VMX is enabled through a bit in the IA32_FEATURE_CONTROL MSR, which can
be locked. The MSR remains locked after a non-power cycle reset, though.
If the MSR is locked, coreboot bails out and leaves VMX in the state it
was found. Because of this, changes to the VMX enable option in the BMC
only take effect after the system is power cycled.
This behaviour is highly undesirable because users are likely not aware
that a power cycle is required for changes to VMX state to take effect.
So, if VMX is supported, the IA32_FEATURE_CONTROL MSR is locked and the
current VMX state does not match the requested state, then issue a full
reset. This will power cycle the system and unlock the MSR, so that the
desired VMX state can be programmed into the MSR. This is checked early
to avoid needlessly doing time-consuming operations (running FSP) twice
if we know we will need to power cycle the system anyway.
Note that a user may change the VMX setting after the newly-added check
but before the setting is read in ramstage to program the MSR, but this
is a non-issue as firmware settings need a reset to take effect anyway.
TEST: Toggle VMX setting in BMC and reboot without power cycle, observe
coreboot automatically issues a power cycle reset because the MSR
is locked and the VMX state differs. Verify that the system boots
properly with VMX in the correct state after having power cycled.
Change-Id: Id9061ba896a7062da45a86fb26eeb58927184dcb
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75141
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Instead of having the maximum number of possible CPU objects defined in
the DSDT, dynamically generate the number of needed CPU devices in the
SSDT like it's done on all other x86 platforms in coreboot.
TEST=APU2 still boots and Linux doesn't show any ACPI errors with this
patch applied and it prints "ACPI: \_SB_.P000: Found 2 idle states".
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Id6f057ad130a27b371722fa66ce0a982afc43c6c
Reviewed-on: https://review.coreboot.org/c/coreboot/+/73073
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Instead of having binaryPI generate a PSTATE SSDT that uses \_PR_ as the
scope for the CPU objects and patching this SSDT in coreboot to use the
\_SB_ scope in patch_ssdt_processor_scope, request binaryPI to use the
\_SB_ scope instead by setting the late platform configuration option
ProcessorScopeInSb to true.
TEST=APU2 still boots and Linux doesn't show any ACPI errors with this
patch applied and it prints "ACPI: \_SB_.P000: Found 2 idle states".
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: I411201b55cfee30ae41da4e6814679bdb49e9bf7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/73386
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
This mainboard contains in addition to its base variant, mc_ehl2, an LCD
panel driven through the PTN3460 eDP-to-LVDS bridge.
This patch enables the PTN3460 support by adding the device to
devicetree.cb and board-specific configuration parameters in
lcd_panel.c.
BUG=none
TEST=Boot with the LCD panel attached and observe whether the picture is
stable and free of artifacts coming from wrong resolution and timing.
Change-Id: I196d7ceeb7ac241c9b95db2ef791a5f3ff7890a7
Signed-off-by: Mario Scheithauer <mario.scheithauer@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74936
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jan Samek <jan.samek@siemens.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
The Realtek RTL8111F NIC is currently not defined at all, nor as a child
device, resulting in the on_board flag not being set to 1. This means
that Linux / udev will call the device enp3s0 rather than eno0, as it's
appropriate for on-board ethernet devices.
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
Change-Id: I95f01a466a59234d1cbe2420f208bf58ae28fcc6
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75106
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Adds the option to set the 'DmaProperty' in the device's _DSD.
This can be done by setting "add_acpi_dma_property"="true". If not set
(or set to false), the device descriptor generation behavior will remain
unchanged. The naming convention for the config option was chosen to
match that of other drivers.
This partially reverts commit 5609f7a684 ("drivers/pcie/generic: Clean
up driver") as the driver is now used on a couple mainboards which need
the DmaProperty.
Change-Id: I996fe4923948d13a20bf8b6b1a93dab0866d0fd4
Signed-off-by: Mark Hasemeyer <markhas@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75094
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jan Samek <jan.samek@siemens.com>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
Create the gothrax variant of the nissa reference board by copying
the template files to a new directory named for the variant.
(Auto-Generated by create_coreboot_variant.sh version 4.5.0.).
BUG=279614675
BRANCH=None
TEST=util/abuild/abuild -p none -t google/brya -x -a
make sure the build includes GOOGLE_GOTHRAX
Signed-off-by: Yunlong Jia <yunlong.jia@ecs.corp-partner.google.com>
Change-Id: I129e4a55e4b87091e425a45392024d04f3977c79
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74748
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Derek Huang <derekhuang@google.com>
The return value of cse_get_bp_info() is an enum integer, where zero
means success and non-zero means failure. The function
store_cse_rw_fw_version() calls the function cse_get_bp_info() and
validates the return value as a boolean causing prematurely returns of
the parent API even if cse_get_bp_info() is successful.
This patch corrects this logical error by returning only if
cse_get_bp_info() fails.
TEST=Build and boot google/nivviks and verify that the ISH version info
command is only being sent during cold boot.
Signed-off-by: Dinesh Gehlot <digehlot@google.com>
Change-Id: Ice278e5ac69ff2f2c9f1936b76d71ae9deb6f855
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74998
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sridhar Siricilla <sridhar.siricilla@intel.com>
At present coreboot_rsdp remains unset for QEMU, which results in
an incomplete LB_TAG_ACPI_RSDP coreboot table generated.
Fix this by assigning coreboot_rsdp properly.
