While MMU_PGPROT is currently only functional for arm64 it can be
extended for other real platforms as well. It does not however make
sense for sandbox to enable and create dummy functions for. Make this
depend on !SANDBOX.
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
In Kconfig syntax, "!=" is a string and not numerical comparison. This
means that to check for a non-zero SPL_SIZE_LIMIT_PROVIDE_STACK value we
need to test that it is "> 0" rather than "!=" 0. This is because "0x0 >
0" is false while "0x0 != 0" is true.
Signed-off-by: Tom Rini <trini@konsulko.com>
We have a large number of library symbols that should not be prompted
for by the user really but rather selected by the platform (or SoC) as
needed. To start with however, make these depend on !COMPILE_TEST.
Signed-off-by: Tom Rini <trini@konsulko.com>
Anshul Dalal <anshuld@ti.com> says:
During the implementation of falcon mode for TI's K3 devices [1], I encountered
several limitations in regards to the current falcon mode support in U-Boot
especially in ensuring a secure boot flow.
Although the current implementation allows for loading of a signed fitImage as
the SPL payload, there are still a few edge cases that might allow bypassing the
verified boot path.
The following issues with current falcon mode need to be resolved:
1) No fallback:
We currently fallback to regular boot flow if falcon mode fails,
this might not be secure.
2) No arguments file:
We currently load a kernel file (which could be a raw image or FIT)
alongside an args file (usually the DT). The args file here doesn't have
any verification mechanism, so should be skipped altogether as the FIT can
contain the DT.
3) No access to env:
In ext and fat fs boot, currently we also reads the environment to get the
names of the kernel and the arg file. This should be disabled in secure
falcon flow as the env might not be secure.
4) No raw image boot:
Boot should fail when the kernel file is a raw kernel image, only FIT should
be allowed.
As per the recommendation of maintainers[2], I have decided to split the above
set of tasks into multiple patch series. This is the first one which fixes the
load order of kernel image and the args file in falcon mode. Along with some
minor cleanup.
[1]: https://lore.kernel.org/u-boot/20250603142452.2707171-1-anshuld@ti.com/
[2]: https://lore.kernel.org/u-boot/20250911172313.GT124814@bill-the-cat/
Link: https://lore.kernel.org/r/20250923124639.667718-1-anshuld@ti.com
Minor cleanup of spl_ext and spl_fat files, removing the outdated
CONFIG_SPL_LIBCOMMON_SUPPORT symbols similar to the commit 1847129025e0
("spl: mmc: Drop checks for CONFIG_SPL_LIBCOMMON_SUPPORT") and adding a
few extra failure reports.
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Anshul Dalal <anshuld@ti.com>
Load the kernel image before args in falcon mode to be consistent with
the load order for other boot media.
Signed-off-by: Anshul Dalal <anshuld@ti.com>
Load the kernel image before args in falcon mode to be consistent with
the load order for other boot media.
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Anshul Dalal <anshuld@ti.com>
Currently in falcon mode, the FS and raw mmc boot loads the args file
first followed by the kernel image whereas others load in the opposite
order. This inconsistency means falcon boot doesn't behave the same
across various boot media.
For example, in the case where the kernel file is a FIT with the kernel
image present alongside the dtb and the args file is another DT, which
DT should be picked? The one form the FIT or the one set by the args
file? Currently this depends entirely on how the boot media handles
falcon mode.
Therefore, this patch enforces the load order of the kernel image first
followed by the args file in FAT FS boot. So in the above example, the
args file would take precedence.
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Anshul Dalal <anshuld@ti.com>
DT binding doesn't mandate i2c-bus as required property because hub itself
doesn't need to have i2c connected.
It can be in standalone mode that only power regulator and reset should be
handled.
Or hub should be configured via spi interface.
