U-Boot sets up the real mode interrupt handler stubs starting from
address 0x1000. In most cases, the first 640K (0x00000 - 0x9ffff)
system memory is reported as system RAM in E820 table to the OS.
(see install_e820_map() implementation for each platform). So OS
can use these memories whatever it wants.
If U-Boot is in an S3 resume path, care must be taken not to corrupt
these memorie otherwise OS data gets lost. Testing shows that, on
Microsoft Windows 10 on Intel Baytrail its wake up vector happens to
be installed at the same address 0x1000. While on Linux its wake up
vector does not overlap this memory range, but after resume kernel
checks low memory range per config option CONFIG_X86_RESERVE_LOW
which is 64K by default to see whether a memory corruption occurs
during the suspend/resume (it's harmless, but warnings are shown
in the kernel dmesg logs).
We cannot simply mark the these memory as reserved in E820 table
because such configuration makes GRUB complain: unable to allocate
real mode page. Hence we choose to back up these memories to the
place where we reserved on our stack for our S3 resume work.
Before jumping to OS wake up vector, we need restore the original
content there.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Stefan Roese <sr@denx.de>
To do something more in acpi_resume() like turning on ACPI mode,
we need locate ACPI FADT table pointer first. But currently this
is done in acpi_find_wakeup_vector().
This changes acpi_resume() signature to accept ACPI FADT pointer
as the parameter. A new API acpi_find_fadt() is introduced, and
acpi_find_wakeup_vector() is updated to use FADT pointer as the
parameter as well.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Stefan Roese <sr@denx.de>
Call board_final_cleanup() before write_tables(), so that anything
done in board_final_cleanup() on a normal boot path is also done
on an S3 resume path.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Stefan Roese <sr@denx.de>
At the end of pre-relocation phase, save the new stack address
to CMOS and use it as the stack on next S3 boot for fsp_init()
continuation function.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Stefan Roese <sr@denx.de>
In an S3 resume path, U-Boot does everything like a cold boot except
in the last_stage_init() it jumps to the OS resume vector.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Stefan Roese <sr@denx.de>
Add one member in the global data to store previous sleep state,
and display the state during boot in print_cpuinfo().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Stefan Roese <sr@denx.de>
Much of the cpu and interrupt code cannot be compiled on 64-bit x86. Move it
into its own directory and build it only in 32-bit mode.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
At present pinctrl driver gets probed in ich6_gpio driver's probe
routine, which has two issues:
- Pin's PADs only gets configured when GPIO driver is probed, which
is not done by default. This leaves the board in a partially
functional state as we must initialize PADs correctly to get
perepherals fully working.
- The probe routine of pinctrl driver is called multiple times, as
normally there are multiple GPIO controllers. It should really
be called just once.
Move the call to syscon_get_by_driver_data() from ich6_gpio driver
to cpu_init_r().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: George McCollister <george.mccollister@gmail.com>
Tested-by: George McCollister <george.mccollister@gmail.com>
When SeaBIOS is on, reserve configuration tables in reserve_arch().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Instead of asking each platform to provide reserve_arch(),
supply it in arch/x86/cpu/cpu.c in a unified way.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
At present board_final_cleanup() is called before booting a Linux
kernel. This actually needs to be done before booting anything,
like SeaBIOS, VxWorks or Windows.
Move the call to last_stage_init() instead.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Each CPU needs to have its microcode loaded. Add support for this so that
all CPUs will have the same version.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This same name is used in USB. Add a prefix to distinguish it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
These two identifiers can be useful for drivers which need to adjust their
behaviour depending on the CPU family or stepping (revision).
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This does not need to be modified at run-time, so make it const.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
These devices currently need to be inited early in boot. Once we have the
init in the right places (with each device doing its own init and no
problems with ordering) we should be able to remove this. For now it is
needed to keep things working.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
There are no other implementations of this function, and boards that need it
can implement a CPU driver.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Right now i8254_init() is called from timer_init() in the tsc timer
driver. But actually i8254 and tsc are completely different things.
