Now that we have a way to find the appropriate EFI system
partition on the internal NVMe storage, nail it down as the
preferred EFI system partition such that we can use it to
store EFI variables. This allows multiple OS installations to
exist on Apple Silicon machines without interfering with eachother.
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
Now that we have a way to find the appropriate EFI system
partition on the internal NVMe storage, store the U-Boot
environment there.
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
The ASMedia XHCI controller found on some of the Apple Silicon
machines needs firmware to operate. Use the file system
firmware loader interface to read the firmware and load it
onto the controller. This allows keyboards connected to the
type-A ports on these machines to function in U-Boot.
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
Add a variant of readl_poll_sleep_timeout that reads a single
byte to match the readb_poll_timeout API that Linux has.
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
Find the appropriate EFI system partition on the internal NVMe
storage and set the U-Boot environment variables such that
the file system firmware loader can load firmware from it.
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
Apple M2 devices have an MTP coprocessor in charge of keyboard/trackpad
handling, communicating over a DockChannel interface. Add a simple
driver for this.
The keyboard does not require any initialization messages, but we have a
problem: we cannot reset the MTP so Linux can start it fresh, and it
delivers a number of informative packets on startup. To work around
this, we buffer those messages and re-inject them into the FIFO (which
is big enough to hold all of them) on shutdown, so Linux finds them when
it initializes its driver. The actual MTP coprocessor is quiesced, which
does work properly.
Signed-off-by: Hector Martin <marcan@marcan.st>
This driver handles the MTP ASC coprocessor, which does not need any
special handling on the RTKit side and communicates out-of-band.
Signed-off-by: Hector Martin <marcan@marcan.st>
For devices without specific buffer methods, just assume we can give
them raw memory pointers when they request a buffer.
Signed-off-by: Hector Martin <marcan@marcan.st>
Commit 68dcbdd594 ("ARM: imx: Add weak default reset_cpu()") caused
the 'reset' command in U-Boot to not cause a board reset.
Fix it by switching to the watchdog driver model via sysreset, which
is the preferred method for implementing the watchdog reset.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Tested-by: Christian Gmeiner <cgmeiner@igalia.com>
Commit 68dcbdd594 ("ARM: imx: Add weak default reset_cpu()") caused
the 'reset' command in U-Boot to not cause a board reset.
Fix it by switching to the watchdog driver model via sysreset, which
is the preferred method for implementing the watchdog reset.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Currently, an i.MX6 Cuboxi board is incorrectly detected
as the HummingBoard model:
U-Boot 2024.04-rc5 (Mar 26 2024 - 15:59:22 +0100)
CPU: Freescale i.MX6Q rev1.3 996 MHz (running at 792 MHz)
CPU: Extended Commercial temperature grade (-20C to 105C) at 26C
Reset cause: POR
Model: SolidRun HummingBoard2 Dual/Quad (1.5som+emmc)
gpio@20a4000: set_dir_flags: error: gpio GPIO3_8 not reserved
gpio@20a4000: get_value: error: gpio GPIO3_8 not reserved
gpio@20a8000: set_dir_flags: error: gpio GPIO4_4 not reserved
gpio@20a8000: get_value: error: gpio GPIO4_4 not reserved
gpio@20b0000: set_dir_flags: error: gpio GPIO6_9 not reserved
gpio@20b0000: get_value: error: gpio GPIO6_9 not reserved
Board: MX6 HummingBoard
DRAM: 2 GiB
...
This error happens because request_detect_gpios() uses the GPIO DM
API, but board_type() still uses the legacy non-DM GPIO API.
Fix it by using the GPIO DM API in board_type() to read the
board revision pins in SPL.
Reported-by: Christian Gmeiner <cgmeiner@igalia.com>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Tested-by: Christian Gmeiner <cgmeiner@igalia.com>
When Falcon Mode is enabled on RISC-V, use CONFIG_VAL
to check PAYLOAD_ARGS_ADDR, not CONFIG_IS_ENABLED.
Fixes: 10c4ab898c ("spl: riscv: falcon: move fdt blob to specified address")
Signed-off-by: Randolph <randolph@andestech.com>
Tested-by: Leo Yu-Chi Liang <ycliang@andestech.com>
While meson_sm_read_efuse() doesn't overflow, the string is not
zero terminated and env_set*() will buffer overflow and add random
characters to environment.
