As of commit ee12634872 (nb/sandybridge,sb/bd82x6x: Configure USB from
southbridge devicetree) and earlier commits, the USB port configuration
should be located in the devicetree instead of the mainboard_usb_ports
array, typically located in the boards early_init.c.
TEST=USB ports still function; and the USBIRx, USBPDO, USBOCM1, and
USBOCM2 RCBA registers in the inteltool dump did not change between
an E6430 build before and after the sb/intel/bd82x6x that moved the
usb config to the devicetree.
Change-Id: Ia5aa03a5894a8ef29e863470925a223f52e0ab70
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83006
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Trying to probe RAM space to figure out top of memory causes an
exception on RISC-V virtual machines with recent versions of QEMU, but
we temporarily enable exception handlers for that and use it to help
detect if a RAM address is usable or not. However, QEMU docs recommend
reading device information from the device-tree blob it provides us at
the start of RAM.
A previous commit adds a library function to parse device-tree blob that
QEMU provides us. Use it to determine top of memory in RISC-V QEMU
virtual machines, but still fall back to the RAM probing approach as a
last-ditch effort.
Change-Id: I9e4a95f49ad373675939329eef40d7423a4132ab
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80363
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Maximilian Brune <maximilian.brune@9elements.com>
The OpenTitan HW implements the same firmware interface as the Ti50
H1D3C hardware variant; it just has a different DID_VID. Allow this new
DID_VID to be recognized correctly.
BUG=b:324940153
Change-Id: Iaacf6d88bc6067948756c465aac1cd8b24ecae1f
Signed-off-by: Jett Rink <jettrink@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83033
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
This patch eliminates coreboot from loading microcode from RW CBFS
(when the RO descriptor is locked, which indicates a fixed RO image)
because the kernel can already patch the microcode on BSPs and APs
while booting to OS.
This may be a chance to lower the burden on the AP FW side because
patching microcode on in-field devices is subject to firmware updates,
which are rarely published and, if required, must go through the
firmware qualification testing procedure (which is costly, unlike
kernel updates for ucode updates).
1. The FIT loads the necessary microcode from the RO during reset.
2. Reloading microcode from RW CBFS impacts boot time
(~60ms, core-dependent).
3. The kernel can still load microcode updates.
ChromeOS devices leverage RO+RW-A/RW-B booting. The RO's microcode is
sufficient for initial boot, and the kernel can apply updates later.
BUG=none
TEST=Verified boot optimization; in-field devices skip RW-CBFS microcode
loading when RO is locked.
Change-Id: I68953d45d3624aba0a3be28bc7b266b7621ddcc4
Signed-off-by: Subrata Banik <subratabanik@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82999
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Reviewed-by: Nick Vaccaro <nvaccaro@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
While both APL and GLK load the CPU microcode from FIT, only GLK
supports the PRMRR/SGX feature. When this feature is supported, the
FIT microcode load will set the msr (0x08b) with the patch id/revision
one less than the revision in the microcode binary. This results in
coreboot attempting (and failing) to reload the microcode again in
ramstage. Avoid the microcode reload attempt for GLK by using a SoC-
specific microcode update check which accounts for the off-by-1 when
comparing versions.
Implementation is based on the one used for SKL and CNL, but modified
based on feedback in comments on Gerrit.
TEST=build/boot google/reef (electro) and google/octopus (ampton),
verify in cbmem console log that CPU microcode update in ramstage is
skipped due to already being up to date, and that GLK uses the
SoC-specific check and APL uses the non-specific/general one.
Change-Id: Iab97f23d4388d5057797bb13f585db821c735bd0
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83037
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Sean Rhodes <sean@starlabs.systems>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
When dynamically generating the DOD (Display Output Device) device
address (_ADR), don't set the DOD constraint flags; only set them when
using the address value to generate the DOD package.
This fixes ACPI brightness control functionality under Windows 11.
Before: Name (_ADR, 0x80010400)
After: Name (_ADR, 0x00000400)
TEST=build/boot Win11 on google brya (banshee), ensure display
brightness controls present and functional.
