Much of the code in zimage.c deals with the zboot command. Move it into
a sepatate zboot.c file within the cmd/ directory. This will eventually
allow use of the zimage logic without the command being enabled.
Signed-off-by: Simon Glass <sjg@chromium.org>
In preparation for splitting the zboot-command code into a separate
file, move the definitions into the header file.
While we are here, mention when load_address and base_ptr are set up
and explain bzimage_addr better. Make cmdline const since it cannot be
changed.
Signed-off-by: Simon Glass <sjg@chromium.org>
Xilinx changes for v2024.07-rc1
xilinx:
- Do not call env_get_location when !ENV_IS_NOWHERE
- Add FDT_FIXUP_PARTITIONS support
- Fix legacy format MAC decoding
zynqmp:
- Enable semihosting SPL support
- DT updates
- Kconfig resort/cleanup
- Don't describe second image/capsule if !SPL
- Add support for dfu/capsule description via MTD
- Support JTAG as alternative boot mode
- Add support for TEG soc variant
zynqmp-kria:
- Wire usb4 boot device
- Update SDIO tristate pin configuration
- Disable SPI_FLASH_BAR to avoid issue with SPI after update
mbv:
- Enable SPL and binman
- Small platform changes
zynqmp-nand:
- Error out in case of unsupported SW ECC
- Clean error path
versal-net:
- Support multiple locations for variables
The str[] buffer declared in make_flamegraph() is used to hold strings
representing the full call-stacks recorded in traces. The size of this
buffer is currently 500 characters and this works well for the documented
examples.
However, it is possible to exhaust this buffer when processing traces
captured when running the UEFI shell on aarch64 sandbox for example.
Indeed, the maximum length needed for such traces can reach 780 characters.
As it is difficult to evaluate the maximum size that would ever be needed
for all the possible traces, let's use a dynamically allocated `abuf'
instead, which we reallocate when needed.
This fixes the following error:
String too short (500 chars)
While at it, fix a few typos in strings and comments.
Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Tom Rini <trini@konsulko.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Michal Simek <michal.simek@amd.com>
Missing line-feeds in error messages lead to output like:
phy_startup() failed: -110FAILED: -110=>
Output like the following is much easier to read:
phy_startup() failed: -110
FAILED: -110
=>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
The NFS protocol uses file handles to refer to file or directory.
In NFSv2 file handles have a fixed size of 32 bytes.
In NFSv3 file handles have a variable length up to 64 bytes. This is
also true for the MOUNT protocol. [1]
When the NFSv3 server replies with a file handle length > 32 bytes, U-Boot
only copies 32 bytes of that file handle and the next LOOKUP Call fails:
BIOS> nfs ${loadaddr} 192.168.1.51:/nfsroot/opos93dev-br/boot/Image
Using ethernet@428a0000 device
File transfer via NFS from server 192.168.1.51; our IP address is 192.168.1.133
Filename '/nfsroot/opos93dev-br/boot/Image'.
Load address: 0x80400000
Loading: *** ERROR: File lookup fail
done
BIOS>
Looking at this transfer in Wireshark, we can see that the server
replies with the following file handle:
length: 36
[hash (CRC-32): 0x230ac67b]
FileHandle: 0100070101005e000000000091763911f87c449fa73c298552db19ba0c9f60002980cfd2
and U-Boot sends the following file handle in the next LOOKUP Call:
length: 32
[hash (CRC-32): 0x6314131b]
FileHandle: 000000240100070101005e000000000091763911f87c449fa73c298552db19ba
Fix this by using a variable length file handle for dirfh.
[1] https://www.rfc-editor.org/rfc/rfc1813.html#page-106
Fixes: b0baca9820 ("net: NFS: Add NFSv3 support")
Signed-off-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Tested-by: Andrew Davis <afd@ti.com>
Open for reading as O_RDONLY instead of O_RDWR:
the only usage of the fd is for the single read() below;
this prevented
mkimage -f auto -A arm64 \
-T kernel -C lz4 -d Image-6.6.15.lz4 \
-b mt8173-elm-hana-6.6.15.dtb outf
when the inputs were unwritable.
