Tom Rini [Wed, 31 Dec 2025 17:13:41 +0000 (11:13 -0600)]
CI: Add "allyesconfig" to one of the build jobs
Now that we can have "make allyesconfig" build and link, add this type
of build to the job which builds host tools as well. In GitLab, make
this job rather than binman testsuite be the job which unblocks the next
stage of the pipeline. This is because we had been using that job for
"sandbox builds", and now that we have an explicit test for that, we
should use it.
Marek Vasut [Fri, 19 Dec 2025 02:40:44 +0000 (03:40 +0100)]
kbuild: Produce diff between base DT and U-Boot augmented DT if DEVICE_TREE_DEBUG=1
In case DEVICE_TREE_DEBUG is set, produce a diff between the base DT and
DT with U-Boot extras, to show how much does the U-Boot DT differ from
the base DT. This is particularly useful together with OF_UPSTREAM, to
minimize the diff between upstream DTs and U-Boot DTs.
This requires DTC 1.7.2 which does not resolve phandles when used in
the 'dtc -I dts -O dts ...' mode. With older DTC, the diff is full of
churn due to the resolved phandles.
Example usage:
$ make r8a779g3_sparrowhawk_defconfig && make DEVICE_TREE_DEBUG=1
$ cat ./dts/upstream/src/arm64/renesas/r8a779g3-sparrow-hawk.dtb.diff
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Synchronize local copy of DTC with Linux 6.17 . This includes the
following picked and squashed commits from Linux kernel. The squash
was necessary, since the DTC here contains changes which were also
part of DTC in Linux alraedy, and the squash helped resolve those
without going back and forth with the changes.
Patrice Chotard [Thu, 18 Dec 2025 17:27:02 +0000 (18:27 +0100)]
clk: scmi: Fix priv initialization in scmi_clk_gate()
In scmi_clk_probe(), in case of CLK_CCF is not enabled, parent private
data is not set, so in scmi_clk_gate(), an uninitialized priv struct is
retrieved.
SCMI request is performed either using scmi_clk_state_in_v1 or
scmi_clk_state_in_v2 struct depending of the unpredictable value of
priv->version which leads to error during SCMI clock enable.
Issue detected on STM32MP157C-DK2 board using the SCMI device tree
stm32mp157c-dk2-scmi.dts.
Fixes: 0619cb32030b ("firmware: scmi: Add clock v3.2 CONFIG_SET support") Reviewed-by: Peng Fan <peng.fan@nxp.com> Reviewed-by: Marek Vasut <marek.vasut+renesas@mailbox.org> Signed-off-by: Patrice Chotard <patrice.chotard@foss.st.com> Signed-off-by: Peng Fan <peng.fan@nxp.com>
The original _bin2bcd() function used / 10 and % 10 operations for
conversion. Although GCC optimizes these operations and does not generate
division or modulus instructions, the new implementation reduces the
number of mov instructions in the generated code for both x86-64 and ARM
architectures.
This optimization calculates the tens digit using (val * 103) >> 10, which
is accurate for values of 'val' in the range [0, 178]. Given that the
valid input range is [0, 99], this method ensures correctness while
simplifying the generated code.
Link: https://lkml.kernel.org/r/20240812170229.229380-1-visitorckw@gmail.com Signed-off-by: Kuan-Wei Chiu <visitorckw@gmail.com> Cc: Ching-Chun Huang (Jim) <jserv@ccns.ncku.edu.tw> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
[visitorckw@gmail.com: Adapt to bin2bcd() in include/bcd.h] Signed-off-by: Kuan-Wei Chiu <visitorckw@gmail.com>
Marek Vasut [Wed, 17 Dec 2025 20:02:50 +0000 (21:02 +0100)]
configs: sandbox: Select environment in FAT FS support
Commit 2a38e712652f ("sandbox: add FAT to the list of usable env drivers")
made environment storage in FAT available on sandbox, but did not enable
the matching ENV_IS_IN_FAT in sandbox configs. This leads to environment
driver lookup failure when env in non-EXT4 is selected using 'env select':
"
env_driver_lookup: No environment driver for location 3
priority not found
"
Enable the missing ENV_IS_IN_FAT to fix this.
