This changes the rockchip-efuse driver to use dev_read_addr instead of
devfdt_get_addr.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
We are about to reuse the rockchip timer (header file) for 64bit ARMv8
chips, so it seems a good time to make the register sizes explicit by
changing from 'unsigned int' to 'u32'.
Reorders the header-includes in rk_timer.c to ensure that 'u32' is
definded before it is used by 'asm/arch/timer.h'.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
In the source_file_format.txt file we talk about how to construct a
valid FIT image. While it already says to look at the source for the
full list, add kernel_noload to the explicit list of types. This is
arguably the most important type to use as most often we are including a
kernel that will run from wherever it is loaded into memory and execute.
This for example, allows you to create a single FIT image for Linux that
can be used on both OMAP and i.MX devices as the kernel will not need to
be moved in memory.
Signed-off-by: Tom Rini <trini@konsulko.com>
STM32F7 and H7 shared the same SDRAM control block.
On STM32H7 few control bits has been added.
The current driver need some minor adaptation as FMC block
enable/disable for H7.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
FMC driver is now able to discover the bank number by
parsing bank subnodes.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
FMC is able to manage 2 SDRAM banks, but the current driver
implementation is only able to manage the first SDRAM bank.
Even if only bank2 is used, some bank1 registers must be
configured.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Replace all fdtdec_get..() calls by ofnode_read...() or dev_read..().
This will allow drivers to support a live device tree.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Retrieve RAM base address from DT instead of using STM32_SDRAM_FMC
For STM32F7, FMC block base address is 0xA0000000, but SDRAM
registers are located at offset 0x140 inside FMC block.
Update the stm32_fmc_regs fields with all FMC registers
to map SDRAM registers at the right address.
These additionals registers will be used later.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Migrate all FMC defines from arch/arm/include/asm/arch-stm32f7/fmc.h
to drivers/ram/stm32_sdram.c
This will avoid to add an additionnal arch-stm32xx/fmc.h file when
a new stm32 family soc will be introduced.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Vikas Manocha <vikas.manocha@st.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
All drivers which was using clock_get() are now using
clk_get_rate() from clock framework, now it's safe to
remove clock_get().
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
Replace proprietary clock_get() by clk_get_rate()
The stm32_qspi is now "generic" and can be used
by other STM32 SoCs.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
Replace proprietary clock_get() by clk_get_rate()
The stm32x7 serial driver is now "generic" and can be used
by other STM32 SoCs.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
This allow to remove include/dm/platform_data/serial_stm32x7.h
which was included in the past by stm32x7 driver and by
stm32f746-disco.c board file.
Since patch 42bf5e7c27 "serial: stm32f7: add device tree support"
this file is no more needed in board file.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
clean the code by removing unused enums, structs and
defines related to clocks
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
Add clock framework .get_rate callback.
This step will allow to convert all drivers which was using
proprietary clock_get() to use clock framework .get_rate().
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
Retrieve RCC base address from DT, this will prepare
the ground for future STM32 SoCs support.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
Also remove its declaration from stm32.h which
is no more needed.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
Use the same clocks macro than the one used by kernel DT.
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Acked-by: Vikas MANOCHA <vikas.manocha@st.com>
The inline assembly functions in mon.c assume that the caller will
check for the return value in r0 according to regular ARM calling
conventions.
However, this assumption breaks down if the compiler inlines the
functions. The caller is then under no obligation to use r0 for the
result.
To fix this disconnect, we must explicitly move the return value
from the smc/bl call to the variable that the function returns.
Signed-off-by: Madan Srinivas <madans@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
The load address of ramdisk, rdaddr is 0x88080000 and fit_loadaddr
is defined as 0x88000000. This leaves only 512Kbytes for the
fit image. When the FIT images are larger than this, it will
overwite the ramdisk and cause the boot to fail.
For eg, The K2 HS fit images are a few MB and end up overwriting
the ramdsk. This patch moves the fit_loadaddr to 0x87000000,
leaving a 16MB window for the fit image. This memory can be
reclaimed once the kernel starts running.
