When ROM boots from secondary container set, SPL should select
correct offset to load u-boot-atf container.
The implementation uses ROM passover information:
1) For non-eMMC boot partition device, use image offset in ROM
passover data to get u-boot-atf container offset.
2) For eMMC boot partition device, use boot stage (secondary)
in ROM passover data to select correct eMMC boot partition
for u-boot-atf container.
Signed-off-by: Ye Li <ye.li@nxp.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Implement the 'getprisec' subcommand of 'bmode' command for i.MX95 by
reading out the ROM log events. This event is set by the BootROM if it
switched to the secondary copy due to primary copy being corrupted.
Signed-off-by: Fedor Ross <fedor.ross@ifm.com>
Reviewed-by: Marek Vasut <marex@nabladev.com>
Query the SM via SCMI, obtain rom_passover_t->img_set_sel and based on
that, add 0 or 0x400000 offset (A or B copy offset) to boot container
read address.
Signed-off-by: Marek Vasut <marex@nabladev.com>
Signed-off-by: Fedor Ross <fedor.ross@ifm.com>
imx8m[m,n]-beacon-baseboard.dtsi was missed to be deleted in
commit f5585124c9 ("arm64: imx: imx8mm-beacon: Migrate to OF_UPSTREAM")
commit a64feb974f ("arm64: imx: imx8mn-beacon: Migrate to OF_UPSTREAM")
arch/arm/dts/imx8mn-evk.dtsi was missed to be deleted in
commit 73d57e0aa4 ("imx: imx8mn-evk: convert to OF_UPSTREAM")
Drop them.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
scripts/Makefile.lib already handles the including path for imx8mq.dtsi
from dts/upstream. No need to keep a copy in arch/arm/dts/, and there
is very minimal changes compared with the one in dts/upstream, so remove
the copy.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
arch/arm/dts/imx8mq-kontron-pitx-imx8m.dts is almost same as upstream Linux
dts, so switch to OF_UPSTREAM by dropping the U-Boot copy of the dts,
enabling OF_UPSTREAM and updating CONFIG_DEFAULT_DEVICE_TREE.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
arch/arm/dts/imx8mq-librem5-r[4,3].dts is almost same as upstream Linux dts,
and arch/arm/dts/imx8mq-librem5.dtsi is out of sync with upstream linux
dts, but it should not break U-Boot after using OF_UPSTREAM. So switch to
OF_UPSTREAM by dropping the U-Boot copy of the dts, enabling OF_UPSTREAM and
updating CONFIG_DEFAULT_DEVICE_TREE.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
arch/arm/dts/imx8mq-pico-pi.dts is almost same as upstream Linux dts,
so switch to OF_UPSTREAM by dropping the U-Boot copy of the dts,
enabling OF_UPSTREAM and updating CONFIG_DEFAULT_DEVICE_TREE.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
arch/arm/dts/imx8mq-phanbell.dts is almost same as upstream Linux dts,
so switch to OF_UPSTREAM by dropping the U-Boot copy of the dts,
enabling OF_UPSTREAM and updating CONFIG_DEFAULT_DEVICE_TREE.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
arch/arm/dts/imx8mq-mnt-reform2.dts are almost same as upstream Linux
imx8mq-mnt-reform2.dts, so switch to OF_USPTREAM for this board, with
only updating imx8mq-mnt-reform2-u-boot.dtsi to keep "simple-panel"
compatible string for display panel.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Following compile error happens for mx6qsabrelite when disable
CONFIG_VIDEO_IPUV3.
board/boundary/nitrogen6x/nitrogen6x.c: In function 'misc_init_r':
board/boundary/nitrogen6x/nitrogen6x.c:912:22: error: 'RGB_BACKLIGHT_GP' undeclared (first use in this function)
912 | gpio_request(RGB_BACKLIGHT_GP, "lvds backlight");
| ^~~~~~~~~~~~~~~~
board/boundary/nitrogen6x/nitrogen6x.c:912:22: note: each undeclared identifier is reported only once for each function it appears in
CC cmd/bind.o
CC drivers/gpio/gpio-uclass.o
CC boot/bootmeth_extlinux.o
board/boundary/nitrogen6x/nitrogen6x.c:913:22: error: 'LVDS_BACKLIGHT_GP' undeclared (first use in this function)
913 | gpio_request(LVDS_BACKLIGHT_GP, "lvds backlight");
| ^~~~~~~~~~~~~~~~~
AR arch/arm/lib/lib.a
make[1]: *** [scripts/Makefile.build:271: board/boundary/nitrogen6x/nitrogen6x.o] Error 1
CC boot/bootmeth_pxe.o
make: *** [Makefile:2205: board/boundary/nitrogen6x] Error 2
make: *** Waiting for unfinished jobs....