Ref: ACPI Spec 6.5 Appendix B.6.1 - _ADR
Change-Id: I1d710c6e55e6cb1d20d580bd784221ee1482b871
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83025
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
OpenSBI got bigger and doesn't fit anymore in 128K which causes coreboot
to not compiler anymore because the region overlaps with ramstage
This patch simply increases the size and uses the OpenSBI linker macro
instead.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: If1ccaafbf91dae986c470020faf9c0b4fba448e5
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83046
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Enable the fixed boot mode option in the VBT and set it to 1920x1080,
so that drobit boards equipped with 4K screens are legible at boot.
TEST=build/boot drobit w/4K screen using edk2 payload, verify boot
resolution set to 1080p and UEFI menus readable without a magnifying
glass.
Change-Id: If1f9e36d9bbdc2955ba890e2832aa64af9ba8f73
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83024
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This commit defines the panel_cfg register for the Razer Blade Stealth
(Kaby Lake). This enables libgfxinit support. These values are derived
from the stock firmware. First, VBIOSes were extracted from the stock
firmware. Then, intelvbtool was used to extract the VBT from each of the
VBIOS tables. Finally, intel_vbt_decode from igt-gpu-tools was used to
extract the register values. Although there were multiple VBIOSes
present in the firmware, all VBIOSes across both firmwares (on version
1.50 for the H2U and 8.02 for the H3Q) had the same register values.
Change-Id: I4c8b26ffb7a70d08655986084a714206d9d0c96a
Signed-off-by: Reagan Bohan <xbjfk.github@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82458
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Commit 88decca14f ("ACPI: Add helper fill_fadt_extended_pm_io()")
moved the population of the extended FADT to a separate function, but
incorrectly placed that function call before various length fields were
populated, leading to spurious validation errors in the cbmem boot log.
Correct this by moving the call to fill_fadt_extended_pm_io() after
the required fields are populated.
TEST=build/boot google/slippy (wolf), verify no FADT errors in cbmem
console log.
Change-Id: I1f8522e4813e6071692206f2b7ad2a2f5086071e
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83035
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
The word "experimental" has been removed from the help text for
HAVE_X86_64_SUPPORT Kconfig. This is because the x86_64 architecture
has now been officially tested and enabled for several x86 SoC
platforms.
This work will provide us with the foundation we need to begin working
with Intel's next-generation SoC platform (which requires to support
64-bit mode of booting by default).
Therefore, we can now remove the word "experimental" from the
"HAVE_X86_64_SUPPORT" Kconfig help text.
TEST=Able to build and boot google/rex64 in 64-bit mode to ChromeOS.
Change-Id: Ibd629f4e2722f3cbabbe297d4481790c9fa9226a
Signed-off-by: Subrata Banik <subratabanik@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83009
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Looks like PCIe root port device IDs for 9-series PCH-H are missing from
commit 434d7d4582 (sb/intel/lynxpoint: Add PCI DIDs for 9 series PCHs)
for some reason. Add them, so that coreboot performs PCIe initialisation
for 9-series PCH-H.
Change-Id: I1589418e5e25daabbf09c66c637e9c4f86aa02a6
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82947
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The H110M does not use memory down, and the spd directory doesn't exist
in the board's directory in the first place. This was probably just copy
and paste leftover from some existing Skylake board in the initial port.
TEST=Timeless build does not change.
Change-Id: I35744310b2bf8a14165dae9808c982e6dc274a74
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83010
Reviewed-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add coreboot support for qemu's sbsa-ref (Server Base System
Architecture) machine (-m sbsa-ref).
The qemu-sbsa coreboot port runs on EL2 and is the payload of the
EL3 firmware (Arm Trusted Firmware).
Note that, coreboot expects a pointer to the FDT in x0. Make sure
to configure TF-A to handoff the FDT pointer.
Example qemu commandline:
qemu-system-aarch64 -nographic -m 2048 -M sbsa-ref \
-pflash <path/to/TFA.fd> \
-pflash <path/to/coreboot.rom>
The Documentation can be found here:
Documentation/mainboard/emulation/qemu-sbsa.md
Change-Id: Iacc9aaf065e0d153336cbef9a9b5b46a9eb24a53
Signed-off-by: David Milosevic <David.Milosevic@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/79086
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
The UsbTcPortEn UPD for FSP-S is being set in ramstage, however the
equivalent FSP-M UPD, the UsbTcPortEnPreMem, was not being set.
Following the Meteor Lake example, set the UsbTcPortEnPreMem UPD
as well for Alder Lake.
