smbios: Fallback to the default DT if sysinfo nodes are missing
In order to fill in the SMBIOS tables U-Boot currently relies on a
"u-boot,sysinfo-smbios" compatible node. This is fine for the boards
that already include such nodes. However with some recent EFI changes,
the majority of boards can boot up distros, which usually rely on
things like dmidecode etc for their reporting. For boards that
lack this special node the SMBIOS output looks like:
System Information
Manufacturer: Unknown
Product Name: Unknown
Version: Unknown
Serial Number: Unknown
UUID: Not Settable
Wake-up Type: Reserved
SKU Number: Unknown
Family: Unknown
This looks problematic since most of the info are "Unknown". The DT spec
specifies standard properties containing relevant information like
'model' and 'compatible' for which the suggested format is
<manufacturer,model>. Unfortunately the 'model' string found in DTs is
usually lacking the manufacturer so we can't use it for both
'Manufacturer' and 'Product Name' SMBIOS entries reliably.
So let's add a last resort to our current smbios parsing. If none of
the sysinfo properties are found, scan for those information in the
root node of the device tree. Use the 'model' to fill the 'Product
Name' and the first value of 'compatible' for the 'Manufacturer', since
that always contains one.
pre-patch:
Handle 0x0001, DMI type 1, 27 bytes
System Information
Manufacturer: Unknown
Product Name: Unknown
Version: Unknown
Serial Number: 100000000bb24ceb
UUID: 30303031-3030-3030-3061-613234636435
Wake-up Type: Reserved
SKU Number: Unknown
Family: Unknown
[...]
and post patch:
Handle 0x0001, DMI type 1, 27 bytes
System Information
Manufacturer: raspberrypi
Product Name: Raspberry Pi 4 Model B Rev 1.1
Version: Unknown
Serial Number: 100000000bb24ceb
UUID: 30303031-3030-3030-3061-613234636435
Wake-up Type: Reserved
SKU Number: Unknown
Family: Unknown
[...]
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Peter Robinson <pbrobinson@gmail.com> Tested-by: Peter Robinson <pbrobinson@gmail.com>
If a value is not valid during the DT or SYSINFO parsing, we explicitly
set that to "Unknown Product" and "Unknown" for the product and
manufacturer respectively. It's cleaner if we move the checks insisde
smbios_add_prop_si() and provide an alternative string in case the
primary is NULL or empty
pre-patch dmidecode
<snip>
Handle 0x0001, DMI type 1, 27 bytes
System Information
Manufacturer: Unknown
Product Name: Unknown Product
Version: Not Specified
Serial Number: Not Specified
UUID: Not Settable
Wake-up Type: Reserved
SKU Number: Not Specified
Family: Not Specified
[...]
post-patch dmidecode:
Handle 0x0001, DMI type 1, 27 bytes
System Information
Manufacturer: Unknown
Product Name: Unknown
Version: Unknown
Serial Number: Unknown
UUID: Not Settable
Wake-up Type: Reserved
SKU Number: Unknown
Family: Unknown
[...]
While at it make smbios_add_prop_si() add a string directly if the prop
node is NULL and replace smbios_add_string() calls with
smbios_add_prop_si(ctx, NULL, ....)
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Peter Robinson <pbrobinson@gmail.com> Tested-by: Peter Robinson <pbrobinson@gmail.com>
pxe_utils: Increase feedback to user when fdt file is not found
extlinux.conf can set fdtdir. fdtdir look for fdt file using
information found in the environment variable. The function does
not report any error in the case the file is not found
Scanning for bootflows in all bootdevs
Seq Method State Uclass Part Name Filename
--- ----------- ------ -------- ---- ------------------------ ----------------
Scanning global bootmeth 'efi_mgr':
No EFI system partition
No EFI system partition
Failed to persist EFI variables
Scanning bootdev 'mmc@fa10000.bootdev':
0 extlinux ready mmc 1 mmc@fa10000.bootdev.part_ /boot/extlinux/extlinux.conf
** Booting bootflow 'mmc@fa10000.bootdev.part_1' with extlinux
1: am62x-sk-buildroot
Retrieving file: /boot/Image
append: console=ttyS2,115200n8 root=PARTUUID=c586a30c-0bf1-4323-aba8-779c814ee135 rw
rootfstype=ext4 rootwait earlycon=ns16550a,mmio32,0x02800000
Retrieving file: /boot/k3-am623_ccm_m3.dtb
Skipping fdtdir /boot/ for failure retrieving dts
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Marek Vasut [Sat, 2 Dec 2023 20:52:31 +0000 (21:52 +0100)]
command: Move command completion temporary buffer to stack
The command completion temporary buffer seems to be only
used by the argv tokenizer, move it to stack. This saves
2 kiB from the binary size (depends on configuration) per:
$ aarch64-linux-gnu-readelf -s u-boot | sort -n -k 3
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org> Reviewed-by: Simon Glass <sjg@chromium.org>
Marek Vasut [Sat, 2 Dec 2023 20:52:30 +0000 (21:52 +0100)]
command: Allocate history buffer using calloc()
The history buffer is currently a static array which can be some
10-40 kiB depending on configuration, and so adds considerably to
the U-Boot binary size. Allocate it dynamically instead to reduce
the U-Boot binary size.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Simon Glass [Wed, 29 Nov 2023 17:31:19 +0000 (10:31 -0700)]
bootflow: Allow setting a cmdline arg with no value
This is supposed to be supported by the 'bootflow cmd' command, at
least according to the help. There is a 'bootflow cmd clear' but it is
often more intuitive to use 'bootcmd cmd set' with an empty value.
