It causes a panic when blob is shrunk and 'new_alloced' is less than
'next_ofs'. The data area that needs to be moved should end up at
'hdr->used_size'.
Fixes: 1fe5937549 ("bloblist: Support resizing a blob")
Signed-off-by: Raymond Mao <raymond.mao@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Tested-by: Michal Simek <michal.simek@amd.com>
Add two helper functions for:
1. marking a blob void
2. getting blob record from a given blob data pointer.
Signed-off-by: Raymond Mao <raymond.mao@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Tested-by: Michal Simek <michal.simek@amd.com>
We recently added IO accessors that will work with KVM for any MMIO
access that casues an exception to the hypervisor. Enable them by
default for QEMU.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tested-by: Mikko Rapeli <mikko.rapeli@linaro.org>
commit 2e2c2a5e72 ("arm: qemu: override flash accessors to use virtualizable instructions")
explains why we can't have instructions with multiple output registers
when running under QEMU + KVM and the instruction leads to an exception
to the hypervisor.
USB XHCI is such a case (MMIO) where a ldr w1, [x0], #4 is emitted for
xhci_start() which works fine with QEMU but crashes for QEMU + KVM.
These instructions cannot be emulated by KVM as they do not produce
syndrome information data that KVM can use to infer the destination
register, the faulting address, whether it was a load or store, or
if it's a 32 or 64 bit general-purpose register.
As a result an external abort is injected from QEMU, via ext_dabt_pending
to KVM and we end up throwing an exception that looks like
U-Boot 2025.07-rc4 (Jun 10 2025 - 12:00:15 +0000)
[...]
Register 8001040 NbrPorts 8
Starting the controller
"Synchronous Abort" handler, esr 0x96000010, far 0x10100040
elr: 000000000005b1c8 lr : 000000000005b1ac (reloc)
elr: 00000000476fc1c8 lr : 00000000476fc1ac
x0 : 0000000010100040 x1 : 0000000000000001
x2 : 0000000000000000 x3 : 0000000000003e80
x4 : 0000000000000000 x5 : 00000000477a5694
x6 : 0000000000000038 x7 : 000000004666f360
x8 : 0000000000000000 x9 : 00000000ffffffd8
x10: 000000000000000d x11: 0000000000000006
x12: 0000000046560a78 x13: 0000000046560dd0
x14: 00000000ffffffff x15: 000000004666eed2
x16: 00000000476ee2f0 x17: 0000000000000000
x18: 0000000046660dd0 x19: 000000004666f480
x20: 0000000000000000 x21: 0000000010100040
x22: 0000000010100000 x23: 0000000000000000
x24: 0000000000000000 x25: 0000000000000000
x26: 0000000000000000 x27: 0000000000000000
x28: 0000000000000000 x29: 000000004666f360
Code: d5033fbf aa1503e0 5287d003 52800002 (b8004401)
Resetting CPU ...
There are two problems making this the default.
- It will emit ldr + add or str + add instead of ldr/str(post increment)
in somne cases
- Some platforms that depend on TPL/SPL grow in size enough so that the
binary doesn't fit anymore.
So let's add proper I/O accessors add a Kconfig option
to turn it off by default apart from our QEMU builds.
Reported-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Tested-by: Mikko Rapeli <mikko.rapeli@linaro.org>
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
A following patch is replacing our IO accessors with
do { ... } while(0) ones in order to make them usable with KVM.
That leads to an error eventually looking like this:
arch/arm/include/asm/io.h:62:9: error: expected expression before 'do'
62 | do { \
| ^~
arch/arm/include/asm/io.h:211:41: note: in expansion of macro '__raw_writel'
211 | #define out_arch(type,endian,a,v) __raw_write##type(cpu_to_##endian(v),a)
| ^~~~~~~~~~~
arch/arm/include/asm/io.h:223:25: note: in expansion of macro 'out_arch'
223 | #define out_be32(a,v) out_arch(l,be32,a,v)
| ^~~~~~~~
drivers/spi/fsl_dspi.c:127:17: note: in expansion of macro 'out_be32'
127 | out_be32(addr, val) : out_le32(addr, val);
| ^~~~~~~~
So adjust the current macros and code to be compatible with the upcoming
change.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Add support of Renesas R-Car Gen4 watchdog timer. Timeouts up to
8184.0s are supported (CKS1 register is not involved). The watchdog
uses the clock of type CLK_TYPE_GEN4_MDSEL.