Fixes: 2a38e712652f ("sandbox: add FAT to the list of usable env drivers") Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Quentin Schulz [Thu, 18 Dec 2025 10:59:58 +0000 (11:59 +0100)]
cmd: bdinfo: fix incorrect Kconfig options check for print_eth()
CMD_NET_LWIP has never existed so it cannot be right. I'm guessing the
intent was to allow print_eth() to be called when NET_LWIP is defined
(NET means "legacy networking stack" as opposed to NET_LWIP which is the
newest (and incompatible) stack). There probably was some mix-up
between CMD_NET and NET options.
The dependency on CMD_NET seems unnecessary as it seems perfectly fine
to run bdinfo without CMD_NET (build and run tested). So let's instead
make the dependency on NET || NET_LWIP.
David Lechner [Wed, 17 Dec 2025 00:56:35 +0000 (18:56 -0600)]
pwm: aspeed: replace %pe in dev_err()
Replace %pe with %d and adjust the argument accordingly in a dev_err()
call in the pwm-aspeed driver. U-boot doesn't support the %pe format
specifier. Likely it was copied from Linux.
Signed-off-by: David Lechner <dlechner@baylibre.com> Reviewed-by: Chia-Wei Wang <chiawei_wang@aspeedtech.com>
Emanuele Ghidoli [Tue, 16 Dec 2025 08:45:28 +0000 (09:45 +0100)]
toradex: tdx-cfg-block: add pid4 support for new modules
Add new PID4 to ConfigBlock handling:
- 0217 Lino iMX93 Dual 2GB IT
- 0218 Lino iMX91 Solo 2GB IT
- 0219 OSM iMX93 Dual 2GB IT
- 0220 OSM iMX91 Solo 2GB IT
- 0221 Verdin AM62 Dual 1GB ET
Lino and OSM are two new SoM families.
The Verdin variant differs from the existing 0073 Verdin AM62 Dual 1GB ET
by the presence of the GPU (AM625 instead of AM623), the absence of
DSI interface (bridge not mounted) and eMMC size increased to 16GB instead
of 4GB.
Marek Vasut [Sun, 28 Dec 2025 03:17:27 +0000 (04:17 +0100)]
m68k: Assure end of U-Boot is at 8-byte aligned offset
Make sure the end of U-Boot is at 8-byte aligned offset, not 4-byte
aligned offset. This allows safely appending DT at the end of U-Boot
with the guarantee that the DT will be at 8-byte aligned offset. This
8-byte alignment is now checked by newer libfdt 1.7.2 .
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Sughosh Ganu [Tue, 16 Dec 2025 09:16:24 +0000 (11:16 +0200)]
kbuild: Bump the build system to 6.1
Our last sync with the kernel was 5.1.
We are so out of sync now, that tracking the patches and backporting
them one by one makes little sense and it's going to take ages.
This is an attempt to sync up Makefiles to 6.1.
Unfortunately due to sheer amount of patches this is not easy to review,
but that's what we decided during a community call for the bump to 5.1,
so we are following the same guidelines here.
Makefile: repair CONFIG_CC_OPTIMIZE_FOR_DEBUG support
Since commit 5f520875bdf0 ("kbuild: Bump the build system to 5.1")
CONFIG_CC_OPTIMIZE_FOR_DEBUG has no effect on the non-host code.
This patch reestablishes the prior logic to add
-Og -Wno-maybe-uninitialized
to KBUILD_CFLAGS.
Fixes: 5f520875bdf0 ("kbuild: Bump the build system to 5.1") Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tom Rini [Wed, 31 Dec 2025 16:13:54 +0000 (10:13 -0600)]
Merge patch series "configs: Remove default malloc length for K3 R5 SPL"
This series from Andrew Davis <afd@ti.com> makes a number of the TI K3
CONFIG symbols have consistent values in SPL, as they are things
determined by the SoC and not the board design.
