This patch provides an option for CPU programming where coreboot
expected to load second microcode patch after BIOS Done bit is set
and before setting the BIOS Reset CPL bit.
BUG=b:233199592
TEST=Build and boot google/taeko to ChromeOS.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I426b38cb1200e60398bc89515838e49ce0a98f06
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64836
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
1. add timeout for root port detection and pass to FSP.
2. add 'slot implemented' flag and pass to FSP.
3. PcieRpSlotImplemented needs to be set when the root port is set to
hotplug. There is an assertion in FSP checking this.
4. PcieRpSlotImplemented is updated only when it is built-in as it is
default to slot implemented in FSP.
BRANCH=firmware-brya-14505.B
Signed-off-by: Cliff Huang <cliff.huang@intel.com>
Change-Id: I13feb1d2d67eaba634a3e700685132fba39e1525
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63942
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Anil Kumar K <anil.kumar.k@intel.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
This patch creates a helper function named `initialize_microcode()`
to load microcode and ease for all function to peform loading
microcode using this helper function.
BUG=b:233199592
TEST=Build and boot google/taeko to ChromeOS.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I7155fc2da7383629930ce147a90ac582782fa5ae
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64835
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
As per Intel Processor EDS, BIOS_DONE bit needs to be set on
all CPUs via MSR.
Also, implement a function to perform any SoC recommended CPU
programming prior to post CPUs init. At present calling
`cpu_soc_bios_done()` for all CPUs from `before_post_cpus_init()`.
Note: It is expected that `before_post_cpus_init()` will be
extended with other CPU programming recommendations in follow up
patches, for example: reload microcode patch etc.
BUG=b:233199592
TEST=Build and boot google/taeko to ChromeOS.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I8066cd724c9f15d259aeb23f3aa71a2d224d5340
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64834
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
This patch implements heci_init() API that perform initialization of
all HECI devices as per MAX_HECI_DEVICES config.
BUG=none
TEST=Able to build and boot google/taeko with this change. No CSE
error observed with `heci_init()` called from romstage.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ia25e18a20cc749fc7eee39b0b591d41540fc14c9
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64855
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
This patch fixes the return type for `devfn` variable inside
heci_set_to_d0i3(). `PCI_DEVFN` macro returns `unsigned int`
instead of `pci_devfn_t`.
TEST=Able to build and boot to ChromeOS without any failure.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ib3a575aa7d71cbe6932e823917b57c5558387433
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64877
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Sean Rhodes <sean@starlabs.systems>
This patch renames heci_init() to cse_init() as HECI initialization
should have a bigger scope than just initializing the CSE
(a.k.a HECI1 alone).
BUG=none
TEST=Able to build and boot google/taeko.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ic7edd55ccdcd70b244615fa06f81803a0ae6ce80
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64854
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Sean Rhodes <sean@starlabs.systems>
This patch removes the MTL CNVi DIDs macros from IA common code and is
added into the generic wifi driver.
As per Intel Connectivity Platform BIOS Guide, Connectivity Controller
IP for MTL-P is `Magnetar` and supported CRF is `Typhoon Peak 2`.
Previously Garfield Peak DIDs for Alder Lake SoC also added similarly
to generic wifi drivers.
BUG=b:224325352
TEST=Able to build and boot on MTL emulator.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ib98762749c71f63df3e8d03be910539469359c68
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64592
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tarun Tuli <taruntuli@google.com>
Reviewed-by: Wonkyu Kim <wonkyu.kim@intel.com>
gpio_configure_pad function gets called for most of the GPIO
configuration for all the boards. This function is not handling NULL
pointers properly which can cause exception in CPU.
This patch fixes the handling and function is able to return early
in case the NULL pointer is passed or any subsequent child function
calls return NULL.
BUG=None
BRANCH=None
TEST=Compilation works fine for all Alder Lake boards.
Change-Id: I97fad72cdd92f70c7c5e6fdd23fbecf535a6e388
Signed-off-by: Maulik V Vaghela <maulik.v.vaghela@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64857
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
This patch brings the feature of disabling RAPL to common code. It
replaces the current solution for APL and EHL.