Link: https://bugs.debian.org/1063097
Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Reviewed-by: Dragan Simic <dsimic@manjaro.org>
When accessing an ext2 system the message "File System is consistent\n" is
shown after each write. This is superfluous noise. Only write a debug
message.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
LDR format files are used primarily by Analog Devices processors but may
be of interest to other vendors. Previously support existed for this
format as part of the U-Boot build, but it has been unmaintained and
unused for a long time. In preparation for adding support for modern ADI
processors that use LDR, modernize the LDR support:
- Introduce CONFIG_LDR_CPU as the CPU string recognized by the LDR tool
may not be the same as CONFIG_CPU
- Add an SPL target that repackages u-boot-spl inside an LDR file
An almost identical target for packaging u-boot into an LDR file already
exists and did not need to be created.
Co-developed-by: Nathan Barrett-Morrison <nathan.morrison@timesys.com>
Signed-off-by: Nathan Barrett-Morrison <nathan.morrison@timesys.com>
Signed-off-by: Ian Roberts <ian.roberts@timesys.com>
Signed-off-by: Greg Malysa <greg.malysa@timesys.com>
With configs such as "am64x_evm_a53" or "imx8mp_venice" which list
multiple device trees to build we get a warning such as:
scripts/Makefile.spl:578: target 'spl/dts/freescale/' given more than once in the same rule
If we sort this list first the warning goes away.
Tested-by: Michal Simek <michal.simek@amd.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
When a file is created in the linux and corresponding file permission
is set, if the file needs to be modified in uboot during the startup
process, the modified file permission will be reset to 755. Therefore,
when the ext4fs_write() function is called, if the file already exists,
the file permission of the new file is equal to the file permission of
the existing file.
Currently if ${bootfile} is unset and sysboot is invoked with no
filename specified then U-Boot will crash will a null-pointer
dereference. Add the missing check and a matching error print.
Fixes: 993c912d30 ("cmd: sysboot: Create a sysboot command dedicated file")
Signed-off-by: Caleb Connolly <caleb.connolly@linaro.org>
The implementation of map_range() creates the requested mapping by
walking the page tables, iterating over multiple PTEs and/or descending
into existing table mappings as needed. When doing so, it assumes any
pre-existing valid PTE to be a table mapping. This assumption is wrong
if the platform code attempts to successively map two overlapping ranges
where the latter intersects a block mapping created for the former.
As a result, map_range() treats the existing block mapping as a table
mapping and descends into it i.e. starts interpreting the
previously-mapped range as an array of PTEs, writing to them and
potentially even descending further (extra fun with MMIO ranges!).
Instead, pass any valid non-table mapping to split_block(), which
ensures that it actually was a block mapping (calls panic() otherwise)
before splitting it.
Fixes: 41e2787f5e ("arm64: Reduce add_map() complexity")
Signed-off-by: Pierre-Clément Tosi <ptosi@google.com>
Tested-by: Fabio Estevam <festevam@gmail.com>
Tested-by: Hiago De Franco <hiago.franco@toradex.com> # Toradex Verdin AM62
Reviewed-by: Marc Zyngier <maz@kernel.org>
The Kria board features a recovery application that activates
when the FW_EN button is pressed.
Upon power-up flash operates in 3B mode, However, the recovery
application changes it back to 4B mode.
Currently, after a reset, u-boot activates CONFIG_SPI_FLASH_BAR
and assumes the flash is in 3B mode. However, there's no code
or reset lines connected to the flash that could return it to
3B mode. To resolve this issue, changes were made to disable
CONFIG_SPI_FLASH_BAR, which activates 4-byte opcodes.
Signed-off-by: Tejas Bhumkar <tejas.arvind.bhumkar@amd.com>
Link: https://lore.kernel.org/r/20240409094826.4131643-1-tejas.arvind.bhumkar@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
The differences between the Milk-V Mars board and the VisionFive 2 board
are small enough that we can support both using the same U-Boot build.
* The model and compatible property are taken from proposed Linux patches.
