Init errors in SPL are currently ignored by this function.
Change the code to init the CPU, reporting an error if something is wrong.
After that, look for the coreboot table.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Enable bootstd options and provide instructions on how to boot a linux
distro using coreboot.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add these options to permit access to more disk types.
Add some documentation as well.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This driver is not actually built since a Kconfig was never created for
it.
Add a Kconfig (which is already implied by COREBOOT) and update the
implementation to avoid using unnecessary memory. Drop the #ifdef at the
top since we can rely on Kconfig to get that right.
To enable it (in addition to serial and video), use:
setenv stdout serial,vidconsole,cbmem
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
[Modified the comment about overflow a little bit]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Coreboot changed a few years ago to include an overflow flag. Update the
structure to match this.
This comes from coreboot commit:
6f5ead14b4 ("mb/google/nissa/var/joxer: Update eMMC DLL settings")
Note: There are several implementations of this in coreboot. I have chosen
to follow the one in src/lib/cbmem_console.c
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When turning off CAR, set the flag to make sure that nothing tries to use
driver model in SPL before jumping to U-Bot proper, since its tables are
in CAR.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The current condition does not handle the samus_tpl case where it sets
up the RAM in SPL but needs to commit the MTRRs in U-Boot proper.
Add another case to handle this and update the comment.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When TPL has already set up the CPU, don't do it again. This existing
code actually has this backwards, so fix it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Make this depend on whether the address matches the offset, rather than
a particular board build. For samus_tpl we don't need to copy, for
example.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Make sure that CONFIG_X86_OFFSET_U_BOOT is the same as CONFIG_TEXT_BASE
as it is changing CONFIG_X86_OFFSET_U_BOOT
Samus boots into U-Boot proper in flash, not RAM. Also expand the SPL
malloc() size a little, to avoid an error.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Samus only takes 7 seconds but it is long enough to think it has hung. Add
a message about what it is doing, similar to the approach on coral.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
On x86 devices we use CAR (Cache-As-RAM) to hold the malloc() region in
SPL, since SDRAM is not set up yet. This means that driver model stores
its tables in this region.
When preparing to jump from SPL to U-Boot proper, we must disable CAR, so
that the CPU can uses the caches normally. This means that driver model
tables become inaccessible. From there until we jump to U-Boot proper, we
must avoid using driver model.
This is only a problem on boards which operate this way, for example
chromebook_link64
Add a flag to indicate that driver model is dead and should not be used.
It can be used in SPL to avoid hanging the machine.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add a note about using -cdrom with QEMU.
Suggested-by: Bin Meng <bmeng@tinylab.org>
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng@tinylab.org>
Recent approach with FADT writer shows that there is
a room for subtle errors. Prevent this from happening
again by introducing acpi_add_fadt() helper.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Each board has its own way of creating this table. Rather than calling the
acpi_create_fadt() function for each one from a common acpi_write_fadt()
function, just move the writer into the board-specific code.
Co-developed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Declare the global_data pointer at the top of the file, to avoid an
error:
arch/x86/include/asm/global_data.h:143:35: error: a label can
only be part of a statement and a declaration is not a statement
board/coreboot/coreboot/coreboot.c:60:2: note: in expansion of macro
‘DECLARE_GLOBAL_DATA_PTR’
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When building in a portage chroot, we do not have the environment needed
to build pylibfdt. It is instead build as a separate package.
Provide a build option to tell U-Boot to skip this part of the build. We
still need it to use binman, etc. but don't need it to build its
dependencies.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
[s/build bytes/builds bytes in tools.rst]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
The problem was, that zboot() didn't work because of missing
ramdisc size.
Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Enable bootstd support for U-Boot at VirtualBox described at
https://source.denx.de/u-boot/u-boot/-/blob/master/doc/develop/bootstd.rst
This is used to boot system images at Virtualbox via
- distroboot (extlinux.conf)
- boot script
Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@bosch.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
[Added 'efi-payload64' tag]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
U-Boot at VirtualBox must load Linux and boot configuration from disk devices.
Here the discs at AHCI (scsi) bus are used to load the needed boot data.
Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
[Added 'efi-payload64' tag and rebased on top of u-boot/master]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
This avoids an error when ifdef CONFIG_PCI is changed to
if CONFIG_IS_ENABLED(PCI)
Signed-off-by: Troy Kisky <troykiskyboundary@gmail.com>
[Rebased on top of u-boot/master]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
clk:
- Dont return error when assigned-clocks is empty or missing
dm:
- Support reading a single indexed u64 value
- Add support for reading bootscript address/flash address from DT
cmd:
- Fix flash_is_unlocked API
fpga:
- Define fpga_load() for debug build
global:
- U-Boot project name cleanup (next2)
net:
- zynq_gem: Use generic_phy_valid() helper
- axienet: Convert to ofnode functions
- gmii2rgmii: Read bridge address from DT
pytest:
- skip tpm2_startup when env__tpm_device_test_skip=True
spi-nor:
- Add mx25u25635f support
- zynqmp_qspi: Tune cache behavior
trace:
- Fix flyrecord alignment issue
xilinx:
- Move scriptaddr to DT as bootscr-address
- Pick script_offset_f/script_size_f from DT as bootscr-flash-offset/size
- Do not generate distro boot variables if disabled
versal:
- Extend memory ranges to cover HBM
- Enable TPM, sha1sum and KASLRSEED
- Fix distroboot prioritization in connection to available devices
- Clean mini targets bootcommand
- Fix clock driver
versal-net:
- Enable TPM, sha1sum and KASLRSEED
- Fix distroboot prioritization in connection to available devices
zynqmp;
- Allow AES to run from SPL
- Enable CMD_KASLRSEED
- Add proper dependencies for USB and remove ZYNQMP_USB
- Fix user si570 default frequency for zcu* boards
- Cover SOM rev2 revision
- Various DT changes
- Add firmware and pinctrl support for tristate configuration
(high impedance/output enable)
- Add output-enable pins to SOMs
- Fix distroboot prioritization in connection to available devices
- Read bootscript address/flash address from DT
- Fix pcap_prog address
-----BEGIN PGP SIGNATURE-----
iF0EABECAB0WIQQbPNTMvXmYlBPRwx7KSWXLKUoMIQUCZQwqWAAKCRDKSWXLKUoM
IQYIAJ9WQrnZoCfi7MuxzYA2t34Dbx4gkwCgk1o6bX+R89jvjBOhgZOlXTQCbb0=
=/rxs
-----END PGP SIGNATURE-----
Merge tag 'xilinx-for-v2024.01-rc1-v2' of https://source.denx.de/u-boot/custodians/u-boot-microblaze into next
Xilinx changes for v2024.01-rc1
clk:
- Dont return error when assigned-clocks is empty or missing
dm:
- Support reading a single indexed u64 value
- Add support for reading bootscript address/flash address from DT
cmd:
- Fix flash_is_unlocked API
fpga:
- Define fpga_load() for debug build
global:
- U-Boot project name cleanup (next2)
net:
- zynq_gem: Use generic_phy_valid() helper
- axienet: Convert to ofnode functions
- gmii2rgmii: Read bridge address from DT
pytest:
- skip tpm2_startup when env__tpm_device_test_skip=True
spi-nor:
- Add mx25u25635f support
- zynqmp_qspi: Tune cache behavior
trace:
- Fix flyrecord alignment issue
xilinx:
- Move scriptaddr to DT as bootscr-address
- Pick script_offset_f/script_size_f from DT as bootscr-flash-offset/size
- Do not generate distro boot variables if disabled
versal:
- Extend memory ranges to cover HBM
- Enable TPM, sha1sum and KASLRSEED
- Fix distroboot prioritization in connection to available devices
- Clean mini targets bootcommand
- Fix clock driver
versal-net:
- Enable TPM, sha1sum and KASLRSEED
- Fix distroboot prioritization in connection to available devices
zynqmp;
- Allow AES to run from SPL
- Enable CMD_KASLRSEED
- Add proper dependencies for USB and remove ZYNQMP_USB
- Fix user si570 default frequency for zcu* boards
- Cover SOM rev2 revision
- Various DT changes
- Add firmware and pinctrl support for tristate configuration
(high impedance/output enable)
- Add output-enable pins to SOMs
- Fix distroboot prioritization in connection to available devices
- Read bootscript address/flash address from DT
- Fix pcap_prog address
Cache related issues are seen with small sized data reads.
Due to this, proper data is not read. Also some times sf probe
fails randomly. To workaround this issue, invalidate dcache after read DMA
is triggered.
Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma@amd.com>
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Link: https://lore.kernel.org/r/20230915031759.28889-3-venkatesh.abbarapu@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
Before DMA read, ideally cache should be invalidated, so that data from
memory will be updated to cache after DMA is completed. But
flush_dcache_range is being used which is incorrect. Change
flush_dcache_range to invalidate_dcache_range.
Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma@amd.com>
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Link: https://lore.kernel.org/r/20230915031759.28889-2-venkatesh.abbarapu@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
Current alignment which is using 16 bytes is not correct in connection to
trace_clocks description and it's length.
That's why use start_addr variable and record proper size based on used
entries.
Fixes: be16fc81b2 ("trace: Update proftool to use new binary format").