Signed-off-by: Michal Simek <michal.simek@amd.com>
Reviewed-by: Marek Vasut <marek.vasut@mailbox.org>
SoCFPGA updates for v2025.10:
CI: https://source.denx.de/u-boot/custodians/u-boot-socfpga/-/pipelines/27762
This pull request brings a set of updates across SoCFPGA platforms
covering Agilex5, Agilex7, N5X, and Stratix10. The changes include:
* Agilex5 enhancements:
- USB3.1 enablement and DWC3 host driver support
- System Manager register configuration for USB3
- Watchdog timeout increase and SDMMC clock API integration
- dcache handling improvements in SMC mailbox path
- Enable SPL_SYS_DCACHE_OFF in defconfig
* Clock driver improvements:
- Introduce dt-bindings header for Agilex clocks
- Add enable/disable API and EMAC clock selection fixes
- Replace manual shifts with FIELD_GET usage
* DDR updates:
- IOSSM mailbox compatibility check
- Correct DDR calibration status handling
* Device tree changes:
- Agilex5: disable cache allocation for reads
- Stratix10: add NAND IP node
- Enable driver model watchdog
- Enable USB3.1 node for Agilex5
* Config cleanups:
- Simplify Agilex7 VAB defconfig
- Remove obsolete SYS_BOOTM_LEN from N5X VAB config
- Enable CRC32 support for SoCFPGA
- Increase USB hub debounce timeout
Overall this set improves reliability of DDR and cache flows,
adds missing USB and MMC features for Agilex5, and refines clock
and configuration handling across platforms.
This patch set has been tested on Agilex 5 devkit, and Agilex devkit.
When the data cache (dcache) is disabled, calling related
status functions can lead to compilation errors due to
undefined references.
Adding a !CONFIG_IS_ENABLED(SYS_DCACHE_OFF) check before
invoking dcache_status() (used in common/memsize.c:get_ram_size())
and mmu_status() (from arch/arm/include/asm/io.h).
Without this check, builds with dcache disabled will fail to compile.
Signed-off-by: Boon Khai Ng <boon.khai.ng@altera.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Rasmus Villemoes <ravi@prevas.dk> says:
This all started from me wondering "why does this standalone app end
up being so huge"? Oh, it essentially links in its own standard C
library to get strlen() and snprintf(). Which then led to asking "why
don't we export all those standard C functions when we have them
anyway?".
CI has chewed on these and seem happy - it was CI which told me about
the necessity of [1/9]: https://github.com/u-boot/u-boot/pull/813
Link: https://lore.kernel.org/r/20250919101002.568488-1-ravi@prevas.dk
The !CONFIG_PHY_AQUANTIA defines were already superfluous since
_exports.h does have a CONFIG_PHY_AQUANTIA, so the entries never
existed. In fact, it couldn't have worked, because the defines would
affect both occurences of the mdio_get_current_dev identifier in the
EXPORT_FUNC(mdio_get_current_dev, struct mii_dev *,
mdio_get_current_dev, void)
so the C code would end up containing four copies of
gd->jt->dummy = dummy
but struct jt_funcs would not and does not have any 'dummy' member.
Now that nothing in _exports.h refers to dummy(), remove the empty
function.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Reviewed-by: Tom Rini <trini@konsulko.com>
If neither CONFIG_BLOBLIST_FIXED NOR CONFIG_BLOBLIST_ALLOC is set,
currently CONFIG_BLOBLIST_SIZE_RELOC defaults to 0 except if
* CONFIG_ARM=y && CONFIG_EFI_LOADER=y && GENERATE_ACPI_TABLE=y.
A size of zero never makes sense for a bloblist.
When using QFW we need more than 64 KiB to host the ACPI table.
In this case CONFIG_BLOBLIST_ALLOC is used.
Set a reasonable default.
Remove the CONFIG_BLOBLIST_SIZE_RELOC in ARM QEMU defconfigs which are
not compatible with ACPI tables passed from QEMU.
Reported-by: Emil Renner Berthing <emil.renner.berthing@canonical.com>
Fixes: 6f9b015c138b ("common: Enable BLOBLIST_TABLES on arm")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
It is expected that bloblists are stored in high memory beyond 2 GiB.
We must not use int as data type for these addresses but phys_addr_t.
Fixes: f9ef9fb033d5 ("bloblist: Handle alignment with a void entry")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
spl_start_uboot is a board overridable function that switches to falcon
boot mode on return value of 0.
