| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Unfortunately these were removed by mistake. This means that adding hash
support to SPL brings in all software algorithms, with a substantial
increase in code size.
The origin of the problem was renaming them to SPL_FIT_xxx and then these
were removed altogether in a later commit.
Add them back. This aligns with CONFIG_MD5, for example, which has an SPL
variant.
Signed-off-by: Simon Glass <sjg@chromium.org>
Fixes: f5bc9c25f31 ("image: Rename SPL_SHAxxx_SUPPORT to SPL_FIT_SHAxxx")
Fixes: eb5171ddec9 ("common: Remove unused CONFIG_FIT_SHAxxx selectors")
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
The gzip API uses the u64 type in it, which is not available in the host
build. This makes it impossible to include the header file.
We could make this type available, but it seems unnecessary. Limiting the
compression size to that of the 'unsigned long' type seems good enough. On
32-bit machines the limit then becomes 4GB, which likely exceeds available
RAM anyway, therefore it should be sufficient. On 64-bit machines this is
effectively u64 anyway.
Update the header file and implementation to use 'ulong' instead of 'u64'.
Add a definition of u32 for the cases that seem to need exactly that
length. This should be safe enough.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
The existing zstd API requires the same sequence of calls to perform its
task. Create a helper for U-Boot, to avoid code duplication, as is done
with other compression algorithms. Make use of of this from the image
code.
Note that the zstd code lacks a test in test/compression.c and this should
be added by the maintainer.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
When passing a data buffer back from a function, it is not always clear
who owns the buffer, i.e. who is responsible for freeing the memory used.
An example of this is where multiple files are decompressed from the
firmware image, using a temporary buffer for reading (since the
compressed data has to live somewhere) and producing a temporary or
permanent buffer with the resuilts.
Where the firmware image can be memory-mapped, as on x86, the compressed
data does not need to be buffered, but the complexity of having a buffer
which is either allocated or not, makes the code hard to understand.
Introduce a new 'abuf' which supports simple buffer operations:
- encapsulating a buffer and its size
- either allocated with malloc() or not
- able to be reliably freed if necessary
- able to be converted to an allocated buffer if needed
This simple API makes it easier to deal with allocated and memory-mapped
buffers.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
Add a function to duplicate a memory region, a little like strdup().
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
The CAAM on IMX7ULP doesn't support public key hardware acceleration
(PKHA), as in other NXP parts. Disable RSA_FREESCALE_EXP for IMX7ULP
too.
Fixed: f4e9ff7135 ("Kconfig: Don't use RSA_FREESCALE_EXP on IMX")
Signed-off-by: Ricardo Salveti <ricardo@foundries.io>
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@foundries.io>
|
|
SPL_RSA_VERIFY requires SPL_RSA to be enabled. Add correct
dependency.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@foundries.io>
|
|
OEM_REVISION is 32-bit unsigned number. It should be increased only when
changing software version. Therefore it should not depend on build time.
Change calculation to use U-Boot version numbers and set this revision
to date number.
Prior this change OEM_REVISION was calculated from build date and stored in
the same format.
After this change macro U_BOOT_BUILD_DATE is not used in other files so
remove it from global autogenerated files and also from Makefile.
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
|
|
This config is only used by three boards with this SOC. Most other
platforms derive this information from devicetree, and are unlikely
to ever need this config.
Moreover, it is confusing when Kconfig asks for this value under
"Support OPTEE images", but does not do anything with the value.
Move it to imx7 for those boards who still make use of it.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
This value is not used by u-boot, and it should not. The load address
of an OPTEE image is defined by said image. Either a uImage or a FIT
will have a defined load address and entry point. Those values are the
correct ones, not CONFIG_OPTEE_LOAD_ADDR.
Commit f25006b96e9f ("optee: Add CONFIG_OPTEE_LOAD_ADDR") justifies
this config by requiring its presence in u-boot's .config for other
images as part of a larger build, claiming it is "the best way".
This argument is not persuasive. U-boot's configuration is driven by
platform requirements, not the other way around. It seems more likely
that the argument is conflating tooling issues with Kconfig. Yocto and
buildroot have excellent mechanisms for defining values across the
board (pun intended). u-boot's Kconfig is the wrong place to do it.
