last sync at commit 5c02350fa03d
("scripts/setlocalversion: sync with linux v6.9") are
e2ff1219a554 setlocalversion: add -e option
523f3dbc187a setlocalversion: work around "git describe" performance
Signed-off-by: Tom Rini <[email protected]>
|
|
Permit to select a bootmenu entry with a key shortcut. This is
especially useful in production or testing scenario to automate flashing
procedure or testing procedure.
The boot entry are changed to append the shortcut key to it.
Example:
1. Run default boot command.
2. Boot system via TFTP.
3. Boot production system from NAND.
4. Boot recovery system from NAND.
5. Load production system via TFTP then write to NAND.
6. Load recovery system via TFTP then write to NAND.
7. Load BL31+U-Boot FIP via TFTP then write to NAND.
8. Load BL2 preloader via TFTP then write to NAND.
9. Reboot.
a. Reset all settings to factory defaults.
0. Exit
0 is always reserved for Exit to console.
On pressing the keyboard key 2, the bootmenu entry 2 is selected and
executed.
Up to 34 key shortcut (0 excluded as reserved) are supported from 1-9
and a-z.
If a shortcut key not present in the bootmenu list is pressed, it is
simply ignored and eventually the autoboot is interrupted.
Capital A-Z are converted to lower a-z and the related option is
selected.
Suggested-by: Weijie Gao <[email protected]>
Signed-off-by: Christian Marangi <[email protected]>
Tested-by: Petr Štetiar <[email protected]>
|
|
Weijie Gao <[email protected]> says:
This patch series adds baudrate accuracy compensation for MediaTek UART
driver in high-speed mode 3.
Link: https://lore.kernel.org/r/[email protected]
|
|
The high-speed UART from MediaTek supports baudrate accuracy
compensation when using high-speed mode 3.
This is done by calculating the first digit of the fraction part of
sample count value. The fraction value will be then used as the
reference to insert 0 to 10 sample cycle(s) to one frame (assume
that frame format is 8n1, i.e. 10 bits per frame).
The fracdiv_[l/m] registers are used to determine whether a bit in one frame
should be inserted with one sample cycle.
With typical 40MHz source clock, the actual baudrates with/without
accuracy compensation are:
Ideal w/o compensation w/ compensation
======== ================ ===============
9600 9603 9600
115200 114942 115207
921600 930232 921659
3000000 3076923 3007519
Signed-off-by: Weijie Gao <[email protected]>
|