Though for SPI, the falcon boot mode was being enabled on return value
of 1 which is not the correct behaviour. Therefore this patch fixes it
to the expected boot flow.
Fixes: 14509a28aa20 ("spl: spi: Consolidate spi_load_image_os into spl_spi_load_image")
Signed-off-by: Anshul Dalal <anshuld@ti.com>
The SPI flash driver does not build without SPI support enabled.
Fixes: 4151f4f822bb ("spl: Rework and tighten some dependencies")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Acked-by: Peng Fan <peng.fan@nxp.com>
Acked-by: Anshul Dalal <anshuld@ti.com>
I had trouble booting some am335x boards (both
beagleboneblack and a custom board). SPL would start just fine, and
apparently load U-Boot proper, but it would hang when jumping to
U-Boot.
While debugging, I stumbled on this memcpy() which from code
inspection very much looked to have overlapping src and dst, and
indeed a simple printf revealed
calling memcpy(0x8087bf68, 0x8087bf80, 0xf7f8)
Now, it will always be with src > dst, our memcpy()
implementations "most likely" do forward-copying, and in the end it
turned out that this wasn't the culprit after all [*].
But to avoid me or others barking up the wrong tree in the future, and
because this use of memcpy() is technically undefined, use memmove()
instead.
[*] That was 358d1cc232c ("spl: Align FDT load address"), which has
since been fixed in master but not the v2025.07 I worked of by
52caad0d14a ("ARM: Align image end to 8 bytes to fit DT alignment").
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@gmx.de>
The CONFIG_SYS_EARLY_PCI_INIT symbol is currently not supported
by Kconfig.
Make it a Kconfig symbol so that users could select it via defconfig.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Explicitly specify the type by replacing macro with variable
to fix the possible compilation warning.
Signed-off-by: Leo Yu-Chi Liang <ycliang@andestech.com>
Explicitly specify the type by replacing macro with variable
to fix the possible compilation warning.
Signed-off-by: Leo Yu-Chi Liang <ycliang@andestech.com>
Lukasz Czechowski <lukasz.czechowski@thaumatec.com> says:
This patch series extends the usb_onboard_hub driver to allow for
support of more types of onboard hub devices, and adds the Cypress HX3
hub family.
First patch in the series updates the bind function, so that it no
longer uses hardcoded compatible strings.
Next patch simplifies the code, by removing unnecessary dm_gpio function
call.
Third patch updates the remove function, which allows the prevent issues
with usb devices reenumeration, in case of calling "usb reset". Although
the issue could still occur in case of invalid initial state of reset
gpio, it is minimized with no impact on main usb_hub driver.
Fourth patch extends the driver with support for multiple power
supplies, the same way it is done in kernel driver. Finally, last patch
provides hub data and of_match table entries for Cypress HX3
Link: https://lore.kernel.org/r/20250722-usb_onboard_hub_cypress_hx3-v4-0-91c3ee958c0e@thaumatec.com
The HX3 is a family of USB3.0 hub controllers that comes in
different variants: CYUSB330x/CYUSB331x/CYUSB332x/CYUSB230x.
To support this hub, controlling of reset pin and two
power supplies is required.
The reset time is set to 10ms, based on the datasheet [1].
Power-on delay time is not required, so it is set to 0.
The compatible strings added to of_match table are
compliant with usb/cypress,hx3.yaml bindings.
[1] https://www.infineon.com/dgdl/Infineon-HX3_USB_3_0_Hub_Consumer_Industrial-DataSheet-v22_00-EN.pdf?fileId=8ac78c8c7d0d8da4017d0ecb53f644b8
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Lukasz Czechowski <lukasz.czechowski@thaumatec.com>
Some of the onboard hubs require multiple power supplies, so extend
the driver to support them.
The implementation is inspired by the kernel driver, as introduced
by commit [1] in the v6.10 kernel.
[1] ec1848cd5d
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Lukasz Czechowski <lukasz.czechowski@thaumatec.com>
In the usb_onboard_hub_remove, the reset gpio, if available, is
freed. The pin state however, remains unchanged, as set in the
usb_onboard_hub_reset. The hub is then left enabled.