Update the command to pass BOOTFLOWCL_EMPTY in this case.
Piotr Kubik [Fri, 24 Nov 2023 16:30:46 +0000 (17:30 +0100)]
bootm: Fix flags used for bootargs string substitution
Commit 51bb33846ad2 ("bootm: Support string substitution in
bootargs") introduced a feature of bootargs string substitution
and changed a flag used in bootm_process_cmdline_env() call
to be either true or false. With this flag value,
condition in bootm_process_cmdline()
`if (flags & BOOTM_CL_SUBST)` is never true
and process_subst() is never called.
Add a simple test to verify if substitution works OK.
Signed-off-by: Piotr Kubik <piotr.kubik@iopsys.eu> Reviewed-by: Simon Glass <sjg@chromium.org>
Robert Catherall [Thu, 23 Nov 2023 14:16:01 +0000 (14:16 +0000)]
arm: afs: Set the filesize env. var. on load
The `afs load` command copies data to RAM. Set the filesize
environment variable to record how much data was 'loaded'
Signed-off-by: Robert Catherall <robert.catherall@arm.com> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Tested-by: Andre Przywara <andre.przywara@arm.com>
Nikita Shubin [Mon, 19 Dec 2022 08:05:27 +0000 (11:05 +0300)]
imxtract: specify max gunzip size
Specify max gunzip size from config to override SYS_XIMG_LEN
default value wich is 0x800000. In case we have a large portion of FIT
image, for example gzipped kernel with decompressed size large than
0x800000 we should enlarge imxract area, otherwise extracting it will
fail.
It used to be a hardcoded define in cmd/ximg.c and we are moving it to
config.
Signed-off-by: Nikita Shubin <n.shubin@yadro.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Tom Rini [Tue, 19 Dec 2023 20:16:13 +0000 (15:16 -0500)]
Merge branch 'staging' of https://source.denx.de/u-boot/custodians/u-boot-tegra into next
This PR contains 4 patchsets:
1. PMIC GPIO cells bringup. Created drivers for MAX7663 and Palmas
PMICs and gpio-uclass patch isolated behind configs for these 2
drivers. No unintentional size increase on any board. (proposed
2023-11-06 without any reaction)
2. Simple PLL clocks support in common tegra clock code which allows
use of simple PLL the same way main PLLs are used (before only
clock_start_pll was available). PLLD2 is an example of simple PLL, it
is used as a video subsystem parent clock and was used to test this
code. So far everything worked as expected. (proposed 2023-11-16
without any reaction)
3. A small patch for tegra emmc to allow pass max frequency from
device tree since some devices may not support full speed.
4. Pinmux DM conversion. Patchset consists of commit with DM wrapper
for existing pinmux code for t20/t30/t114, pinmux and funcmux files
relocation into a dedicated folder inside pinctrl, conversion of some
tegra boards to device tree pinmux setup.
Since implementation of pinctrl driver for T20 Paz00 can switch
to device tree pinmux setup along with remove of board pinmux
and some minor device tree and defconfig tweaks.
Svyatoslav Ryhel [Tue, 28 Nov 2023 11:43:31 +0000 (13:43 +0200)]
board: htc: endeavoru: switch to DM pinmux
Drop the pinmux setup in the board in favor of setting it up in
the device tree. Device tree nodes match nodes used for the Linux
device tree and are set according to the downstream kernel.