TEST=Build coreboot for QEMU x86 i440fx (default) with U-Boot x86
as the payload, boot coreboot.rom with QEMU, and run 'acpi list'
from U-Boot shell to show the ACPI tables.
Signed-off-by: Bin Meng <bmeng@tinylab.org>
Change-Id: I5bc3f0528d4431fd388ca52b8865f9be0e1faf92
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75088
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Tim Wawrzynczak <inforichland@gmail.com>
For the sequenced controlled shutdown path, there's a 10ms delay
after the PEXVDD rail is disabled to permit discharge needed on
Agah/Proxima.
This can be dropped to 3ms for Hades designs Proto0 and forward.
Once Agah board is dropped, "if CONFIG" can be cleaned up/removed.
BUG=b:271167335
TEST=builds
Signed-off-by: Tarun Tuli <taruntuli@google.com>
Change-Id: I8a0d62ec76caff861adce2d6c0ba2d4e4064affa
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75051
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
With the introduction of a new Linux version a problem has appeared
after a software initiated reset via CF9h register. The problem
manifests itself in the fact that the Linux kernel does not start after
the reboot. The problem is solved by setting bit 3 to 1 in Reset Control
Register (I/O port CF9h). This leads to the fact that the PCH will drive
SLP_S3 active low in the reset sequence. It leads to the same behavior
as in commit 04ea73ee78 ("siemens/mc_apl3: Set Full Reset Bit into
Reset Control Register") explained.
Change-Id: Ia8b7f997ca6234add569da751e1070144790e258
Signed-off-by: Mario Scheithauer <mario.scheithauer@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75041
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jan Samek <jan.samek@siemens.com>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
- VBT information: The link from 01.org is dead, but appears to have
been identical to the i915 page in the Linux kernel docs based on
snapshots on archive.org.
- Cgit: coreboot no longer has cgit running for the repos it hosts,
and these links redirect to the Gitiles list of repos hosted on
review.coreboot.org. Based on snapshots on archive.org, these used
to link to the individual repo or tree. Replace these with an
equivalent Gitiles link.
Change-Id: Id0bfee7b806c851fbe1dcf357e14d9b593e8569a
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74188
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Reviewed-by: Patrick Georgi <patrick@coreboot.org>
If the ME is disabled with the `me_state` CMOS setting, boot
times are approximately 5 seconds longer:
942:before sending EOP to ME 1,240,773 (5,599)
943:after sending EOP to ME 6,263,951 (5,023,177)
Total Time: 6,167,443
This is because the current code only checks if the ME is
disabled for CSE LITE SKUs. With this patch, boot times are
approximately 5 seconds quicker:
Total Time: 1,143,932
Tested on `starbook/adl` and `starbook/tgl`.
Signed-off-by: Sean Rhodes <sean@starlabs.systems>
Change-Id: I182f30d4fbf43955747c6a7a0b284a43f9c5e4ef
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74435
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sridhar Siricilla <sridhar.siricilla@intel.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Update header files for FSP for Meteor Lake platform to version 3165_81, previous version being 3084_85.
FSPM:
1. Change UPD name from 'GtExtraTurboVoltage' to 'GtAdaptiveVoltage'
2. Change UPD name from 'CoreVoltageAdaptive' to 'CoreAdaptiveVoltage'
3. Change UPD name from 'RingVoltageAdaptive' to 'RingAdaptiveVoltage'
4. Address offset changes
FSPS:
1. Remove deprecated UPD 'PcieDpc'
2. Address offset changes
BUG=b:280005256
TEST=Able to build and boot google/rex to ChromeOS.
Signed-off-by: Kilari Raasi <kilari.raasi@intel.com>
Change-Id: I67939ecf71166fca4f3d2d6cd4622215bebc5718
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74803
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Ronak Kanabar <ronak.kanabar@intel.com>
Reviewed-by: Dinesh Gehlot <digehlot@google.com>
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Based on feedback and experiences from new coreboot users, it isn't
clear that Tutorial 1 is mainly intended to set up the toolchain and
will not produce a bootable ROM for their board. Thus, add a note
explicitly mentioning this with a short explanation.
The process of manually building and adding the payload is also unusual,
since payloads are usually handled automatically by the build system.
This adds a note in the summary to provide an explanation of this.
The savedefconfig output is also outdated, as Kconfig now outputs
additional lines (even though many of those are the same as the
defaults). This has caused confusion, leading users to think that they
may have configured coreboot incorrectly. Update this to the current
defconfig contents and add a note that this may change depending on the
coreboot version.
Change-Id: I13206aa05a425ddfe33ee35feff0db490585a59f
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/73816
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The Linux kernel recently updated the wording of their sign-off
procedure, changing the ambiguous "real name" requirement to "a known
identity" and dropping "no pseudonyms". Anonymous contributions remain
uncommittable [1]. As discussed in the April 19, 2023 leadership
meeting, update our policy to go along with Linux and flashrom (who also
updated their policy).
[1] Linux kernel commit d4563201f3
(Documentation: simplify and clarify DCO contribution example language)
Change-Id: Ie676334f7c1509524adcb8dbb78495fb4da35ede
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74851
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <patrick@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Currently, this only exists on the old wiki and the developers.html page
on coreboot.org, but it really ought to be somewhere in the new docs
alongside the other contribution guidelines. This was largely copied
from the text from the developers.html page.
Change-Id: If50b3827ab36234719f9a90239caec4612eb6762
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74825
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