Furthermore, it is not "best" for u-boot because it hardcodes a value
which is then not used. In fact the load address that u-boot uses is
the one derived from the OPTEE image.
Confused yet? I sure was. To prevent future confusion, remove
CONFIG_OPTEE_LOAD_ADDR.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
It is no longer used in u-boot. Information about the TZDRAM location
is usually available in the devicetree as "/reserved-memory/" nodes.
Because this isn't used, remove it.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
The configs TZDRAM_BASE and TZDRAM_SIZE are expected to describe the
memory allocated to the OPTEE region. according to according to commit
c5a6e8bd00cc ("optee: Add optee_verify_bootm_image()"). The TZDRAM is
with some limitations, described by "/reserved-memory" nodes in the
devicetree.
Consequently TZDRAM_BASE and TZDRAM_SIZE can point to imaginary
regions which have nothing to do with actual DRAM. They are not used
to configure the hardware or set up the Trust Zone Controller (TZC)
for OP-TEE -- the devicetree values are used instead.
When a valid OP-TEE image does not fall within the region described by
these configs, u-boot will refuse to load it. In fact, it mostly
serves to cause "bootm" to reject perfectly good OP-TEE images.
Ironically, someone has to correctly configure the devicetree for
TZDRAM, then go back and enter the same information in Kconfig for
"bootm". To remedy this, do not use TZDRAM_BASE and TZDRAM_SIZE in the
verification of OPTEE images.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
The configuration CONFIG_OPTEE is defined 2 times:
1- in lib/optee/Kconfig for support of OPTEE images loaded by bootm command
2- in drivers/tee/optee/Kconfig for support of OP-TEE driver.
It is abnormal to have the same CONFIG define for 2 purpose;
and it is difficult to managed correctly their dependencies.
Moreover CONFIG_SPL_OPTEE is defined in common/spl/Kconfig
to manage OPTEE image load in SPL.
This definition causes an issue with the macro CONFIG_IS_ENABLED(OPTEE)
to test the availability of the OP-TEE driver.
This patch cleans the configuration dependency with:
- CONFIG_OPTEE_IMAGE (renamed) => support of OP-TEE image in U-Boot
- CONFIG_SPL_OPTEE_IMAGE (renamed) => support of OP-TEE image in SPL
- CONFIG_OPTEE (same) => support of OP-TEE driver in U-Boot
- CONFIG_OPTEE_LIB (new) => support of OP-TEE library
After this patch, the macro have the correct behavior:
- CONFIG_IS_ENABLED(OPTEE_IMAGE) => Load of OP-TEE image is supported
- CONFIG_IS_ENABLED(OPTEE) => OP-TEE driver is supported
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
|
|
https://source.denx.de/u-boot/custodians/u-boot-dm into next
Various of-platdata improvements, including CONFIG_OF_REAL
|
|
Prepare v2021.10-rc5
|
|
Simplify some of the Makefile rules using this Kconfig.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
The current API is outdated as it requires a devicetree pointer.
Move these functions to use the ofnode API and update this globally. Add
some tests while we are here.
Correct the call in exynos_dsim_config_parse_dt() which is obviously
wrong.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
TCG EFI Protocol Specification defines the number_of_algorithms
field in spec ID event to be equal to the number of active
algorithms supported by the TPM device. In current implementation,
this field is populated with the count of all algorithms supported
by the TPM which leads to incorrect spec ID event creation.
Similarly, the algorithm array in spec ID event should be a variable
length array with length being equal to the number_of_algorithms field.
In current implementation this is defined as a fixed length array
which has been fixed.
Signed-off-by: Ruchika Gupta <ruchika.gupta@linaro.org>
CC: Masahisa Kojima <masahisa.kojima@linaro.org>
CC: Ilias Apalodimas <ilias.apalodimas@linaro.org>
CC: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
|
|
The arc/arm/m68k/microblaze/mips/ppc arch_lmb_reserve() implementations
are all mostly the same, except for a couple of details. Implement a
generic arch_lmb_reserve_generic() function which can be parametrized
enough to cater for those differences between architectures. This can
also be parametrized enough so it can handle cases where U-Boot is not
relocated to the end of DRAM e.g. because there is some other reserved
memory past U-Boot (e.g. unmovable firmware for coprocessor), it is not
relocated at all, and other such use cases.