For special case if RAPL disabling is only working via changes in MCHBAR
a new config switch was introduced.
Test: Boot mc_apl4/5 with this patch and ensure that the
relevant bits in MSR 0x610 are the same as before the
patch.
Change-Id: I2098ddcd2f19e3ebd87ef00c544e1427674f5e84
Signed-off-by: Uwe Poeche <uwe.poeche@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64596
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
This patch implements a SoC overrides to set CPU privilege level
as the MSR is not consistent across platforms. For example:
On APL/GLK/DNV, it's MSR 0x120 and CNL onwards it's MSR 0x151.
BUG=b:233199592
TEST=Build and boot google/taeko to ChromeOS.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I4584516102560e6bb2a4ae8c0d036be40ed96990
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64806
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
mtrr_use_temp_range is a lot smarter than the plain set_var_mtrr. It
will compute a new optimal solution with the temp ranges included
while also taking care of the cleanup before loading the payload/s3
resume.
Change-Id: I283ba07fc12c410be39dfdc828657598237247c1
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63550
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The patch uses coreboot error codes instead of uint8_t data type in the
pre_mem_debug_init function.
TEST=build code for Gimble
Signed-off-by: Sridhar Siricilla <sridhar.siricilla@intel.com>
Change-Id: I957ff764900cb789bf2aacf0472dcb281f48af07
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64301
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Maulik V Vaghela <maulik.v.vaghela@intel.com>
Move the Kconfig option `FSP_HYPERTHREADING` to common Intel Kconfig so
that it can be reused by other SoCs. Since not all SoCs support
hyperthreading, make it conditional on `HAVE_HYPERTHREADING`. SoCs
supporting hyperthreading need to select it so that `FSP_HYPERTHREADING`
is available.
Change-Id: I892d48b488cbf828057f0e9be9edc4352c58bbe7
Signed-off-by: Felix Singer <felixsinger@posteo.net>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60543
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sean Rhodes <sean@starlabs.systems>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
The patch skips sending DISCONNECT IPC command to PMC if system resumes
from S3.
coreboot notice DISCONNECT IPC command getting timedout during S3
resume if system has AC connected behind Type-C hub. This impacts
system resume time. Please refer TA# 730910 for more information.
coreboot need not send the DISCONNECT IPC command when system resumes
from S3 state.
TEST=Verified system boots to OS and verfied below tests on Gimble
1. coreboot doesn't send the DISCONNECT during S3 resume
2. After S3 resume, system detects the pen drive with Superspeed
3. After system resumes from S3, hot-plug the pen drive, system detects
the pen drive
3. System sends IPC commands when system boots from S0 or S5.
Signed-off-by: Sridhar Siricilla <sridhar.siricilla@intel.com>
Change-Id: I6ad006ae8677919c7dfeca8eec0af11454a2e89d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63932
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
All targets now use cbmem for the BERT region, so the implementation can
be common.
This also drops the obsolete comment about the need to have bert in a
reserved region (cbmem gets fixed to be in a reserved region).
Change-Id: I6f33d9e05a02492a1c91fb7af94aadaa9acd2931
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64602
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Add the boot flash MMIO window to the resources to report this region as
reserved to the OS. This is done to stay consistent with the reserved
memory ranges by coreboot and make the OS aware of them.
As x86 systems preserves the upper 16 MiB below 4G for BIOS flash
decoding use the complete window for reporting independent of the
actually used SPI flash size. This will block the preserved MMIO window.
Change-Id: Ib3a77e9233c3c63bad4de926670edb4545ceaddf
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64077
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
If the SPI controller is hidden from the OS (which is default on Apollo
Lake) then OS has no chance to probe the device and therefore can not be
aware of the resources this PCI device occupies. If the OS needs to move
some resources for a reason it can happen that the new allocated window
will be shadowed by the hidden PCI device resource and hence causing a
conflict. As a result this MMIO window will be inaccessible from the OS
which will cause issues in applications. For instance on Apollo Lake
this causes flashrom to stop working.
