AMD/Xilinx changes for v2024.10-rc1
common:
- spl: Introduce SoC specific init function
xilinx:
- Enable FF-A and NVMEM
- Rename spl_board_init() to spl_soc_init()
zynqmp:
- DT alignments
- Enable reset from SPL
- Enable USB3 for KD240
- Align multiboot register on Kria for proper reboot
- Allow multiboot environment write even in saved environment
- Move zynqmp commands from board/ to arch/
- Clean up xilinx_zynqmp.h
versal:
- Do not prioritize boot device if driver is not enabled
versal-net:
- Setup location for redundant variables in SPI
versal2:
- Add support for new SOC
mmc:
- Fix tap delay for SD on Versal NET
spi:
- Add SPI_NOR_OCTAL_READ flag for mx66uw2g345gx0 flash part
gpio:
- Cover MODEPIN firmware dependency
Pull request for u-boot-nand-20240617
The first patch is by Arseniy Krasnov and adds support for OTP area
access on MX30LFxG18AC chip series.
The second patch is by John Watts and adds MTD dependency in Kconfig
for UBI.
The last patch is by Ravi Minnikanti and fixes bitflip return value on
page read.
This patch corrects the mio and pll configuration registers for using usb3
on the kd240 starter kit. Without this patch, the usb3 to sd card bridge does
not initialize correctly and u-boot is unable to find the OS located on the
kd240 starter kit sd card.
In addition, this patch correctly configures mio76 and mio77 as gpio pins
which are used as reset gpio pins on the kd240 starter kit.
Signed-off-by: Neal Frager <neal.frager@amd.com>
Link: https://lore.kernel.org/r/20240604083854.2033917-1-neal.frager@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
On Kria when board starts from Image A or Image B partition multiboot
register is already setup to that location. When reset command is called
board is issuing soft reset which start SW at already used location (offset
of multiboot * 32k).
But board should continue to run from multiboot offset 0 (start of QSPI)
and call early bootloader every reboot that's why clear multiboot register
to 0 by default to go that route all the time.
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/edaf714a778bdd7447533a77b3455e4fd623f9da.1717420131.git.michal.simek@amd.com
SOC can boot out of the device which is not accessible from APU and running
this is detected as a warning, as the device is not accessible.For example
getting below warning when the boot mode is OSPI and OSPI is not enabled in
device tree.
Invalid bus 0 (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)
Ignoring the prioritization of the boot device which driver is not enabled
and continue with the default boot_targets. Recommendation is to use custom
boot_targets via environment file as is done for example for Kria via
zynqmp_kria.env file.
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/8b7cca5c7b84cb4854104e0c48f8aa63c4ec5ace.1715322156.git.michal.simek@amd.com
Once the environment was saved, the current multiboot image information
became unreachable. When dealing with firmware updates, this information
is necessary alongside the saved environment to know the booted image.
Move the multiboot environment set operation before the saved environment
check to ensure this information is always available.
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
Link: https://lore.kernel.org/r/20240529100107.137159-1-kory.maincent@bootlin.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
The zynqmp cmds.c is currently tied to the board but the commands
contained within are more closely tied to the architecture. To
allow usage of those commands when the architecture is ZynqMP but
the board is not, this change moves the cmds into the arch/ tree.
The source file is renamed to zynqmp.c to reflect the command name
as well.
Signed-off-by: Charlie Johnston <charlie.johnston@loftorbital.com>
Link: https://lore.kernel.org/r/20240410195008.405061-2-charlie.johnston@loftorbital.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
Boards which have for example MAC address in eeprom but not in Xilinx
format (legacy or FRU) could reference it via nvmem cells.
For example:
&gem0 {
nvmem-cells = <&mac>;
nvmem-cell-names = "mac-address";
};
&eeprom {
#address-cells = <1>;
#size-cells = <1>;
mac: mac-address@f0 {
reg = <0xf0 6>;
};
};
For getting it work above DT changes are required but also CONFIG_NVMEM
should be enabled. That's why enable it by default in generic defconfigs
to be able to use it directly by changing DT only.