Since tsc timer has been converted to driver model, we should find
a new place that is appropriate for U-Boot to call i8254_init(),
which is now x86_cpu_init_f().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
This is not referenced anywhere. Remove it, as well as
tsc_base_kclocks and tsc_prev in the global data.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Fix 'Reomve' typo:
Signed-off-by: Simon Glass <sjg@chromium.org>
Some OS (like VxWorks) requires GDT entry 1 to be the 32-bit CS.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Jian Luo <jian.luo4@boschrexroth.de>
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>
Rather than keeping track of the Global Descriptor Table in its own memory
we may as well put it in global_data with everything else. As a first step,
stop using the separately allocated GDT.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When U-Boot runs as an EFI payload it needs to avoid setting up the CPU
again. Also U-Boot currently does not handle interrupts for many devices, so
run with interrupts disabled.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Currently cpu-x86 driver is probed only for SMP. We add the same
support for UP when there is only one cpu node in the deive tree.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Turn on cache on the pci option rom area to improve the performance.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Commit afbbd413a fixed this for non-driver-model. Make sure that the driver
model code handles this also.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
We should setup fixed range MTRRs for some legacy regions like VGA
RAM and PCI ROM areas as uncacheable. Note FSP may setup these to
other cache settings, but we can override this in x86_cpu_init_f().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Currently lapic_setup() is called before calling mp_init(), which
then calls init_bsp() where it calls enable_lapic(), which was
already enabled in lapic_setup(). Hence move lapic_setup() call
into init_bsp() to avoid the duplication.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Most of the MP initialization codes in arch/x86/cpu/baytrail/cpu.c is
common to all x86 processors, except detect_num_cpus() which varies
from cpu to cpu. Move these to arch/x86/cpu/cpu.c and implement the
new 'get_count' method for baytrail and cpu_x86 drivers. Now we call
cpu_get_count() in mp_init() to get the number of CPUs.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
This commit does the following to clean up x86 cpu dm drivers:
- Move cpu_x86 driver codes from arch/x86/cpu/cpu.c to a dedicated
file arch/x86/cpu/cpu_x86.c
- Rename x86_cpu_get_desc() to cpu_x86_get_desc() to keep consistent
naming with other dm drivers
- Add a new cpu_x86_bind() in the cpu_x86 driver which does exactly
the same as the one in the intel baytrail cpu driver
- Update intel baytrail cpu driver to use cpu_x86_get_desc() and
cpu_x86_bind()
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Currently the FSP execution environment GDT is setup by U-Boot in
arch/x86/cpu/start16.S, which works pretty well. But if we try to
move the FspInitEntry call a little bit later to better fit into
U-Boot's initialization sequence, FSP will fail to bring up the AP
due to #GP fault as AP's GDT is duplicated from BSP whose GDT is
now moved into CAR, and unfortunately FSP calls AP initialization
after it disables the CAR. So basically the BSP's GDT still refers
to the one in the CAR, whose content is no longer available, so
when AP starts up and loads its segment register, it blows up.
To resolve this, we load GDT before calling into FspInitEntry.
The GDT is the same one used in arch/x86/cpu/start16.S, which is
in the ROM and exists forever.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Andrew Bradford <andrew.bradford@kodakalaris.com>
Tested-by: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
This permits init of additional CPU cores after relocation and when driver
model is ready.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When we start up additional CPUs we want them to use the same Global
Descriptor Table. Store the address of this in global_data so we can
reference it later.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The existing code is pretty ancient and is unreliable on modern hardware.
Generally it will hang.
We can use port 0xcf9 to initiate reset on more modern hardware (say in the
last 10 years). Update the reset_cpu() function to do this, and add a new
'full reset' function to perform a full power cycle.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
We can write the configuration table in last_stage_init() for all x86
boards, but not with coreboot since coreboot already has them.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
CPUID (EAX 01H) returns MTRR support flag in EDX bit 12. Probe this
flag in x86_cpu_init_f() and save it in global data.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Option ROMs require a few additional descriptors. Add these, and remove the
enum since we now have to access several descriptors from assembler.
Signed-off-by: Simon Glass <sjg@chromium.org>
Since cpu_init_interrupts() was moved out of cpu_init_r(), it is
useless to keep cpu_init_r() for x86, thus remove it.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Currently cpu_init_interrupts() is called from cpu_init_r() to
setup the interrupt and exception of the cpu core, but at that
time the i8259 has not been initialized to mask all the irqs
and remap the master i8259 interrupt vector base, so the whole
system is at risk of being interrupted, and if interrupted,
wrong interrupt/exception message is shown.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
This implementation has a 'cpu' prefix and returns a pointer to the string,
avoiding the need for copying.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The CPU identification happens in x86_cpu_init_f() and corresponding
fields are saved in the global data for later use.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Some functions are missing prototypes. Fix those that are specific to x86.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Instead of an x86-specific cpu_init_f() function, use the normal U-Boot one
for this purpose. Also remove a useless/misleading comment.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add code to jump to a 64-bit Linux kernel. We need to set up a flat page
table structure, a new GDT and then go through a few hoops in the right
order.
Signed-off-by: Simon Glass <sjg@chromium.org>
