The current implementation makes it look like the 'if (from_spl)' part is
dead code because these features are not enabled for sandbox. We could
enable it for sandbox_spl, but this is not done yet (it requires sharing
memory between SPL and U-Boot proper which is in fact supported).
It is probably nicer to avoid #ifdef anyway. Change it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Coverity (CID: 163244)
Fixes: 824bb1b (bootstage: Support SPL)
There was for long time no activity in the 5xx area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 5xx,
so remove it.
Signed-off-by: Heiko Schocher <hs@denx.de>
There was for long time no activity in the 8260 area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 8260,
so remove it.
Signed-off-by: Heiko Schocher <hs@denx.de>
There was for long time no activity in the 8xx area.
We need to go further and convert to Kconfig, but it
turned out, nobody is interested anymore in 8xx,
so remove it (with a heavy heart, knowing that I remove
here the root of U-Boot).
Signed-off-by: Heiko Schocher <hs@denx.de>
At present bootstage only supports U-Boot proper. But SPL can also consume
boot time so it is useful to have the record start there.
Add bootstage support to SPL. Also support stashing the timing information
when SPL finishes so that it can be picked up and reported by U-Boot
proper. This provides a full boot time record, excluding only the time
taken by the boot ROM.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present we don't allow use of bootstage before driver model is running.
This means we cannot time the init of driver model itself.
Now that bootstage requires its own board-specific timer, we can move its
init to earlier in the sequence, both before and after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some boards cannot access pre-relocation data after relocation. Reserve
space for this and copy it during preparation for relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present bootstage uses the data section of the image to store its
information. There are a few problems with this:
- It does not work on all boards (e.g. those which run from flash before
relocation)
- Allocated strings still point back to the pre-relocation data after
relocation
Now that U-Boot has a pre-relocation malloc() we can use this instead,
with a pointer to the data in global_data. Update bootstage to do this and
set up an init routine to allocate the memory.
Now that we have a real init function, we can drop the fake 'reset' record
and add a normal one instead.
Note that part of the problem with allocated strings remains. They are
reallocated but this will only work where pre-relocation memory is
accessible after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
With a small tweak we can avoid including these files for all boards.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Coldfire arch is not happy with timer_init since interrupt handlers
are still not set at that stage, and the boot hangs silently.
Signed-off-by: Angelo Dureghello <angelo@sysam.it>
This allows us to use the same DRAM init function on all archs. Add a
dummy function for arc, which does not use DRAM init here.
Signed-off-by: Simon Glass <sjg@chromium.org>
[trini: Dummy function on nios2]
Signed-off-by: Tom Rini <trini@konsulko.com>
With d53ecad92f some unused interrupt related code was removed.
However all of these options are currently unused. Rather than migrate
some of these options to Kconfig we just remove the code in question.
The only related code changes here are that in some cases we use
CONFIG_STACKSIZE in non-IRQ related context. In these cases we rename
and move the value local to the code in question.
Fixes: d53ecad92f ("Merge branch 'master' of git://git.denx.de/u-boot-sunxi")
Signed-off-by: Tom Rini <trini@konsulko.com>
These includes don't seem to be needed now. Drop them. Reserve the
mp.h header for PowerPC for now.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: York Sun <york.sun@nxp.com>
This header file is used by three archs. It could be used by all of them
since relocation is a common function. Move it into a generic file.
Signed-off-by: Simon Glass <sjg@chromium.org>
This header file is used by two archs. It could be used by all of them
since it allows the cache to be on during relocation. Move it into a
generic file.
Signed-off-by: Simon Glass <sjg@chromium.org>
We don't need this PPC-specific function in generic code. Move it to
the powerpc directory.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Drop headers which are not used or needed in this file. The compiler.h
header is included by common.h.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
By making dram_init_banksize() return an error code we can drop the
wrapper. Adjust this and clean up all implementations.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Move the ugly #ifdefs inside the reserve_video() function so we can
collect all this init into one place.
Signed-off-by: Simon Glass <sjg@chromium.org>
The arch-specific details of the cache being off are best handled inside
the reserve_mmu(). This cleans up the init sequence a little.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
CONFIG_ALT_LB_ADDR is really a detail of how this logbuffer is allocated
rather than whether to do it at all. So move the #ifdef into the function.
Signed-off-by: Simon Glass <sjg@chromium.org>
All archs put U-Boot at the bottom of the relocated region. Xtensa does
not, but perhaps not for any good reason. Adjust it to see if things
still work OK.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present we cannot use this function as an init sequence call without a
wrapper, since it returns the RAM size. Adjust it to set the RAM size in
global_data instead, and return 0 on success.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
It looks like only cm5200 and tqm8xx use this feature, so we don't really
need it in generic code. Drop it and have the users access gd->board_type
directly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
At present we misuse print_cpuinfo() do so CPU init on x86. This is done
because it is the next available call after the console is enabled. But
several arches use checkcpu() instead. Despite the horrible name (which
we can fix), it seems a better choice.
