Simon Glass [Wed, 24 Nov 2021 16:26:46 +0000 (09:26 -0700)]
sandbox: Drop CONFIG_SYS_TIMER_RATE
This is not used by sandbox since it uses driver model for the timer.
Drop it.
Also update the tools_only build to avoid build errors, since it does
actually build U-Boot too. Enable DM so we can use CONFIG_TIMER,
disable EFI_LOADER to avoid an error about board_quiesce_devices() and
disable NET to avoid having to define CONFIG_AVB_BUF_ADDR
Simon Glass [Wed, 24 Nov 2021 16:26:39 +0000 (09:26 -0700)]
trace: sandbox: Use only the Kconfig options
At present there are Kconfig options for tracing, but sandbox uses
plain #defines to set them. Correct this and make the tracing command
default to enabled so that this is not needed.
Simon Glass [Wed, 24 Nov 2021 04:09:52 +0000 (21:09 -0700)]
binman: Support lists of external blobs
Sometimes it is useful to have a list of related external blobs in a
single entry. An example is the DDR binaries used by meson. There are
9 files in total. Add support for this, so we don't have to have a
separate entry for each.
Simon Glass [Wed, 24 Nov 2021 04:09:50 +0000 (21:09 -0700)]
binman: Allow extracting a file in an alternative format
In some cases entries encapsulate other data and it is useful to access
the data within. An example is the fdtmap which consists of a 16-byte
header, followed by a devicetree.
Provide an option to specify an alternative format when extracting files.
In the case of fdtmap, this is 'fdt', which produces an FDT file which can
be viewed with fdtdump.
Simon Glass [Wed, 24 Nov 2021 04:09:49 +0000 (21:09 -0700)]
binman: Allow listing an image created by a newer version
If an older version of binman is used to list images created by a newer
one, it is possible that it will contain entry types that are not
supported. At present this produces an error.
Adjust binman to use a plain 'blob' entry type to cope with this, so the
image can at least be listed.
Simon Glass [Wed, 24 Nov 2021 04:09:48 +0000 (21:09 -0700)]
binman: Allow providing tools and blob directories
At present it is necessary to symlink files containing external blobs into
the U-Boot tree in order for binman to find them. This is not very
convenient.
Add two new environment/Makefile variables to help with this. Add
documentation as well, fixing a related nit.
Simon Glass [Thu, 4 Nov 2021 03:09:20 +0000 (21:09 -0600)]
bloblist: Support allocating the bloblist
Typically the bloblist is positioned at a fixed address in memory until
relocation. This is convenient when it is set up in SPL or before
relocation.
But for EFI we want to set it up only when U-Boot proper is running. Add
a way to allocate it using malloc() and update the documentation to cover
this aspect of bloblist.
Note there are no tests of this feature at present, nor any direct testing
of bloblist_init().
This can be added, e.g. by making this option controllable at runtime.
Simon Glass [Tue, 23 Nov 2021 18:03:52 +0000 (11:03 -0700)]
binman: cbfs: Refactor the init process
Update the constructor to work in the recommended way, where the node
properties are read in a separate function. This makes it more similar to
entry_Section.
Simon Glass [Tue, 23 Nov 2021 18:03:51 +0000 (11:03 -0700)]
binman: Use normal entries in cbfs
This currently uses _cbfs_entries[] to store entries. Since the entries
are in fact valid etypes, we may as well use the same name as
entry_Section uses, which is _entries. This allows reusing more of the
code there (in a future patch).
Simon Glass [Tue, 23 Nov 2021 18:03:50 +0000 (11:03 -0700)]
binman: Move cbfs.ObtainContents() down a bit
It is easier to understand this file if reading the entries comes before
obtaining the contents, since that is the order in which Binman proceeds.
Move the function down a bit.
Simon Glass [Tue, 23 Nov 2021 18:03:48 +0000 (11:03 -0700)]
binman: Allow control of which entries to read
The ObtainContents() and GetEntryContents() methods in this file read
every single entry in the section. This is the common case.
However when one of the entries has had its data updated (e.g. with
'binman replace') we don't want to read it again from the file. Allow
the entry to be skipped, for this purpose. This is currently done in the
CBFS implementation, so adding it here will allow that to use more of
the entry_Section code.
Simon Glass [Tue, 23 Nov 2021 18:03:47 +0000 (11:03 -0700)]
binman: Allow overriding BuildSectionData()
This method is currently marked private. However it is useful to be able
to subclass it, since much of the entry_Section code can be reused. Rename
it.