* The EEPROM is atmel,24c02 according to the vendor U-Boot.
* The second Ethernet port is not available.
usb@10100000 does not exist in U-Boot yet. So we don't have to reflect
differences in usage here.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
Export a function get_product_id_from_eeprom() to read the product ID.
This value can be used for fixing up the device-tree on JH7110 based
products.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
Currently in set_fdtfile() we set the value of environment variable fdtfile
unconditionally. The implies that a value in the environment will be
ignored.
With the patch environment variable fdtfile will only be set if it does not
yet exist. This requires that CONFIG_DEFAULT_FDT_FILE is not set.
Now the user can either set and save fdtfile interactively or in the U-Boot
configuration to overrule the device-tree name chosen based on the
hardware in set_fdtfile().
Reported-by: E Shattow <lucent@gmail.com>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
When virtio_init() gets called from board_init() PCI isn't ready. Thus,
virtio-over-PCI (e.g. network interfaces) devices can't be detected and
used without additional `virtio scan` scan in the shell or a script.
Signed-off-by: Łukasz Stelmach <l.stelmach@samsung.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
A new property has been added, with an extensive rationale at [1], that
can be used in place of "riscv,isa" to indicate what extensions are
supported by a given platform that is a list of strings rather than a
single string. There are some differences between the new property,
"riscv,isa-extensions" and the incumbent "riscv,isa" - chief among them
for the sake of parsing being the list of strings, as opposed to a
string. Another advantage is strictly defined meanings for each string
in a dt-binding, rather than deriving meaning from RVI standards. This
will likely to some divergence over time, but U-Boot's current use of
extension detection is very limited - there are just four callsites of
supports_extension() in mainline U-Boot.
These checks are limited to two checks for FPU support and two checks
for "s" and "u". "s" and "u" are not supported by the new property, but
they were also not permitted in "riscv,isa". These checks are only
meaningful (or run) in M-Mode, in which case supports_extension() does
not parse the devicetree anyway.
Add support for the new property in U-Boot, prioritising it, before
falling back to the, now deprecated, "riscv,isa" property if it is not
present.
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
cpu_get_desc() for the RISC-V CPU currently reads "riscv,isa" to get
the description, but it is no longer a required property and cannot be
assummed to always be present, as the new "riscv,isa-extensions" and
"riscv,isa-base" properties may be present instead.
On RISC-V, cpu_get_desc() has two main uses - firstly providing an
informational name for the CPU for smbios or at boot with
DISPLAY_CPUINFO etc and secondly it forms the basis of ISA extension
detection in supports_extension() as it returns (a portion of) an ISA
string.
cpu_get_desc() returns a string, which aligned with "riscv,isa" but
the new property is a list of strings. Rather than add support for
the list of strings property, which would require creating an isa
string from "riscv,isa-extensions", modify the RISC-V CPU's
implementaion of cpu_get_desc() return the first compatible as the
cpu description instead. This may be fine for the informational cases,
but it would break extension dtection, given supports_extension()
expects cpu_get_desc() to return an ISA string.
Call dev_read_string() directly in supports_extension() to get the
contents of "riscv,isa" so that extension detection remains functional.
As a knock-on affect of this change, extension detection is no longer
broken for long ISA strings. Previously if the ISA string exceeded the
32 element array that supports_extension() passed to cpu_get_desc(),
it would return ENOSPC and no extensions would be detected.
This bug probably had no impact as U-Boot does not currently do anything
meaningful with the results of supports_extension() and most SoCs
supported by U-Boot don't have anywhere near that complex of an ISA
string. The QEMU virt machine's CPUs do however, so extension detection
doesn't work there.
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
Add dcache operations invalidate_dcache_range and flush_dcache_range for
cv1800b.
Signed-off-by: Kongyang Liu <seashell11234455@gmail.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang@andestech.com>
When debugging, it is useful to have a backtrace to find
out what is in the call stack as the previous function (RA)
may not have been the culprit.
Since this adds size to the build, do not add it by default
and avoid putting it in the SPL build if not needed.