Adjust the various x86 CPU implementations to move their init code into
checkcpu() and use print_cpuinfo() only for printing CPU info.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Move these two function calls into checkcpu(), which is called on this
arch immediately after these two.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
We don't need a special hook for sandbox as one of the later ones will do
just as well. We can print error messages about bad options after we
print the banner. In fact, it seems better.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Combine the conditions so this appears in the init list only once.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
Reviewed-by: Stefan Roese <sr@denx.de>
More than half of the architectures use this function so let's make them
all use it.
For those which don't actually define it, we can rely on the weak function
in lib/time.c
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
Reviewed-by: Stefan Roese <sr@denx.de>
There is no good reason to use a different name on PowerPC. Change it to
timer_init() like the others.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Now that both branches of the #if do the same thing, we can unify them.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
Reviewed-by: Stefan Roese <sr@denx.de>
We really don't need to have a name like this in the generic init
sequence. Use the generic get_clocks() name so that we can merge these
two at some point.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <york.sun@nxp.com>
Reviewed-by: Stefan Roese <sr@denx.de>
We have two chunks of code which depend on this CONFIG options. There is
likely no need to keep them apart, so join them.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
While x86 is the only user and this could in principle be moved to
arch_cpu_init() there is some justification for this being a separate
call. It provides a way to handle init which is not CPU-specific, but
must happen before the CPU can be set up.
Rename the function to be more generic.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
There is no need to have this call in the generic init sequence and no
other architecture has needed it in the time it has been there. Move it
into sandbox's private code.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: Stefan Roese <sr@denx.de>
Certain boards come in different variations by way of utilizing daughter
boards, for example. These boards might contain additional chips, which
are added to the main board's busses, e.g. I2C.
The device tree support for such boards would either, quite naturally,
employ the overlay mechanism to add such chips to the tree, or would use
one large default device tree, and delete the devices that are actually
not present.
Regardless of approach, even on the U-Boot level, a modification of the
device tree is a prerequisite to have such modular families of boards
supported properly.
Therefore, we add an option to make the U-Boot device tree (the actual
copy later used by the driver model) writeable, and add a callback
method that allows boards to modify the device tree at an early stage,
at which, hopefully, also the application of device tree overlays will
be possible.
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Stefan Roese <sr@denx.de>
For ARMv8 Layerscape SoCs, secure memory and MC memorey are reserved
at the end of DDR. DDR is spit into two or three banks. This patch
reverts commit aabd7ddb and simplifies the calculation of reserved
memory, and moves the code into common SoC file. Secure memory is
carved out first. DDR bank size is reduced. Reserved memory is then
allocated on the top of available memory. U-Boot still has access
to reserved memory as data transferring is needed. Device tree is
fixed with reduced memory size to hide the reserved memory from OS.
The same region is reserved for efi_loader.
Signed-off-by: York Sun <york.sun@nxp.com>
The BSS region may overlap with relocations. If we clear BSS we will
overwrite the start of the relocation area. This doesn't matter when running
from SPI flash, since it is read-only. But when relocating 64-bit U-Boot
from one place in RAM to another, relocation will fail because some of its
relocations have been zeroed.
To fix this, put the ELF fixup call before the BSS clearing call.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
These tables should be declared static const. Unfortunately the table in
board_r is updated on machines with manual relocation.
Update them.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Adjust the code so that 64-bit startup works. Since we don't need to do CAR
changes in U-Boot proper anymore (they are done in SPL) we can simplify the
flow and return normally from board_init_f().
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Since a platform conversion to generic board support has not been
accomplished some architecture specific bits are missing from board_f
init sequence, the change adds a number of basic expected callbacks
into early init sequence.
Signed-off-by: Vladimir Zapolskiy <vz@mleia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This is already called earlier, from fdtdec_setup(), so drop this unnecessary
call from the init sequence.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Fix various misspellings of:
* deprecated
* partition
* preceding,preceded
* preparation
* its versus it's
* export
* existing
* scenario
* redundant
* remaining
* value
* architecture
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Reviewed-by: Jagan Teki <jteki@openedev.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Now, arch/${ARCH}/include/asm/errno.h and include/linux/errno.h have
the same content. (both just wrap <asm-generic/errno.h>)
Replace all include directives for <asm/errno.h> with <linux/errno.h>.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
[trini: Fixup include/clk.]
Signed-off-by: Tom Rini <trini@konsulko.com>
Currently we have a mismash of architectures which use arch_cpu_init
from architecture-wide code (arc, avr32, blackfin, mips, nios2, xtensa)
and architectures which use arch_cpu_init from machine/SoC level code
(arm, x86).