Svyatoslav Ryhel [Mon, 27 Nov 2023 17:08:32 +0000 (19:08 +0200)]
board: lg: x3-t30: switch to DM pinmux
Drop the pinmux setup in the board in favor of setting it up in
the device tree. Device tree nodes match nodes used for the Linux
device tree and are set according to the service manual.
Svyatoslav Ryhel [Sun, 26 Nov 2023 15:54:03 +0000 (17:54 +0200)]
drivers: pinctrl: create Tegra DM pinctrl driver
The existing pinctrl driver available for Tegra SOC is well
designed, but it lacks DM support. Let's add a DM compatible
overlay, which allows use of the device tree, along with preserving
backward compatibility with all existing setups and the ability
to use it in SPL board configuration stage.
SPL configs are used only by the ARMv7-based Tegra SOC's, so move
SPL_SYSRESET under TEGRA_ARMV7_COMMON selection and enable SPL_DM
since SPL_SYSRESET depends on it.
Peter Geis [Tue, 19 Dec 2023 13:35:52 +0000 (15:35 +0200)]
mmc: tegra: use max-frequency from device tree if provided
The driver currently hard-codes the max freqency for the sdhci
controllers. If the controller is unable to operate at the max
frequency, the mmc card will not be available on the first scan.
Subsequent scans will eventually find a working combination.
Fix this by allowing the driver to check for the max-frequency
property and default to the original value if it doesn't exist.
Signed-off-by: Peter Geis <pgwipeout@gmail.com> Signed-off-by: Svyatoslav Ryhel <clamor95@gmail.com>
ARM: tegra: clock: support get and set rate for simple PLL
Simple PLL clocks like PLLD2 were omitted since they do not share common
4 register structure with main clocks. Such clocks are containd in simple
PLL group. Only clock_start_pll function supported them. This patch expands
this support on clock_set_rate and clock_get_rate which should make
simple PLL clocks equal to main PLL clocks.
Tom Rini [Tue, 19 Dec 2023 15:15:57 +0000 (10:15 -0500)]
Merge patch series "Fix J7200 kernel boot when using upstream u-boot"
Reid Tonking <reidt@ti.com> says:
Since the 09.01.00.002 release of ti-linux-firmware [0] upstream uboot
has led to the kernel hanging during boot [1] for the TI J7200 S0C. The
issue was found to be a few patches that had be added to ti-u-boot, but not
yet upstreamed. This series adds the missing two patches to allow upstream
u-boot to boot the kernel properly [2].
Bryan Brattlof [Thu, 7 Dec 2023 16:52:10 +0000 (10:52 -0600)]
arm: mach-k3: j72xx: add new 'virtual' mux
In order for the Cortex-A72s to operate at different frequencies other
than the default 2GHz, add in a new 'virtual' mux (a mux that does not
physically exist in the clock tree) that can be selected.
Julien Masson [Mon, 4 Dec 2023 10:48:55 +0000 (11:48 +0100)]
arm: mediatek: add support for MediaTek MT8365 SoC
This patch adds basic support for MediaTek MT8365 SoC.
The dtsi has been copied from Linux source code tag v6.7-rc2.
(commit 9b5d64654ea8f51fe1e8e29ca1777b620be8fb7c)
spl_mmc_emmc_boot_partition return a number different from 0
if the partition is a boot one. We can have the uboot img
for instance in a raw offset in emmc partition 0 so we would
like to continue to load the next stage. If the user want
to use EMMC as boot device allow him to use any part of the
emmc and not only boot partition
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Etienne Carriere [Wed, 29 Nov 2023 12:37:53 +0000 (13:37 +0100)]
tee: optee: don't enumerate services if there ain't any
Change optee driver service enumeration to not enumerate (and
allocate a zero sized shared memory buffer) when OP-TEE
reports that there is no service to enumerate.
This change fixes an existing issue that occurs when the such zero
sized shared memory buffer allocated from malloc() has a physical
address of offset 0 of a physical 4kB page. In such case, OP-TEE
secure world refuses to register the zero-sized shared memory
area and makes U-Boot optee service enumeration to fail.
Etienne Carriere [Wed, 29 Nov 2023 12:37:52 +0000 (13:37 +0100)]
tee: optee: don't fail on services enumeration failure
Change optee probe function to only warn when service enumeration
sequence fails instead of reporting an optee driver probe failure.
Indeed U-Boot can still use OP-TEE even if some OP-TEE services are
not discovered.
