Bin Meng [Fri, 18 Jun 2021 07:09:30 +0000 (15:09 +0800)]
x86: Discard .note.gnu.property sections
When switching to kernel.org x86_64 gcc 11.1.0 toolchain, u-boot.rom
built from qemu-x86_defconfig no longer boots anymore. Investigation
shows that U-Boot fails at a very early stage during the boot process,
in fdtdec_prepare_fdt() where fdt_check_header() complains that there
is not a valid device tree found at gd->fdt_blob which points to _end.
Now _end points to an allocated section .note.gnu.property which of
course is wrong.
This issue is however not seen when using the default Ubuntu 20.04 gnu
toolchain (gcc 9.3.0 with binutils 2.34). Further investigation shows
that it is caused by a behavior change of binutils v2.36 which is part
of the kernel.org gcc 11.1.0 toolchain, via the following commit:
939b95c77bf2 ("Linux/x86: Configure gas with --enable-x86-used-note by default")
In fact, there was already a regression bug report [1] for binutils two
months ago, but the binutils folks did not think it is a bug :(
To resolve this, there are several options:
* pass -Wa,-mx86-used-note=no to gas
* pass -R .note.gnu.property to objcopy
* discard the section in the linker script
Linux kernel uses the discard way [2], so let's do the same for U-Boot.
cmd: pxe_utils: sysboot: fix crash if either board or soc are not set.
If the environment does not have "soc" or "board" set, and fdtdir
option is specified in extlinux.conf, the bootloader will crash whilst
dereferencing a null pointer. Add a guard against null soc or
board. Fixes a crash of qemu-riscv64_smode configuration, which does
not have CONFIG_SYS_SOC defined.
Signed-off-by: Dimitri John Ledkov <xnox@ubuntu.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
In u-boot device trees both prefixes are used. Until recently the only
user of the ethernet alias was the sandbox test device tree. This
changed with commit fc054d563bfb ("net: Introduce DSA class for Ethernet
switches"). There, the MAC addresses are inherited based on the devices
sequence IDs which is in turn given by the device tree.
Before there are more users in u-boot and both worlds will differ even
more, rename the alias prefix to "ethernet" to match the linux ones.
Also adapt the test cases and rename any old aliases in the u-boot
device trees.
Cc: David Wu <david.wu@rock-chips.com> Signed-off-by: Michael Walle <michael@walle.cc> Reviewed-by: Fabio Estevam <festevam@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Michael Walle [Wed, 24 Feb 2021 16:30:44 +0000 (17:30 +0100)]
net: use a more deterministic approach to get the active ethernet device
If the environment variable "ethact" is not set, the first device in the
uclass is returned. This depends on the probing order of the ethernet
devices. Moreover it is not not configurable at all.
Try to return the ethernet device with sequence id 0 first which then
can be configured by the aliases in a device tree. Fall back to the old
mechanism in case of an error.
Signed-off-by: Michael Walle <michael@walle.cc> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Daniil Stas [Sun, 30 May 2021 13:34:09 +0000 (13:34 +0000)]
net: dwc_eth_qos: Revert some changes of commit 3a97da12ee7b
Revert some changes of commit 3a97da12ee7b ("net: dwc_eth_qos: add dwc
eqos for imx support") that were probably added by mistake.
One of these changes can lead to received data corruption (enabling
FUP and FEP bits). Another causes invalid register rxq_ctrl0 settings
for some platforms. And another makes some writes at unknown memory
location.
Fixes: 3a97da12ee7b ("net: dwc_eth_qos: add dwc eqos for imx support") Signed-off-by: Daniil Stas <daniil.stas@posteo.net> Cc: Ye Li <ye.li@nxp.com> Cc: Fugang Duan <fugang.duan@nxp.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Ramon Fried <rfried.dev@gmail.com> Cc: Joe Hershberger <joe.hershberger@ni.com> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Tim Harvey [Fri, 16 Apr 2021 20:25:09 +0000 (13:25 -0700)]
net: e1000: do not attempt to set hwaddr for i210 without FLASH
commit f1bcad22dd19 ("net: e1000: add support for writing to EEPROM")
adds support for storing hwaddr in EEPROM however i210 devices do not
support this and thus results in errors such as:
Warning: e1000#0 failed to set MAC address'
Check if a flash device is present and if not return -ENOSYS indicating
this is not supported.
