This reverts commit 1bb9786da3 ("cpu/x86/mp_init.c: Set a bogus
initial lapic_id"), since it breaks MP init on amd/mandolin:
[INFO ] CPU #0 initialized
[INFO ] Initializing CPU #3
[INFO ] Initializing CPU #1
[INFO ] Initializing CPU #2
[EMERG] CPU: missing CPU device structureCPU: vendor AMD device 810f81
[DEBUG] CPU: family 17, model 18, stepping 01
[DEBUG] microcode: patch id to apply = 0x08108109
[INFO ] microcode: being updated to patch id = 0x08108109 succeeded
[INFO ] CPU #1 initialized
[ERROR] MP record 3 timeout.
[INFO ] bsp_do_flight_plan done after 1206 msecs.
[ERROR] MP initialization failure.
[EMERG] mp_init_with_smm failed. Halting.
TEST=The board boots again with the revert applied
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Ic1cae88f7345f9ff79e8f6e574521095b57c8cb7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/69186
Reviewed-by: Jason Glenesk <jason.glenesk@amd.corp-partner.google.com>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Reviewed-by: Fred Reitberger <reitbergerfred@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Add function needed to generate ACPI backlight control SSDT, along with
Kconfig values for accessing the registers.
Tested by adding gfx register on system76/lemp11. Backlight controls
work on Windows 10 and Linux 6.1.
Change-Id: I1cc33bf0121ff44aea68a7e3615c5e58e2ab6ce2
Signed-off-by: Tim Crawford <tcrawford@system76.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/69076
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Replace `while (...);` with `do {} while (...);` so that it's easier to
distinguish polling loops from something else, like function calls. The
`{}` can be understood as "nothing", so that the construct is naturally
read as "do nothing while (...)".
Another reason to prefer this method is that Jenkins does not complain.
Change-Id: Ifbf3cf072f8b817b2fdeece4ef89bae0822bb6e6
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/69077
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Now that coreboot performs the necessary power sequencing, switch
from using the 'probed' flag to 'detect' for all I2C touchscreens.
This alleviates ChromeOS from having to probe to see which
touchscreen model is actually present, prevents breaking ACPI spec
by generating device entries with status 'enabled and present'
which aren't actually present, and improves compatibility with
upstream Linux and Windows.
BUG=b:121309055
TEST=build/boot ChromeOS and Linux on skyrim, ensure touchscreen is
functional, and ACPI device entry generated for correct touchscreen
model.
Change-Id: Id9e3089decf0f94a1358929684ce248e52cbe41f
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/67779
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Frank Wu <frank_wu@compal.corp-partner.google.com>
Leakage from the SPI CS line onto the FPMCU VDD rail was preventing
the FPMCU from fully shutting down on AP reset.
Instead of simply turning off the power rail, now ensure the CS
line is not driven high until late in coreboot.
This ensures it is completely off for the requisite minimum of 200ms
(now measured at approx 1100ms).
BUG=b:245953688
TEST=Confirmed FPMCU is still functional on Kohaku.
Confirmed FpRebootPowerCycle unit test now passes
BRANCH=Hatch
Signed-off-by: Tarun Tuli <taruntuli@google.com>
Change-Id: I1e7e32f61c3ac1b3154d42821cc1dd4c5d3de303
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68819
Reviewed-by: Nick Vaccaro <nvaccaro@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
According to Intel doc# 575683 the PECI bus should be low when idle and
is pulled up by clients with strong drive. However, for unknown reasons
the bus stays high on this board, blocking s0ix entry.
The PECI reference schematic in the ASPEED AST2400 BMC datasheet
(actually not related to this board) says that a pull-down is *required*
for the idle state.
This might be just a requirement of this BMC, since this is nowhere
documented in Intel datasheets, schematics or elsewhere. However,
configuring a weak pull-down (20 k) on the PECI pad indeed solves this
problem for now.
Change-Id: I85193000af67cd2c0465bdbb58cdd51b68fd5b4f
Signed-off-by: Michael Niewöhner <foss@mniewoehner.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68794
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
According to Intel doc# 575683 the PECI bus should be low when idle and
is pulled up by clients with strong drive. However, for unknown reasons
the bus stays high on this board, blocking s0ix entry.
The PECI reference schematic in the ASPEED AST2400 BMC datasheet
(actually not related to this board) says that a pull-down is *required*
for the idle state.