Udit Kumar [Sat, 25 Nov 2023 18:29:39 +0000 (23:59 +0530)]
remoteproc: k3-dsp: Avoid reloading of firmware
DSP core is going into abnormal state when load callback is called
after starting of DSP core.
Reload of firmware needs core to be stopped first, followed by
load.
So avoid loading of firmware, when core is started.
Robert Catherall [Thu, 23 Nov 2023 18:22:58 +0000 (18:22 +0000)]
arm: vexpress64: juno: Allow boot from VirtIO
The AEM and Juno FVPs (Fixed Virtual Platforms) support a VirtIO
disc interface. Adding VIRTIO to the list of boot devices allows
these FastModel platforms to boot from 'disc' in the same way
the hardware counterpart can boot from SATA or USB.
This is a NOP if CONFIG_CMD_VIRTIO is not enabled, so no impact
on Juno hardware (which is built with vexpress_aemv8a_juno_defconfig)
Signed-off-by: Robert Catherall <robert.catherall@arm.com> Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Neil Armstrong [Wed, 13 Dec 2023 09:30:13 +0000 (10:30 +0100)]
ARM: dts: sync Amlogic GX DT to Linux v6.4
Sync Amlogic GXBB, GXL & GXM DTs from Linux v6.4, and also
switch to GXL MDIO MUX driver to adapt to DT change,
Most of the changes are only cosmetic or doesn't concern U-Boot,
the most important change for U-Boot is the GXL mdio mux compatible
switch to amlogic,gxl-mdio-mux.
Michal Simek [Mon, 6 Nov 2023 11:56:47 +0000 (12:56 +0100)]
riscv: Add support for AMD/Xilinx MicroBlaze V
MicroBlaze V is new AMD/Xilinx soft-core 32bit RISC-V processor IP.
It is hardware compatible with classic MicroBlaze processor.
The patch contains initial wiring and configuration for initial HW design
with memory, cpu, interrupt controller, timers and uartlite console
(interrupt controller is listed but U-Boot is not using it).
Provided DT is just describing one configuration and should be taken only
as example.
Signed-off-by: Michal Simek <michal.simek@amd.com> Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com> Reviewed-by: Padmarao Begari <padmarao.begari@microchip.com>
Sean Anderson [Sat, 2 Dec 2023 19:33:14 +0000 (14:33 -0500)]
doc: Replace examples of MD5 and SHA1 with SHA256
Both SHA1 and (especially) MD5 are no longer as safe as they once were for
cryptographic use. Replaces examples which use them with examples using
SHA256 instead. This will provide more-secure defaults for users who use
documentation examples as a base for their own use. This is not too
necessary for non-verified-boot scenarios (since someone could just replace
the checksum), but I wanted to be complete.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Peter Robinson <pbrobinson@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Tom Rini <trini@konsulko.com>
AKASHI Takahiro [Tue, 21 Nov 2023 01:29:44 +0000 (10:29 +0900)]
cmd: bootefi: move library interfaces under lib/efi_loader
In the prior commits, interfaces for executing EFI binary and boot manager
were carved out. Move them under efi_loader directory so that they can
be called from other places without depending on bootefi command.
Only efi_selftest-related code will be left in bootefi.c.
AKASHI Takahiro [Tue, 21 Nov 2023 01:29:43 +0000 (10:29 +0900)]
cmd: bootefi: localize global device paths for efi_selftest
Device paths allocated in bootefi_test_prepare() will be immediately
consumed by do_efi_selftest() and there is no need to keep them for later
use. Introduce test-specific varialbles to make it easier to move other
bootmgr functions into library directory in the next commit.
Andrew Davis [Thu, 30 Nov 2023 14:49:11 +0000 (08:49 -0600)]
board: ti: k3: Remove need for CFG_SYS_SDRAM_BASE
The base address of extended DDR does not change across the K3 family.
Setting this per SoC is not needed. Remove this definition to help
remove the last bits from K3 include/configs/*.h files.
Vishal Mahaveer [Tue, 28 Nov 2023 19:40:24 +0000 (13:40 -0600)]
board: ti: am62x/am62ax: Update virtual interrupt allocations in board config
Updates as a result of TIFS core now reserving a virtual interrupt
for enabling interrupts between DM to TIFS core. Because of this
change other virtual interrupt counts decrease by one.
Update am62ax rm-cfg with allocation entries for C7x core. Following
updates are added for C7x:
- Share split BCDMA tx and rx channels between DM R5 and C7x
- Share rings for split BCDMA tx and rx channels between DM R5 and C7x
- Add Global events and Virtual interrupts for C7x
Andrew Davis [Tue, 28 Nov 2023 17:05:27 +0000 (11:05 -0600)]
arm: mach-k3: Remove non-cached memory map areas
All normal memory areas should be mapped as such.