Also document one confusing part of this code, so people can understand
how to add a test for this case.
Fix up a few pylint warnings to avoid regressing the score.
Simon Glass [Tue, 23 Nov 2021 18:03:42 +0000 (11:03 -0700)]
binman: Add a way to obtain the version
Add a -V option which shows the version number of binman. For now this
just uses a local 'version' file. Once the tool is packaged in some way
we can figure out an approach that suits.
Simon Glass [Tue, 23 Nov 2021 18:03:38 +0000 (11:03 -0700)]
dtoc: Bring in the libfdt module automatically
Use the same technique as with binman to load this module from the U-Boot
tree if available. This allows running tests without having to specify
the PYTHONPATH variable.
* fixes the bug in function bind_drivers_pass that for
CONFIG_CC_OPTIMIZE_FOR_SIZE=n and no entries in the driver_info list,
i.e. n_ents == 0, the processor steps into the first loop iteration
despite the loop condition being false.
* the Xilinx Zynq-7000 device would eventually hang due to an attempted
access to an invalid memory address
* the bug is fixed by changing the type of idx from uint to int
Board: zynq-zybo
Target: ARM
Compiler: arm-none-eabi-gcc 9.2.1
Signed-off-by: Alexander Preissner <fpga-garage@preissner-muc.de> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Radu Bulie [Sat, 27 Nov 2021 12:52:35 +0000 (14:52 +0200)]
drivers: net: Soft reset felix switch core
It turns out that in custom designs if the system is reset
multiple times in conjunction with a slight increase in external
temperature, the felix switch starts to behave in a strange way:
packets are no longer received on the ENECT interface connected
to the L2switch internal port (the TX side of internal port stops working
or the packets do not reach there. It is not very clear where
the packets remain blocked. None of the counters points to a disruption
in the L2switch)
The issue is not reproducible on NXP reference designs.
It was observed that by adding the switch core reset, the problem
goes aways, even if intensive testing in temperature chambers
is applied.
The current patch performs soft reset on the switch core to ensure proper
operation of the L2switch.
Signed-off-by: Radu Bulie <radu-andrei.bulie@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Nishanth Menon [Tue, 19 Oct 2021 17:32:29 +0000 (12:32 -0500)]
spl: fit: Skip attempting to load 0 length image
When, for various reasons, a bad FIT image is used where a loadable
image is marked as 0 length, attempt is made for a 0 length allocation and
read of 0 byte read operation.
Instead provide warning in log and skip attempting to do such a load.
scripts: remove CONFIG_IS_ENABLED and CONFIG_VAL in config_whitelist.txt
The helper macro CONFIG_IS_ENABLED and CONFIG_VAL are not real
configurations and they are no more present in u-boot.cfg so they can
be removed in config_whitelist.txt.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Acked-by: Simon Glass <sjg@chromium.org>
scripts: remove CONFIG_IS_ENABLED and CONFIG_VAL in generated u_boot.cfg
The two helpers macros CONFIG_IS_ENABLED and CONFIG_VAL are defined in
include/linux/kconfig.h but they are not real configurations; they can
be safely removed in the generated configuration file "u-boot.cfg".
This patch simplifies the comparison of this U-Boot configuration file.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Acked-by: Simon Glass <sjg@chromium.org>
Tom Rini [Wed, 1 Dec 2021 18:30:59 +0000 (13:30 -0500)]
Merge branch '2021-12-01-Kconfig-migrations' into next
- Finish converting CONFIG_USE_BOOTCOMMAND, CONFIG_BOOTCOMMAND,
CONFIG_RAMBOOTCOMMAND, CONFIG_NFSBOOTCOMMAND, all of
CONFIG_SYS_[BO]R[0-7]_PRELIM, CONFIG_FSL_DDR_BIST and
CONFIG_FSL_DDR_INTERACTIVE.
