When switching on or off the ARM caches some care must be taken to ensure
existing cache line allocations are not left in an inconsistent state.
An example of this is when cache lines are considered non-shared by
and L3 controller even though the lines are shared. To prevent these
and other issues all cache lines should be cleared before enabling
or disabling a coherent master's cache. ARM cores and many L3 controllers
provide a way to efficiently clean out all cache lines to allow for
this, unfortunately there is no such easy way to do this on current K3
MSMC based systems.
We could explicitly clean out every valid external address tracked by
MSMC (all of DRAM), or we could attempt to identify only the set of
addresses accessed by a given boot stage and flush only those
specifically. This patch attempts the latter. We start with cleaning the
SPL load address. More addresses can be added here later as they are
identified.
Note that we perform a flush operation for both the flush and invalidate
operations, this is not a typo. We do this to avoid the situation that
some ARM cores will promote an invalidate to a clean+invalidate, but only
emit the invalidation operation externally, leading to a loss of data.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Tested-by: Faiz Abbas <faiz_abbas@ti.com>
Add DT entries for main domain watchdog0 and 1 instances on the J721e
well as RTI1-based watchdog on the AM65x. RTI0 does not work for this
purpose on the AM65x, so leave it out.
On AM65x, we mark the power-domain as shared because RTI firmware such
as https://github.com/siemens/k3-rti-wdt may request it as well in order
to prevent accidental shutdown of the watchdog.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
This is based on the Linux kernel driver for the RTI watchdog.
To actually reset the system on an AM65x, it requires firmware running
on the R5 that accepts the NMI and issues the actual system reset via
TISCI. Kind of an iTCO, except that this watchdog hardware has support
for no-way-out, and only for that.
On the J721E, reset works without extra firmware help when routing the
RTI interrupt via the ESM.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Use "cons.config.build_dir" instead of writing to the source directory
(read-only). This will fix the test failures in Azure.
Signed-off-by: Joao Marcos Costa <joaomarcos.costa@bootlin.com>
Up to now we used the value of the bootargs environment variable as load
options in the boot manager. This is not correct. The data has to be taken
from the Boot#### variable.
Let the boot manager copy the optional data of the EFI_LOAD_OPTION as load
options to the loaded image protocol.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The bootefi bootmgr command has to set the load options for a loaded image
from the value of BootXXXX variable. If the boot manager is not used, the
value is set from the environment variable bootargs (or efi_selftest).
Factor out a common function efi_set_load_options().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Some amount of code was moved/derived from efi_variable.c regarding
UEFI secure boot, in particluar in the commit 012c56ac76 ("efi_loader:
restructure code for TEE variables").
So add the orignal author's copyright notice.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
This converts the following to Kconfig:
CONFIG_SYS_MMC_ENV_DEV
CONFIG_SYS_MMC_ENV_PART
Note that with this conversion we now have consistent behavior with
respect to ensuring that we have always selected the correct MMC
device and hardware partition.
Signed-off-by: Tom Rini <trini@konsulko.com>
At present with sunxi 64-bit, the Makefile builds
u-boot-sunxi-with-spl.bin and then binman overwrites it with its own
version. But the binman definition lacks some parts, in particular
BL31.
For now, work around this with a hack.
Tested-by: Andre Przywara <andre.przywara@arm.com>
Tested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
Fixes: 42b18df80f ("x86: Makefile: Drop explicit targets built by binman")
On some devices the environment variable serial# cannot be overwritten.
Set the variable only if it is not set.
For our unit test it is sufficient to test if any value for serial-number
is set.
Fixes: 8a5cdf601f ("test: efi_selftest: Do not force serial# setting")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The hush parser not enabled for some boards, e.g.
sipeed_maix_bitm_defconfig.
With CONFIG_HUSH_PARSER=n a double quotation mark is not interpreted as the
beginning of a string. Use a single quotation mark instead.
Furthermore without the hush parser variables have to be referenced as
${varname}. Add the missing braces.