In order to clean this mess up & allow for both use cases, introduce a
new mach_cpu_init callback which is run immediately after arch_cpu_init.
This will allow for architectures to have arch-wide code without needing
individual machines to all implement their own arch_cpu_init with a call
to some common function.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The Xtensa processor architecture is a configurable, extensible,
and synthesizable 32-bit RISC processor core provided by Cadence.
This is the first part of the basic architecture port with changes to
common files. The 'arch/xtensa' directory, and boards and additional
drivers will be in separate commits.
Signed-off-by: Chris Zankel <chris@zankel.net>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
When secure ram is used, MMU tables have to be put into secure ram.
To use common MMU code, gd->arch.tlb_addr will be used to host TLB
entry pointer. To save allocated memory for later use, tlb_allocated
variable is added to global data structure.
Signed-off-by: York Sun <york.sun@nxp.com>
Secure_ram variable was put in generic global data. But only ARMv8
uses this variable. Move it to ARM specific data structure.
Signed-off-by: York Sun <york.sun@nxp.com>
Allow this to be unset, such that gd->mon_len is invalid. This seems to be
what the sh architecture does.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present on x86 machines with use cache-as-RAM, the memory goes away just
before board_init_r() is called. This means that serial drivers are
no-longer unavailable, until initr_dm() it called, etc.
Any attempt to use printf() within this period will cause a hang.
To fix this, mark the serial devices as 'unavailable' when it is no-longer
available. Bring it back when serial_initialize() is called. This means that
the debug UART will be used instead for this period.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When using the early timer, we need to manually trigger setting up the
real timer. This will not happen automatically. Do this immediately after
starting driver model.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Before relocation we need to reserve memory for the video driver frame
buffers so that they can use this memory when they start up (after
relocation). Add a call to the uclass to permit this.
The current top and bottom of the region is stored in global_data so that
it can be checked post-relocation to ensure enough memory is available. No
video device should be probed before relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Anatolij Gustschin <agust@denx.de>
These two functions are conceptually the same. Move them together in the
pre-relocation init.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Anatolij Gustschin <agust@denx.de>
As the name may be confusing, the CONFIG_SYS_MEM_TOP_HIDE reserves
some memory from the end of ram, tracked by gd->ram_size. It is not
always the top of u-boot visible memory. Rewrite the macro with a
weak function to provide flexibility for complex calcuation. Legacy
use of this macro is still supported.
Signed-off-by: York Sun <yorksun@freescale.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Secure memory is at the end of memory, separated and reserved
from OS, tracked by gd->secure_ram. Secure memory can host
MMU tables, security monitor, etc. This is different from PRAM
used to reserve private memory. PRAM offers memory at the top
of u-boot memory, not necessarily the real end of memory for
systems with very large DDR. Using the end of memory simplifies
MMU setup and avoid memory fragmentation.
"bdinfo" command shows gd->secure_ram value if this memory is
marked as secured.
Signed-off-by: York Sun <yorksun@freescale.com>
With driver model timer support, there should not be an explict
call to timer_init(). Remove this call for x86.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Don't relocate fdt in case of CONFIG_OF EMBED as the fdt is
already embedded with u-boot image.
Update fdt_blob after relocation as the fdt will be copied
during u-boot relocation.
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Alexey Brodkin <abrodkin@synopsys.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com> (QEMU x86)
Tested-by: Thomas Chou <thomas@wytron.com.tw> (Nios2)
Acked-by: Thomas Chou <thomas@wytron.com.tw>
Acked-by: Simon Glass <sjg@chromium.org>
Fixed the prom_relocate() function in start.S file by reserving memory in
the board_init_f sequence and saving the offset to the __prom_start_reloc
variable. This value is used as the destination when relocating the PROM.
Add the prom_init() function to the end of the board_init_r sequence.
Signed-off-by: Francois Retief <fgretief@spaceteq.co.za>
Introduce the CONFIG_SYS_TIMER_* macros in include/asm/config.h to make use
of the generic timer infrastructure in lib/time.c.
Created a timer_init() function to initialize the timer hardware and update
the #ifdef in board_init_f to allow this function to be called during the
start-up sequence.
Signed-off-by: Francois Retief <fgretief@spaceteq.co.za>
It is useful to be able to record console output and provide console input
via a buffer. This provides sandbox with the ability to run a command and
check its output. If the console is set to silent then no visible output
is generated.
This also provides a means to fix the problem where tests produce unwanted
output, such as errors or warnings. This can be confusing. We can instead
set the console to silent and record this output. It can be checked later
in the test if required.
It is possible that this may prove useful for non-test situations. For
example the console output may be suppressed for normal operations, but
recorded and stored for access by the OS. That feature is not implemented
at present.