This patch adds a SSDT extension for the PCI device if it is hidden from
the OS and reports the occupied resource via ACPI to the OS. For the
cases where the device is hidden later at coreboot runtime and therefore
is not marked as hidden in the PCI device itself a Kconfig switch called
'FAST_SPI_GENERATE_SSDT' is introduced. It defaults to 'no' and can be
set from SOC code to override it.
Since there is no defined ACPI ID for the fast SPI controller available
now, the generic one (PNP0C02) is used.
Test: Boot mc_apl4 and make sure flashrom works again.
Change-Id: Ia16dfe6e001188aad26418afe0f04c53ecfd56f1
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64076
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sean Rhodes <sean@starlabs.systems>
Reviewed-by: Subrata Banik <subratabanik@google.com>
There are efforts to have bootflows that do not follow a traditional
bootblock-romstage-postcar-ramstage model. As part of that CBMEM
initialisation hooks will need to move from romstage to bootblock.
The interface towards platforms and drivers will change to use one of
CBMEM_CREATION_HOOK() or CBMEM_READY_HOOK(). Former will only be called
in the first stage with CBMEM available.
Change-Id: Ie24bf4e818ca69f539196c3a814f3c52d4103d7e
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63375
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
There are two APL specific config switches for RAPL and min. cpu clock
(APL_SKIP_SET_POWER_LIMITS, APL_SET_MIN_CLOCK_RATIO). These switches
could be used in future in other CPU platforms. Move them to common code
instead of having them just for one SOC.
Test: Make sure that the clock ratio (MSR 0x198) and the RAPL settings
(MSR0x610) do not change with this patch applied on mc_apl{1,4,5}
mainboard.
Change-Id: I3d63d1b9b6c96586a3c20bf8c1d8001b1d7c4bed
Signed-off-by: Uwe Poeche <uwe.poeche@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63546
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
Starting with Meteor Lake SoC, the PCR/DMI interface to program GPMR
is replaced with IOC (I/O Cache), hence, this patch implements IOC
driver to support that migration.
Reference: 643504 MTL FAS section 7.5.2
TEST=Build and boot to OS for TGL RVP and MTL PSS
Signed-off-by: Wonkyu Kim <wonkyu.kim@intel.com>
Change-Id: I768027c2ca78310c03845f70f17df19dc8cd0982
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63198
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Nick Vaccaro <nvaccaro@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The fast SPI controller (usually handling the boot NOR flash) is a
different controller type than the generic SPI controllers as it
provides access to the boot flash and usually is not used for generic
SPI slave connections.
Though there is common code for the fast SPI controller it currently do
not uses the PCI driver structure. This patch adds the PCI driver
envelope to the fast SPI driver and moves Apollo Lake as the first
platform to this driver.
Change-Id: I31bf39ec1c622db887dec9ca8623a7f282402849
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64075
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
GNVS is more fragile as you need to keep struct elements in sync with
ASL code.
Change-Id: I2cd5e6b56e4a0dbbb11f4a0ac97e8f84d53b90ec
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64216
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
The "No memory dimm at address" line in get_spd_sn and get_spd fucntion
have different format of SPD address.
get_spd_sn shows a 8-bit address format but get_spd shows a 7-bit
address format when there is no DIMM connected. It can be confusing
when debugging.
Change-Id: I46a006f4024b12d27ae0a933b7c40515034d5d64
Signed-off-by: Kane Chen <kane.chen@intel.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64290
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jamie Chen <jamie.chen@intel.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
There were 3 different functions in gpio.c file which used to
get gpio group and pin information separately through function
calls.
Since these are static function, we can modify argument to
pass group and pin information from parent/calling function.
This will reduce redundant work of getting information 3 times
separately.
BUG=None
BRANCH=None
TEST=code compiles and correct information is passed to functions.
Check by using pin information on Brya.
Change-Id: Ie92be8c22838ebc5e831be58545e2023eecfff24
Signed-off-by: Maulik V Vaghela <maulik.v.vaghela@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64231
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Since coreboot locks GPIO registers after GPIO configuration, OS is not
able to program GPE_EN register to program wake events. This causes
the issue of event not getting logged into event log (since GPE_EN bit
is not set).