Reported-by: Sean Anderson <seanga2@gmail.com>
Fixes: 8b86c609b8 ("test/py: add test of basic shell functionality")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Modify various test/py filesystem creation routines to support systems
that don't implement the metadata_csum ext4 feature.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Add LED support for Cortina Access Presidio Engineering Board
Signed-off-by: Jway Lin <jway.lin@cortina-access.com>
Signed-off-by: Alex Nemirovsky <alex.nemirovsky@cortina-access.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
CC: Simon Glass <sjg@chromium.org>
Add Cortina Access LED controller support for CAxxxx SOCs
Signed-off-by: Jway Lin <jway.lin@cortina-access.com>
Signed-off-by: Alex Nemirovsky <alex.nemirovsky@cortina-access.com>
CC: Simon Glass <sjg@chromium.org>
Add head file fixed link error and remove unused flashing function
Reviewed-by: Simon Glass <sjg@chromium.org>
Add Python scripts to test 'ls' and 'load' commands. The scripts
generate a SquashFS image and clean the directory after the assertions,
or if an exception is raised.
Signed-off-by: Joao Marcos Costa <joaomarcos.costa@bootlin.com>
Add call to zlib's 'uncompress' function. Add function to display the
right error message depending on the decompression's return value.
Signed-off-by: Joao Marcos Costa <joaomarcos.costa@bootlin.com>
Add zlib (v1.2.11) uncompr() function to U-Boot. SquashFS depends on
this function to decompress data from a raw disk image. The actual
support for zlib into SquashFS sources will be added in a follow-up
commit.
Signed-off-by: Joao Marcos Costa <joaomarcos.costa@bootlin.com>
Add support for SquashFS filesystem. Right now, it does not support
compression but support for zlib will be added in a follow-up commit.
Signed-off-by: Joao Marcos Costa <joaomarcos.costa@bootlin.com>
Add various fixes and improvements to this command that were missed in
the original version. Unfortunately I forgot to send v2.
- Fix Kconfig name
- Use a separate variable for the remaining search length
- Correct a minor bug
- Move into a separate test suite
- Add -q flag to the 'quiet' test to test operation when console is enabled
- Enable the feature for sandbox
Signed-off-by: Simon Glass <sjg@chromium.org>
The test flags used by driver model are currently not available to other
tests. Rather than creating two sets of flags, make these flags generic
by changing the DM_ prefix to UT_ and moving them to the test.h header.
This will allow adding other test flags without confusion.
Signed-off-by: Simon Glass <sjg@chromium.org>
On boards without console recording these function are currently missing.
It is more convenient for them to be present but to return dummy values.
That way if we know that a test needs recording, we can check if it is
available, and skip the test if not, while avoiding #ifdefs.
Update the header file according and adjust console_record_reset_enable()
to return an error if recording is not available.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some lines of the output are not worth testing, or not worth testing in
their entirety. For example, when checking a hex dump we know that the
hex-dump routine can display ASCII so we only need to check the hex bytes,
not the ASCII dump. Add a new test macros which can check only part of
a console line.
Sometimes it is useful to skip a line altogether, so add a macro for that
also.
Signed-off-by: Simon Glass <sjg@chromium.org>
The external data is located after the mmapped FDT pointed to by
'old_fdt', not in the newly created FDT we are importing into at 'fdt'.
Signed-off-by: Patrick Oppenlander <patrick.oppenlander@gmail.com>
This patch addresses issue #2 for signed configurations.
-----8<-----
Including the image cipher properties in the configuration signature
prevents an attacker from modifying cipher, key or iv properties.
Signed-off-by: Patrick Oppenlander <patrick.oppenlander@gmail.com>
Reviewed-by: Philippe Reynes <philippe.reynes@softathome.com>
Previously, mkimage -F could be run multiple times causing already
ciphered image data to be ciphered again.
Signed-off-by: Patrick Oppenlander <patrick.oppenlander@gmail.com>
Reviewed-by: Philippe Reynes <philippe.reynes@softathome.com>
Also replace fdt_delprop/fdt_setprop with fdt_setprop as fdt_setprop can
replace an existing property value.
