This adds a call to dm_remove_devices_flags() to do_bootm_linux()
so that drivers that have one of the removal flags set (e.g.
DM_FLAG_ACTIVE_DMA_REMOVE) in their driver struct, may do some
last-stage cleanup before the OS is started.
arm and x86 already did such, and we should do the same for riscv.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
MIPS is the only architecture currently supported by U-Boot that does
not implement any of the in/out register access functions.
To have a interface that is useable across architectures, add the
functions to the MIPS architecture (implemented using the __raw_write
and __raw_read functions).
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Use a starting address of 256MB which should be available. This helps to
make sandbox RAM buffers pointers more recognisable.
Signed-off-by: Simon Glass <sjg@chromium.org>
The upcoming changes to the regmap interface will contain a proper check
for plausibility when reading/writing from/to a register map. To still
have the current tests pass, increase the size of the memory region for
the syscon0 device, since one of the tests reads and writes beyond this
range.
Reviewed-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Now that we fixed the pre-relocation driver binding for driver marked
with DM_FLAG_PRE_RELOC flag, add a test case to cover that scenario.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The make macro to check if the binary exceeds the board size limit is
taken straight from the root Makefile.
Without this and e.g. enabled EFI Vybrid fails booting as the regular
size limit check does not take the final u-boot.imx binary size into
account which is bigger due to alignment as well as IMX header stuff.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
If the function gpr_init is used in a common MX6 spl
implementation we have to ensure that it is only called for
suitable cpu types, otherwise it breaks hardware parts like
enet1, can1, can2, etc.
Signed-off-by: Christoph Niedermaier <cniedermaier@dh-electronics.de>
The new board version has the 2nd FPGA connected via CS# 0 instead of
2 on SPI bus 1. Change this setup in the DT accordingly. Please note
that this change does still work on the old board version because the
CS signal is not used on this board.
Signed-off-by: Stefan Roese <sr@denx.de>
Add the "spi-flash" compatible string so that the generic sf_probe
driver can probe the SPI flash on the theadorable Armada-XP board.
Signed-off-by: Stefan Roese <sr@denx.de>
With patch 49b23e035d (pci: mvebu: Increase size of PCIe default mapping)
the mapping size for each PCI(e) controller was increased from 32MiB to
128MiB. This leads to problems on boards with multiple PCIe slots / ports
which are unable to map all PCIe ports, e.g. the Armada-XP theadorable:
DRAM: 2 GiB (667 MHz, 64-bit, ECC not enabled)
SF: Detected m25p128 with page size 256 Bytes, erase size 256 KiB, total 16 MiB
Cannot add window '4:f8', conflicts with another window
PCIe unable to add mbus window for mem at f0000000+08000000
Model: Marvell Armada XP theadorable
This patch moves the base address for the PCI(e) memory spaces from
0xe8000000 to the end of SDRAM (clipped to a max of 0xc0000000 right now).
This gives move room and flexibility for PCI(e) mappings.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: VlaoMao <vlaomao@gmail.com>
Tested-by: VlaoMao <vlaomao at gmail.com>
Commit 3c28576bb0 ("arm: dts: imx8qxp: fix build warining")
fixed the dts warning by removing the unnecessary
#address-cells/#size-cells, but the recommendation for regulators is not
to place them under "simple-bus", so move the reg_usdhc2_vmmc regulator
accordingly.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
New ACPI assembler issues a warning:
board/intel/edison/dsdt.asl.tmp 13: Offset (0x00),
Remark 2158 - ^ Unnecessary/redundant use of Offset operator
Indeed, in the OperationRegion the offset is 0x00 by default.
Thus, drop unneeded Offset() use as suggested by ACPI assembler.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Update TI Keystone 2 driver to re-use common mdio lib.
Reviewed-by: Tom Rini <trini@konsulko.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Networking support for all TI K2 boards converted to use DM model and
CONFIG_DM_ETH enabled in all corresponding defconfig files, hence drop
unused non DM K2 networking code.
Reviewed-by: Tom Rini <trini@konsulko.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Add drivers/net/ti/ folder and move all TI's code in this folder for better
maintenance.
Reviewed-by: Tom Rini <trini@konsulko.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This is a large update of the AST2500 SoC DTS file bringing it to the
level of commit 927c2fc2db19 :
Author: Joel Stanley <joel@jms.id.au>
Date: Sat Jun 2 01:18:53 2018 -0700
ARM: dts: aspeed: Fix hwrng register address
There are some differences on the compatibility property names. scu,
reset and clock drivers are also different.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Currently the size of the relocation table will be shrunk
to the actual size needed. Although this gives a maximal
space saving, it messes up the _end symbol. This breaks
features like appended DTBs because the _end symbol doesn't
point to the real end of the U-Boot binary.
