This change affects Intel CPUs only. As most platforms are doing
uCode update using FIT, they aren't affected by this code either.
Update microcode in MP-init using a single spinlock when running on
a Hyper-Threading enabled CPU on pre FIT platforms.
This will slow down the MP-init boot flow.
Intel SDM and various BWGs specify to use a semaphore to update
microcode on one thread per core on Hyper-Threading enabled CPUs.
Due to this complex code would be necessary to determine the core #ID,
initializing and picking the right semaphore out of CONFIG_MAX_CPUS / 2.
Instead use the existing global spinlock already present in MPinit code.
Assuming that only pre-FIT platforms with Hyper-Threading enabled and at
most 8 threads will ever run into this condition, the boot delay is
negligible.
This change is a counterproposal to the previous published patch series
being much more unsophisticated.
Change-Id: I27bf5177859c12e92d6ce7a2966c965d7262b472
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/49303
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Target added to INTERMEDIATE all operate on coreboot.pre, each modifying
the file in some way. When running them in parallel, coreboot.pre can be
read from and written to in parallel which can corrupt the result.
Add a function to create those rules that also adds existing
INTERMEDIATE targets to enforce an order (as established by evaluation
order of Makefile.inc files).
While at it, also add the addition to the PHONY target so we don't
forget it.
BUG=chromium:1154313, b:174585424
TEST=Built a configuration with SeaBIOS + SeaBIOS config files (ps2
timeout and sercon) and saw that they were executed.
Change-Id: Ia5803806e6c33083dfe5dec8904a65c46436e756
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/49358
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Commit 542307b815 (broadwell: Add small delay before Flex Ratio reboot)
introduced a workaround for Broadwell. Implement it on Haswell as well.
Since this is only necessary when a TPM is present on a system, only do
the delay (which is not that small, to be honest) on TPM-enabled builds.
Change-Id: Id8b58e9fa2a1c81989305f5b4b765b82c01e1596
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/46941
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The MSR only needs to be set when IO MWAIT redirection is to be enabled.
This was copied from Sandy Bridge, which already had this inconsistency.
Change-Id: I424333afd654db9a7e180e9a2c31d369e3d92fd6
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/46917
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To work around various bugs running KVM enabled, copy page tables to
DRAM in assembly before jumping to x86_64 mode.
Tested on QEMU using KVM, no more stange bugs happen:
Tested on host
- CPU Intel(R) Core(TM) i7-7700HQ
- Linux 5.9
- qemu 4.2.1
Used to crash on emulating MMX instructions and failed to translate
some addresses using the virtual MMU when running in long mode.
Tested on host
- CPU AMD EPYC 7401P 24-Core Processor
- Linux 5.4
- qemu 4.2.1
Used to crash on jumping to long mode.
Change-Id: Ic0bdd2bef7197edd2e7488a8efdeba7eb4ab0dd4
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/49228
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
The _CST method is supposed to return a package. If a mainboard used
zero for all ACPI C-states, the generated _CST would return nothing,
which is invalid. Instead, return a package with no C-state entries.
This change is a no-op, since all mainboards have at least one valid
ACPI C-state. This is what `acpigen_write_CST_package()` does, too.
Change-Id: I1f531e168683ed108a8d6d03dee6f5415fd15587
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/49092
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
For arch/x86 the realmode part has to be located within the same 64
KiB as the reset vector. Some older intel platforms also require 4 KiB
alignment for _start16bit.
To enforce the above, and to separate required parts of .text without
matching *(.text.*) rules in linker scripts, tag the pre-C environment
assembly code with section .init directive.
Description of .init section for ELF:
This section holds executable instructions that contribute to the
process initialization code. When a program starts to run, the
system arranges to execute the code in this section before calling the
main program entry point (called main for C programs).
Change-Id: If32518b1c19d08935727330314904b52a246af3c
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/47599
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
All mainboards use the same values for AC and battery, even desktop
boards without a battery. Use the AC values everywhere and drop the
battery values. Subsequent commits will rename the AC power options
accordingly, and will also clean up the corresponding acpigen code.