Tom Rini [Sat, 13 Nov 2021 23:10:40 +0000 (18:10 -0500)]
Convert CONFIG_SYS_BR0_PRELIM et al to Kconfig
This converts the following to Kconfig:
CONFIG_SYS_BR0_PRELIM
CONFIG_SYS_OR1_PRELIM
CONFIG_SYS_BR1_PRELIM
CONFIG_SYS_OR2_PRELIM
CONFIG_SYS_BR2_PRELIM
CONFIG_SYS_OR2_PRELIM
CONFIG_SYS_BR3_PRELIM
CONFIG_SYS_OR3_PRELIM
CONFIG_SYS_BR4_PRELIM
CONFIG_SYS_OR4_PRELIM
CONFIG_SYS_BR5_PRELIM
CONFIG_SYS_OR5_PRELIM
CONFIG_SYS_BR6_PRELIM
CONFIG_SYS_OR6_PRELIM
CONFIG_SYS_BR7_PRELIM
CONFIG_SYS_OR7_PRELIM
This also introduces CONFIG_SYS_BR0_PRELIM_BOOL as not all platforms
that can set these values do so. Add the relevant SYS_BRx_PRELIM_BOOL
to platforms that had not been previously migrated.
Tom Rini [Wed, 1 Dec 2021 12:22:25 +0000 (07:22 -0500)]
Merge tag 'u-boot-stm32-20211130' of https://source.denx.de/u-boot/custodians/u-boot-stm into next
- add nor1 device support for DFU command
- remove CONFIG_STM32_IPCC from stm32mp15 defconfigs
- enable simple framebuffer node for splashscreen for stm32mp1
- use lower-case hex for address for stm32 MCU and MPU's device tree
- define LOG_CATEGORY for stmfx pinctrl driver
- add support for probing bus voltage level translator
- add custom PHY reset bindings on AV96
- enable KSZ90x1 PHY driver on DHCOR
- stm32mp1 DDR update:
- add DDR read data eye training
- remove DDR calibration result
- remove DDR tuning support
- compute DDR size from DDRCTL registers
- DHSOM boards:
- increase USB power-good delay
- add update_sf script to install U-Boot into SF
- increase PHY auto-negotiation timeout to 20 seconds
- fix SoM and board coding strap GPIO handling
- auto-detect uSD level translator
Patrick Delaunay [Thu, 25 Nov 2021 10:54:53 +0000 (11:54 +0100)]
board: stm32mp1: add support of nor1 device in dfu command
Add support of mtd backend for nor1 when this device is present on the
board, on STM32MP157C-EV1 for example, as the support of several MTD
spi-nor instance are now supported with commit b7f060565e31 ("mtd:
spi-nor: allow registering multiple MTDs when DM is enabled").
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Wed, 24 Nov 2021 09:52:19 +0000 (10:52 +0100)]
ram: stm32mp1: remove __maybe_unused on stm32mp1_ddr_setup
Since the commit f42045b2e750 ("stm32mp15: replace CONFIG_TFABOOT when
it is possible") the function stm32mp1_ddr_setup is always called so the
__maybe_unused can be removed.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Wed, 24 Nov 2021 09:52:18 +0000 (10:52 +0100)]
ram: stm32mp1: compute DDR size from DDRCTL registers
Compute the DDR size from DDR controller register (mstr and addrmap)
in U-Boot proper as the DDR information are useful only for SPL
but not for U-Boot proper, for example with TFABOOT.
This patch simplify U-Boot DT when several DDR size are supported
and support of next SOC in STM32MP family.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com> Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Patrick Delaunay [Fri, 19 Nov 2021 09:07:32 +0000 (10:07 +0100)]
configs: stm32mp15: deactivate the CONFIG_STM32_IPCC
The IPCC mailbox is only used for communication with M4 firmware but
it is not used in the stm32 remoteproc driver; it was planed but the
support of this mailbox in remoteproc for STM32MP15x is dropped.
So the associated drivers and config CONFIG_STM32_IPCC can be
deactivated to reduce the U-Boot size; the CONFIG_DM_MAILBOX can be
also deactivated as the mailbox UCLASS is no more used.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Mon, 15 Nov 2021 15:32:23 +0000 (16:32 +0100)]
board: stm32mp1: enable simple framebuffer node for splashscreen
Enable an existing simple framebuffer node in the Linux kernel device
tree and the add the associated reserved memory node to preserved the
resources (clock, memory) used by the stm32 video driver to display
the splashscreen = background in exlinux.conf file.
These resources will be released by the Linux driver only when the
associated driver is ready to avoid transition issues during the Linux
kernel initialization between U-Boot splash screen and the final display.
See Linux documentation for details:
Documentation/devicetree/bindings/display/simple-framebuffer.yaml
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Mon, 15 Nov 2021 15:32:22 +0000 (16:32 +0100)]
video: stm32: stm32_ltdc: align framebuffer on 2MB
Align the framebuffer size on MMU_SECTION_SIZE in kernel, = max 2MB for
LPAE for armV7, to avoid issue with the simple frame buffer activation,
when U-Boot add a reserved memory in the kernel device tree to preserve
the splash screen until Linux driver initialization.
