Tom Rini [Mon, 20 Jun 2022 12:09:24 +0000 (08:09 -0400)]
Merge tag 'u-boot-stm32-20220620' of https://source.denx.de/u-boot/custodians/u-boot-stm into next
- Add STM32MP13 SoCs support with associated board STM32M135F-DK
- Correct livetree support in stm32mp1 boards
- Activate livetree for stm32mp15 DHSOM boards
Michael Walle [Mon, 30 May 2022 21:02:09 +0000 (23:02 +0200)]
board: sl28: support 8 GiB memory
The board supports up to 8 GiB memory. The memory is soldered on the
board but the configuration is equivalent to a dual chip select, dual
rank DIMM module.
Michael Walle [Mon, 30 May 2022 21:02:07 +0000 (23:02 +0200)]
board: sl28: set CPO value
With a 8GiB memory board, it seems that the "very unlikely event" of a
DDR initialization with non-optimal values are not really that unlikely.
It happens in about every other reboot. As described in erratum
A-009942, preset the DEBUG_28 register with an optimal value. The value
iself depends on the memory configuration of the board, but the used
value seems to work well for all variants.
Michael Walle [Mon, 30 May 2022 21:02:05 +0000 (23:02 +0200)]
armv8: layerscape: add missing RCW source defines
A board might need to get the source of the RCW word, which is also the
boot source in most cases.
These defines are taken from the LS1028A and I expect they are the same
across the SoCs with the same chassis, after all, there was already a
reset source for NOR flash.
Pali Rohár [Thu, 26 May 2022 08:52:27 +0000 (10:52 +0200)]
board: freescale: p1_p2_rdb_pc: Move boot reset macros to p1_p2_bootsrc.h
Code for changing boot source is platform generic and can be used by any
P1* and P2* compatible RDB board. Not only by boards which use config
header file p1_p2_rdb_pc.h.
So move this code from p1_p2_rdb_pc.h to p1_p2_bootsrc.h and cleanup macros
for generating boot source env variables in CONFIG_EXTRA_ENV_SETTINGS.
This allows to use code for resetting board and rebooting to other boot
source also by other boards in future.
Pali Rohár [Wed, 11 May 2022 18:27:13 +0000 (20:27 +0200)]
mmc: fsl_esdhc: Add new config option for default fallback mode
Currently default fallback SDHC mode is 1-bit. Add new config option
CONFIG_SYS_FSL_ESDHC_DEFAULT_BUS_WIDTH to allow specifying default fallback
mode. This is useful e.g. for SPL builds which loads other parts from SD
card during boot process.
Pali Rohár [Wed, 11 May 2022 18:27:12 +0000 (20:27 +0200)]
mmc: fsl_esdhc: Set fallback mode to 1-bit
8-bit mode is not supported by SD cards and on P2020 are four SDHC pins
shared with SPI (so if P2020 board have also SPI then only 4-bit SDHC mode
is provided). So 8-bit SDHC mode is really bad default.
When max bus width is not provided then set mode to 1-bit. This mode is
supported by all cards, so it is the best option for fallback mode.
Also P2020 bootrom sets mode to 1-bit when booting from SD/MMC card.
Alison Wang [Tue, 10 May 2022 10:29:10 +0000 (18:29 +0800)]
ls1028a: hdp: Add config support for HDP firmware loading
This patch adds config support for HDP firmware loading on LS1028A.
Signed-off-by: Oliver Brown <oliver.brown@nxp.com> Signed-off-by: Alison Wang <alison.wang@nxp.com> Signed-off-by: Ye Li <ye.li@nxp.com> Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Pali Rohár [Sun, 1 May 2022 15:45:58 +0000 (17:45 +0200)]
powerpc: fsl_law: Add definition for first PCIe target interface
Header file asm/fsl_law.h already provides correct definition for second
and third PCIe controller (LAW_TRGT_IF_PCIE_2 and LAW_TRGT_IF_PCIE_3). But
is missing definition for the first PCIe controller (LAW_TRGT_IF_PCIE_1).