This is intentional so as to ease reviewing the devicetree changes.
Also update util/autoport accordingly.
Change-Id: I581dc9b733d1f3006a4dc81d8a2fec255d2a0a0f
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/49088
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
We have identical gdtptr16 and gdtptr. The reference in
gdtptr_offset calculation is not accounted for when
considering --gc-sections, so to support linking
gdt_init.S separately add dummy use of gdtptr symbol.
Realmode execution already accessed gdt that was located
outside [_start16bit,_estart16bit] region. Remove latter
symbol as the former was not really a start of region,
but entry point symbol.
With the romcc bootblock solution, entry32.inc may have
been linked into romstage before, but the !ENV_BOOTBLOCK
case seems obsolete now.
Change-Id: I0a3f6aeb217ca4e38b936b8c9ec8b0b69732cbb9
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/47964
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Do not combine the host bridge device ID with the CPU stepping because
it is confusing. Although Sandy/Ivy Bridge processors incorporate both
CPU and northbridge components into the same die, it is best to treat
them separately. Plus, this change enables moving CPU stepping macros
from northbridge code into the CPU scope, which is done in a follow-up.
Change-Id: I27ad609eb53b96987ad5445301b5392055fa4ea1
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/48408
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Since most assembly files are no longer concatenated together
but built separately, section changes with .previous at the
end of the files have become spurious.
TEST=BUILD_TIMELESS
Change-Id: I2970eed2b114a53475ba385eec4e97bb7ae7095c
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/47963
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
This adds a helper function for long mode to call some code in protected
mode and return back to long mode.
The primary use case is to run binaries that have been compiled for
protected mode, like the FSP or MRC binaries.
Tested on Intel Skylake. The FSP-M runs and returns without error while
coreboot runs in long mode.
Change-Id: I22af2d224b546c0be9e7295330b4b6602df106d6
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/48175
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
This patch renames cbfs_boot_map_with_leak() and cbfs_boot_load_file()
to cbfs_map() and cbfs_load() respectively. This is supposed to be the
start of a new, better organized CBFS API where the most common
operations have the most simple and straight-forward names. Less
commonly used variants of these operations (e.g. cbfs_ro_load() or
cbfs_region_load()) can be introduced later. It seems unnecessary to
keep carrying around "boot" in the names of most CBFS APIs if the vast
majority of accesses go to the boot CBFS (instead, more unusual
operations should have longer names that describe how they diverge from
the common ones).
cbfs_map() is paired with a new cbfs_unmap() to allow callers to cleanly
reap mappings when desired. A few new cbfs_unmap() calls are added to
generic code where it makes sense, but it seems unnecessary to introduce
this everywhere in platform or architecture specific code where the boot
medium is known to be memory-mapped anyway. In fact, even for
non-memory-mapped platforms, sometimes leaking a mapping to the CBFS
cache is a much cleaner solution than jumping through hoops to provide
some other storage for some long-lived file object, and it shouldn't be
outright forbidden when it makes sense.
Additionally, remove the type arguments from these function signatures.
The goal is to eventually remove type arguments for lookup from the
whole CBFS API. Filenames already uniquely identify CBFS files. The type
field is just informational, and there should be APIs to allow callers
to check it when desired, but it's not clear what we gain from forcing
this as a parameter into every single CBFS access when the vast majority
of the time it provides no additional value and is just clutter.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: Ib24325400815a9c3d25f66c61829a24a239bb88e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/39304
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Wim Vervoorn <wvervoorn@eltan.com>
Reviewed-by: Mariusz Szafrański <mariuszx.szafranski@intel.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Introduce a weak function to let the platform code provide the processor
voltage in 100mV units.
Implement the function on Intel platforms using the MSR_PERF_STATUS msr.
On other platforms the processor voltage still reads as unknown.
Tested on Intel CFL. The CPU voltage is correctly advertised.
Change-Id: I31a7efcbeede50d986a1c096a4a59a316e09f825
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/43904
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Jonathan Zhang <jonzhang@fb.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