The timeout is set in
dts/upstream/src/arm64/renesas/r8a779f0-spider-cpu.dtsi section &rwdt.
This driver is based on upstream linux commit:
e70140ba0d2b("Get rid of 'remove_new' relic from platform driver struct")
Signed-off-by: Shmuel Leib Melamud <smelamud@redhat.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Add support of CLK_TYPE_GEN4_MDSEL clock type to gen3_clk_get_rate64()
function. In particular, this type of clock is used by Renesas R-Car
Gen4 watchdog. It operates similarly to CLK_TYPE_GEN3_MDSEL clock.
Signed-off-by: Shmuel Leib Melamud <smelamud@redhat.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Reviewed-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Reviewed-by: Stefan Roese <sr@denx.de>
The bootph-* tags are now part of mainline Linux DTs as commits:
f1a1268572cc ("ARM: dts: renesas: rcar-gen2: Add boot phase tags")
399f14ff6625 ("arm64: dts: renesas: rcar: Add boot phase tags")
624b2a23d46a ("arm64: dts: renesas: rzg2: Add boot phase tags")
Drop local duplicates of the same tags, as well as duplicate
property assignments which are part of the core DTs themselves.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Reviewed-by: Sumit Garg <sumit.garg@oss.qualcomm.com>
Contrary to the datasheet, setting both DETACH and PULLD_DIS bits to 1
does not always drive the DP and DM lines to high-impedance. This
prevents the host from reliably detecting a USB disconnect and subsequent
reconnect.
The symptom is that the first gadget command (e.g., dhcp) succeeds, while
subsequent commands (e.g., nfs) fail.
Disabling and re-enabling the controller entirely, instead of toggling the
PULLD_DIS bit, reliably generates a disconnect event.
The Linux driver works correctly because gadget_disconnect/gadget_connect
are always followed by gadget_udc_start/gadget_udc_stop. In U-Boot
pullup() is used solely.
This behavior has been observed on the SAM9X60-Curiosity and
AT91SAM9G25-EK boards and has been reported to Microchip.
Signed-off-by: Zixun LI <admin@hifiphile.com>
Link: https://lore.kernel.org/r/20250602-pullup-v1-1-edcde5a050dd@hifiphile.com
[mkorpershoek: reworded commit title + comment to usba_udc_pullup()]
Signed-off-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Remove duplicate .ops assignment that was overriding the correct
ti_musb_gadget_ops with musb_usb_ops (host ops) in the ti_musb_peripheral
driver. This was causing U-Boot crashes when trying to call the
handle_interrupts operation since the wrong ops structure was being used.
Fixes: 7d98dbcc3d ("usb: musb-new: Add support for DM_USB")
Fixes: 281eaf1ed8 ("usb: gadget: musb: Convert interrupt handling to usb_gadget_generic_ops")
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Link: https://lore.kernel.org/r/20250611171031.840277-1-kory.maincent@bootlin.com
Signed-off-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
The third_arg for a dfu alt is read as an integer and is overloaded for
different supported backends. For ext4 and fat, this third_arg
represents the partition and forms the dev part string, which should
have its partition in hex. This commit fixes dfu ext4/fat usage for
devices with ten or more partitions.
Signed-off-by: Ivan Pang <ipman@amazon.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Reviewed-by: Lukasz Majewski <lukma@denx.de>
Link: https://lore.kernel.org/r/20250611050127.38011-1-ipman@amazon.com
Signed-off-by: Mattijs Korpershoek <mkorpershoek@kernel.org>
Writing to eMMC on RK3528 is affected with the same or a similar issue
as on RK3588, where eMMC must init to HS200 at least once to fully work.
Trying to write u-boot-rockchip.bin to eMMC fails with:
=> mmc write $fileaddr 40 5000
MMC write: dev # 0, block # 64, count 20480 ... mmc write failed
0 blocks written: ERROR
For U-Boot to enable HS200 mode the mmc-hs200-1_8v prop must be defined
in the device tree. Linux does not seem to be affected and is able to
detect and use HS200 without this prop.