Note that existing definition for LAW_TRGT_IF_PCIE_2 and LAW_TRGT_IF_PCIE_3
are slightly complicated, but are really correct for P2020 platform.
When CONFIG_WDT is enabled then non-DM watchdog code cannot be used due to
conflicting functions like watchdog_reset(). So disable compilation of
mpc85xx watchdog_reset() function when CONFIG_WDT is enabled.
Sean Anderson [Fri, 22 Apr 2022 18:34:19 +0000 (14:34 -0400)]
ARM: dts: ls1021a: update the clockgen node
QorIQ platforms now use different clock bindings. Although we don't use
the device tree for clocks on this platform, it is helpful to sync it
because then the bindings will more closely match Linux. Additionally,
it allows for using more clock fractions (such as platform/4).
This corresponds to Linux commit b6f5e7019391 ("ARM: dts: ls1021a:
update the clockgen node").
Signed-off-by: Sean Anderson <sean.anderson@seco.com>
Sean Anderson [Fri, 22 Apr 2022 18:34:18 +0000 (14:34 -0400)]
arm: layerscape: Add sfp driver
This adds a driver for the Security Fuse Processor (SFP) present on
LS1012A, LS1021A, LS1043A, and LS1046A processors. It holds the
Super-Root Key (SRK), One-Time-Programmable Master Key (OTPMK), and
other "security" related fuses. Similar devices (sharing the same name)
are present on other processors, but for the moment this just supports
the LS2 variants.
The mirror registers are loaded during power-on reset. All mirror
registers must be programmed or read at once. Because of this, `fuse
prog` will program all fuses, even though only one might be specified.
To prevent accidentally burning through all your fuse programming cycles
with something like `fuse prog 0 0 A B C D`, we limit ourselves to one
programming cycle per reset. Fuses are numbered based on their address.
The fuse at 0x1e80200 is 0, the fuse at 0x1e80204 is 1, etc.
The TA_PROG_SFP supply must be enabled when programming fuses, but must
be disabled when reading them. Typically this supply is enabled by
inserting a jumper or by setting a register in the board's FPGA. I've
also added support for using a regulator. This could be helpful for
automatically issuing the FPGA write, or for toggling a GPIO controlling
the supply.
I suggest using the following procedure for programming:
1. Override the fuses you wish to program
=> fuse override 0 2 A B C D
2. Inspect the values and ensure that they are what you expect
=> fuse sense 0 2 4
3. Enable TA_PROG_SFP
4. Issue a program command using OSPR0 as a dummy. Since it contains the
write-protect bit you will usually want to write it last anyway.
=> fuse prog 0 0 0
5. Disable TA_PROG_SFP
6. Read back the fuses and ensure they are correct
=> fuse read 0 2 4
Signed-off-by: Sean Anderson <sean.anderson@seco.com>
If CONFIG_VIDEO_DM=n we query the display size from the serial console.
Especially when using a remote console the response can be so late that
it interferes with autoboot.
Only query the console size when running an EFI binary.
Add debug output showing the determined console size.
Reported-by: Fabio Estevam <festevam@gmail.com> Fixes: a57ad20d07e8 ("efi_loader: split efi_init_obj_list() into two stages") Fixes: a9bf024b2933 ("efi_loader: disk: a helper function to create efi_disk objects from udevice") Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> Tested-by: Fabio Estevam <festevam@denx.de> Tested-by: Heiko Thiery <heiko.thiery@gmail.com>
When the UEFI sub-system is initialized it sends an escape sequence to the
serial console to determine the terminal size. This stops the
run_command_list() function of the console emulation from recognizing the
U-Boot command line prompt.
Add a 'print -e' command as first command in the command list to work
around this issue.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Patrick Delaunay [Wed, 15 Jun 2022 17:41:48 +0000 (19:41 +0200)]
configs: stm32mp: cleanup the stm32mp15 file
Remove STM32_SYSRAM_END and clean the comments in stm32mp15_common.h file
after moving some CONFIG to Kconfig: CONFIG_SYS_CBSIZE,
CONFIG_SPL_MAX_FOOTPRINT, CONFIG_SYS_SPL_MALLOC_START and
CONFIG_SYS_SPL_MALLOC_SIZE.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
stm32mp: stpmic1: remove the debug unit request by debugger
Depending on backup register value, U-Boot SPL maintains the debug unit
powered-on for debugging purpose; only BUCK1 is required for powering
the debug unit, so revert the setting for all the other power lanes,
except BUCK3 that has to be always on.