During second onboard hub probing, the hub is initially enabled
(reset pin in state "0"), and then it is being reset by the reset
function (transition to "1" and then to "0"). Because of this,
the hub first disconnects from root hub, and then it connects again.
When the devices are being discovered in the usb_scan_port in the
usb_hub driver, initially there is the USB_PORT_STAT_CONNECTION bit
not set in portstatus, but USB_PORT_STAT_C_CONNECTION set in
portchange data (which is because disconnect event occurred first).
In this condition, the driver does not wait for devices to appear.
This can cause the hub (and all child devices) to be not enumerated
when rescanning on "usb reset" command.
To fix this, set the reset gpio to active in usb_onboard_hub_remove,
to put the hub into reset state.
However, in case the hub reset pin is by default held in high
state by HW before U-Boot takes over, in which case the USB hub is
active, then during the first probe it gets reset and we might get
into the same issue of the hub being not enumerated.
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Lukasz Czechowski <lukasz.czechowski@thaumatec.com>
The reset GPIO is obtained during driver probing by the function
devm_gpiod_get_optional, which means the GPIO will be automatically
freed when the device is removed. Because of this, explicit call
to free the reset GPIO in hub remove function is not needed.
To support the Managed device resources, the DEVRES config must
be enabled, otherwise the devres functions fall back to non-managed
variants. Set the necessary dependency to DEVRES in Kconfig.
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Lukasz Czechowski <lukasz.czechowski@thaumatec.com>
Currently the check in usb_onboard_hub_bind is relying on specific
compatible string for the Michrochip USB5744. Replace this with
more generic approach that will allow to add new types of devices
to the of_match table. Because the driver only needs to bind one
"half" of the hub, the peer-hub node is used to find out if it
was already done. In case peer-hub was bound, -ENODEV is returned.
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Lukasz Czechowski <lukasz.czechowski@thaumatec.com>
In boot_from_devices the variable loader is not NULL checked after
assignment and before first use but later code does check it for NULL.
Add a NULL check before first use.
This issue was found by Smatch.
Signed-off-by: Andrew Goodbody <andrew.goodbody@linaro.org>
Add the basic Kconfig options, addresses and other values for the
existing Kconfig settings for the new Allwinner A523/T527/H728 SoC.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
The 'overhead' variable is uninitialized and actually shall not be used.
Delete it to fix coverity CID 37041718 - Uninitialized scalar variable.
Fixes: 73c40fcb7367 ("spl: Refactor spl_load_info->read to use units of bytes")
Reported-by: Andrew Goodbody <andrew.goodbody@linaro.org>
Signed-off-by: Ye Li <ye.li@nxp.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Andrew Goodbody <andrew.goodbody@linaro.org>
To support PQC container format which is used for post quantum
authentication on new i.MX parts like i.MX94
The major changes compared to legacy container format is in
signature block, new container tag and version, and new alignment
of container header.
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Jacky Bai <ping.bai@nxp.com>
Signed-off-by: Alice Guo <alice.guo@nxp.com>
Acked-by: Peng Fan <peng.fan@nxp.com>
When printing the contents of an lbaint_t variable we need to use LBAF
to print it in order to get the correct format type depending on 32 or
64bit-ness.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Link: https://lore.kernel.org/r/20250702010603.19354-2-trini@konsulko.com
Signed-off-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Originally when bootdelay >99, the countdown breaks. Setting bootdelay
at 101 or bigger will trigger this error.
Signed-off-by: David Zang <davidzangcs@gmail.com>
We have a few options that we cannot enable in a "allyesconfig" type
build because we cannot use zero as a default value.
- The logic around HAS_BOARD_SIZE_LIMIT assumes that if we have set this
then we compare with it. Similarly, we need to set SPL_NO_BSS_LIMIT as
the default there.
- Both SYS_CUSTOM_LDSCRIPT and ENV_USE_DEFAULT_ENV_TEXT_FILE then prompt
for a file name to use.
- The SYS_I2C_SOFT driver is a legacy driver which requires a lot of
configuration within the board config. file instead, so disable it.