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Tested-by: Leo Yu-Chi Liang <ycliang@andestech.com>
Pull request efi-2024-07-rc1
Documentation:
* improve description of FAT partition name generation
* add missing :: in doc/usage/cmd/itest.rst
UEFI:
* fix address mode for __efi_runtime_start/stop,
__efi_runtime_rel_start/stop
* fix size of variable attribute constants
* enable booting via EFI boot manager by default
* correct the sequence of the EFI boot methods
* correct finding the default EFI binary
* don't delete variable from memory if update failed
* fix append write behavior to non-existent variable
* Use binman for testing capsule updates on the sandbox
* Consider capsule test files in .gitignore and make clean
A symbol defined in a linker script (e.g. __efi_runtime_rel_start = .;)
is only a symbol, not a variable and should not be dereferenced.
The common practice is either define it as
extern uint32_t __efi_runtime_rel_start or
extern char __efi_runtime_rel_start[] and access it as
&__efi_runtime_rel_start or __efi_runtime_rel_start respectively.
So let's access it properly since we define it as an array
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
A symbol defined in a linker script (e.g. __efi_runtime_start = .;) is
only a symbol, not a variable and should not be dereferenced.
The common practice is either define it as
extern uint32_t __efi_runtime_start or
extern char __efi_runtime_start[] and access it as
&__efi_runtime_start or __efi_runtime_start respectively.
So let's access it properly since we define it as an array
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
* The sandbox must not use an arbitrary file name bootsbox.efi but the
file name matching the host architecture to properly boot the respective
file. We already have an include which provides a macro with the name of
the EFI binary. Use it.
* The path to the EFI binary should be absolute.
* The path and the file name must be capitalized to conform to the UEFI
specification. This is important when reading from case sensitive
file systems.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tested-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
efi_default_filename.h requires HOST_ARCH to be defined. Up to now we
defined it via a CFLAGS. This does not scale. Add the symbol to
version_autogenerated.h instead.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
If UEFI is enabled in U-Boot, we want it to conform to the UEFI
specification. This requires enabling the boot manager boot method.
Reported-by: E Shattow <lucent@gmail.com>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
The default sequence of boot methods is determined by alphabetical sorting
during linkage.
* efi_mgr must run before efi to be UEFI compliant
* pxe should run as last resort
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Our efi_var_mem_xxx() functions don't have a replace variant. Instead we
add a new variable and delete the old instance when trying to replace a
variable. Currently we delete the old version without checking the new
one got added
Signed-off-by: Ilias Apalodimas <apalos@gmail.com>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
We don't yet support EFI_VARIABLE_ENHANCED_AUTHENTICATED_ACCESS for file
based variables, but we should pass it to TEE based variable stores.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
GetVariable() and SetVariable() only accept a 32bit value for attributes.
It makes not sense to define EFI_VARIABLE_READ_ONLY as unsigned long.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
GetVariable() and SetVariable() use an uint32_t value for attributes.
The UEFI specification defines the related constants as 32bit.
Add the missing EFI_VARIABLE_ENHANCED_AUTHENTICATED_ACCESS constant.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Current "variables" efi_selftest result is inconsistent
between the U-Boot file storage and the tee-based StandaloneMM
RPMB secure storage.
U-Boot file storage implementation does not accept SetVariale
call to non-existent variable with EFI_VARIABLE_APPEND_WRITE,
it return EFI_NOT_FOUND.
However it is accepted and new variable is created in EDK II
StandaloneMM implementation if valid data and size are specified.
If data size is 0, EFI_SUCCESS is returned.
Since UEFI specification does not clearly describe the behavior
of the append write to non-existent variable, let's update
the U-Boot file storage implementation to get aligned with
the EDK II reference implementation.
Signed-off-by: Masahisa Kojima <kojima.masahisa@socionext.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tested-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
List all prefix currently used for generating FAT partition names.
Describe which device class uses which prefix.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
A certain set of capsule files are now generated as part of the
sandbox build. Add these files to the CLEAN_FILES list for deletion on
invoking any of the cleanup targets.
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canoncal.com>
Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org>