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/9c8ee7a4c7a16367438a92a4c9581bac9d968f84.1712815454.git.michal.simek@amd.com
Make sure the OS would not get any spurious IWDG pretimeout IRQ
right after the system wakes up. This may happen in case the SoC
got woken up by another source than the IWDG pretimeout and the
pretimeout IRQ arrived immediately afterward, but too late to be
handled by the suspend main loop. In case either of the IWDG is
enabled, ping it first and then return to the OS.
Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@foundries.io>
Once a page is read with higher bitflips all subsequent reads
are returning the same bitflip value even though they have none.
max_bitflip variable is not being reset to 0 across page reads.
This is causing problems like incorrectly
marking erase blocks bad by UBI and causing read failures.
Verified the change with both MTD reads and UBI.
This change is inline with other NFC drivers.
Sample error log where a block is marked bad incorrectly:
ubi0: fixable bit-flip detected at PEB 125
ubi0: run torture test for PEB 125
ubi0: fixable bit-flip detected at PEB 125
ubi0 error: torture_peb: read problems on freshly erased PEB 125,
must be bad
ubi0 error: erase_worker: failed to erase PEB 125, error -5
ubi0: mark PEB 125 as bad
Link: https://lore.kernel.org/all/ea0422cd-a8e6-3c36-f551-a0142893301b@marvell.com
Signed-off-by: rminnikanti <rminnikanti@marvell.com>
Reviewed-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: rminnikanti <rminnikanti@marvell.com>
Acked-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Nothing in-tree calls watchdog_reset() anymore (that stopped two years
ago with the removal of the WATCHDOG_RESET macro). So that function is
dead code.
That was the only caller of reset_85xx_watchdog(), so that
can obviously also be removed.
Finally, init_85xx_watchdog() is/was also not called from anywhere, so
that can go away as well, which nicely also removes a bit of
arch-specific code from the generic watchdog.h header.
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
There is no longer any code in tree that calls a watchdog_reset()
function. The macro WATCHDOG_RESET, which used to emit a call to
watchdog_reset(), got removed two years ago.
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
The external functions defined here are not called from anywhere. So
they, and consequently the whole file, can be dropped.
Cc: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
The next patch will remove all the other code from watchdog.c, which
would leave just this function in there. It seems just as natural for
this function to be defined in cpu.c, allowing us to delete watchdog.c
completely.
Cc: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
watchdog_reset() is no more. Make the comments match the code and
today's reality.
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
watchdog_reset() is no longer called from anywhere, so we do not need
to define a dummy no-op function. Remove that definition, and update
references to say schedule() instead.
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
There are no calls of "watchdog_reset()" anymore anywhere in the tree
since the WATCHDOG_RESET macro got removed in 942d07df0e ("watchdog:
Remove WATCHDOG_RESET macro").
The only places the identifiers watchdog_disable and watchdog_init are
called are in arch/arm/mach-omap2/, so those can obviously not refer
to these instances.
Hence these functions are not actually used at all and can be
removed. As a bonus, this also removes two leftover references to
WATCHDOG_RESET.
Cc: Huan Wang <alison.wang@nxp.com>
Cc: Angelo Dureghello <angelo@kernel-space.org>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Stefan Roese <sr@denx.de>
There are of course not a whole lot of examples in-tree yet, but
before they appear, let's make this API change: Instead of separately
allocating a 'struct cyclic_info', make the users embed such an
instance in their own structure, and make the convention that the
callback simply receives the 'struct cyclic_info *', from which the
clients can get their own data using the container_of() macro.
This has a number of advantages.
First, it means cyclic_register() simply cannot fail, simplifying the
code. The necessary storage will simply be allocated automatically
when the client's own structure is allocated (often via
uclass_priv_auto or similar).
Second, code for which CONFIG_CYCLIC is just an option can more easily
be written without #ifdefs, if we just provide an empty struct
cyclic_info {}. For example, the nested CONFIG_IS_ENABLED()s in
https://lore.kernel.org/u-boot/20240316201416.211480-1-marek.vasut+renesas@mailbox.org/
are mostly due to the existence of the 'struct cyclic_info *' member
being guarded by #ifdef CONFIG_CYCLIC.