Enable use of HS200 and fix eMMC write on RK3528 by adding the missing
mmc-hs200-1_8v prop for affected boards:
=> mmc write $fileaddr 40 5000
MMC write: dev # 0, block # 64, count 20480 ... 20480 blocks written: OK
Fixes: b112a44531 ("board: rockchip: Add minimal generic RK3528 board")
Fixes: ccbddf6453 ("board: rockchip: Add Radxa E20C")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Raspberry Pi 5 can now have up to 16 GiB of RAM where the memory spans 8
DRAM banks in total. Increase the config value to 8 to initialize the
whole RAM. Without this change, kernel only sees 8 GiB of RAM on the 16
GiB CM5 as reported in [1].
[1] https://github.com/home-assistant/operating-system/issues/3989
Signed-off-by: Jan Čermák <sairon@sairon.cz>
Reviewed-by: Matthias Brugger <mbrugger@suse.com>
Sync bitfield.h header with Linux 6.15 version. Mainly is to permit the
introduction of FIELD_PREP_CONST. The bug.h header changed to
build_bug.h doesn't cause any regression as we also ship split header
similar to how it's done with in Linux.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The GENMASK/GENMASK_ULL macro requires inclusion of linux/bitops.h
header. It does currently work as bitfield.h includes it indirectly
but this will change when bitfield.h will be synced with new Linux
version.
Also raw printf require linux/bug.h header (also currently included
indirectly by bitfield.h)
Explicitly include the headers to fix future compilation error.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The WARN macro requires inclusion of linux/bug.h header. It does
currently work as bitfield.h includes it indirectly but this will
change when bitfield.h will be synced with new Linux version.
Explicitly include the header to fix future compilation error.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The WARN macro requires inclusion of linux/bug.h header. It does
currently work as bitfield.h includes it indirectly but this will
change when bitfield.h will be synced with new Linux version.
Explicitly include the header to fix future compilation error.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
E Shattow <e@freeshell.de> says:
Make consistent use of lowercase hexadecimal prefix '0x' throughout U-Boot.
There are a few remaining uses of uppercase 'X' to denote hexadecimal prefix
or placeholder in documentation and error messages.
External devicetree-rebasing dts/upstream and the generated code of
xilinx/zynq are ignored for the series.
Link: https://lore.kernel.org/r/20250606224558.1117422-1-e@freeshell.de
Use consistent lowercase hex prefix style in drivers/*
Does not change hex prefix case in allcaps uppercase style error messages
Signed-off-by: E Shattow <e@freeshell.de>
Use consistent lowercase hex prefix style in board/*
Does not change auto-generated files in xilinx/zynq/*/ps7_init_gpl.c
Signed-off-by: E Shattow <e@freeshell.de>
Rasmus Villemoes <ravi@prevas.dk> says:
There's really no reason for the gd pointer to have the volatile
qualifier.
In fact, I claim that it's completely unnecessary and just pessimizes
code generation and forces ugly casts in lots of places. For example,
see the casts in drivers/core/tag.c where elements are added to
gd->dm_taglist, or a helper such as cyclic_get_list() that should not
be needed.
Also, it is what ends up standing in the way of an otherwise
innocent code cleanup of list.h:
https://lore.kernel.org/u-boot/20250522165656.GB2179216@bill-the-cat/
Note that riscv, x86 as well as arm64 with LTO enabled has not had
this volatile qualifier, so it's unlikely there's any generic code
that depends on it.
Link: https://lore.kernel.org/r/20250604195612.2312979-1-ravi@prevas.dk
It's slightly ironic that riscv at very first had the gd pointer
volatile qualified [6020faf62c ("riscv: nx25: include: Add header
files to support RISC-V")], removed that back in 2018
[40717eb849 ("riscv: checkpatch: Fix use of volatile")], and then in
2020 this helper was added [6b9966e1aa ("riscv: define function
set_gd()")] which needlessly had volatile in the prototype.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
The global gd pointer is no longer volatile-qualified. Callers of this
helper function have been updated to no longer use volatile-qualifed
temporary variables, so update the prototype accordingly.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
The global gd pointer is no longer volatile-qualified, so drop that
qualifier from these bookkeeping variables.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
The global gd pointer is no longer volatile-qualified, so drop that
qualifier from these bookkeeping variables.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Now that the global gd pointer is no longer volatile-qualified,
there's no reason for the temporary variables used for
saving/restoring it to have that qualifier.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Tested-by: Anshul Dalal <anshuld@ti.com>
The way DECLARE_GLOBAL_DATA_PTR is used, the stuff under #if 0 can
never compile as you cannot have a non-constant initializer at global
scope (and one would get linker errors as well because the 'gd' symbol
would be defined in multiple TUs).