Remove the size shrinking and make the size of the relocation
table fixed but configurable. This follows the Linux approach
and the user can adjust the size to his needs.
Also rename the relocation table section from .rel to .data.reloc
to follow the Linux approach and to avoid ambiguities with the
.rel.* sections added by the linker.
Reported-by: Lars Povlsen <lars.povlsen@microsemi.com>
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Patch "ARM: rmobile: Mark 4-64GiB as DRAM on Gen3" marked the entire
64bit DRAM space as cachable. On CortexA57, this might result in odd
side effects, where the CPU tries to prefetch from those areas and if
there is no DRAM backing them, CPU bus hang can happen.
This patch fixes it by generating the mem_map structure based on the
actual memory layout obtained from the DT, thus not marking areas
without any DRAM behind them as cachable.
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Fixes: c1ec347638 ("ARM: rmobile: Mark 4-64GiB as DRAM on Gen3")
Cc: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Real marketing name of the board was recently updated so
to accommodate that change renaming the board and all
related to it.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Using imply for SYS_MMCSD_RAW_MODE_U_BOOT_USE_PARTITION instead of
select ensures we can build without partition support (used to build
a network boot only version of SPL and U-Boot).
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Per T1040RM (Rev. 1, 08/2015), there are 2 issues with the RCW EC2
settings.
- The value of FSL_CORENET_RCWSR13_EC2_FM1_GPIO is wrong and should
be 0x04000000 (value of 1 in RCW bit [420:421])
- Value of 2/3 are reserved in RCW bit [420:421], hence there is no
macro FSL_CORENET_RCWSR13_EC2_FM1_DTSEC5_MII.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Poonam Aggrwal <poonam.aggrwal@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
Per T1040RM (Rev. 1, 08/2015), the value of
FSL_CORENET_RCWSR13_MAC2_GMII_SEL_ENET_PORT is wrong
and should be 0x00000080 (bit 440 in the RCW).
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Poonam Aggrwal <poonam.aggrwal@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>
Workaround makes FS as default mode on all affected socs.
Add support to check erratum-A005275 validity for an soc. This info is
required to determine whether a given soc is affected by this erratum.
Add quirk for this erratum "has_fsl_erratum_a005275" . This quirk is used
to enable workaround for the errata
Force FS mode as default by:
- making EPS as FS
- setting PFSC bit to disable HS chirping
This workaround can be disabled by mentioning "no_erratum_a005275" in
hwconfig string
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Reviewed-by: York Sun <york.sun@nxp.com>
At the moment we rely on the infamous get_ram_size() function to learn
the actual DRAM size in U-Boot proper. This function has two issues:
1) It only works if the DRAM size is a power of two. We start to see
boards which have 3GB of (usable) DRAM, so this does not fit anymore.
2) As U-Boot has no notion of reserved memory so far, it will happily
ride through the DRAM, possibly stepping on secure-only memory. This
could be a region of DRAM reserved for OP-TEE or some other secure
payload, for instance. It will most likely crash in that case.
As the SPL DRAM init routine has very accurate knowledge of the actual
DRAM size, lets propagate this wisdom to U-Boot proper.
We re-purpose a currently reserved word in our SPL header for that.
The SPL itself stores the detected DRAM size there, and bumps the SPL
header version number in that case. U-Boot proper checks for a valid
SPL header and a high enough version number, then uses the DRAM size
from there. If the SPL header field is not sufficient, we fall back to
the old DRAM scanning routine.
Part of the DRAM might be present and probed by SPL, but not accessible
by the CPU. They're restricted in the main U-Boot binary, when accessing
the DRAM size from SPL header.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Reviewed-by: Jagan Teki <jagan@openedev.com>
Allwinner 64-bit SoCs can use 4GiB DRAM chip, however their memory map
has only allocated 3GiB for DRAM, so only 3GiB of the DRAM is
accessible.
Add a Kconfig option for the maximum accessible DRAM.
For A80 it should be a much higher value (8GiB), but as I have no A80
device to test and originally U-Boot only supports 2GiB DRAM on A80, it
currently still falls under the 2GiB situation.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Reviewed-by: Jagan Teki <jagan@openedev.com>
All Allwinner 64-bit SoCs now are known to be able to access 3GiB of
external DRAM, however the size of DRAM part in the MMU translation
table is still 2GiB.
Change the size of DRAM part in MMU table to 3GiB.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Reviewed-by: Jagan Teki <jagan@openedev.com>
On Allwinner SoCs we use some free bytes at the beginning of the SPL image
to store various information. We have a version byte to allow updates,
but changing this always requires all tools to be updated as well.
Introduce the concept of semantic versioning [1] to the SPL header:
The major part of the version number only changes on incompatible
updates, a minor number bump indicates backward compatibility.