Signed-off-by: Tom Rini <trini@konsulko.com>
This symbol is not something that the user should be enabling or
disabling but rather the developer for a particular board should select
it when required.
This is mostly size neutral, however a few places do have changes. In
the case of i.MX6ULL systems, it is always the case that
arch_misc_init() could call setup_serial_number() and do useful work,
but was not enabled widely, but now is. In the case of i.MX23/28
systems, we should be able to call mx28_fixup_vt() again here, so do so.
Finally, some platforms were calling arch_misc_init() and then not doing
anything and this results in removing the option.
Acked-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Add Renesas R-Car Gen4 SCIF/HSCIF loader compatible SREC generation.
This is a regular U-Boot SPL SREC augmented with a short header which
describes where to store the received data and how much data to store.
This header is interpreted by the R-Car Gen4 BootROM SCIF/HSCIF loader.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
This file was making environment calls without including <env.h> and so
relying on an indirect inclusion from elsewhere. Add the missing include
directly.
Signed-off-by: Tom Rini <trini@konsulko.com>
The variable base is unsigned so >= 0 is always true. Fix this test
so that it is actually useful. The fix prevents the code from causing
a segfault in the case where Ctrl-w is pressed on a line consisting
only of spaces.
Fixes: dcc18ce0dbaf ("cli: Implement delete-word in cread_line()")
Signed-off-by: Andrew Goodbody <andrew.goodbody@linaro.org>
Heinrich Schuchardt <heinrich.schuchardt@canonical.com> says:
spl_fit_get_image_name() used to lack a detection of malformed image name
properties in FIT images. The change in commit 3704b888a4ca ("common/spl:
fix potential out of buffer access in spl_fit_get_image_name function")
tried to fix this but led to function spl_fit_get_image_name() no longer
detecting if a property at index > 1 does not exist.
This patch is reverted.
An explicit check for malformed image name properties is introduced.
Link: https://lore.kernel.org/u-boot/38f5d078-3328-4bdb-9c95-4fb5fe89ddc2@gmx.de/T/#u
Link: https://lore.kernel.org/r/20250624153431.46986-1-heinrich.schuchardt@canonical.com
A malformed FIT image could have an image name property that is not NUL
terminated. Reject such images.
Reported-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Tested-by: E Shattow <e@freeshell.de>
The change in commit 3704b888a4ca ("common/spl: fix potential out of buffer
access in spl_fit_get_image_name function") led to function
spl_fit_get_image_name() no longer detecting if a property does not exist
at a non-zero buffer.
Link: https://lore.kernel.org/u-boot/38f5d078-3328-4bdb-9c95-4fb5fe89ddc2@gmx.de/T/#m59f3a23e675daa992c28d12236de71cae2ca2bb9
Fixes: 3704b888a4ca ("common/spl: fix potential out of buffer access in spl_fit_get_image_name function")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Tested-by: E Shattow <e@freeshell.de>
The Allwinner A100 SoC has been around for a while, mostly on cheap
tablets, but didn't generate much interest in the community so far.
There were some efforts by two Allwinner employees in 2020, which led
to basic upstream Linux support for that SoC, although this momentum
dried up pretty quickly, leaving a lot of peripherals unsupported.
The A100 was silently replaced with the seemingly identical Allwinner
A133, which is reportedly a better bin of the A100. So far we assume
that both are compatible from a software perspective. There are some
more devices with the A133 out there now, so people are working on
filling the gaps, and adding U-Boot (and TF-A) support.
Based on the just added pinctrl, clock and DRAM support, this adds the
missing bits, mostly addresses and values for the SPL.