This might be just a requirement of this BMC, since this is nowhere
documented in Intel datasheets, schematics or elsewhere. However,
configuring a weak pull-down (20 k) on the PECI pad indeed solves this
problem for now.
Change-Id: Ib5a6b0ad3553c2cf795037d6a1982102bcb04644
Signed-off-by: Michael Niewöhner <foss@mniewoehner.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68793
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To support an RPL SKU on gladios, gladios must use the FSP for RPL.
Select SOC_INTEL_RAPTORLAKE for gladios so that it will use the RPL
FSP headers for gladios.
BUG=b:239513596
BRANCH=None
TEST=FW_NAME=gladios emerge-brask intel-rplfsp
coreboot-private-files-baseboard-brya coreboot chromeos-bootimage
Signed-off-by: Kevin Chiu <kevin.chiu.17802@gmail.com>
Change-Id: Ic30f7fe30eb0a3151cdf46fff609819056b2fbfe
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68570
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nick Vaccaro <nvaccaro@google.com>
Enable the MEI in device trees of some Ibex Peak, Cougar Point and
Panther Point boards where they have been disabled.
Change-Id: I4327d19d3ed1a93a6466057f6eceed49ab9441c5
Signed-off-by: Evgeny Zinoviev <me@ch1p.io>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42412
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Alexander Couzens <lynxis@fe80.eu>
Assuming variants have a touchscreen by default, set the enable GPIO
high and hold in reset during romstage, then release reset in ramstage.
This will allow the touchscreen to make use of the runtime I2C detect
feature (enabled in a subsequent commit) so that an ACPI device entry
is created only for the touchscreen actually present.
Variants/SKUs which do not have a touchscreen (if any) can use the
romstage/ramstage GPIO override tables to set the associated enable/
reset GPIOs to NC.
BUG=b:121309055
TEST=build/boot skyrim with rest of patch series
Change-Id: Ic4d7ac8f951bb94da2216a24dc85a96275c9d449
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/67778
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Don't set bit 2 in _STA in order for Windows not to show a warning about
an unknown device in the device manager for this device. Since the _STA
object just returns a constant, a name definition can be used instead of
a method definition.
TEST=The unknown device with device instance path ACPI\AAHB0000\0
disappeared from the device manager in Windows 10 build 19045 on a
Mandolin board with a Picasso APU.
Just shutting down and then booting it again won't clear some internal
state in Windows, so a reboot is needed instead for the change to become
visible.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: I8cb1712756c3623cc3ea16210af69cde0fa18f62
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68961
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
The SPD data for DRAM init has moved into the hwinfo data structure and
is therefore not used from spd.bin anymore. spd.bin will not receive any
updates, changes will only be done in hwinfo. There is no reason to keep
spd.bin around so remove it for both variants.
Change-Id: Ie6091b655ba7ff2e01b684266ce34b85593b8623
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68908
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
This patch ensures the P2SB PCI device resource is getting reserved
so that the resource allocator is not assigning this resource to any
other PCI device during the PCI enumeration.
BUG=b:254207628
TEST=Able to ensure on the Google/Rex device, the PCI enumeration
is not assigning the P2SB BAR (0xE000_0000) to TBT Root Port3.
Instead the 0xE000_0000 address is being assigned to the P2SB
PCI device.
Without this patch:
[SPEW ] PCI: 00:07.3 resource base e0000000 size c200000 align
20 gran 20 limit ec1fffff flags 60080202 index 20
[DEBUG] GENERIC: 1.0
[DEBUG] NONE
[SPEW ] NONE resource base e0000000 size c200000 align 12 gran
12 limit ec1fffff flags 40000200 index 10
With this patch:
[SPEW ] PCI: 00:07.3 resource base e1000000 size c200000 align
20 gran 20 limit ed1fffff flags 60080202 index 20
[DEBUG] GENERIC: 1.0
[DEBUG] NONE
[SPEW ] NONE resource base e1000000 size c200000 align 12 gran
12 limit ed1fffff flags 40000200 index 10
......
[DEBUG] PCI: 00:1f.1
[SPEW ] PCI: 00:1f.1 resource base e0000000 size 1000000 align
0 gran 0 limit 0 flags f0000200 index 10
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ib0789b442af23f6be81c666e284633ef342dffe0
Reviewed-on: https://review.coreboot.org/c/coreboot/+/68909
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Ivy Jian <ivy.jian@quanta.corp-partner.google.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Tarun Tuli <taruntuli@google.com>