David Lechner [Wed, 10 Dec 2025 23:40:14 +0000 (17:40 -0600)]
arm: dts: mediatek: switch mt8365 to OF_UPSTREAM
Change mt8365_evk_defconfig to use CONFIG_OF_UPSTREAM=y and delete the
U-Boot copy of the devicetree source files for mt8365.
The upstream devicetree is identical to the U-Boot one being removed
(other than having more nodes for devices not used by U-Boot and
upstream fixed a compatible string in &scpsys, also not affecting
U-Boot).
There was one minor glitch with upstream missing a few topckgen macro
definitions, so those are added to the clock driver directly as a
workaround.
Reviewed-by: Macpaul Lin <macpaul.lin@mediatek.com> Signed-off-by: David Lechner <dlechner@baylibre.com>
Andrew Davis [Mon, 8 Dec 2025 19:06:33 +0000 (13:06 -0600)]
spl: Kconfig: k3: Increase malloc size after relocation for R5
Seems the "generous 2MB space" is no longer enough for SPL on some K3 R5
platforms so let's increase this to 4MB. That matches what we give to
ARM64 SPL, so combine these.
Signed-off-by: Andrew Davis <afd@ti.com> Reviewed-by: Bryan Brattlof <bb@ti.com>
Tom Rini [Tue, 30 Dec 2025 16:22:06 +0000 (10:22 -0600)]
Merge patch series "video: Remove unused drivers, clean up dependencies"
Tom Rini <trini@konsulko.com> says:
This is v2 of the series I originally posted back in August[1]. The
changes here are that I've dropped the first patch as TI has recently
posted their rework of the driver in question, and I added Svyatoslav's
Reviewed-by tag. The end goal here is that "allyesconfig" will be able
to build (on sandbox).
Swamil Jain [Thu, 6 Nov 2025 08:58:41 +0000 (14:28 +0530)]
drivers: video: tidss: Refactor tidss_drv
- Refactor tidss_drv to improve modularity, enabling support for more
display interfaces beyond OLDI in the future
- Add detection and initialization of active OLDI panels using the DT
- Port tidss_oldi.c from the upstream Linux kernel oldi series[0] and
derive several APIs from it to determine the dual link pixel order
- Add tidss_oldi_init() and helper routines to handle OLDI-specific
setup and move related helper routines to tidss_oldi.c
Igor Belwon [Tue, 4 Nov 2025 14:50:36 +0000 (15:50 +0100)]
video: simplefb: Add stride handling
Some framebuffers (i.e MediaTek) do not have regular stride - its line
length is more than the display width by 8 pixels (on MT6878). As such,
introduce the optional stride property, which fixes these framebuffers.
Signed-off-by: Igor Belwon <igor.belwon@mentallysanemainliners.org>
Tom Rini [Wed, 2 Jul 2025 01:05:18 +0000 (19:05 -0600)]
common/splash_source.c: Change bmp_load_addr to ulong from u32
The variable bmp_load_addr is used to hold the address in memory of
where to put the splash image (as a bmp). For 32/64bit correctness, this
needs to be a ulong and not u32 today.
Tom Rini [Tue, 30 Dec 2025 16:18:39 +0000 (10:18 -0600)]
Merge patch series "video: display: refactor display_read_timing to avoid code duplication"
Julien Stephan <jstephan@baylibre.com> says:
Commit 2dcf143398ad ("dm: video: Repurpose the 'displayport' uclass to 'display'")
left the display_read_edid() function unused by mistake.
This series addresses that oversight and introduces a new useful cmd.
Patch 1:
- Refactors display_read_timing() to use the existing
display_read_edid() function, eliminating redundant code.
- Marks display_read_edid() as static since it is not used outside of
the file.
Patch 2:
- Adds a new read_edid command, which can be very useful for debugging
or developing new display drivers.
- As this command uses display_read_edid(), the function is made
non-static again.