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Alexey Brodkin <alexey.brodkin@synopsys.com>
Cc: Angelo Dureghello <angelo@sysam.it>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Cc: Hai Pham <hai.pham.ud@renesas.com>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tom Rini <trini@konsulko.com>
Cc: Wolfgang Denk <wd@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
|
|
ARMv8.0 has optional crc32 instruction for crc32 calculation. The
instruction is mandatory since ARMv8.1. The crc32 calculation is
faster using the dedicated instruction, e.g. 1.4 GHz iMX8MN gives:
=> time crc32 0x50000000 0x2000000
time: 0.126 seconds # crc32 instruction
time: 0.213 seconds # software crc32
Add implementation using the compiler builtin wrapper for the crc32
instruction and enable it by default, since we don't support any
platforms which do not implement this instruction.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Simon Glass <sjg@chromium.org>
[trini: Make crc32_table guarded by CONFIG_ARM64_CRC32]
Signed-off-by: Tom Rini <trini@konsulko.com>
|
|
This was unfortunately applied despite much discussion about it beiong
the wrong way to implement this feature.
Revert it before too many other things are built on top of it.
This reverts commit ddf67daac39de76d2697d587148f4c2cb768f492.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
Header file version.h does not use anything from timestamp.h. Including of
timestamp.h has side effect which cause recompiling object file at every
make run because timestamp.h changes at every run.
So remove timestamp.h from version.h and include timestamp.h in files
which needs it.
This change reduce recompilation time of final U-Boot binary when U-Boot
source files were not changed as less source files needs to be recompiled.
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
[trini: Add in lib/acpi/acpi_table.c and test/dm/acpi.c, rework a few others]
Signed-off-by: Tom Rini <trini@konsulko.com>
|
|
More C files do not use compile time timestamp macros and do not have to be
recompiled every time when SOURCE_DATE_EPOCH changes.
This patch moves version_string[] from version.h to version_string.h and
updates other C files which only needs version_string[] string to include
version_string.h instead of version.h. After applying this patch these
files are not recompiled every time when SOURCE_DATE_EPOCH changes.
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
|
|
Macro U_BOOT_VERSION_STRING is already stored in variable version_string.
So use directly this variable instead of storing U_BOOT_VERSION_STRING into
temporary variable.
Signed-off-by: Pali Rohár <pali@kernel.org>
[trini: This area was reworked since posted, what is here is now really
inspired by the previous version, so drop Ilias' Acked-by]
Signed-off-by: Tom Rini <trini@konsulko.com>
|
|
Prepare v2021.10-rc4
Signed-off-by: Tom Rini <trini@konsulko.com>
# gpg: Signature made Tue 14 Sep 2021 06:58:32 PM EDT
# gpg: using RSA key 1A3C7F70E08FAB1707809BBF147C39FF9634B72C
# gpg: Good signature from "Thomas Rini <trini@konsulko.com>" [ultimate]
# Conflicts:
# board/Arcturus/ucp1020/spl.c
# cmd/mvebu/Kconfig
# common/Kconfig.boot
# common/image-fit.c
# configs/UCP1020_defconfig
# configs/sifive_unmatched_defconfig
# drivers/pci/Kconfig
# include/configs/UCP1020.h
# include/configs/sifive-unmatched.h
# lib/Makefile
# scripts/config_whitelist.txt
|
|
We are not guaranteed to have the padding_pkcs_15_verify symbol since
commit 92c960bc1d ("lib: rsa: Remove #ifdefs from rsa.h"), and
commit 61416fe9df ("Kconfig: FIT_SIGNATURE should not select RSA_VERIFY")
The padding_algos only make sense with RSA verification, which can now
be disabled in lieu of ECDSA. In fact this will lead to build failures
because of the missing symbol mentioned earlier.