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/691dad64df80993ca4cfb6d0e33964ed26f50bee.1694779918.git.michal.simek@amd.com
Flyrecord tracing data are page aligned that's why it is necessary to
calculate alignment properly. Because trace_clocks description is the part
of record length it is necessary to have information about length earlier.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/d3853d91b6fa7e3a1e5c24dd3c17335cf0041b5b.1694779918.git.michal.simek@amd.com
When flash erase is called with size parameter, code is checking
if sectors are locked or not. But for checking, the whole device
length minus offset is used instead of actual size which should
be erased. That's why when only some sectors are locked it is
not possible to erase unlocked sectors.
The length is calculated as "length = max_chipsize - offset",
flash_is_unlocked() api is getting updated with length which is
incorrect. Fix this flash_is_unlocked() api by passing the size
parameter.
ZynqMP> sf erase 0 100000
len=0x8000000 which is flash size
size=0x100000
We need to update the size in the flash_is_unlocked() api and not
the length.
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Link: https://lore.kernel.org/r/20230920025450.6281-1-venkatesh.abbarapu@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
When CONFIG_CMD_USB and CONFIG_USB are disabled, still some compilation
errors are seen as below.
In file included from include/configs/xilinx_zynqmp.h:173,
from include/config.h:3,
from include/common.h:16,
from env/common.c:10:
include/config_distro_bootcmd.h:302:9: error: expected '}' before 'BOOT_TARGET_DEVICES_references_USB_without_CONFIG_CMD_USB'
302 | BOOT_TARGET_DEVICES_references_USB_without_CONFIG_CMD_USB
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/config_distro_bootcmd.h:302:9: note: in definition of macro
'BOOTENV_DEV_NAME_USB'
302 | BOOT_TARGET_DEVICES_references_USB_without_CONFIG_CMD_USB
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/configs/xilinx_zynqmp.h:77:41: note: in expansion of macro
'BOOTENV_DEV_NAME'
77 | # define BOOT_TARGET_DEVICES_USB(func) func(USB, usb, 0)
func(USB, usb, 1)
| ^~~~
include/configs/xilinx_zynqmp.h:168:9: note: in expansion of macro
'BOOT_TARGET_DEVICES_USB'
168 | BOOT_TARGET_DEVICES_USB(func) \
| ^~~~~~~~~~~~~~~~~~~~~~~
include/config_distro_bootcmd.h:454:25: note: in expansion of macro
'BOOT_TARGET_DEVICES'
454 | "boot_targets=" BOOT_TARGET_DEVICES(BOOTENV_DEV_NAME) "\0"
| ^~~~~~~~~~~~~~~~~~~
include/config_distro_bootcmd.h:474:9: note: in expansion of macro
'BOOTENV_BOOT_TARGETS'
474 | BOOTENV_BOOT_TARGETS \
| ^~~~~~~~~~~~~~~~~~~~
include/configs/xilinx_zynqmp.h:179:9: note: in expansion of macro
'BOOTENV'
179 | BOOTENV
| ^~~~~~~
include/env_default.h:120:9: note: in expansion of macro
'CFG_EXTRA_ENV_SETTINGS'
120 | CFG_EXTRA_ENV_SETTINGS
| ^~~~~~~~~~~~~~~~~~~~~~
In file included from env/common.c:32:
include/env_default.h:27:36: note: to match this '{'
27 | const char default_environment[] = {
| ^
scripts/Makefile.build:256: recipe for target 'env/common.o' failed
make[1]: *** [env/common.o] Error 1
Makefile:1853: recipe for target 'env' failed
make: *** [env] Error 2
make: *** Waiting for unfinished jobs....
Add CONFIG_USB_STORAGE as dependency for USB related macro's such as
BOOT_TARGET_DEVICES_USB() and DFU_DEFAULT_POLL_TIMEOUT and
CONFIG_THOR_RESET_OFF.
Remove CONFIG_ZYNQMP_USB from Kconfig and also from defconfig since it
is not used anywhere else.
Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma@amd.com>
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Link: https://lore.kernel.org/r/20230904031528.11817-3-venkatesh.abbarapu@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
When CONFIG_CMD_USB and CONFIG_USB are disabled some compilation errors are seen as below.
cmd/thordown.o: in function `usb_gadget_initialize':
include/linux/usb/gadget.h:981: undefined reference to `board_usb_init'
cmd/thordown.o: in function `do_thor_down':
cmd/thordown.c:68: undefined reference to `g_dnl_unregister'
cmd/thordown.o: in function `usb_gadget_release':
include/linux/usb/gadget.h:986: undefined reference to `board_usb_cleanup'
cmd/thordown.o: in function `do_thor_down':
cmd/thordown.c:41: undefined reference to `g_dnl_register'
cmd/thordown.c:48: undefined reference to `thor_init'
cmd/thordown.c:56: undefined reference to `thor_handle'
gnu/aarch64/lin/aarch64-linux/bin/aarch64-linux-gnu-ld.bfd: line 4: 8485
Segmentation fault (core dumped) $CC --sysroot=$LIBC
--no-warn-rwx-segment "$@"
Makefile:1779: recipe for target 'u-boot' failed
make: *** [u-boot] Error 139
make: *** Deleting file 'u-boot'
Add dependency of USB_GADGET_DOWNLOAD for CONFIG_CMD_THOR_DOWNLOAD to fix the errors.
Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma@amd.com>
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Link: https://lore.kernel.org/r/20230904031528.11817-2-venkatesh.abbarapu@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
For reference clocks, PM_CLK_GET_PARENT call is not allowed.
PM_CLK_GET_PARENT only allowed for MUX clocks. Rename the
versal_clock_ref() with versal_clock_get_ref_rate() for better
readability. Fix the versal_clock_get_ref_rate function by
passing the parent_id, and check whether the parent_id
belongs to ref_clk or pl_alt_ref_clk.
Also adding the function versal_clock_get_fixed_factor_rate()
if the clk_id belongs to the fixed factor clock.
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Link: https://lore.kernel.org/r/20230912033055.2549-1-venkatesh.abbarapu@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
Correct the i2c address for si570 oscillator that generates the si570_user1
clock. i2c address was changed by commit b6a8c603d680 ("arm64: zynqmp: Fix
i2c addresses for vck190 SC") because address in node name wasn't aligned
with reg property. But actual 0x5f address is correct which is quite rare
because all other si570s are at 0x5d.
Signed-off-by: Saeed Nowshadi <saeed.nowshadi@amd.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/6f31881b0e2dd657f0d4ff0869c009c2e1224f22.1694441445.git.michal.simek@amd.com
Added no-wp DT property in OSPI flash node for all board dts & dtsi files
on which the WP# signal of the OSPI flash device is not connected. If this
property is set, then the software will avoid setting the status register
write disable (SRWD) bit in status register during status register
write operation.
Signed-off-by: Amit Kumar Mahapatra <amit.kumar-mahapatra@amd.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/7e88dd7b9306bdf0738b2248bf9017e1997d25dc.1694441445.git.michal.simek@amd.com
xlnx,mio_bank was used in past but it was renamed to xlnx,mio-bank because
'_' in property shoudln't be used. There is no impact on the platform
because if the properly is not defined bank 0 is default. Bank 0 and 1 have
the same configuration that's why there shouldn't be any issue.
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/ace68d4b7701d1606a85cb18242409fce941b363.1694441445.git.michal.simek@amd.com
Current code expects bridge phy address at 0 which is not correct
expectation because bridge phy address is configurable.
That's why update the code to read reg property to figure it out
where bridge is and use it in phy creation code.
Signed-off-by: Tejas Bhumkar <tejas.arvind.bhumkar@amd.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/20230915045043.4167628-1-tejas.arvind.bhumkar@amd.com
SOC can boot in the device which is not accessible from APU and running
this is detected as error which ends up in stopping boot process.
Boot mode detection and logic around is present to setup priority on boot
devices that SOC boot device is likely also used for booting OS.
Change logic to detect this case with showing message about it but don't fail
in boot process and don't prioritize boot device in this case.
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/20230904032035.11926-4-venkatesh.abbarapu@amd.com
SOC can boot in the device which is not accessible from APU and running
this is detected as error which ends up in stopping boot process.
Boot mode detection and logic around is present to setup priority on boot
devices that SOC boot device is likely also used for booting OS.
Change logic to detect this case with showing message about it but don't fail
in boot process and don't prioritize boot device in this case.
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/20230904032035.11926-3-venkatesh.abbarapu@amd.com
SOC can boot in the device which is not accessible from APU and running
this is detected as error which ends up in stopping boot process.
Boot mode detection and logic around is present to setup priority
on boot devices that SOC boot device is likely also used for booting OS.
Change logic to detect this case with showing message about it but don't
fail in boot process and don't prioritize boot device in this case.
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/20230904032035.11926-2-venkatesh.abbarapu@amd.com
Location of bootscript in flash can be specified via /options/u-boot DT
node by using bootscr-flash-offset and bootscr-flash-size properties.
Values should be saved to script_offset_f and script_size_f variables.
Variables are described in doc/develop/bootstd.rst as:
script_offset_f
SPI flash offset from which to load the U-Boot script, e.g. 0xffe000
script_size_f
Size of the script to load, e.g. 0x2000
Both of them are used by sf_get_bootflow() in drivers/mtd/spi/sf_bootdev.c
to identify bootscript location inside flash.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/60a84405f3fefabb8b48a4e1ce84431483a729f3.1693465465.git.michal.simek@amd.com