Tom Rini [Wed, 12 Nov 2025 20:00:05 +0000 (14:00 -0600)]
video: anx9804: Only build when needed
The logic for how to handle this video driver is slightly odd. Only in
the case of when CONFIG_VIDEO_LCD_PANEL_EDP_4_LANE_1620M_VIA_ANX9804 is
enabled do we need to have this file built, and otherwise we have a
dummy function in use. Correct the logic by only building this file when
needed.
Tom Rini [Wed, 12 Nov 2025 20:00:03 +0000 (14:00 -0600)]
video: tegra: Rework some of the driver dependencies
Looking these drivers over, all of them cannot build without access to
some platform specific header files. Express those requirements in
Kconfig as well. Furthermore, update the logic a bit more to reflect
which parts are optional when other drivers are enabled and which parts
cannot be enabled (meaningfully) by themselves.
Reviewed-by: Svyatoslav Ryhel <clamor95@gmail.com> Signed-off-by: Tom Rini <trini@konsulko.com>
clk: renesas: Do not disable realtime modules on R8A77995 D3
Later versions of the datasheet makes it clear D3 do not have any
realtime module stop control registers (RMSTPCRx). Remove the
manipulation of them from the module clock table to match this.
Suggested-by: Marek Vasut <marek.vasut+renesas@mailbox.org> Signed-off-by: Niklas Sรถderlund <niklas.soderlund+renesas@ragnatech.se>
clk: renesas: Do not enable MSTP4 extra modules on R8A77995 D3
Since commit a2bd99549c61 ("clk: renesas: Tear clock controller down
last before booting OS") enabling the module gated by bit 8 in MSTP4
prevents Linux from booting. The bits 8 and 7 of MSTP4 where only
documented in early versions of the datasheet and have since been
removed.
To allow Linux to boot update the MSTP4 enable value to reflect the
hardware default, 0x80.
Suggested-by: Marek Vasut <marek.vasut+renesas@mailbox.org> Signed-off-by: Niklas Sรถderlund <niklas.soderlund+renesas@ragnatech.se>
- Swicth imx8ulp-evk to standard boot and OF_UPSTREAM.
- Cleanup of the IPUv3 video driver.
- Add support for the NXP FRDM-IMX91 board.
- Make flash.bin target available on i.MX9.
- Fix mxsfb pixel clock polarity.
Marek Vasut [Sun, 28 Dec 2025 19:43:21 +0000 (20:43 +0100)]
Makefile: Make flash.bin target available on i.MX9
The current implementation of flash.bin generation with
CONFIG_SPL_LOAD_IMX_CONTAINER=y requires build of u-boot.cnt
which is i.MX8 specific. Reinstate the i.MX8 check to avoid
this dependency for i.MX9 .
Fill in flash.bin target for i.MX9 into imx specific Makefile.
Fixes: c3587197c0c9 ("Makefile: Make flash.bin target available for all platforms") Signed-off-by: Marek Vasut <marex@nabladev.com>
Brian Ruley [Mon, 29 Dec 2025 10:48:07 +0000 (12:48 +0200)]
video: imx: ipuv3: refactor to use dm-managed state
Get rid of most globals that are spread around between TU's and place
them in their own structs managed by dm. Device state is now owned by
each driver instance. This design mirrors the Linux IPUv3 driver
architecture.
This work is done in preparation to migrate the driver to the clock
framework. While not the primary intent, this change also enables
multiple IPU instances to exist contemporarily.
Signed-off-by: Brian Ruley <brian.ruley@gehealthcare.com>
Brian Ruley [Mon, 29 Dec 2025 10:48:02 +0000 (12:48 +0200)]
video: imx: ipuv3: apply clang-format
Bring the code into compliance with U-Boot's coding style guidelines for
upcoming changes. Sort includes to tidy things up and apply
{ RemoveBracesLLVM: true } to remove unnecessary blocks.
Signed-off-by: Brian Ruley <brian.ruley@gehealthcare.com>
Joseph Guo [Fri, 12 Dec 2025 06:20:20 +0000 (15:20 +0900)]
arm64: dts: add NXP FRDM-IMX91 device tree
Add the device tree files for the FRDM-IMX91 board.