See Linux documentation for details:
Documentation/devicetree/bindings/display/simple-framebuffer.yaml
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Mon, 15 Nov 2021 15:32:18 +0000 (16:32 +0100)]
common: rename lcd_simplefb.c file to fdt_simplefb.c
Rename the file lcd_simplefb.c to fdt_simplefb.c to be aligned
with the configuration name and with the associated include file
./include/fdt_simplefb.h
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Mon, 15 Nov 2021 14:32:30 +0000 (15:32 +0100)]
stm32mp1: ram: remove tuning support
Remove the DDR interactive command tuning, as the support of a predefined
DDR PHY tuning is removed for STM32MP1 driver in SPL and in TF-A
and the result of this tuning will be never used.
Moreover this SW tuning procedure can failed on some hardware
configuration (to many BIST errors and no convergence); it will be no
more supported in the next delivery of the DDR utilities included in
the CubeMX tool of STMicroelectronics.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Mon, 15 Nov 2021 14:32:29 +0000 (15:32 +0100)]
stm32mp1: ram: remove the support of calibration result
The support of a predefined DDR PHY tuning result is removed for
STM32MP1 driver because it is not needed at the supported frequency
when built-in calibration is executed.
The calibration parameters were provided in the device tree by the
optional node "st,phy-cal", activated in ddr helper file by the
compilation flag DDR_PHY_CAL_SKIP and filled with values generated
by the CubeMX DDR utilities.
This patch
- updates the binding file to remove "st,phy-cal" support
- updates the device trees and remove the associated defines
- simplifies the STM32MP1 DDR driver and remove the support of
the optional parameter "st,phy-cal"
After this patch, the built-in calibration is always executed
and the calibration registers are moved in the phy dynamic part;
that allows manual tests.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:29:44 +0000 (03:29 +0100)]
ARM: dts: stm32: Auto-detect DHSOM with uSD level translator
The uSD level translator on DHSOM and Avenger96 are optional, however it
is possible to auto-detect it. This is done by setting SD CMD line high,
and then testing whether signal level on CK line matches the signal level
on CKIN line. If so, the uSD level translator is present, otherwise it is
not populated.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:29:43 +0000 (03:29 +0100)]
mmc: stm32_sdmmc2: Add support for probing bus voltage level translator
Add support for testing whether bus voltage level translator is present
and operational. This is useful on systems where the bus voltage level
translator is optional, as the translator can be auto-detected by the
driver and the feedback clock functionality can be disabled if it is
not present.
The translator test sets CMD high to avoid interfering with a card, and
then verifies whether signal set on CK is detected on CKIN. If the signal
is detected, translator is present, otherwise the CKIN feedback clock are
disabled.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Yann Gautier <yann.gautier@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:27:37 +0000 (03:27 +0100)]
ARM: dts: stm32: Add custom PHY reset bindings on AV96
The ethernet PHY must be reset on AV96, however DWMAC currently does
not support the MDIO-bus PHY GPIO reset bindings and the ethernet MAC
PHY reset property is going away on next DT sync. Add PHY specific
reset bindings to trigger the PHY reset and fix sporadic ethernet
malfunctions, until the next DT sync.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:26:39 +0000 (03:26 +0100)]
ARM: stm32: Fix SoM and board coding strap GPIO handling on DHSOM
The variables retaining the strap values have to be initialized, always,
make it so. Moreover, free the requested GPIO list at the end to avoid
wasting memory.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:26:05 +0000 (03:26 +0100)]
ARM: stm32: Increase PHY auto-negotiation timeout to 20s on DHSOM
The Micrel PHYs on known DHSOM based boards take a while to come out
of reset, increase the auto-negotiation timeout to prevent it from
timing out in case the ethernet is used right after the board was
reset.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:25:13 +0000 (03:25 +0100)]
ARM: stm32: Add update_sf script to install U-Boot into SF on DHSOM
Add script to read U-Boot from SD card and write it to matching
locations in the SPI NOR, thus making the SPI NOR bootable. The
script erases the entire SPI NOR, including U-Boot environment,
to make sure the installation is clean. To retain environment
from current running U-Boot, run 'saveenv' after running the
'update_sf' script.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Marek Vasut [Sat, 13 Nov 2021 02:24:44 +0000 (03:24 +0100)]
ARM: stm32: Increase USB power-good delay on DHSOM
The USB hub on STM32MP1 DHCOM boards needs to wait a bit longer until
the USB Vbus is stable. Increase the USB power-good delay to 1 s.