To be functional this patch requires a modification in the debugger
,openocd for example, to update the STM32MP15 backup register when it is
required to debug SPL after reset. After deeper analysis this behavior
will be never supported in tools so the associated code, will be never
used and the associated code can be removed.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Patrick Delaunay [Thu, 19 May 2022 15:56:46 +0000 (17:56 +0200)]
misc: stm32mp13: introduce STM32MP13 RCC driver
Add the MISC RCC driver for STM32MP13, and bind it to the RCC reset
driver, required for initial support.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Acked-by: Sean Anderson <seanga2@gmail.com>
Change-Id: Ida11c15462caf140f87b1e3239efa2b8a689acb9
Patrick Delaunay [Thu, 19 May 2022 15:56:45 +0000 (17:56 +0200)]
clk: Add directory for STM32 clock drivers
Add a directory in drivers/clk to regroup the clock drivers for all
STM32 Soc with CONFIG_ARCH_STM32 (MCUs with cortex M) or
CONFIG_ARCH_STM32MP (MPUs with cortex A).
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com> Reviewed-by: Grzegorz Szymaszek <gszymaszek@short.pl> Acked-by: Sean Anderson <seanga2@gmail.com>
Change-Id: I955af307963f732167396f0157a30cf2fc91f150
Patrick Delaunay [Fri, 20 May 2022 16:24:51 +0000 (18:24 +0200)]
mmc: stm32_sdmmc2: make reset property optional
Although not recommended, the reset property could be made optional.
This way the driver will probe even if no reset property is provided
in an sdmmc node in DT. This reset is already optional in Linux.
Patrick Delaunay [Fri, 20 May 2022 16:24:50 +0000 (18:24 +0200)]
ram: stm32mp1: add support of STM32MP13x
Add support for new compatible "st,stm32mp13-ddr" to manage the
DDR sub system (Controller and PHY) in STM32MP13x SOC:
- only one AXI port
- support of 16 port output (MEMC_DRAM_DATA_WIDTH = 2)
The STM32MP15x SOC have 2 AXI ports and 32 bits support.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Dinesh Maniyam [Wed, 1 Jun 2022 07:57:25 +0000 (15:57 +0800)]
drivers: cache: ncore: Disable snoop filter
There is hardware bug in NCORE CCU IP and it is causing an issue in the
coherent directory tracking of outstanding cache lines.
The workaround is disabling snoop filter.
Signed-off-by: Dinesh Maniyam <dinesh.maniyam@intel.com> Reviewed-by: Tien Fong Chee <tien.fong.chee@intel.com>
Dinesh Maniyam [Wed, 1 Jun 2022 07:54:59 +0000 (15:54 +0800)]
arch: arm: socfpga: timer_s10: Override udelay for secure section
Override __udelay() as 'always inlined' function so that PSCI code
run in '__secure' section can call this delay function as well.
Signed-off-by: Chee Hong Ang <chee.hong.ang@intel.com> Signed-off-by: Dinesh Maniyam <dinesh.maniyam@intel.com> Reviewed-by: Tien Fong Chee <tien.fong.chee@intel.com>
Marek Vasut [Mon, 13 Jun 2022 09:55:21 +0000 (11:55 +0200)]
ARM: dts: stm32: Add DHCOR based DRC Compact board
Add DT for DH DRC Compact unit, which is a universal controller device.
The system has two ethernet ports, one CAN, RS485 and RS232, USB, uSD
card slot, eMMC and SDIO Wi-Fi.
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>
Patrick Delaunay [Thu, 16 Jun 2022 16:37:59 +0000 (18:37 +0200)]
stm32mp: stm32prog: fix the last character of dfu_alt_add third parameter
The third parameter of dfu_alt_add(), the string description of alternate,
is build in stm32prog_alt_add() with a unnecessary character ';' at the
end of the string.