The A133 seems to be an predecessor to the H6, so we can share a lot of
code with that (and the H616 code), and just need to adjust some details.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
-----BEGIN PGP SIGNATURE-----
iQGzBAABCgAdFiEEGjx/cOCPqxcHgJu/FHw5/5Y0tywFAmhZtLAACgkQFHw5/5Y0
tyyl/gv+LLsF/PdBp93gXaUVFQwh3lrAWltjK6n+gKDpIazlX8Ajiu2tFBzYgBdP
JNp3PsHJ8ER/GpU4nG/13KfWg8c4CGvAqY0zLxKufMckBdZAE5EwztvSNzZKZA6f
5FQCecidpwlWBNZRWVbxAAtl83nGBwUmgPAoxDBv4I7DQBJD7mJfzxVnWZuwSioP
ix3aingjRSIUpldHR5IAQfEmnPPfv7IS6EYb2lOoV22WtTfnF6bpteR/2OUxnW6z
YxXHaZrDkIfE+gXYOQu5Tau8R0ca4l+ws8GaocTD9VVCDem8JU7b0GCfoXg+ry2Q
dLYuROREeiXEi14I6arZznpPQM8iIAeDOy0w5b3qXbav0+GWB4rmJKLMniD5JT2r
r8Kle/QeagILMVPOaUVgjmkEnft4thD14ipvr8kUCurUh1VtFYWTGzKVS47yHaxY
I1zQaz1NNdtyQNWO/Nwbs+88C7v8q7DeLIIjUt2/J01Bu8ouf2WyfG04iBeZOitt
CxBZ++2z
=e4ez
-----END PGP SIGNATURE-----
Merge tag 'v2025.07-rc5' into next
Prepare v2025.07-rc5
With this merge, tighten up the LTO_FLAGS removal we added to not
trigger on ARMv7 (which is Thumb-2 and should be fine).
Commit 6fe50e395080 ("arm: asm/system.h: mrc and mcr need .arm if
__thumb2__ is not set") is not a proper fix for the LTO link error
mentioned in its description. It causes 32-bit arm instructions to be
mixed with thumb instructions, which the Kirkwood SoCs do not support.
For example, board_init_r() is mostly generated in Thumb-1 mode as
expected since the build flags contain -mthumb -mthumb-interwork. The
MCR instruction corresponding to writefr_extra_feature_reg() is also
correcly emitted as a 32-bit ARM instruction (it cannot be encoded in
Thumb-1 anyways). The problem is, the compiler inlines the MCR without
generating the BX or BLX instruction which are needed to transition
between the ARM and the Thumb-1 states. From the objdump output:
006186a0 <board_init_r>:
board_init_r():
/home/jerome/work/u-boot/common/board_r.c:799
6186a0: b5f0 push {r4, r5, r6, r7, lr}
6186a2: b0ab sub sp, #172 @ 0xac
get_gd():
/home/jerome/work/u-boot/./arch/arm/include/asm/global_data.h:127
6186a4: 464a mov r2, r9
...
/home/jerome/work/u-boot/arch/arm/mach-kirkwood/cpu.c:242
619aae: 9b15 ldr r3, [sp, #84] @ 0x54
writefr_extra_feature_reg():
/home/jerome/work/u-boot/./arch/arm/include/asm/arch/cpu.h:100
619ab0: ee2f3f11 mcr 15, 1, r3, cr15, cr1, {0}
^^^^^^^^
32-bit ARM instruction
Further investigation is needed to understand how to fix the issue so
that the code size is minimal for all boards. In the mean time, this
fix disables LTO for the two problematic files (common/board_f.c and
common/board_r.c). This makes the Kirkwood-based boards bootable again.
The binary size is increased by 1048 bytes which is perfectly
acceptable.
Fixes: 6fe50e395080 ("arm: asm/system.h: mrc and mcr need .arm if __thumb2__ is not set")
Reported-by: Tony Dinh <mibodhi@gmail.com>
Tested-by: Tony Dinh <mibodhi@gmail.com>
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Marek Vasut <marek.vasut+renesas@mailbox.org> says:
Rename the environment related variables and add ENV_ prefix, so that
all configuration options which are related to environment would have
an CONFIG_ENV_ prefix. No functional change.
Link: https://lore.kernel.org/r/20250609192701.20260-1-marek.vasut+renesas@mailbox.org
There are no users of DELAY_ENVIRONMENT and the same effect can
be achieved either using DT /config/load-environment property,
or by using ENV_IS_NOWHERE . Remove this configuration option
and matching functionality.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