This adds default-undefined STM32MP_BOARD_EXTRA_ENV variable into
stm32mp15_common.h to reduce duplication in board-specific config
files adding custom environment.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Vincent Stehlé [Wed, 24 Nov 2021 14:54:20 +0000 (15:54 +0100)]
qemu: common: Fix build with update capsule
The common emulation Makefile has a dependency on a non-existent
qemu_capsule.o when building with support for capsule update enabled
(CONFIG_EFI_RUNTIME_UPDATE_CAPSULE=y).
The code which was in qemu_capsule.c has been completely moved to
lib/efi_loader/efi_capsule.c by commit 7a6fb28c8e4b ("efi_loader: capsule:
add back efi_get_public_key_data()").
Remove the false dependency.
This fixes the following build error:
make[1]: *** No rule to make target 'board/emulation/common/qemu_capsule.o', needed by 'board/emulation/common/built-in.o'. Stop.
Fixes: commit 47a25e81d35c ("Revert "efi_capsule: Move signature from DTB to .rodata"") Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com> Cc: Simon Glass <sjg@chromium.org>
Simon Glass [Sat, 23 Oct 2021 23:26:12 +0000 (17:26 -0600)]
test/py: Raise a ValueError if a command fails
At present an Exception is raised if a command fails. This is a very broad
class and makes it difficult for callers to catch the error without also
catching other things, like programming bugs.
Simon Glass [Sat, 23 Oct 2021 23:26:11 +0000 (17:26 -0600)]
test/py: Relax the naming rules for unit tests
At present the collection function used by pytest is quite strict on the
naming of the functions it detects. In particular it requires the name of
the test to be repeated in the function name.
This is not enforced anywhere else, but instead the tests are silently
omitted from the pytest run. This affects a few dozen tests.
The rule does not seem to have any particular purpose. Relax it, so that
all tests that use the UNIT_TEST() macro will run, regardless of the name
of the test function.
Simon Glass [Sat, 23 Oct 2021 23:26:10 +0000 (17:26 -0600)]
sandbox: Enable HEXDUMP for sandbox_flattree
At present the hexdump tests are disabled in sandbox_flattree. This is
good, because they do not pass. Enable the required Kconfig so that they
will, when enabled.
Simon Glass [Sat, 23 Oct 2021 23:26:08 +0000 (17:26 -0600)]
dm: core: Allow finding children / uclasses by partial name
In some cases it is useful to search just by a partial name, such as
when looking for a sibling device that has a common name substring. Add
helper functions to handle these requirements.
Simon Glass [Sat, 23 Oct 2021 23:26:07 +0000 (17:26 -0600)]
dm: core: Add a way to obtain a string list
At present we support reading a string list a string at a time. Apart
from being inefficient, this makes it impossible to separate reading of
the devicetree into the of_to_plat() method where it belongs, since any
code which needs access to the string must read it from the devicetree.
Add a function which returns the string property as an array of pointers
to the strings, which is easily used by clients.
Simon Glass [Sat, 23 Oct 2021 23:26:05 +0000 (17:26 -0600)]
dm: core: Fix handling of uclass pre_unbind method
This method is currently called after the platform data has been freed.
But the pre_unbind() method may wish to access this, e.g. to free some
data structures stored there.
Split the unbinding of devices into two pieces, as is done with removal.
This corrects the problem.
Also tidy a code-style issue in device_remove() while we are here.
Simon Glass [Sat, 23 Oct 2021 23:26:04 +0000 (17:26 -0600)]
dm: core: Add tests for stringlist functions
These functions currently lack tests so add some. The error handling
differs betwee livetree and flattree at present, so only check the error
codes with livetree.
Simon Glass [Sat, 23 Oct 2021 23:26:03 +0000 (17:26 -0600)]
common: Allow a smaller console-recording pre-reloc
Before relocation there is generally not as much available memory and not
that much console output. At present the console-output buffer is the same
side before and after relocation. Add a separate Kconfig option to remove
this limitation.
Simon Glass [Sat, 23 Oct 2021 23:26:02 +0000 (17:26 -0600)]
disk: part_dos: Fix a NULL pointer error
When ext is NULL we cannot dereference it. Update the code flow to avoid
this, so that layout_mbr_partitions() can be used with partition tables
that do not include an extended partition.