We added these un-cached holes in our memory map to hack around the
remoteproc driver missing the proper cache maintenance operations.
The problem is having these non-cached memory map areas causes stability
issues later in system operation due to the nature of the K3 coherency
architecture. Plus these are board specific carveouts and instead
should have been added at the board level, not here in the SoC common
code area.
Remove these non-cached memory map areas.
Signed-off-by: Andrew Davis <afd@ti.com> Reviewed-by: Nishanth Menon <nm@ti.com> Tested-by: Nishanth Menon <nm@ti.com>
Andrew Davis [Tue, 28 Nov 2023 17:05:26 +0000 (11:05 -0600)]
arm: mach-k3: Do not map ATF and OPTEE regions in MMU
ATF and OPTEE regions may be firewalled from non-secure entities. To
prevent access to this area we leave a hole there in the MMU map. This
is the same idea as [0] but we complete that patch by adding the same
for AM65, J721e, J7200, and J721s2 here.
[0] commit 0688ff3ae23c ("arm: mach-k3: arm64-mmu: do not map ATF and OPTEE regions in A53 MMU")
Andrew Davis [Tue, 28 Nov 2023 17:05:25 +0000 (11:05 -0600)]
arm: mach-k3: Let the compiler size the mem_map lists
NR_MMU_REGIONS is a copy/paste from another platform that extends
this list later. We do not do that, so let the list be the size
of the initializer list.
Signed-off-by: Andrew Davis <afd@ti.com> Reviewed-by: Nishanth Menon <nm@ti.com> Tested-by: Nishanth Menon <nm@ti.com>
Andrew Davis [Wed, 22 Nov 2023 21:30:05 +0000 (15:30 -0600)]
arm: mach-k3: Move K3 common schema.yaml out of board directory
This file is common for all K3, move it out of board/ directory and
into mach-k3. As we need to change the path in k3-binman.dtsi let's
take this opportunity to switch to absolute paths which makes adding
non-TI boards (like Toradex Verdin) not need to override these paths.
Tom Rini [Fri, 15 Dec 2023 18:33:11 +0000 (13:33 -0500)]
Merge tag 'u-boot-stm32-20231215' of https://source.denx.de/u-boot/custodians/u-boot-stm into next
_ run savedefconfig on all STM32 defconfig
STM32 MCU:
_ Sync stm32f469-disco DT with Linux 6.5
_ rework ltdc node for stm32f769-disco
_ clk: stm32f: Fix settings for LCD_CLK
_ Support display on stm32f469-disco board
Jim Liu [Tue, 14 Nov 2023 09:00:04 +0000 (17:00 +0800)]
clk: nuvoton: add read only feature for clk driver
Add a flag to set ahb/apb/fiu/spi clock divider as read-only
The spi clock setting is related to booting flash, it is setup by early
bootloader.
It just protects the clock source and can't modify it in uboot.
Igor Prusov [Thu, 9 Nov 2023 10:55:13 +0000 (13:55 +0300)]
clk: Add dump operation to clk_ops
This adds dump function to struct clk_ops which should replace
soc_clk_dump. It allows clock drivers to provide custom dump
implementation without overriding generic CCF dump function.
Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Tested-by: Patrice Chotard <patrice.chotard@foss.st.com> Reviewed-by: Sean Anderson <seanga2@gmail.com> Signed-off-by: Igor Prusov <ivprusov@sberdevices.ru>
[ Fixed parameter name in documentation ] Signed-off-by: Sean Anderson <seanga2@gmail.com> Link: https://lore.kernel.org/r/20231109105516.24892-6-ivprusov@sberdevices.ru
Igor Prusov [Tue, 5 Dec 2023 23:23:33 +0000 (02:23 +0300)]
clk: Check that composite clock's div has set_rate()
It's possible for composite clocks to have a divider that does not
implement set_rate() operation. For example, sandbox_clk_composite()
registers composite clock with a divider that only has get_rate().
Currently clk_composite_set_rate() only checks thate rate_ops are
present, so for sandbox it will cause NULL dereference during
clk_set_rate().
This patch adds rate_ops->set_rate check tp clk_composite_set_rate().