Signed-off-by: Simon Glass <sjg@chromium.org>
The console includes a global variable and several functions that are only
used by a small subset of U-Boot files. Before adding more functions, move
the definitions into their own header file.
Signed-off-by: Simon Glass <sjg@chromium.org>
As mark_bootstage() uses timer, it should go after driver model
is initialized.
Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Currently timer_init() is called in board_r.c which is quite late.
Some vgabios execution requires we set up the i8254 timer correctly,
but video initialization comes before timer_init(). Move the call
to board_f.c.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
get_clocks is wrapped by CONFIG_FSL_CLK and CONFIG_M68K in seperate
piece code. They can be merged into one snippet.
Signed-off-by: Peng Fan <Peng.Fan@freescale.com>
Cc: Tom Rini <trini@konsulko.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Alexey Brodkin <abrodkin@synopsys.com>
Cc: "angelo@sysam.it" <angelo@sysam.it>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: Stephen Warren <swarren@nvidia.com>
Cc: "Andreas Bießmann" <andreas.devel@googlemail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Angelo Dureghello <angelo@sysam.it>
The variables bd_t:bi_memstart and bd_t:bi_memsize have to be
initialized also on MIPS. Otherwise LMB and cmd_bdinfo do not
correctly work. This currently breaks the booting of FIT images
on MIPS. Enable the board_init_f hook setup_board_part1()
for MIPS to fix this.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
For most PPC platforms, they will call the first get_clocks() in
init_sequence_f[] as they define CONFIG_PPC. CONFIG_SYS_FSL_CLK is
then defined to call the second get_clocks(), which should be
redundant for PPC.
Signed-off-by: Gong Qianyu <Qianyu.Gong@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
This function will be used by both SPL and U-Boot proper. So move it into
a common place. Also change the #ifdef so that the early malloc() area is
not set up in SPL if CONFIG_SYS_SPL_MALLOC_START is defined. In that case
it would never actually be used, and just chews up stack space.
Signed-off-by: Simon Glass <sjg@chromium.org>
Zap almost all of the ad-hoc timer code from interrupts.c and
use the code in lib/time.c instead.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Move x86_fsp_init() call after initf_malloc() so that we can fix up
the gd->malloc_limit later.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
This causes widespread breakage due to the operation of the low-level code
in crt0.S and cro0_64.S for ARM at least.
The fix is not complicated but it seems safer to revert this for now.
This reverts commit 2afddae075.
Signed-off-by: Simon Glass <sjg@chromium.org>
There is quite a bit of assembler code that can be removed if we use the
generic global_data setup. Less arch-specific code makes it easier to add
new features and maintain the start-up code.
Drop the unneeded code and adjust the hooks in board_f.c to cope.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
At present we have a simple assignment to gd. With some archs this is
implemented as a register or through some other means; a simple assignment
does not suit in all cases.
Change this to a function and add documentation to describe how this all
works.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: York Sun <yorksun@freescale.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Some archs like to have larger alignment for their global data. Use 16 bytes
which suits all current archs.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
bd->bi_dram[] has both start address and size defined as 32-bit,
which is not the case on some platforms where >=4GiB memory bank
is used. Change them to support such memory banks.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
When running U-Boot as an EFI application we cannot relocate since we do not
have relocation information. U-Boot has already been relocated to a suitable
address.
Add a global_data flag to control skipping relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When U-Boot runs as an EFI application is does not have a definition of
CONFIG_SYS_TEXT_BASE. U-Boot is a relocatable application and the relocation
is done by EFI. U-Boot can be loaded at any address.
This is similar to how sandbox works. Adjust the early board init to deal
with this.
Signed-off-by: Ben Stoltz <stoltz@google.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The call to FspInitEntry is done in arch/x86/lib/fsp/fsp_car.S so far.
It worked pretty well but looks not that good. Apart from doing too
much work than just enabling CAR, it cannot read the configuration
data from device tree at that time. Now we want to move it a little
bit later as part of init_sequence_f[] being called by board_init_f().
This way it looks and works better in the U-Boot initialization path.
Due to FSP's design, after calling FspInitEntry it will not return to
its caller, instead it jumps to a continuation function which is given
by bootloader with a new stack in system memory. The original stack in
the CAR is gone, but its content is perserved by FSP and described by
a bootloader temporary memory HOB. Technically we can recover anything
we had before in the previous stack, but that is way too complicated.
To make life much easier, in the FSP continuation routine we just
simply call fsp_init_done() and jump back to car_init_ret() to redo
the whole board_init_f() initialization, but this time with a non-zero
HOB list pointer saved in U-Boot's global data so that we can bypass
the FspInitEntry for the second time.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Andrew Bradford <andrew.bradford@kodakalaris.com>
Tested-by: Simon Glass <sjg@chromium.org>