Provide the initial DT support for FRDM-IMX91.
The board devicetree already attempted to upstream, but not been
accepted yet:
https://lore.kernel.org/all/20251114-imx91_frdm-v1-0-e5763bdf9336@nxp.com/
Alice Guo [Tue, 9 Dec 2025 02:19:19 +0000 (10:19 +0800)]
imx8ulp_evk: Switch to use devicetree imported from Linux kernel release
Enable OF_UPSTREAM for i.MX8ULP EVK so that devicetree imported from
Linux kernel release can be used.
If mailbox@29220000 is enabled, gd->arch.ele_dev will be set to this
device for communication with ELE firmware. This is incorrect because
mu@27020000 is the MU used for communication with the ELE firmware. To
prevent misconfiguration, disable mailbox@29220000.
The driver model for watchdog timer is not enabled yet, so disable wdog3
temporarily.
Alice Guo [Tue, 9 Dec 2025 02:19:18 +0000 (10:19 +0800)]
imx8ulp_evk: Move environment variables to .env file
Add board-specific environment variables to imx8ulp_evk.env for better
maintainability. Define bsp_bootcmd in the environment to resolve the
runtime error: "bsp_bootcmd" not defined.
Alice Guo [Tue, 9 Dec 2025 02:19:17 +0000 (10:19 +0800)]
imx8ulp_evk: Convert to standard boot
Replace CONFIG_DISTRO_DEFAULTS with CONFIG_BOOTSTD_FULL to enable the
standard boot framework and use standard boot on i.MX8ULP.
Update CONFIG_BOOTCOMMAND to run bootflow scan before falling back to
board-specific bootcmd, and remove legacy distro boot environment from
imx8ulp_evk.h since bootstd now handles boot targets.
Without the patch a splash screen displays with aliasing-like jagged edges.
Signed-off-by: Sam Meredith <sam@aandtinstruments.com>
[fabio: Put more information into the commit log] Signed-off-by: Fabio Estevam <festevam@gmail.com>
Marek Vasut [Wed, 19 Nov 2025 17:43:53 +0000 (18:43 +0100)]
boot: Warn users about fdt_high=~0 usage
In case the 'fdt_high' environment variable is set to ~0, warn users
about the dangers of the fdt_high usage. This will hopefully lead to
removal of most of the fdt_high ~0 usage over time.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org> Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Marek Vasut [Mon, 24 Nov 2025 14:51:01 +0000 (15:51 +0100)]
i2c: Avoid calling dev_read_*() if CONFIG_OF_PLATDATA=y
If CONFIG_OF_PLATDATA=y , then the udevice has no valid OF node associated
with it and ofnode_valid(node) evaluates to 0. The dev_read_u32_default()
call ultimately reaches ofnode_read_u32_index() which invokes fdt_getprop()
and passes result of ofnode_to_offset(node) as an offset parameter into it.
The ofnode_to_offset(node) returns -1 for invalid node, which leads to an
fdt_getprop(..., -1, ...) invocation, which will crash sandbox with SIGSEGV
because libfdt can not handle negative node offsets without full tree check,
which U-Boot inhibits to keep size lower.
Since i2c_child_post_bind() already calls dev_has_ofnode(dev), reuse the
same call and assign i2c->speed_hz = I2C_SPEED_STANDARD_RATE in case the
device has no valid node associated with it, and do not call any of the
dev_read_*() functions for devices without valid nodes.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Marek Vasut [Mon, 24 Nov 2025 14:51:00 +0000 (15:51 +0100)]
i2c: Inline i2c_chip_of_to_plat() into i2c_child_post_bind()
The i2c_chip_of_to_plat() is called only from i2c_child_post_bind(),
inline i2c_chip_of_to_plat() into i2c_child_post_bind(). Drop the
if CONFIG_IS_ENABLED(OF_REAL) and depend on if (!dev_has_ofnode(dev))
which does check CONFIG_IS_ENABLED(OF_REAL) internally too.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Marek Vasut [Fri, 21 Nov 2025 16:17:55 +0000 (17:17 +0100)]
gpio: sandbox: Avoid calling dev_read_*() if CONFIG_OF_PLATDATA=y
If CONFIG_OF_PLATDATA=y , then the udevice has no valid OF node associated
with it and ofnode_valid(node) evaluates to 0. The dev_read_u32_default()
call ultimately reaches ofnode_read_u32_index() which invokes fdt_getprop()
and passes result of ofnode_to_offset(node) as an offset parameter into it.