To fix this, use reported macros included in CONFIG_VIDEO_IPUV3.
Fixes: 1b51e5f4cd ("nitrogen6x: reserve used gpios")
Signed-off-by: Jacky Cao <Jacky.Cao@sony.com>
Reviewed-by: Simon Gaynor <simon.gaynor@ezurio.com>
Reduce the NXP MAINTAINERS entry by using N: entry glob.
Signed-off-by: Marek Vasut <marex@nabladev.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
This driver depends on DM_I2C and DM_SPI, add it.
Fixes: 3b639f6438 ("gpio: mcp230xx: Add support for models with SPI interface.")
Signed-off-by: Charles Perry <charles.perry@microchip.com>
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
This patch removes the static MDIO pinmux configuration from
rmii1_pin_mux[] and instead configures the MDIO pins conditionally
during board_init(). Previously, the MDIO_CLK and MDIO_DATA pins
were always configured for CPSW in mux.c, which could lead to
unnecessary pin ownership and conflicts in scenarios where CPSW
is not used.
With this change, the MDIO pins are configured only when required,
ensuring that CPSW Ethernet functionality in U-Boot remains unaffected.
This approach keeps Ethernet boot behavior intact and provides cleaner
separation between CPSW and other Ethernet use cases.
Reviewed-by: Markus Schneider-Pargmann (TI) <msp@baylibre.com>
Signed-off-by: Parvathi Pudi <parvathi@couthit.com>
Currently the oscillator stop flag (OSF) bit is never checked or cleared
on the DS1339 RTC chip.
On getting the time from the RTC, check if the OSF bit is set, log a
warning, and clear the flag. This matches the behavior of the DS1337
chip.
Note that the `date` command always reads from the RTC even when
setting or resetting the date, so the OSF flag is cleared in those cases
as well.
Signed-off-by: Ronan Dalton <ronan.dalton@alliedtelesis.co.nz>
Cc: Tom Rini <trini@konsulko.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Francesco Dolcini <francesco.dolcini@toradex.com>
Cc: Mark Tomlinson <mark.tomlinson@alliedtelesis.co.nz>
Cc: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Simon Glass <sjg@chromium.org>
Simon Glass <sjg@chromium.org> says:
Several boards use 'test -n ${var}' (unquoted) in env scripts to detect
an empty variable and trigger a fallback or skip a hook. That works only
because of a U-Boot 'test' quirk where calls with argc < 3 returned
false; an empty variable made the expression 'test -n' (one operand) and
hit that path.
Commit 8b0619579b ("cmd: test: fix handling of single-argument form of
test") aligned 'test' with POSIX so those expressions flipped to true.
f7e7c55e53 ("cmd: test: add bug-compatibility special case for 'test
-n'") restored the old behaviour for the exact 'test -n' (one arg) case,
so the boards are not broken at runtime today, but the reliance on a
bug-compat path isn't great.
This series updates various boards to quote each affected variable so
the emptiness check is explicit and does not depend on the
special-case path.
Each commit carries a Fixes: tag pointing at 8b0619579b
Link: https://lore.kernel.org/r/20260503204936.3151124-1-sjg@chromium.org
The rescueboot script optionally runs a per-board rescue_reason hook
with:
if test -n ${rescue_reason}; then run rescue_reason; fi;
The default state is "no rescue reason script", i.e. rescue_reason
unset. The expression then expands to 'test -n' with no operand and
relies on a U-Boot 'test' quirk that treats a missing operand as
false to skip the run.
Quote the variable so an unset rescue_reason expands to 'test -n ""'
and the emptiness check is explicit.
Fixes: 8b0619579b ("cmd: test: fix handling of single-argument form of test")
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@nabladev.com>
The mccmon6 bootcmd starts with:
if test -n ${recovery_status}; then run boot_recovery; ...
The default state is "no recovery requested", i.e. recovery_status
unset. The expression then expands to 'test -n' with no operand and
relies on a U-Boot 'test' quirk that treats a missing operand as
false to skip recovery.
Quote the variable so an unset recovery_status expands to 'test -n ""'
and the emptiness check is explicit.
Fixes: 8b0619579b ("cmd: test: fix handling of single-argument form of test")
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Łukasz Majewski <lukma@nabladev.com>
On beaglev_fire, design_overlays gates an overlay-application loop
on:
if test -n ${no_of_overlays}; then ...
The default state is "no overlays", i.e. no_of_overlays unset. The
expression then expands to 'test -n' with no operand and relies on a
U-Boot 'test' quirk that treats a missing operand as false to skip
the loop.
Quote the variable so an unset no_of_overlays expands to 'test -n ""'
and the emptiness check is explicit.