To resolve this, move the padding_algos to a linker list, with
declarations moved to rsa_verify.c. This is consistent with commit
6909edb4ce ("image: rsa: Move verification algorithm to a linker list")
One could argue that the added #ifdef USE_HOSTCC is ugly, and should
be hidden within the U_BOOT_PADDING_ALGO() macro. However, this would
be inconsistent with the "cryptos" list. This logic for was not
previously explored:
Without knowledge of the U_BOOT_PADDING_ALGO() macro, its use is
similar to something being declared. However, should #ifndef
USE_HOSTCC be part of the macro, it would not be obvious that it
behaves differently on host code and target code. Having the #ifndef
outside the macro makes this obvious.
Also, the #ifdef is not always necessary. For example ecda-verify
makes use of U_BOOT_CRYPTO_ALGO() without any accompanying #ifdefs.
The fundamental issue is a lack of separation of host and target code
in rsa_verify. Therefore, the declaration of a padding algo with the
external #ifdef is more readable and consistent.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
There are trivial typos in the Kconfig file. Fixed them.
Also, fixed grammar in the descriptions with typos.
Fixes: d56b4b1974 ("configs: Migrate RBTREE, LZO, CMD_MTDPARTS, CMD_UBI and CMD_UBIFS")
Fixes: 7264f2928b ("spl: fit: Eanble GZIP support for image decompression")
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@foundries.io>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
|
|
If keydir is not provided but name is we want to use name as key_id.
But with the current coding name is only used on its own if it is NULL
and keydir is provided which never occurs.
Fixes: 824ee745fbca ("lib/rsa: Use the 'keyfile' argument from mkimage")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
|
|
We don't need this check anymore since when PCI is enabled, driver model
is always used.
Sadly this doesn't work with nds32 for some reason to do with the
toolchain. Add a work-around for that.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
https://source.denx.de/u-boot/custodians/u-boot-efi
Pull request for efi-2021-10-rc4-2
Documentation:
* improve documentation of U-Boot for /config DT node
* integrate bloblist documentation
UEFI:
* correct usage of EFI_CALL()
* code tidy up
|
|
We can call do_reset() directly without invoking the UEFI API.
This decreases the code size.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
|
|
There is two unneeded EFI_CALL references in tcg2_measure_pe_image().
The first one in efi_search_protocol() and the second on in the device path
calculation. The second isn't even a function we should be calling, but a
pointer assignment, which happens to work with the existing macro.
While at it switch the malloc call to a calloc, remove the unnecessary cast
and get rid of an unneeded if statement before copying the device path
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
|
|
The move to driver model should by now be completed. To be able to remove
pre-driver model code from our block IO code require CONFIG_BLK=y for UEFI
support.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
|
|
boot_service_capability_min is constant, it should be capitalized.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
|
|
When dropping SHA512_ALGO in general, we didn't catch some cases where
an option was selecting both SHA512 and SHA512_ALGO and caused them to
select SHA512 twice. Kconfig doesn't complain, but this is still wrong
and should be corrected.
Fixes: e60e44993120 ("lib: Drop SHA512_ALGO in lieu of SHA512")
Reported-by: Andreas Schwab <schwab@suse.de>
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
|
|
MD5 is being called directly in some places, but it is not available
via hash_lookup_algo("md5"). This is inconsistent with other hasing
routines. To resolve this, add an "md5" entry to hash_algos[].
The #ifdef clause looks funnier than those for other entries. This is
because both MD5 and SPL_MD5 configs exist, whereas the other hashes
do not have "SPL_" entries. The long term plan is to get rid of the
ifdefs, so those should not be expected to survive much longer.
The md5 entry does not have .hash_init/update/finish members. That's
okay because hash_progressive_lookup_algo() will catch that, and
return -EPROTONOSUPPORT, while hash_lookup_algo() will return the
correct pointer.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
[trini: Use CONFIG_IS_ENABLED not IS_ENABLED for MD5 check]
Signed-off-by: Tom Rini <trini@konsulko.com>
|
|
All of these configs exist. Stick to using CONFIG_[ST]PL_HASH, and drop all
references to CONFIG_[ST]PL_HASH_SUPPORT. This means we need for
CHAIN_OF_TRUST to select SPL_HASH now.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
[trini: Add TPL case, fix CHAIN_OF_TRUST, other tweaks]
Signed-off-by: Tom Rini <trini@konsulko.com>
|
|
SHA512_ALGO was used as a "either SHA512 or SHA384", although the
implementations of these two algorithms share a majority of code.