This patch just documents the major/minor split, adds some comments
to the header file and uses the versioning information for the existing
users.
[1] https://semver.org
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Acked-by: Jagan Teki <jagan@openedev.com>
The Pine A64 Plus/non-Plus model detection code is now built on all
64-bit ARM SoCs, even if the code cannot be triggered when H5/H6 is in
use.
Disable them when the board is Pine A64 by adding a Kconfig option that
is only selected on Pine A64.
On GCC 7.3.1 this makes the size of the function reduces 184 bytes, and
saves a 104 byte strstr() function, then makes SPL on H6 succeed to
build.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Reviewed-by: Jagan Teki <jagan@openedev.com>
Since commit 80df194f01 ("x86: detect unsupported relocation types"),
an error message is seen on QEMU x86 target during boot:
do_elf_reloc_fixups32: unsupported relocation type 0x1 at fff841f0, offset = 0xfff00087
do_elf_reloc_fixups32: unsupported relocation type 0x2 at fff841f8, offset = 0xfff00091
Check offset 0xfff00087 and 0xfff00091 in the u-boot ELF image,
fff00087 000df401 R_386_32 00000000 car_uninit
fff00091 000df402 R_386_PC32 00000000 car_uninit
we see R_386_32 and R_386_PC32 relocation type is generated for
symbol car_uninit, which is declared as a weak symbol in start.S.
However the actual weak symbol implementation ends up nowhere. As
we can see below, it's *UND*.
$ objdump -t u-boot | grep car_uninit
00000000 w *UND* 00000000 car_uninit
With this fix, it is normal now.
$ objdump -t u-boot | grep car_uninit
fff00094 w F .text.start 00000001 car_uninit
Reported-by: Hannes Schmelzer <hannes@schmelzer.or.at>
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Tested-by: Hannes Schmelzer <oe5hpm@oevsv.at>
To allow bigger 64 bit prefetchable PCI regions in Linux, this patch
changes the base address and range of the ACPI area passed to Linux.
BayTrail can only physically access 36 bit of PCI address space. So
just chaning the range without changing the base address won't work
here, as 0xf.ffff.ffff is already the maximum address.
With this patch, a maximum of 16 GiB of local DDR is supported. This
should be enough for all BayTrail boards though.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Introduce a new script to check whether file exists and
use that check in Makefile to avoid break CI system.
The script return 1 when the required files not exists, return 0
when files exists. The script will ignore check to u-boot-dtb.bin,
because if there is something wrong to generate u-boot-dtb.bin,
there must be some code error.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
This is required for the current Linux kernel to reboot. It should also
probably be fixed in Linux.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
We can have the case where u-boot is launched after some other low level
enabler, like for example when u-boot runs after arm-trusted-firmware
and/or optee. So, because of that we may need to jump the initialization of
some IP blocks even because we may no longer have the permission for that.
So, if the config option to skip low level init is set disable also timer,
board and csu initialization.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Fabio Estevam <fabio.estevam@nxp.com>
Cc: Albert Aribaud <albert.u.boot@aribaud.net>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: u-boot@lists.denx.de
Reviewed-by: Peng Fan <peng.fan@nxp.com>
When building i.MX8/8X board, use imx8image type.
`-e $(CONFIG_SYS_TEXT_BASE)` is not needed, but
no harm to keep it for i.MX8/8X
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Introduce dtsi for i.MX8QXP, since there is other variants i.MX8DX(P),
so add them there, because i.MX8QXP includes the dtsi of them.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Anatolij Gustschin <agust@denx.de>
Add the power domain DM driver for i.MX8, that it depends on the DTB
power domain trees to generate the power domain provider devices. Users
need to add power domain trees with property "compatible = "nxp,imx8-pd";"
When power on a PD device, the driver will power on its ancestor PD
devices in power domain tree.
When power off a PD device, the driver will check its child PD devices
first. Only if all child PD devices are off, then power off the current PD
device. Then the driver checks sibling PD devices. If sibling PD devices
are off, then it will power off parent PD device.
There is no counter maintained in this driver, but a state to hold current
on/off state. So the request and free functions are empty.
The power domain implementation in i.MX8 DTB set the "#power-domain-cells"
to 0, so there is no ID binding with each PD device. We don't use "id"
variable in struct power_domain. At the same time, we have to set of_xlate
to empty to bypass standard of_xlate in uclass driver.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Anatolij Gustschin <agust@denx.de>
Cc: Stefano Babic <sbabic@denx.de>
This driver is mostly used to avoid build errors.
We use uclass clk driver for clk related operations.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Anatolij Gustschin <agust@denx.de>
Cc: Stefano Babic <sbabic@denx.de>