Fixes: 8b0619579b ("cmd: test: fix handling of single-argument form of test")
Signed-off-by: Simon Glass <sjg@chromium.org>
The Siemens am33x-common, env-common and draco-etamin headers gate
boot-partition selection logic on:
if test -n ${A}; then ...
if test -n ${B}; then ...
A and B are flags that the upgrade machinery sets to mark "the other
partition just became active". The default state is unset, in which
case the expression expands to 'test -n' with no operand and relies
on a U-Boot 'test' quirk that treats a missing operand as false to
skip the branch.
Quote each variable so an unset A or B expands to 'test -n ""' and
the emptiness check is explicit.
Fixes: 8b0619579b ("cmd: test: fix handling of single-argument form of test")
Signed-off-by: Simon Glass <sjg@chromium.org>
bootcmd_ti_mmc skips a per-board init hook with:
if test -n ${board_init}; then run board_init; fi;
The default case is "no board override", i.e. board_init unset. The
expression then expands to 'test -n' with no operand and relies on a
U-Boot 'test' quirk that treats a missing operand as false to skip the
run.
Quote the variable so an unset board_init expands to 'test -n ""' and
the emptiness check is explicit.
Fixes: 8b0619579b ("cmd: test: fix handling of single-argument form of test")
Signed-off-by: Simon Glass <sjg@chromium.org>
Rasmus Villemoes <ravi@prevas.dk> says:
There are quite a few places where we allocate X+1 bytes, initialize
the first X bytes via memcpy() and then set the last byte to 0.
The kernel has a helper for that, kmemdup_nul(). Introduce a similar
one, and start making use of it in a few places. Also the existing
memdup() helper can be put to more use.
There are lots more places one could modify. But for code shared with
host tools, one would need to do some refactoring, putting memdup()
and memdup_nul() in their own str-util.c TU which could then also be
included in the tools build.
Link: https://lore.kernel.org/r/20260421075439.16696-1-ravi@prevas.dk
Use memdup() instead of open-coding it.
In the dm_setup_inst() case, there was never any reason to use
calloc(), as the whole allocation is definitely initialized via the
immediately following memcpy().
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
We have memdup_nul() for exactly this pattern of duplicating a block
of memory and ensuring there's a nul byte after the copy.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
With the addition of memdup_nul(), strdup() and strndup() can be
implemented as one-liners.
While not required by POSIX or C, do keep the behaviour of gracefully
accepting a NULL source and simply return NULL.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
This is completely analogous to the linux kernel's kmemdup_nul()
helper, apart from the lack of the gfp_t argument: Allocate a buffer
of size {len}+1, copy {len} bytes from the given buffer, and add a
final nul byte.
This pattern exists in a number of places, so this helper can reduce
some boilerplate code.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
There has never been an arch-specific optimized implementation of
str[n]dup, nor is there likely to ever be one, because unlike their
cousins strlen(), strcpy() and similar that simply read/write the
src/dst, the dup functions by definition involve memory allocation. So
drop this irrelevant cpp guard.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
The helper stdio_clone only has a single caller, so it certainly
doesn't need to be public. But in fact, it is merely an open-coded
memdup() - which for some reason uses calloc() even if the whole
allocation is obviously immediately overwritten.
Drop it and just use memdup() directly.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
It doesn't make sense to restrict memdup() to only return char*
pointers, especially when it is already defined to accept void*. This
makes it uglier to use to e.g. duplicate a struct.
Make it return void*, just as kmemdup() does in the kernel (and which
our kmemdup() in fact also does).
While in here, make a small optimization: memcpy() is defined to
return the destination register, so we write this in a way that the
compiler may do a tail call.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Add the workaround proposed for J784S4 errata i2437 (link) for SE
clock-gating turning off too early. Without this, a hardware bug present
in C7120 leads to C7120 CPU hanging.
Link: https://www.ti.com/lit/pdf/sprz536
Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
Signed-off-by: Udit Kumar <u-kumar1@ti.com>
Add the workaround proposed for J721S2 errata i2437 (link) for SE
clock-gating turning off too early. Without this, a hardware bug present
in C7120 leads to C7120 CPU hanging.
Link: https://www.ti.com/lit/pdf/sprz530
Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
Signed-off-by: Udit Kumar <u-kumar1@ti.com>
Add a helper macro to write and verify a 32-bit value to a memory-mapped
register. This is essential for hardware errata workarounds that require
confirmation that register writes have taken effect before proceeding with
initialization.
Signed-off-by: Udit Kumar <u-kumar1@ti.com>
When encountering a symlink pointing to an absolute path, ubifs_findfile
would return the target of the symlink as the result instead of resolving
any following components in the original path. Fix it by following the
same code path that is used for relative paths except that we set the
next inode to the root if we see a leading slash.