The ofnode_to_offset(node) returns -1 for invalid node, which leads to an
fdt_getprop(..., -1, ...) invocation, which will crash sandbox with SIGSEGV
because libfdt can not handle negative node offsets without full tree check,
which U-Boot inhibits to keep size lower.
Since gpio_sandbox_probe() already calls dev_has_ofnode(dev) and assigns
uc_priv->gpio_count to CONFIG_SANDBOX_GPIO_COUNT accordingly, add matching
dev_has_ofnode(dev) check into sandbox_gpio_of_to_plat() and do not call
any of the dev_read_*() functions for devices without valid nodes there
either.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Marek Vasut [Fri, 21 Nov 2025 16:17:37 +0000 (17:17 +0100)]
clk: fixed_rate: Avoid calling dev_read_*() if CONFIG_OF_PLATDATA=y
If CONFIG_OF_PLATDATA=y , then the udevice has no valid OF node associated
with it and ofnode_valid(node) evaluates to 0. The dev_read_u32_default()
call ultimately reaches ofnode_read_u32_index() which invokes fdt_getprop()
and passes result of ofnode_to_offset(node) as an offset parameter into it.
The ofnode_to_offset(node) returns -1 for invalid node, which leads to an
fdt_getprop(..., -1, ...) invocation, which will crash sandbox with SIGSEGV
because libfdt can not handle negative node offsets without full tree check,
which U-Boot inhibits to keep size lower.
Add dev_has_ofnode(dev) check and do not assign clock rate in case the
device has no valid node associated with it, and do not call any of the
dev_read_*() functions for devices without valid nodes.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
* Update StarFive Jh7110 common description.
* Describe command line options of the bdinfo command.
* Describe configuration dependencies of the bdinfo command.
UEFI:
* Trigger capsule updates with automatically generated boot options.
* In the LoadImage unit test add a check that device-paths are correctly
used.
* In the variables at runtime test remove an unnecessary
__efi_runtime_data attribute.
Others:
* Let the bdinfo command output device-tree information even if LMB is
no used.
* Add long help texts for all options of the bdinfo command.
E Shattow [Sun, 21 Dec 2025 11:36:04 +0000 (03:36 -0800)]
doc: board: starfive: update jh7110 common description
Updates to the JH7110 common description:
- add detailed overview of JH-7110 SoC and boot process
- revise descriptions of deprecated StarFive loader modes
- refresh build directions grouped with SPL debug advice
- reduce usage instructions into common methods shared by supported boards
- cite starfive_visionfive2 board maintainer description of StarFive loader
- cite published datasheets for ambient operating temperature data
Redundant/deprecated sections of each board doc are dropped accordingly:
- deepcomputing fml13v01
- milk-v mars
- pine64 star64 (also add inclusion of JH7110 common description)
- visionfive2
Signed-off-by: E Shattow <e@freeshell.de> Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Quentin Schulz [Thu, 18 Dec 2025 11:44:43 +0000 (12:44 +0100)]
doc: cmd: bdinfo: document options
bdinfo may also have -a, -e and -m options depending on some symbols
being set. Document all this and provide an example on how to use them
and what they typically output.
Quentin Schulz [Thu, 18 Dec 2025 11:44:41 +0000 (12:44 +0100)]
cmd: bdinfo: provide long help with all options
Document the bdinfo -a, -e and -m options in the long help, but only
when they can be used. The string concatenation is a bit odd with two
newlines, but it does render properly once in U-Boot CLI.