Signed-off-by: Patrick Oppenlander <patrick.oppenlander@gmail.com>
Reviewed-by: Philippe Reynes <philippe.reynes@softathome.com>
Previously mkimage would process any node matching the regex cipher.*
and apply the ciphers to the image data in the order they appeared in
the FDT. This meant that data could be inadvertently ciphered multiple
times.
Switch to processing a single cipher node which exactly matches
FIT_CIPHER_NODENAME.
Signed-off-by: Patrick Oppenlander <patrick.oppenlander@gmail.com>
Reviewed-by: Philippe Reynes <philippe.reynes@softathome.com>
blkcache_init manually relocates blkcache list pointers when
CONFIG_NEEDS_MANUAL_RELOC is enabled. However, it is called very late in
the boot sequence, which could be a problem if previous boot calls execute
blkcache operations with the non-relocated pointers. For example, mmc is
initialized earlier and might call blkcache_invalidate (in
mmc_select_hwpart()) when trying to load the environment from mmc via
env_load().
To fix this issue, move blkcache_init boot call earlier, before mmc gets
initialized.
Acked-by: Angelo Dureghello <angelo.dureghello@timesys.com>
Tested-by: Angelo Dureghello <angelo.dureghello@timesys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Extend manual relocation of block_cache list pointers to all platforms that
enable CONFIG_NEEDS_MANUAL_RELOC. Remove m68k-specific checks and provide a
single implementation that adds gd->reloc_off to the pre-relocation
pointers.
Acked-by: Angelo Dureghello <angelo.dureghello@timesys.com>
Tested-by: Angelo Dureghello <angelo.dureghello@timesys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Eric Nelson <eric@nelint.com>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
[trini: Add guard around DECLARE_GLOBAL_DATA_PTR to avoid size growth]
Signed-off-by: Tom Rini <trini@konsulko.com>
serial_initialize is called only during the common init sequence, after
relocation (in common/board_r.c). Because it has a void return value, it
has to wrapped in initr_serial. In order to be able to get rid of this
indirection, make serial_initialize return int.
Remove extern from prototype in order to silence the following checkpatch
warning:
check: extern prototypes should be avoided in .h files
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Use IS_ENABLED() instead of #ifdef in blk_post_probe function.
No functional change intended.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
[trini: Fix thinko and use CONFIG_HAVE_BLOCK_DEVICE in IS_ENABLED()]
Signed-off-by: Tom Rini <trini@konsulko.com>
Now that all arch specific code was converted to setup_bdinfo, we can
remove setup_board_part1.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Factor out ppc-specific bdinfo setup from generic init sequence to
arch_setup_bdinfo in arch/powerpc/lib/bdinfo.c.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Factor out m68k-specific bdinfo setup to arch_setup_bdinfo in
arch/m68k/lib/bdinfo.c. Also, use if(IS_ENABLED()) instead of #ifdef where
possible.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Move sram related bdinfo from arch-specific setup_board_part1 to generic
code in setup_bdinfo. Also use "if (IS_ENABLED(CONFIG_SYS_HAS_SRAM))"
instead of "#ifdef CONFIG_SYS_SRAM_BASE".
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Move all assignments to gd->bd->bi_mem{start,size} to generic code in
setup_bdinfo.
Xtensa architecture is special in this regard as it defines its own
handling of gd->bd->bi_mem{start,size} fields. In order to avoid defining
a weak SDRAM function, let arch_setup_bdinfo overwrite the generic flags.
For ARC architecture, remove ARCH_EARLY_INIT_R from Kconfig since it is
not needed anymore.
Also, use gd->ram_base to populate bi_memstart to avoid an ifdef.
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Acked-by: Alexey Brodkin <abrokdin@synopsys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Certain architectures (ppc, mips, sh, m68k) use setup board_part1 and
setup_board_part2 calls during pre-relocation init to populate gd->bd
boardinfo fields. This makes the generic init sequence cluttered with
arch-specific ifdefs.
In order to clean these arch-specific sequences from generic init,
introduce arch_setup_bdinfo weak initcall so that everyone can define their
own bdinfo setup routines.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Introduce setup_bdinfo initcall as a generic routine to populate bdinfo
fields.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>