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Acked-by: Angelo Dureghello <angelo@kernel-space.org>
There's a bunch of other places where this qualifier should be
dropped, e.g. in the set_gd() prototype and for various variables used
for stashing the value in the mach-imx/ directory and elsewhere. But
that will be done in follow-up patches.
Signed-off-by: Rasmus Villemoes <ravi@prevas.dk>
Tested-by: Anshul Dalal <anshuld@ti.com>
This patch is to solve the sandbox building error:
$ make O=build-sandbox -s sandbox_defconfig
$ make O=build-sandbox -s -j2
/usr/bin/ld: /tmp/u-boot.27rzOu.ltrans58.ltrans.o: undefined reference to symbol 'pthread_kill@@GLIBC_2.2.5'
/usr/bin/ld: /lib/x86_64-linux-gnu/libpthread.so.0: error adding symbols: DSO missing from command line
collect2: error: ld returned 1 exit status
[...]
Fixes: b989f9ed9f ("test: lib: add initjmp() test")
Signed-off-by: Raymond Mao <raymond.mao@linaro.org>
Reviewed-by: Jerome Forissier <jerome.forissier@linaro.org>
Tested-by: Patrice Chotard <patrice.chotard@foss.st.com>
This function is currently a misnomer at times as we have cases where it
passes arguments to the image. In preparation for making that be a more
common case rename this function to jump_to_image(...). In order to do
this, rename jump_to_image in board_init_r(...) to jumper so that we do
not have a conflict.
Signed-off-by: Simon Glass <sjg@chromium.org>
[trini: Reword the commit message, adding missing cases of
jump_to_image_no_args()]
Signed-off-by: Tom Rini <trini@konsulko.com>
Sukrut Bellary <sbellary@baylibre.com> says:
This patch series adds the support for
1. In am33xx SoC[1], enhanced capture (eCAP) supports auxiliary PWM (APWM).
This series adds the PWM driver support for the APWM feature for eCAP on
AM33xx. AM335X_ECAP0_IN_PWM0_OUT is used to enable the backlight.
2. Fix build warning in ti-ehrpwm driver in dev_deb().
3. Enable eCAP0 PWM and LCD pin muxing to support splash screen on
AM335x EVM[2].
[1] AM335x TRM - https://www.ti.com/lit/ug/spruh73q/spruh73q.pdf
[2] AM335x EVM - https://www.ti.com/tool/TMDXEVM3358
per discussion on the earlier patch series,
https://lore.kernel.org/all/20250319202516.3300444-1-sbellary@baylibre.com/
dropping the device tree changes in this series due to OF_UPSTREAM
conflict.
As we are dropping the DT patch, started with the new series instead of
V2.
This series doesn't contain defconfig changes since we don't want to
enable LCD, splash screen and PWM support by default.
Enabling splash screen and PWM support in defconfig causes u-boot crash
on AM335x based beaglebone black. This will be handled in a separate
patch.
Link: https://lore.kernel.org/r/20250530212232.1686613-1-sbellary@baylibre.com
Raymond MaoandTom Rini
Ilias ApalodimasandTom Rini
Shmuel Leib MelamudandMarek Vasut
Lukasz MajewskiandFabio Estevam
Zixun LIandMattijs Korpershoek
Kory MaincentandMattijs Korpershoek
Ivan PangandMattijs Korpershoek
Jonas KarlmanandTom Rini
Jan ČermákandPeter Robinson
Christian MarangiandTom Rini
E ShattowandTom Rini
Rasmus VillemoesandTom Rini
Simon GlassandTom Rini