Tested-by: Michal Simek <michal.simek@amd.com> Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Ilias Apalodimas [Wed, 17 Dec 2025 09:40:04 +0000 (11:40 +0200)]
efi_loader: Trigger capsule updates with automatically generated boot options
The EFI spec in ยง8.5.5 says
"The directory \EFI\UpdateCapsule is checked for capsules only within
the EFI system partition on the device specified in the active boot
option determine by reference to BootNext variable or BootOrder variable
processing."
Automatically generated boot options don't point to the ESP, they point to
the disk itself and find_handle() won't match when searching for an ESP
during a capsule update.
This happens because find_handle() only matches device paths that are
shorter or equal to the device path passed as an argument.
Since the EFI spec allows it we want to allow capsule updates, when the
boot option points to a disk, but that disk contains an ESP with a
\EFI\UpdateCapsule directory.
So, let's change device_is_present_and_system_part() and check if the
supplied device path contains an ESP. If it does return the handle of
the device. Otherwise, iterate over child devices and return the handle
of the first child that contains an ESP.
The returned handle can then be reused later. Rather than calling
efi_fs_from_path(), we can simply look up the EFI_SIMPLE_FILE_SYSTEM_PROTOCOL
on the discovered handle, avoiding the need to re-parse device paths.
Reported-by: Balaji Selvanathan <balaji.selvanathan@oss.qualcomm.com> Reported-by: John Toomey <john.toomey@amd.com> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Quentin Schulz [Fri, 12 Dec 2025 18:13:20 +0000 (19:13 +0100)]
power: regulator: remove SPL_REGULATOR_PWM due to unmeetable SPL_DM_PWM dependency
SPL_DM_PWM option simply doesn't exist. Moreover, drivers/pwm is only
included by drivers/Makefile for non-xPL stages so making
SPL_REGULATOR_PWM properly build for SPL/xPL is more involved than just
adding an SPL_DM_PWM option.
Reading the original commit (ddc824f89aa8 ("power: regulator: Allow PWM
regulator to be omitted from SPL."), the intent seemingly wasn't to
allow building support in XPL but rather to allow removing it which is
done by using $(PHASE_) ($(SPL_) at that time) in the Makefile. If
anyone needs that, let them figure out what they need to do without
misleading potential users of this symbol by simply removing it.
Quentin Schulz [Fri, 12 Dec 2025 18:13:19 +0000 (19:13 +0100)]
power: regulator: fix dependency for REGULATOR_PWM
The PWM regulator driver is a uclass driver, thus requiring DM_PWM to be
enabled to be actually usable (and with the appropriate PWM controller
driver enabled as well, but that we cannot enforce easily), so let's add
this missing dependency.
Quentin Schulz [Fri, 12 Dec 2025 18:13:18 +0000 (19:13 +0100)]
spl: fix prompt for SPL_BOOTROM_SUPPORT
SPL_BOOTROM_SUPPORT currently doesn't specify it enables returning to
BootROM *from SPL*, which TPL_BOOTROM_SUPPORT does say. So let's align
the prompts so that both say from which stage you can return to the
BootROM.
Fixes: 225d30b70846 ("spl: add a 'return to bootrom' boot method") Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de> Reviewed-by: Anshul Dalal <anshuld@ti.com> Reviewed-by: Kory Maincent <kory.maincent@bootlin.com> Signed-off-by: Peng Fan <peng.fan@nxp.com>
Mathieu Othacehe [Wed, 10 Dec 2025 14:17:04 +0000 (15:17 +0100)]
net: ravb: Configure CXR31 and CXR35 on rzg2l
As in Linux with d78c0ced60 ("net: ravb: Make write access to CXR35 first
before accessing other EMAC register"), configure CXR31 and CXR35 correctly
on rzg2. MII mode does not work correctly unless those registers are
properly configured.
Signed-off-by: Mathieu Othacehe <othacehe@gnu.org> Reviewed-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
projects / thirdparty / u-boot.git / log