Acked-by: Viacheslav Bocharov <adeep@lexina.in>
Link: https://lore.kernel.org/r/20240320-u-boot-fix-p200-serial-v2-1-972be646a301@linaro.org
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Add a fall-back method to generate ethaddr from CPU serial on p200 boards
if the MAC cannot be read from efuse. This prevents random MAC addresses
on the WeTek Hub/Play2 boards.
Signed-off-by: Christian Hewitt <christianshewitt@gmail.com>
Link: https://lore.kernel.org/r/20240324151905.3817732-3-christianshewitt@gmail.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Placing the snps,reset content needed for Ethernet to probe in a common
uboot.dtsi results in the content not being used and broken Ethernet. Fix
this by creating two board specific dtsi files with the right content.
Fixes: 67d5128df9 ("ARM: dts: add support for WeTek Hub and WeTek Play2")
Signed-off-by: Christian Hewitt <christianshewitt@gmail.com>
Link: https://lore.kernel.org/r/20240324151905.3817732-2-christianshewitt@gmail.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Increase the SPL_STACK_R_MALLOC_SIMPLE_LEN to 0x200000 to accommodate the
size of tispl.bin fit image. With the recent upgrade of ti-linux-firmware
from version v9.1.0 to v9.2.5, the size of tispl.bin fit image has
increased to 1.4MB, causing allocation errors in the R5 SPL:
```
alloc space exhausted
Could not get FIT buffer of 1325056 bytes
check CONFIG_SPL_SYS_MALLOC_SIZE
```
Signed-off-by: Hiago De Franco <hiago.franco@toradex.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
The Linaro mail address is not valid anymore.
Cc: kojima.masahisa@socionext.com
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Masahisa Kojima <kojima.masahisa@socionext.com>
The lowest speed grade of Toradex AM62 SoMs is K speed, resulting in a
max value of 800MHz for the CPU operating frequency. A solution with
runtime selection of the CPU frequency is already planned to avoid these
kinds of problems in the future.
Fixes: 8fb8a6d499 ("arm: dts: k3-am625-verdin-r5:Change CPU frequency to 1000MHz")
Signed-off-by: Joao Paulo Goncalves <joao.goncalves@toradex.com>
Since commit cc7df0b9e8 ("serial: lpuart: Enable IPG clock")
the apalis-imx8qm board no longer boots.
The reason is that the imx8qm clock driver does not handle the
LPUART IPG clocks inside get_rate(), set_rate() and enable() functions.
Fix the boot regression by adding the LPUART IPG entries.
Fixes: cc7df0b9e8 ("serial: lpuart: Enable IPG clock")
Reported-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Since commit cc7df0b9e8 ("serial: lpuart: Enable IPG clock")
the colibri-imx8qxp board no longer boots.
The reason is that the imx8qxp clock driver does not handle the
LPUART IPG clocks inside get_rate(), set_rate() and enable() functions.
Fix the boot regression by adding the LPUART IPG entries.
Fixes: cc7df0b9e8 ("serial: lpuart: Enable IPG clock")
Reported-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Tested-by: Hiago De Franco <hiago.franco@toradex.com> # Toradex Colibri iMX8X
Acked-by: Sean Anderson <seanga2@gmail.com>
The "Series-patchwork-url:" tag description says that it overrides the
settings file but doesn't specify the name of the setting. Elsewhere
in the documentation about the "useful" settings we see a setting that
sounds promising called "patchwork_server" that's actually not a valid
setting. It should be "patchwork_url".
Fix these problems so the doc is right and more useful.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
If bootefi selftest is executed and a problem with the device-tree
installation occurs, efi_install_fdt() writes sensible error messages.
It never returns EFI_INVALID_PARAMETER. It neither makes sense to check
for EFI_INVALID_PARAMETER nor to show the usage help for the bootefi
command in this case.
Fixes: 296faf4f7e ("cmd: bootefi: re-organize do_bootefi()")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
If an EFI binary returns an error code EFI_INVALID_PARAMETER, we show the
usage help for the bootefi command:
Shell> exit 0x8000000000000002
## Application failed, r = 2
bootefi - Boots an EFI payload from memory
Usage:
bootefi <image address>[:<image size>] [<fdt address>]
- boot EFI payload
bootefi bootmgr [fdt address]
- load and boot EFI payload based on BootOrder/BootXXXX variables.
If specified, the device tree located at <fdt address> gets
exposed as EFI configuration table.