This separator was required in the first implementation of
dfu_alt_add() but is no more needed in the current implementation;
this separator is managed only in dfu_config_interfaces() which call
dfu_alt_add() for this parameter without this separator.
And since the commit 53b406369e9d ("DFU: Check the number of arguments
and argument string strictly"), this added character cause an error when
the stm32prog command is executed because the third parameter of
dfu_alt_add() must be a string with a numerical value; 's' must be NULL
in the result of call in dfu_fill_entity_mmc():
third_arg = simple_strtoul(argv[2], &s, 0);
Fixes: 53b406369e9d ("DFU: Check the number of arguments and argument string strictly") Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Marek Vasut [Sun, 1 May 2022 16:43:55 +0000 (18:43 +0200)]
net: Fix discuss discard typo
Replace discuss with discard, that is what happens with packet with
incorrect checksum. Fix the typo.
Fixes: 4b37fd146bb ("Convert CONFIG_UDP_CHECKSUM to Kconfig") Signed-off-by: Marek Vasut <marex@denx.de> Cc: Ramon Fried <rfried.dev@gmail.com> Cc: Simon Glass <sjg@chromium.org>
Corentin LABBE [Tue, 14 Jun 2022 08:44:07 +0000 (08:44 +0000)]
board: ti: am335x: eth_cpsw should depend on CONFIG_NET
The origin of this patch is the breaking of am335x-hs boot
due to commit e41651fffda7 ("dm: Support parent devices with of-platdata")
HS boards have less SRAM for SPL and so this commit increased memory usage beyond am335x limit.
This commit added 10 driver binding pass and am335x boot only if one pass is done.
SPL try to do more than one pass due to eth_cpsw failing.
Since HS SPL does not need network (and NET is already disabled in config),
the easiest fix is to "remove" eth_cpsw from SPL by testing if NET is enabled.
Signed-off-by: Corentin LABBE <clabbe@baylibre.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Andrew Davis <afd@ti.com>
Andre Przywara [Mon, 13 Jun 2022 23:11:10 +0000 (00:11 +0100)]
armv8: always use current exception level for TCR_ELx access
Currently get_tcr() takes an "el" parameter, to select the proper
version of the TCR_ELx system register.
This is problematic in case of the Apple M1, since it runs with
HCR_EL2.E2H fixed to 1, so TCR_EL2 is actually using the TCR_EL1 layout,
and we get the wrong version.
For U-Boot's purposes the only sensible choice here is the current
exception level, and indeed most callers treat it like that, so let's
remove that parameter and read the current EL inside the function.
This allows us to check for the E2H bit, and pretend it's EL1 in this
case.
There are two callers which don't care about the EL, and they pass 0,
which looks wrong, but is irrelevant in these two cases, since we don't
use the return value there. So the change cannot affect those two.
Signed-off-by: Andre Przywara <andre.przywara@arm.com> Reviewed-by: Mark Kettenis <kettenis@openbsd.org> Tested-by: Mark Kettenis <kettenis@openbsd.org>
With the move to use DM_CLK the boards uart stops working. The used
properties are not supported by the imx8mq clock driver. Thus
the correct baudrate cannot be selected. Remove this properties here and
the board can start with working uart. Keep it in the main dts because
linux handles these porperties fine.
Andre Przywara [Tue, 7 Jun 2022 23:42:22 +0000 (00:42 +0100)]
usb: host: ehci-generic: Make resets and clocks optional
The generic EHCI binding does not *require* resets and clocks
properties, and indeed for instance the Allwinner A20 SoCs does not
need or define any resets in its DT.
Don't easily give up if clk_get_bulk() or reset_get_bulk() return an
error, but check if that is due to the DT simply having no entries for
either of them.
This fixes USB operation on all boards with an Allwinner A10 or A20 SoC,
which were reporting an error after commit ba96176ab70e2999:
=======================
Bus usb@1c14000: ehci_generic usb@1c14000: Failed to get resets (err=-2)
probe failed, error -2
=======================
Signed-off-by: Andre Przywara <andre.przywara@arm.com> Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Following Jincheng's report, an out-of-band write leading to arbitrary
code execution is possible because on one side the squashfs logic
accepts directory names up to 65535 bytes (u16), while U-Boot fs logic
accepts directory names up to 255 bytes long.