Signed-off-by: Tim Harvey <tharvey@gateworks.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Priyanka Jain [Thu, 17 Jun 2021 07:22:46 +0000 (12:52 +0530)]
board: freescale: t1028xrdb: Add MAINTAINER for revD
Add Priyanka Jain as MAINTAINER for
T2080RDB_revD_defconfig,
T2080RDB_revD_NAND_defconfig,
T2080RDB_revD_SDCARD_defconfig and
T2080RDB_revD_SPIFLASH_defconfig
Ran Wang [Wed, 16 Jun 2021 12:23:19 +0000 (17:53 +0530)]
armv8: Update erratum number to align with doc
Change the USB erratum number A-050106 to A-050204 as A-050106 is
a duplicate and never be published.
Fixes 0cfa00cdb94 (“armv8: Add workaround for USB erratum A-050106”) Signed-off-by: Ran Wang <ran.wang_1@nxp.com>
[Rebased] Signed-off-by: Priyanka Jain <priyanka.jain@nxp.com>
Camelia Groza [Fri, 11 Jun 2021 12:28:08 +0000 (15:28 +0300)]
board: freescale: t208xrdb: add a config option for rev D dts fixups
Under DM, we rely on u-boot's device tree to provide the correct PHY
addresses. The board_fix_fdt callback is intended to be used for
device tree fixups before relocation. Unfortunately, this isn't an
option when booting from flash since the device tree isn't writable
before relocation.
This patch introduces the CONFIG_T2080RDB_REV_D option to signal that a
board revision D or up is the target. The config option is used to set
the correct Aquantia PHY address in the board's u-boot device tree.
Defconfig files with the option enable explicitly are added for
convenience.
Camelia Groza [Wed, 16 Jun 2021 12:17:31 +0000 (17:47 +0530)]
board: freescale: t208xrdb: add Linux fdt fix-ups for rev D
The T2080RDB boards revisions D and up have updated 10G Aquantia PHYs
connected to MAC1 and MAC2. The second Aquantia PHY is located at a
different address on the MDIO bus compared to rev C (0x8 instead of 0x1).
Fix-up the Linux device tree to update the PHY address for the second
Aquantia PHY on boards revisions D and up. Also rename the PHY node to
reflect the changes.
Camelia Groza [Fri, 11 Jun 2021 12:28:05 +0000 (15:28 +0300)]
board: freescale: t208xrdb: remove the legacy non-DM_ETH code
Both DM_ETH and DM_PCI are enabled for the T2080RDB board. Remove the
board_eth_init() callback and the non-DM_ETH code paths since they are not
needed anymore.
A system clock of 66MHz was chosen for the pg-wcom-ls102xa.
Compared to the Evalboard, this corresponds to a reduction of 1/3.
The system counter clock should have been reduced accordingly,
but that was not the case, so we had a system time that was
1/3 behind the real time.
This patch corrects the system counter clock to
8.333MHz = 66.667MHz / 8.
In order to improve power consumption ls102x allows to disable peripherals
that are not in use.
This patch follows SELI8 HW design description and disables peripherals
that are not in use in our designs, the same configuration is applicable
and for EXPU1.
This patch uses available hwconfig option for updating ls102x device
disable configuration.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
The default behavior in the latest u-boot revisions is to rotate the
active net device to the next available if the requested link is not
established.
For our ls102x based devices this would mean that if active debug net
device is not available, u-boot will rotate and set the next net device
that is one of the estar adapters.
To return from this situation manual action to set correct ethact
adapter will be needed and this can be annoying when working in
debug mode.
Setting ethrotate=no will disable net adapter rotation and will make sure
that the primary adapter is always used.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
This is most probably a typo, and in older u-uboot versions is same as
'saveenv', in the newer uboot versions there is a separate 'save' command
that is different from 'saveenv'.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
In our designs we reserve PRAM area at the end of the RAM, and in order
this area to be visible and taken into account by the u-boot memory mgmt
CONFIG_PRAM has to be defined.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
km/ls102xa: add support for u-boot POST memory test
From production view this is standard test executed during production on
all linux based foxmc cards.
On CENT2 HW defined memory region is zero means that some DDR accesses are
done by memory_post_dataline and memory_post_addrline but pattern tests
are skipped that's why mem_regions is fast there.
On ls102x for the complete DDR region of 1GiB memory_regions_post_test
takes approx. 4min and this is too much for production, so this patch
defines only 1MiB region as compromise.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
board/km: add support for expu1 design based on nxp
The EXPU1 design is a new 40G capable ethernet service unit card for
Hitachi-Powergrids wired-com product lines.