Yang Xiwen [Fri, 10 Nov 2023 19:19:52 +0000 (03:19 +0800)]
clk: check parent_name in clk_register to avoid confusing log_error() output
For some gate clocks and fixed clocks without a parent, calling
clk_register will print an useless error message indicating that parent
is missing. Fix that by gaurding log_xxx() with an if-statement.
Tom Rini [Fri, 15 Dec 2023 14:41:44 +0000 (09:41 -0500)]
Merge patch series "bootm: Handle compressed arm64 images with bootm"
To quote the author:
This little series corrects a problem I noticed with arm64 images,
where the kernel is not recognised if compression is used:
U-Boot> tftp image.fit
Using ethernet@7d580000 device
TFTP from server 192.168.4.7; our IP address is 192.168.4.147
Filename 'image.fit'.
Load address: 0x1000000
Loading: ################################################## 23 MiB
20.5 MiB/s
done
Bytes transferred = 24118272 (1700400 hex)
U-Boot> bootm
## Loading kernel from FIT Image at 01000000 ...
Using 'conf-768' configuration
Trying 'kernel' kernel subimage
Description: Linux
Type: Kernel Image (no loading done)
Compression: gzip compressed
Data Start: 0x01000120
Data Size: 13662338 Bytes = 13 MiB
Verifying Hash Integrity ... OK
Bad Linux ARM64 Image magic!
With this series:
U-Boot> tftp 20000000 image.fit
Using ethernet@7d580000 device
TFTP from server 192.168.4.7; our IP address is 192.168.4.147
Filename 'image.fit'.
Load address: 0x20000000
Loading: ################################################## 23.5 MiB
20.8 MiB/s
done
Bytes transferred = 24642560 (1780400 hex)
U-Boot> bootm 0x20000000
## Loading kernel from FIT Image at 20000000 ...
Using 'conf-768' configuration
Trying 'kernel' kernel subimage
Description: Linux
Type: Kernel Image (no loading done)
Compression: zstd compressed
Data Start: 0x20000120
Data Size: 14333475 Bytes = 13.7 MiB
Verifying Hash Integrity ... OK
Using kernel load address 80000
## Loading fdt from FIT Image at 20000000 ...
Using 'conf-768' configuration
Trying 'fdt-768' fdt subimage
Description: Raspberry Pi 4 Model B
Type: Flat Device Tree
Compression: zstd compressed
Data Start: 0x215f820c
Data Size: 9137 Bytes = 8.9 KiB
Architecture: AArch64
Verifying Hash Integrity ... OK
Uncompressing Flat Device Tree to 3aff3010
Booting using the fdt blob at 0x3aff3010
Working FDT set to 3aff3010
Uncompressing Kernel Image (no loading done) to 80000
Moving Image from 0x80000 to 0x200000, end=2b00000
Using Device Tree in place at 000000003aff3010, end 000000003afff4c4
Working FDT set to 3aff3010
Starting kernel ...
[ 0.000000] Booting Linux on physical CPU 0x0000000000 [0x410fd083]
The problem is that the arm64 magic is checked before the image is
decompressed. However this is only part of it. The kernel_noload image
type doesn't work with compression, since the kernel is not loaded. So
this series deals with that by using an lmb-allocated buffer for the
uncompressed kernel.
Another issue is that the arm64 handling is done too early, before the
image is loaded. This series moves it to after loading, so that
compression can be handled.
A patch is included to show the kernel load-address, so it is easy to
see what is going on.
One annoying feature of arm64 is that the image is often copied to
another address. It might be possible for U-Boot to figure that out
earlier and decompress it to the right place, but perhaps not.
With all of this it should be possible to boot a compressed kernel on
any of the 990 arm64 boards supported by Linux, although I have only
tested two.
Simon Glass [Sun, 19 Nov 2023 14:43:34 +0000 (07:43 -0700)]
bootm: Support kernel_noload with compression
It is not currently possible to execute the kernel in-place without
loading it. Use lmb to allocate memory for it.
Co-developed-by: Tom Rini <trini@konsulko.com> Signed-off-by: Simon Glass <sjg@chromium.org> Suggested-by: Tom Rini <trini@konsulko.com> Signed-off-by: Tom Rini <trini@konsulko.com>
Simon Glass [Sun, 19 Nov 2023 14:43:33 +0000 (07:43 -0700)]
bootm: Move arm64-image processing later
If the image is compressed, then the existing check fails, since the
header is wrong.
Move the check later in the boot process, after the kernel is
decompressed. This allows use of bootm with compressed kernels, while
still permitting an uncompressed kernel to be used.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Tom Rini <trini@konsulko.com>