Prevent such an exploit from happening by capping directory name sizes
to 255. Use a define for this purpose so that developers can link the
limitation to its source and eventually kill it some day by dynamically
allocating this array (if ever desired).
Tom Rini [Wed, 8 Jun 2022 18:30:14 +0000 (14:30 -0400)]
odroid_xu3: Fix board environment variable
When migrating CONFIG_CONS_INDEX to Kconfig, on this platform we changed
what "board" evaluated to in the environment. This in turn meant that
we would no longer try and find the correct fdtfile via the normal
distro boot logic. Fix this by overriding board in the default
environment, as done on other platforms where CONFIG_SYS_BOARD is not
what we want to be in the board environment variable.
Fixes: f76750d11133 ("Convert CONFIG_CONS_INDEX et al to Kconfig") Reported-by: Gabriel Hojda <ghojda@yo2urs.ro> Tested-by: Gabriel Hojda <ghojda@yo2urs.ro> Signed-off-by: Tom Rini <trini@konsulko.com>
b4 utility [1] is introduced by Linux Kernel developers and used to
fetch patches and patch series from lore.kernel.org and is proven
to be useful for U-Boot development. Detailed usage of the tool can be
read under post from the original author [2].
This tool fetches files from the list and populates the source folder
with additional files (*.cover and *.mbx) which are not ignored by git
and shown as newly added files.
Add those file patterns into .gitignore file, so they can be safely
skipped during changes attestation.
Stefan Roese [Fri, 29 Apr 2022 13:34:44 +0000 (15:34 +0200)]
crypto: fsl_hash: Remove unnecessary alignment check in caam_hash()
While working on an LX2160 based board and updating to latest mainline
I noticed problems using the HW accelerated hash functions on this
platform, when trying to boot a FIT Kernel image. Here the resulting
error message:
Using 'conf-freescale_lx2160a.dtb' configuration
Trying 'kernel-1' kernel subimage
Verifying Hash Integrity ... sha256Error: Address arguments are not aligned
CAAM was not setup properly or it is faulty
error!
Bad hash value for 'hash-1' hash node in 'kernel-1' image node
Bad Data Hash
ERROR: can't get kernel image!
Testing and checking with Gaurav Jain from NXP has revealed, that this
alignment check is not necessary here at all. So let's remove this
check completely.
Marek Vasut [Mon, 20 Dec 2021 21:57:57 +0000 (22:57 +0100)]
arm: socfpga: vining: Unmount UBIFS and detach UBI in ubi_load script
Clean up in ubiload script. Unmount UBIFS from which kernel image was
loaded and detach UBI on which the UBIFS is located, otherwise message
similar to the following is printed just before booting kernel:
Removing MTD device #7 (rootfs) with use count 1
Error when deleting partition "rootfs" (-16)
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Siew Chin Lim <elly.siew.chin.lim@intel.com> Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com> Cc: Tien Fong Chee <tien.fong.chee@intel.com> Reviewed-by: Tien Fong Chee <tien.fong.chee@intel.com>
ddr: altera: Stratix10: Use phys_size_t for memory size
Replace with phys_size_t for all memory size variables declaration
for the sake of scalability. phys_size_t is defined in
/arch/arm/include/asm/types.h.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
ddr: altera: Ignore bit[7-4] for both seq2core & core2seq handshake in HPS
Bit[7-4] for both register seq2core and core2seq handshake in HPS are not
required for triggering DDR re-calibration or resetting EMIF. So, ignoring
these bits just for playing it safe.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
projects / u-boot.git / log
![2 years agoddr: altera: Ignore bit[7-4] for both seq2core & core2seq handshake in HPS](/img/static/git-logo.png?p=u-boot.git;a=commit;h=ee06c5390f2f1e2f1bc23e14a7cd8665c1e42ff4){kind=link}