The base SoC is same as for already added SELI8 card, consequently the
already added u-boot support for SELI8 is reused.
Signed-off-by: Rainer Boschung <rainer.boschung@hitachi-powergrids.com> Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com>
[Fixed new line error at EOF] Signed-off-by: Priyanka Jain <priyanka.jain@nxp.com>
net: tsec: add option to set device max-speed via dts
Current tsec adapter sets adapter gigabit capabilities by default, and in
reality this must not always be the case.
It is possible that tsec adapter is used for 100Mbps connection, and in
this case setting 1000Mbps capabilities can lead to some side effects such
longer autoneg process.
In our ls102x designs this problem leads to long autoneg times (> 4 sec)
in case board rgmii link is 100Mbps capable only.
Limiting the rgmii link capabilities provides faster and smoother
link establishment.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
During the mainlining of the board this was by mistake set to sector size.
Our user space env scripts are expecting envsize of 0x4000, and setting
this differently will break our cross-platform compatibility.
Signed-off-by: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
Yangbo Lu [Thu, 3 Jun 2021 02:51:17 +0000 (10:51 +0800)]
mmc: fsl_esdhc: convert to CONFIG_FSL_ESDHC_VS33_NOT_SUPPORT
For eSDHC, power supply is through peripheral circuit.
Some eSDHC versions have value 0 of the bit but that
does not reflect the truth. 3.3V is common for SD/MMC,
and is supported for all boards with eSDHC in current
u-boot. So, make 3.3V is supported in default in code.
CONFIG_FSL_ESDHC_VS33_NOT_SUPPORT can be enabled if
future board does not support 3.3V.
This is also a fix-up for one previous patch, which converted
to use IS_ENABLED() for CONFIG_SYS_FSL_MMC_HAS_CAPBLT_VS33
that is not a Kconfig option.
Fixes: 52faec31827e ("mmc: fsl_esdhc: replace most #ifdefs by IS_ENABLED()") Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com> Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
Biwen Li [Fri, 21 May 2021 07:15:06 +0000 (15:15 +0800)]
configs: ls2088aqds: fix synchronous exception
IFC NOR flash base address of ls2088a is 0x580000000,
and offset of env crc is 0x500000, so fix the macro
CONFIG_ENV_ADDR to fix synchronous exception(access illegal address)
armv8: ls1012a: Pass PPFE firmware to Linux through FDT
Read Linux PPFE firmware from flash partition and pass it to Linux through
FDT entry. So that we can avoid placing PPFE firmware in Linux rootfs.
(FDT may increase at max by 64KB)
Bin Meng [Tue, 15 Jun 2021 05:45:57 +0000 (13:45 +0800)]
riscv: andes_plic: Fix riscv_get_ipi() mask
Current logic in riscv_get_ipi() for Andes PLICSW does not look
correct. The mask to test IPI pending bits for a hart should be
left shifted by (8 * gd->arch.boot_hart), just the same as what
is done in riscv_send_ipi().
Fixes: 8b3e97badf97 ("riscv: add functions for reading the IPI status") Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Rick Chen <rick@andestech.com> Tested-by: Rick Chen <rick@andestech.com>
Bin Meng [Fri, 4 Jun 2021 05:51:13 +0000 (13:51 +0800)]
riscv: ae350: dts: Add missing "u-boot, dm-spl" for SPL config
At present the AE350 SPL defconfig is using OF_PRIOR_STAGE. The
intention was to use gdb to load device tree before running U-Boot
SPL/proper from RAM. When we switch to OF_SEPARATE we will have to
use our own DT but without "u-boot,dm-spl" in several essential
nodes, SPL does not boot.
Let's add all the required "u-boot,dm-spl" for SPL config.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Rick Chen <rick@andestech.com>
Bin Meng [Fri, 4 Jun 2021 05:51:12 +0000 (13:51 +0800)]
riscv: ae350: dts: Fix #interrupt-cells for plic0 in 32-bit
All the device nodes that refer to plic0 as their interrupt parent
have 2 cells encoded in their interrupts property, but plic0 only
provides 1 cell in #interrupt-cells which is incorrect.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Rick Chen <rick@andestech.com> Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
When using bootefi to boot a EFI binary, u-boot is supposed to
provide networking service for EFI application. Currently, 'pfe stop'
command is called from bootcmd before running bootefi. As a result
network stops working for EFI applications and console is flooded with
"Rx pkt not on expected port" messages.