Signed-off-by: Madan Srinivas <madans@ti.com>
On early K2 devices (eg. K2HK) the secure ROM code does not support
loading secure code to firewall protected memory, before decrypting,
authenticating and executing it.
To load the boot monitor on these devices, it is necessary to first
authenticate and run a copy loop from non-secure memory that copies
the boot monitor behind firewall protected memory, before decrypting
and executing it.
On K2G, the secure ROM does not allow secure code executing from
unprotected memory. Further, ROM first copies the signed and encrypted
image into firewall protected memory, then decrypts, authenticates
and executes it.
As a result of this, we cannot use the copy loop for K2G. The
mon_install has to be modified to pass the address the signed and
encrypted secure boot monitor image to the authentication API.
For backward compatibility with other K2 devices and K2G GP,
the mon_install API still supports a single argument. In this case
the second argument is set to 0 by u-boot and is ignored by ROM
Signed-off-by: Thanh Tran <thanh-tran@ti.com>
Signed-off-by: Madan Srinivas <madans@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
We first split the CONFIG_BOOTCOMMAND into its components to improve
readability. We then make the following order changes:
- Run findfdt first so the fdt name can be used in envboot like OMAP
- Install the boot monitor before running the PMMC so we can make any
needed secure changes before PMMC, do this on both HS and non-HS
- Move set_name_pmmc to just before get_pmmc_${boot}
Signed-off-by: Andrew F. Davis <afd@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Updates the default u-boot environment variables to support secure
boot. On secure devices, a secure boot monitor (sec-bm) needs to
be installed by u-boot.
Signed-off-by: Madan Srinivas <madans@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Updates the default u-boot environment variables to support FIT image
loading.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
BeagleBoard X15 revC board is similar to X15 revB1 except
with a SR2.0 where revB1 uses a SR1.1. Add board detection
support for revC.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
board_is*("rev", board_ti_get_rev()) uses strncmp() for
revison detection and assumes it is success if return value
is <= 0. This will fail in case of multiple versions, as
revb will be true for board_is_*revb() and board_is_*reva().
Fix it by looking for exact match of the string.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
GPIOs are now supported on Meson GXBB, enable driver and command in
the config.
Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This commit adds GPIO support to the Amlogic Meson pin controller
driver, based on code from Linux kernel.
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Import Amlogic Meson DTS files from Linux kernel version 4.12
Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
As is often the case with SoC development, slightly different
products (i.e. different part number) are developed based on the
same silicon-die. Such fine grained information is unmaintainable.
Also, "SoC:" is a better fit that "CPU:".
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
It has been a while since ARM Trusted Firmware supported UniPhier SoC
family. U-Boot SPL was intended as a temporary loader that runs in
secure world. It is a maintenance headache to support two different
boot mechanisms. Secure firmware is realm of ARM Trusted Firmware
and now U-Boot only serves as a non-secure boot loader for UniPhier
ARMv8 SoCs.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
The CONFIG_ENV_IS_IN_... options which have already been converted to
Kconfig only have a small amount of help. Move the rest of it over from
the README.
Signed-off-by: Simon Glass <sjg@chromium.org>
This converts the following to Kconfig:
CONFIG_ENV_IS_IN_MMC
CONFIG_ENV_IS_IN_NAND
CONFIG_ENV_IS_IN_UBI
CONFIG_ENV_IS_NOWHERE
In fact this already exists for sunxi as a 'choice' config. However not
all the choices are available in Kconfig yet so we cannot use that. It
would lead to more than one option being set.
In addition, one purpose of this series is to allow the environment to be
stored in more than one place. So the existing choice is converted to a
normal config allowing each option to be set independently.
There are not many opportunities for Kconfig updates to reduce the size of
this patch. This was tested with
./tools/moveconfig.py -i CONFIG_ENV_IS_IN_MMC
And then manual updates. This is because for CHAIN_OF_TRUST boards they
can only have ENV_IS_NOWHERE set, so we enforce that via Kconfig logic
now.