GPE_EN register programming is required for the GPIO pins which are
capable of generating SCI for the system wake. Elog mechanism relies
on GPE_EN and GPE_STS bit to log correct wake signal.
This patch add supports to program GPE_EN register before coreboot locks
the GPIO registers. Note that coreboot will only program GPE_EN bits for
GPIO capable of generating SCI.
This will help resolve issue where we don't see wake event GPIO in event
log.
BUG=b:222375516
BRANCH=firmware-brya-14505.B
TEST=Compile code for Brya and see GPE_EN bits set from the kernel console
Change-Id: I27e525f50c374c2cc9675e77eaa7774683a6e7c2
Signed-off-by: Maulik V Vaghela <maulik.v.vaghela@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64089
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Frank Wu <frank_wu@compal.corp-partner.google.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
coreboot needs to set GPE_EN bit for the GPIOs which are wake capable
from s0ix/sleep. Due to GPIO locking mechanism, coreboot/OS will not
be able to write GPE_EN register post GPIO has been locked.
This patch adds support in SoC code to provide correct offset for
GPE_EN and GPE_STS registers to the common code.
Plan is to use this offsets to set GPE_EN bits before GPIO locking
in coreboot which will be part of subsequent CL.
BUG=b:222375516
BRANCH=firmware-brya-14505.B
TEST=Check if code compiles for Brya and correct offset values are printed.
Change-Id: I6b813b30b8b360f8eccbf539b57387310e380560
Signed-off-by: Maulik V Vaghela <maulik.v.vaghela@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64088
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Frank Wu <frank_wu@compal.corp-partner.google.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
By using mp_run_on_all_cpus_synchronously to run APs MTRR init, it
gurantees the BSP will run post_cpus_add_romcache until all APs finishes
_x86_setup_mtrrs task.
BUG=b:225766934
TEST=Test on redrix and found the MTRR race condition on AP/BSP is gone.
Change-Id: I1fd889f880a0c605e6c739423a434d2adbc12d26
Signed-off-by: Kane Chen <kane.chen@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63567
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Implement sub-function 1 (Get Device Constraints)
of the Low Power S0 Idle Device-Specific Method (_DSM).
This provides a way in which to describe various devices required
D-states to enter LPM (S0ix). The information can be used to help
in diagnostics and understanding of S0ix entry failure.
Values were derived from Intel document 595644 (rev 0.45) and
the ADL FSP sample ASL.
This implementation adds support for ADL. Other SoC's could be
ported to be included as well. If they aren't, they will default
to the existing behavior of a single hardcoded device to ensure
compatibility with Windows.
TEST=Built and tested on brya by verifying SSDT contents
Change-Id: Ibe46a0583c522a8adf0a015cd3a698f694482437
Signed-off-by: Tarun Tuli <taruntuli@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63969
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Lance Zhao
A few ACPI device path name handlers are missing. Add handling
to ensure that these names are returned during acpi_device_path()
calls.
TEST=Built and tested on brya
Signed-off-by: Tarun Tuli <taruntuli@google.com>
Change-Id: I37d6dd5df921c931af72dd469c3f4067c61b0df3
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63984
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
This patch sends an IPC to PMC to inform about PCI enumeration.
BUG=b:211954778
TEST=Able to build and boot google/redrix to OS.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I77d428f9501feaccab8bb431090d10ce8d3af9b2
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63953
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
pmc_prev_sleep_state() isn't handling the case where acpi_sleep_from_pm1()
returns ACPI_S4. Pass that value along so it can get set as a
prev_sleep_state. Without this, consumers see prev_sleep_state as 0
and always treat resume as a cold boot. With this, consumers can
correctly do behavior specific to S4 resume, like skipping the
disconnect IPC command to the PMC on Alderlake systems.
BUG=b:230031158
TEST=Resume from S4 on Primus4es
Signed-off-by: Evan Green <evgreen@chromium.org>
Change-Id: I3fb3dc428a749db80293e51a04a2096514a7b689
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64002
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Move ASL POST code declarations into a common file to avoid redundancy.
Also, provide a dummy implementation when `POST_IO` is not enabled, as
the value of `CONFIG_POST_IO_PORT` can't be used.