From a Kconfig interface perspective, it makes sense to present two
distinct options. This requires #ifdefing out the SHA512
implementation from sha512.c. The latter doesn't make any sense.
It's reasonable to say in Kconfig that SHA384 depends on SHA512, and
seems to be the more polite way to handle the selection.
Thus, automatically select SHA512 when SHA384 is enabled.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
|
|
Rename this option so that CONFIG_IS_ENABLED can be used with it.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
Rename these options so that CONFIG_IS_ENABLED can be used with them.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
TCG EFI Protocol Specification defines that PCRIndex parameter
passed from caller must be 0 to 23.
TPM2_MAX_PCRS is currently used to check the range of PCRIndex,
but TPM2_MAX_PCRS is tpm2 device dependent and may have larger value.
This commit newly adds EFI_TCG2_MAX_PCR_INDEX macro, it is used to
check the range of PCRIndex parameter.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Acked-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
|
|
TCG EFI Protocol Specification defines the required parameter
checking and return value for each API.
This commit adds the missing parameter check and
fixes the wrong return value to comply the specification.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
|
|
When U-Boot is started we have to use the existing variables to determine
in which secure boot state we are.
* If a platform key PK is present and DeployedMode=1, we are in deployed
mode.
* If no platform key PK is present and AuditMode=1, we are in audit mode.
* Otherwise if a platform key is present, we are in user mode.
* Otherwise if no platform key is present, we are in setup mode.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
|
|
Writing variables AuditMode and DeployedMode serves to switch between
Secure Boot modes. Provide a separate value for these in efi_auth_var_type.
With this patch the variables will not be read from from file even if they
are marked as non-volatile by mistake.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
|
|
The UEFI specification requires that the signature database may only be
stored in tamper-resistant storage. So these variable may not be read
from an unsigned file.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
|
|
We should not first allocate memory and then report a rounded up value as
image size. Instead first round up according to section allocation and then
allocate the memory.
Fixes: 82786754b9d2 ("efi_loader: ImageSize must be multiple of SectionAlignment")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
|
|
In a section header VirtualSize may be zero. This is for instance seen in
the .sbat section of shim. In this case use SizeOfRawData as section size.
Fixes: 9d30a941cce5 ("efi_loader: don't load beyond VirtualSize")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Asherah Connor <ashe@kivikakk.ee>
|
|
CONFIG_MMC_SPI_CRC_ON needs the crc16 functions, but it was not included
in an SPL build. For non-SPL builds, crc16.o is already added
unconditionally. This also removes CONFIG_SPL_YMODEM_SUPPORT from the
sifive board configs, which is only relevant for some ARM boards and was
only set for its side effect of adding crc16.o.
|
|
With glibc 2.33 (Ubuntu package glibc6 2.33-0ubuntu9) building
sifive_unmatched_defconfig results in:
In file included from /usr/include/stdio.h:866,
from ././include/compiler.h:26,
from <command-line>:
In function ‘snprintf’,
inlined from ‘rsa_engine_get_priv_key’ at ./tools/../^:273:4:
/usr/include/riscv64-linux-gnu/bits/stdio2.h:71:10: warning:
‘%s’ directive argument is null [-Wformat-truncation=]
71 | return __builtin___snprintf_chk (__s, __n,
__USE_FORTIFY_LEVEL - 1,
|
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
72 | __glibc_objsize (__s), __fmt,
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
73 | __va_arg_pack ());
| ~~~~~~~~~~~~~~~~~
Avoid passing a NULL string.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
|
|
Export the MD5 hash init/update/finish progressive APIs
for better flexibility.
Signed-off-by: Chia-Wei Wang <chiawei_wang@aspeedtech.com>
|