Implement board_quiesce_devices() for ls1012a boards and call
pfe_command_stop() from it instead of calling 'pfe stop' from
*_bootcmd and bootcmd.
Michael Walle [Tue, 13 Apr 2021 15:54:17 +0000 (17:54 +0200)]
board: sl28: fix RGMII clock and voltage
It was noticed that the clock isn't continuously enabled when there is
no link. This is because the 125MHz clock is derived from the internal
PLL which seems to go into some kind of power-down mode every once in a
while. The LS1028A expects a contiuous clock. Thus enable the PLL all
the time.
Also, the RGMII pad voltage is wrong, it was configured to 2.5V (that is
the VDDH regulator). The correct voltage is 1.8V, i.e. the VDDIO
regulator.
Signed-off-by: Michael Walle <michael@walle.cc> Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
Pali Rohár [Mon, 14 Jun 2021 14:45:58 +0000 (16:45 +0200)]
arm: mvebu: turris_{omnia, mox}: ensure running bootcmd_rescue always works
One of the points of putting the rescue boot command into default
environment is that user can invoke it without physical access to the
board (without having to press the factory reset button), by running
run bootcmd_rescue
in U-Boot's console.
Therefore we have to ensure that bootcmd_rescue is always set to default
value, regardless of whether the factory reset button was pressed.
Otherwise the variable will be empty for example after upgrade from
previous U-Boot.
Fixes: ec3784d62646 ("arm: mvebu: turris_mox: add support for board rescue mode") Fixes: 176c3e7760a2 ("arm: mvebu: turris_omnia: support invoking rescue boot from console") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Michal Simek [Thu, 10 Jun 2021 12:33:55 +0000 (14:33 +0200)]
arm64: zynqmp: Disable EFI_CAPSULE_ON_DISK_EARLY
Sata/scsi and usb based devices are not started at this stage that's why
disks are not found and early update can't happen.
It is because of deficiency in the UEFI implementation which is not able to
deal with block devices which are added or removed after initialization.
EFI capsule on disk early feature needs to be fixed first to be able to
enable this feature properly.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
arm64: zynqmp: Writing correct value to ANALOG_BUS
The default register configuration after powerup for PSSYSMON_ANALOG_BUS
register is incorrect. Hence, fix this in SPL by writing correct fixed
value. It follows UG1085 chapter 'PS SYSMON Analog_Bus' and reflects commit
sw_apps:zynq ("056ca65d44549ce27f716d423e8dfdefeee7440c")
in Xilinx:embeddedsw[1].
[1] https://github.com/Xilinx/embeddedsw
Signed-off-by: Adrian Fiergolski <adrian.fiergolski@fastree3d.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Marek Behún [Mon, 7 Jun 2021 14:34:47 +0000 (16:34 +0200)]
arm: mvebu: turris_mox: add support for board rescue mode
Add necessary config options and board code to support board factory
reset / rescue mode on Turris MOX.
In order to also support invoking rescue mode from U-Boot console,
without having to press the factory reset button, put the rescue command
into `bootcmd_rescue` default environment variable. When factory reset
button is pressed, invoke rescue mode via distroboot by setting
`boot_targets` to `rescue`.
Rescue boot from console can be invoked by running
run bootcmd_rescue
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de>
The fragmented files were not correctly read because of two issues:
- The squashfs_file_info struct has a field named 'comp', which tells if
the file's fragment is compressed or not. This field was always set to
'true' in sqfs_get_regfile_info and sqfs_get_lregfile_info. It should
actually take sqfs_frag_lookup's return value. This patch addresses
these two assignments.
- In sqfs_read, the fragments (compressed or not) were copied to the
output buffer through a for loop which was reading data at the wrong
offset. Replace these loops by equivalent calls to memcpy, with the
right parameters.
I tested this patch by comparing the MD5 checksum of a few fragmented
files with the respective md5sum output in sandbox, considering both
compressed and uncompressed fragments.
Signed-off-by: Joao Marcos Costa <jmcosta944@gmail.com> Tested-by: Richard Genoud <richard.genoud@posteo.net> Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com>
imx8mn: configs: add support for distro boot commands
Supported boot device types in iMX8MN: MMC, DHCP.
Add DISTRO_DEFAULTS config option and include the distro boot command
header file to enable full support of distro boot on i.MX8M Nano EVK (both
DDR and LPDDR derivatives).