The existing code used memcpy and sprintf to copy the symlink target
into a fixed size stack buffer and was therefore vulnerable to buffer
overflows with a sufficiently long symlink target. Fix it by using a
heap buffer for the temporary path during path resolution.
Signed-off-by: Peter Collingbourne <peter@pcc.me.uk>
Fixes: 9d7952e4c6 ("ubifs: Add support for looking up directory and relative symlinks")
The PKCS#11 signing tests in ftest.py call tools.run('softhsm2-util', ...)
directly (and the equivalent for pkcs11-tool and p11-kit), even though
the test setup has already constructed the corresponding Bintool
instances. As Quentin Schulz observed on v1, the bintool wrapper for
these tools is currently used only as an "is this installed?" probe.
Route the eight remaining call sites in ftest.py through
<bintool>.run_cmd(...), which the Bintool base class already provides.
The change is test-side only; no production binman code calls these
tools.
Suggested-by: Quentin Schulz <quentin.schulz@cherry.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
The Sphinx-generated bintools.rst currently produces an empty section
for this bintool, since its class docstring is only a single line and
so the body under the heading is blank.
Extend the docstring with a short description of what softhsm2-util
does and how binman uses it, so the generated documentation has useful
content.
Suggested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
The Sphinx-generated bintools.rst currently produces an empty section
for this bintool, since its class docstring is only a single line and
so the body under the heading is blank.
Extend the docstring with a short description of what pkcs11-tool does
and how binman uses it, so the generated documentation has useful
content.
Suggested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
The Sphinx-generated bintools.rst currently produces an empty section
for this bintool, since its class docstring is only a single line and
so the body under the heading is blank.
Extend the docstring with a short description of what p11-kit does and
how binman uses it, so the generated documentation has useful content.
Suggested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
When using a worktree '.git' actually is a file and pygit2
init_repository() does not like that. The intention is not to create
a new git repo anyway so use normal constructor instead of
init_repository().
Signed-off-by: Ludwig Nussel <ludwig.nussel@siemens.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
When creating the RSDT and the XSDT table they contain no entries.
The table size therefore must equal the header size.
Without this change a NULL deference has been observed in
acpi_find_table() when running `ut dm` on sandbox64_defconfig
executed via `sudo ./u-boot -D`.
Fixes: 94ba15a3f1 ("x86: Move base tables to a writer function")
Fixes: 7e586f6907 ("acpi: Put table-setup code in its own function")
Fixes: ab5efd576c ("x86: acpi: Adjust order in acpi_table.c")
Fixes: 867bcb63e7 ("x86: Generate a valid ACPI table")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Update dm_test_acpi_ctx_and_base_tables() in test/dm/acpi.c to expect
sizeof(struct acpi_table_header) for the initial table length (instead
of sizeof(*rsdt) / sizeof(*xsdt)), and to compute the checksum over
header->length bytes rather than the full struct size:
Signed-off-by: Simon Glass <sjg@chromium.org>
The dwc3_free_one_event_buffer() function incorrectly called free()
on event buffer structures allocated with devm_kzalloc(). This
caused heap corruption and a synchronous abort when exiting
fastboot mode via "fastboot continue".
Device-managed memory is automatically freed when the device is
removed, so manual deallocation causes the heap allocator to access
corrupted metadata.
Fixes: 884b10e86a ("usb: dwc3: core: fix memory leaks in event buffer cleanup")
Signed-off-by: Balaji Selvanathan <balaji.selvanathan@oss.qualcomm.com>
Reviewed-by: Marek Vasut <marek.vasut+usb@mailbox.org>
The symbol USB_EMUL is how sandbox has access to USB. It's
implementation however enforces a few other requirements. It must have
SCSI enabled, and in turn that means it must have BLK enabled. Finally,
we should not be using SANDBOX itself as a symbol to decide what to
build or not build here, as SANDBOX is selected for COMPILE_TEST builds
as well and so may not have enabled the sandbox specific USB support.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Marek Vasut <marek.vasut+usb@mailbox.org>
As exposed by "make randconfig", we have an issue with the dependencies
for RESET_RZG2L_USBPHY_CTRL. As this functionally depends on
REGULATOR_RZG2L_USBPHY, express this dependency directly in Kconfig as
well.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Ye LiandFabio Estevam
Fedor RossandFabio Estevam
Marek VasutandFabio Estevam
Peng FanandFabio Estevam
Jacky CaoandFabio Estevam
Charles PerryandTom Rini
Parvathi PudiandTom Rini
Ronan DaltonandTom Rini
Simon GlassandTom Rini
Rasmus VillemoesandTom Rini
Neha Malcom FrancisandTom Rini
Udit KumarandTom Rini
Peter CollingbourneandTom Rini
Ludwig NusselandSimon Glass
Heinrich SchuchardtandSimon Glass
Balaji SelvanathanandMarek Vasut