And we do probably want to avoid the extra memory overhead of that
member when !CONFIG_CYCLIC. But that is automatic if, instead of a
'struct cyclic_info *', one simply embeds a 'struct cyclic_info',
which will have size 0 when !CONFIG_CYCLIC. Also, the no-op
cyclic_register() function can just unconditionally be called, and the
compiler will see that (1) the callback is referenced, so not emit a
warning for a maybe-unused function and (2) see that it can actually
never be reached, so not emit any code for it.
Reviewed-by: Stefan Roese <sr@denx.de>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Currently, the cyclic_register() done in wdt_start() is not undone in
wdt_stop(). Moreover, calling wdt_start multiple times (which is
perfectly allowed on an already started device, e.g. to change the
timeout value) will result in another struct cyclic_info being
registered, referring to the same watchdog device.
This can easily be seen on e.g. a wandboard:
=> cyclic list
function: watchdog@20bc000, cpu-time: 22 us, frequency: 1.01 times/s
=> wdt list
watchdog@20bc000 (imx_wdt)
=> wdt dev watchdog@20bc000
=> wdt start 50000
WDT: Started watchdog@20bc000 with servicing every 1000ms (50s timeout)
=> cyclic list
function: watchdog@20bc000, cpu-time: 37 us, frequency: 1.03 times/s
function: watchdog@20bc000, cpu-time: 241 us, frequency: 1.01 times/s
=> wdt start 12345
WDT: Started watchdog@20bc000 with servicing every 1000ms (12s timeout)
=> cyclic list
function: watchdog@20bc000, cpu-time: 36 us, frequency: 1.03 times/s
function: watchdog@20bc000, cpu-time: 100 us, frequency: 1.04 times/s
function: watchdog@20bc000, cpu-time: 299 us, frequency: 1.00 times/s
So properly unregister the watchdog device from the cyclic framework
in wdt_stop(). In wdt_start(), we cannot just skip the registration,
as the (new) timeout value may mean that we have to ask the cyclic
framework to call us more often. So if we're already running,
first unregister the old cyclic instance.
Reviewed-by: Stefan Roese <sr@denx.de>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
We are not checking the return value of strdup(), nor
freeing the string in cyclic_unregister().
However, all current users either pass a string literal or the
dev->name of the client device. So in all cases the name string will
live at least as long as the cyclic_info is registered, so just make
that a requirement.
Reviewed-by: Stefan Roese <sr@denx.de>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
We currently only describe the process to enable measured boot using
bootm. Describe the UEFI requirements as well which predate bootm.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
The PC client spec [0], doesn't describe measurements for DTBs. It does
describe what do to for ACPI tables though.
There is a description for ACPI in 3.3.4.1 PCR[0] – SRTM, POST BIOS,
and Embedded Drivers and they explicitly mention ACPI in there. There's
no mention of ACPI in 3.3.4.2 PCR[1] – Host Platform Configuration.
However, in Figure 6 -- PCR Mapping of UEFI Components ACPI is shown
in PCR1. The general description also mentions PCR0 is for code and PCR1
is for data such as ACPI and SMBIOS.
So let's switch over the DTB measurements to PCR1 which seems a better
fit.
[0] https://trustedcomputinggroup.org/resource/pc-client-specific-platform-firmware-profile-specification
Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Eddie James <eajames@linux.ibm.com>
We moved our documentation to another hoster and therefore the URL
changed. Point to the latest documentation instead of release versions
to not link out-dated documentation.
Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
We moved our documentation to another hoster and therefore the URL
changed. Point to the latest documentation instead of release versions
to not link out-dated documentation.
Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Reviewed-by: Wadim Egorov <w.egorov@phytec.de>
Loading and running the ELF image is the responsibility of the
library and should not be associated with the command line interface.
It is also required to run ELF images from FIT with the bootm command
so as not to depend on the command line interface.
Signed-off-by: Maxim Moskalets <maximmosk4@gmail.com>