Drop previous environment, which was targeting customized boot commands and
boot order.
arm: imx: imx8mm: correct unrecognized fracpll frequency
Frequency requested by ddrphy_init_set_dfi_clk from fracpll uses MHZ()
macro, which expands the value provided to the Hz range without taking into
account the precise Hz setting. This causes the frequency of 266 MHz not ot
be found in the imx8mm_fracpll_tbl, since it is entered there with a
precise Hz value. This in turn causes the boot hang in SPL, as proper DDR
fracpll frequency cannot be determined.
Correct the value in imx8mm_fracpll_tbl to match the one expanded by
MHZ(266) macro, rounding it down to MHz range only.
Signed-off-by: Andrey Zhizhikin <andrey.zhizhikin@leica-geosystems.com> Cc: Stefano Babic <sbabic@denx.de> Cc: Fabio Estevam <festevam@gmail.com> Cc: "NXP i.MX U-Boot Team" <uboot-imx@nxp.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Simon Glass <sjg@chromium.org> Cc: Ye Li <ye.li@nxp.com> Fixes: 825ab6b406 ("driver: ddr: Refine the ddr init driver on imx8m") Reviewed-by: Fabio Estevam <festevam@gmail.com>
Tim Harvey [Mon, 3 May 2021 18:21:27 +0000 (11:21 -0700)]
imx: ventana: enable dm support for PCI and FEC ethernet
Enable driver model support for FEC ethernet which allows us to remove
the iomux and board_eth_init function. Replace the toggling of the ethernet
phy reset with dt configuration.
Enable driver model support for PCI which allows us to remove the
eth1000_initialize() call. Additionally enable PCI_INIT_R to scan for
PCI devices on init such as the e1000 that is present on the GW552x.
Convert board_pci_fixup to use dm callback and remove pcidisable env
variable which is not supported for DM_PCI and thus leave PCI always
enabled during init.
arm: imx: imx8mm: clock: make debug output more descriptive
Clock initialization functionality has ambitious debug messages, which are
printed out when failures are triggered during execution:
- Separate frequency table lookup functions have the the same output that
makes it impossible to understand which function failed and produced the
output
- PLL decoding routine has a generic debug statement printed, which does
not state the actual value failed to be found
Extend the output for both cases with prefixing table lookup functions
output with function name, and report the failed value in PLL decoding
routine.
Signed-off-by: Andrey Zhizhikin <andrey.zhizhikin@leica-geosystems.com> Cc: Stefano Babic <sbabic@denx.de> Cc: Fabio Estevam <festevam@gmail.com> Cc: "NXP i.MX U-Boot Team" <uboot-imx@nxp.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Simon Glass <sjg@chromium.org> Cc: Ye Li <ye.li@nxp.com> Reviewed-by: Fabio Estevam <festevam@gmail.com>
Oleh Kravchenko [Fri, 14 May 2021 21:18:31 +0000 (00:18 +0300)]
Add out4.ru O4-iMX-NANO board
Board designed for quick prototyping and has one microSD port,
2 Ethernet ports, 2 USB ports, I2C, SPI, CAN, RS-485, GPIO,
UART interfaces, and 2 RGB LEDs.
Signed-off-by: Oleh Kravchenko <oleg@kaa.org.ua> Cc: Stefano Babic <sbabic@denx.de>
Giulio Benetti [Sun, 16 May 2021 21:57:01 +0000 (23:57 +0200)]
ARM: dts: imxrt1020-evk: move all u-boot, dm-spl to imxrt1020-evk-u-boot.dtsi file
At the moment a lot of u-boot,dm-spl properties are present in board .dts
file but this is not correct since u-boot,dm-spl property is u-boot
specific and must be listed into the separate imrt1020-evk-u-boot.dtsi
file. So let's move every u-boot,dm-spl property present in
imxrt1020-evk.dts to imxrt1020-evk-u-boot.dtsi file.
Giulio Benetti [Sun, 16 May 2021 21:57:00 +0000 (23:57 +0200)]
ARM: dts: imxrt1050-evk: move all u-boot, dm-spl to imxrt1050-evk-u-boot.dtsi file
At the moment a lot of u-boot,dm-spl properties are present in board .dts
file but this is not correct since u-boot,dm-spl property is u-boot
specific and must be listed into the separate imrt1050-evk-u-boot.dtsi
file. So let's move every u-boot,dm-spl property present in
imxrt1050-evk.dts to imxrt1050-evk-u-boot.dtsi file.
projects / u-boot.git / log