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
It seems that gcc 6.3 at least is smart enough to warn about the _val
variable being unassigned in the default case in the set_hdr_field()
macro, but not smart enough to figure out that the default case is never
taken. This results in warnings such as the following:
pfx##hdr32[idx].field = _val; \
^
../tools/mips-relocs.c:51:11: note: _val was declared here
uint64_t _val; \
^
../tools/mips-relocs.c:88:2: note: in expansion of macro set_hdr_field
set_hdr_field(p, idx, field, val)
^~~~~~~~~~~~~
../tools/mips-relocs.c:408:3: note: in expansion of macro set_phdr_field
set_phdr_field(i, p_filesz, load_sz);
^~~~~~~~~~~~~~
../tools/mips-relocs.c: In function main:
../tools/mips-relocs.c:77:25: warning: _val may be used uninitialized
in this function [-Wmaybe-uninitialized]
Avoid this by assigning _val = 0 in the default case, and asserting that
we didn't actually hit it for good measure.
For reference gcc 7.1.1 seems to be smart enough to not hit the above
warning without this patch.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Fixes: 011dd93ca97a ("MIPS: Stop building position independent code")
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: Tom Rini <trini@konsulko.com>
Cc: u-boot@lists.denx.de
This patch fixes 2 bugs introduced by the following commit
2bb5b63 MIPS: bootm: rework and fix broken bootm code
The CONFIG_IS_ENABLED macro prepends 'CONFIG_' Hence, remove CONFIG_
from CONFIG_MIPS_BOOT_ENV_LEGACY usage.
Also, 2bb5b63 reworks bootm so that linux_env_legacy runs before
linux_cmdline_legacy. However, linux_env_legacy depends on
linux_cmdline_legacy running first as linux_cmdline_init initialilzes
linux_argp which linux_env_legacy later depends on during its
initialization.
Reorder the code so that linux_cmdline_legacy runs before
linux_env_legacy.
Signed-off-by: Zubair Lutfullah Kakakhel <Zubair.Kakakhel@imgtec.com>
U-Boot has up until now built with -fpic for the MIPS architecture,
producing position independent code which uses indirection through a
global offset table, making relocation fairly straightforward as it
simply involves patching up GOT entries.
Using -fpic does however have some downsides. The biggest of these is
that generated code is bloated in various ways. For example, function
calls are indirected through the GOT & the t9 register:
8f998064 lw t9,-32668(gp)
0320f809 jalr t9
Without -fpic the call is simply:
0f803f01 jal be00fc04 <puts>
This is more compact & faster (due to the lack of the load & the
dependency the jump has on its result). It is also easier to read &
debug because the disassembly shows what function is being called,
rather than just an offset from gp which would then have to be looked up
in the ELF to discover the target function.
Another disadvantage of -fpic is that each function begins with a
sequence to calculate the value of the gp register, for example:
3c1c0004 lui gp,0x4
279c3384 addiu gp,gp,13188
0399e021 addu gp,gp,t9
Without using -fpic this sequence no longer appears at the start of each
function, reducing code size considerably.
This patch switches U-Boot from building with -fpic to building with
-fno-pic, in order to gain the benefits described above. The cost of
this is an extra step during the build process to extract relocation
data from the ELF & write it into a new .rel section in a compact
format, plus the added complexity of dealing with multiple types of
relocation rather than the single type that applied to the GOT. The
benefit is smaller, cleaner, more debuggable code. The relocate_code()
function is reimplemented in C to handle the new relocation scheme,
which also makes it easier to read & debug.
Taking maltael_defconfig as an example the size of u-boot.bin built
using the Codescape MIPS 2016.05-06 toolchain (gcc 4.9.2, binutils
2.24.90) shrinks from 254KiB to 224KiB.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: u-boot@lists.denx.de
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>