Change-Id: I891bd8754f10f16d618e76e1ab88c26164776a50
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63988
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Add display on and off notifications which call mainboard hooks if
present. This allows to handle some board specific functions in user
absence or presence (when display goes off from inactivity or on from
activity).
TEST=Use Display on/off notification on Clevo NV41 to tell EC about
laptop inactivity. It is necessary to properly handle S0ix entry (stop
the fans and start blinking the power led).
Signed-off-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Signed-off-by: Michał Kopeć <michal.kopec@3mdeb.com>
Change-Id: Ie80f631ecffa74467ab6d6162e552ba977f7e3f4
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62494
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Krystian Hebel <krystian.hebel@3mdeb.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This patch enforces disabling of the CSE device if CSE stays in
SOFT TEMP DISABLE state. The recommendation is to make CSE function
disable to avoid receiving any CSE commands from the OS layer.
BUG=b:228789015
TEST=None
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I77c254195895a93a5606adee8b6f43d8b7100848
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63822
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sridhar Siricilla <sridhar.siricilla@intel.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
This patch removes redundant DISABLE_HECI1_AT_PRE_BOOT config check for
heci1_disable(), once by caller (from various SoC) and again inside the
callee (heci1_disable) function.
As all callers of heci1_disable() function are doing
DISABLE_HECI1_AT_PRE_BOOT config enabled check, hence, the second check
inside the callee can be dropped.
BUG=b:228789015
TEST=Able to build and boot google/redrix with this change. CSE PCI
device is getting function disabled upon selecting
DISABLE_HECI1_AT_PRE_BOOT from SoC config.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I47d7a9989e355987618d089f79c3340fcf4953ad
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63821
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Sridhar Siricilla <sridhar.siricilla@intel.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
`sa_lockdown_cfg` function ensures locking the PAM register hence,
skip dedicated calling into `sa_lock_pam()` from the SoC
`finalize.c` file. Dropped sa_lock_pam() call from ADL/CNL/EHL/JSL
and TGL.
Additionally, this patch enforces SA lockdown configuration for SKL
and ICL as well.
BUG=b:211954778
TEST=Able to build google/brya with these changes.
> localhost ~ # lspci -xxx | less
00:00.0 Host bridge: Device 8086:4601 (rev 04)
Bit 0 for all PAM registers a.k.a, PAMx_0_0_0_PCI.LOCK bit is set
(meaning locked).
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ibd464d2507393ed0c746eb1fbd10e36092ed5599
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63518
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Commit 211be9c03 (soc/intel/cmn/{block, pch}: Migrate GPMR driver)
drops `dmi` prefix from `lockdown_cfg` function name.
This patch adds the `gpmr` prefix to the lockdown_cfg function to make
it meaningful.
BUG=b:211954778
TEST=Able to build google/brya.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Idaa0e089131ab125348e2430355041c4ee7971de
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63789
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
There is a bug floating around where communciation with the PMC fails
after transitions through S3/S4/S5. This CL does not address that issue.
However in working with error cases presented by a failing PMC, we're
forced into an early return in read_pmc_lpm_requirements(), which sets
up _DSM buffers in the SSDT for the PEP device.
The function itself returns void, so the error is swallowed regardless.
However returning early is not the appropriate action because it causes
the size of the buffer written into the _DSM method to change. This causes
the SSDT to change size and layout across an S3 or S4 transition, which
results in mayhem in the kernel, as the kernel is not expecting these
tables to change out from under it.
Instead of returning early, it's better to simply print the error and
keep going, attaching a zeroed out buffer for the substate requirements.
This results in an empty requirements mask for all states. From what I
can see in the kernel this is no more broken than today's behavior, as
this buffer seems to only be used for printing a debugfs file.
In fact in this particular case the kernel doesn't even notice, as this
buffer is copied out at boot, and not refreshed at resume.
BUG=b:230031158
TEST=hibernate and resume on Primus4ES
Signed-off-by: Evan Green <evgreen@chromium.org>
Change-Id: Ibe35d50b350b1b96dea313dfcbd00745970c16ab
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63790
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