This makes no sense.
Fixes: 296faf4f7e ("cmd: bootefi: re-organize do_bootefi()")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
We should not try to remove protocol interfaces from a NULL handle.
efi_run_image() should always return the return code of the executed EFI
binary.
Fixes: 6422820ac3 ("efi_loader: split unrelated code from efi_bootmgr.c")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Update and improve the building documentation of Apalis iMX8.
The following changes were made:
- imx-atf repository changed to nxp-imx GitHub.
- imx-atf branch updated to 'lf_v2.6'.
- imx-seco updated to version 5.8.7.
- nxp-imx mfgtools link updated to GitHub releases.
- General writing improvements.
Signed-off-by: Hiago De Franco <hiago.franco@toradex.com>
Update and improve the building documentation of Colibri iMX8X.
The following changes were made:
- imx-atf repository changed to nxp-imx GitHub.
- imx-atf branch updated to 'lf_v2.6'.
- imx-seco updated to version 5.8.7.
- nxp-imx mfgtools link updated to GitHub releases.
- General writing improvements.
Signed-off-by: Hiago De Franco <hiago.franco@toradex.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
On imx8m[m|p|q].dtsi, upstream Linux uses different names for NPU/VPU
IP block nodes. It leads variants without such HW block having it
enabled by default.
This patch adds the upstream Linux node's paths to the disable list while
keep the compatibility with downstream Linux.
Signed-off-by: Vitor Soares <vitor.soares@toradex.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Commit bfb3409d67 ("imx: toradex/apalis-imx8: correct SCU API usage")
made an incorrect logic change in the error code check of
sc_misc_otp_fuse_read():
- if (scierr == SC_ERR_NONE) {
+ if (scierr) {
/* QP has one A72 core disabled */
is_quadplus = ((val >> 4) & 0x3) != 0x0;
}
The other changes in this commit are correct.
sc_misc_otp_fuse_read() returns 0 on a successful fuse read.
This inversion causes board_mem_get_layout() to report incorrect RAM size.
Go back the original error check logic to fix the problem.
Fixes: bfb3409d67 ("imx: toradex/apalis-imx8: correct SCU API usage")
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Acked-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Commit aa6e698a7a ("imx: toradex/colibri-imx8x: correct SCU API usage")
made an incorrect logic change in the error code check of
sc_misc_otp_fuse_read():
- if (sc_err == SC_ERR_NONE) {
+ if (sc_err) {
/* DX has two A35 cores disabled */
return (val & 0xf) != 0x0;
}
The other changes in this commit are correct.
sc_misc_otp_fuse_read() returns 0 on a successful fuse read.
This inversion causes board_mem_get_layout() to report incorrect RAM size.
Go back the original error check logic to fix the problem.
Fixes: aa6e698a7a ("imx: toradex/colibri-imx8x: correct SCU API usage")
Reported-by: Hiago De Franco <hiago.franco@toradex.com>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Acked-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Tested-by: Hiago De Franco <hiago.franco@toradex.com> # Toradex Colibri iMX8X 1GB
define the phytec_get_som_type function also when the SoM detection is
disabled.
Fixes:
commit 110d321a56 ("board: phytec: common: phytec_som_detection: Add phytec_get_som_type")
Signed-off-by: Benjamin Hahn <B.Hahn@phytec.de>
- mips: fix bug in cache init on MIPS32r2 or later
-----BEGIN PGP SIGNATURE-----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=Qt/+
-----END PGP SIGNATURE-----
Merge tag 'mips-fixes-for-v2024.04' of https://source.denx.de/u-boot/custodians/u-boot-mips
- mips: implement __udivdi3 to fix building of SquashFS
- mips: fix bug in cache init on MIPS32r2 or later
The intention of change_k0_cca() is to read the C0.Config register into
register $t0, update $t0 with the new cache coherency mode passed in $a0
and write back $t0 to C0.Config. With MIPS32 R2 or later instruction
sets, this can be achieved with a single instruction with INS. The
source and destination register of the INS instruction is passed as
first parameter. In case of change_k0_cca() it is register $t0. But
for writing back the updated value to C0.Config, the incorrect $a0
register is used. This is only correct in the MIPS32 R1 code path.
Fix the `mtc0` instruction to write back the value of the $t0 register.
Fix the MIPS32 R1 code path to also store the updated value in $t0.
Reported by user ddqxy138 on Github.
b838586086
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