Setting this FSP-M UPD will cause FSP to properly program sideband
use BSSB_LSx pins for the enabled Type-C ports. Required for proper
DCI debug and TCSS initialization flow.
Change-Id: If3b79167ec1769ddfb7d28a6c78a3e80bd10afe7
Signed-off-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80500
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Mainboard is QAL80/LA-7781P (UMA). The version with an Nvidia dGPU was
not tested. This is based on the autoport output with some manual fixes.
The VBT was obtained using `intelvbttool --inlegacy --outvbt data.vbt`
while running version A24 (latest version) of the vendor firmware.
The flash is 8MiB + 4MiB, and can be easily accessed by removing the
keyboard. It can also be internally flashed by sending a command to the
EC, which causes the EC to pull the FDO pin low and the firmware to skip
setting up any chipset based write protections [1]. The EC is the SMSC
MEC5055, which seems to be compatible with the existing MEC5035 code.
Working:
- Libgfxinit
- USB EHCI debug (left side usb port is HCD index 2, middle port on the
right side is HCD index 1) with the CH347
- Keyboard
- Touchpad/trackpoint
- ExpressCard (tested with USB 3.0 card)
- Audio
- Ethernet
- SD card reader
- mPCIe WiFi
- SeaBIOS 1.16.3
- edk2 (MrChromebox's fork, uefipayload_202309)
- Internal flashing using dell-flash-unlock
Not working:
- S3 suspend: Possibly EC related, DRAM power is getting cut when
entering S3
- Physical wireless switch: this triggers an SMI handler in the vendor
firmware which sends commands to the EC to enable/disable wireless
devices, and has not been reimplemented
- Battery reporting: needs ACPI code for the EC
- Brightness hotkeys: probably EC related
- The system reports that the power button was pressed and shuts down
when the CPU hits around 86 degrees Celsius, before the CPU can
thermal throttle. Likely EC and possibly PECI related.
- Integrated keyboard with upstream GRUB 2.12 payload: Upstream GRUB
initializes the 8042 PS/2 controller in a way that is incompatible
with how the EC firmware emulates it. GRUB tries to initialize the
controller with scan code set 2 without translation, but the EC only
ever returns set 1 scan codes to the system and thus is only works as
an untranslated set 1 keyboard or a translated set 2 keyboard,
regardless of commands to set the scan code. A USB keyboard works
fine.
Unknown/untested:
- Dock
- eSATA
- TPM
- dGPU on non-UMA model
- Bluetooth module (not included on my system)
[1] https://gitlab.com/nic3-14159/dell-flash-unlock
Change-Id: I93c6622fc5da1d0d61a5b2c197ac7227d9525908
Signed-off-by: Nicholas Chin <nic.c3.14@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/77444
Reviewed-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Introduce HAVE_SHARED_PS2_PORT Kconfig for this Super I/O to have
mainboards indicate if they have one shared PS/2 port on the rear
panel. On these boards (where a Y-cable cannot allow both
keyboard and mouse to work off the same port), if a PS/2 keyboard is
not present, SIO should be configured to swap its role to mouse, to
allow the OS to find and initialize any mouse connected.
Supporting code will come in a separate patch. Idea is to condition
them on this Kconfig.
Change-Id: I156b15c6ba233cbe8b9ba4d2cfbca6836ad7483a
Signed-off-by: Keith Hui <buurin@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82631
Reviewed-by: Eric Lai <ericllai@google.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Remove USB configurations and data structures from northbridge
devicetree (SNB+MRC boards) and bootblock/romstage C code
(native-only SNB boards). All USB configurations are drawn from
southbridge devicetree going forward.
Change-Id: Ie1cd21077136998a6e90050c95263f2efed68a67
Signed-off-by: Keith Hui <buurin@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81882
Reviewed-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Transfer all USB responsibilities to southbridge/intel/bd82x6x,
using one set of USB port configuration supplied by mainboards
in the southbridge section of their devicetree.
For MRC raminit, export southbridge_fill_pei_data() as a hook for
southbridge code to implement. With new code via this hook, bd82x6x
fills pei_data based on this one set of USB port config.
For native raminit, early_usb_init() now goes directly to the devicetree
and no longer get passed an address to it.
TEST=abuild passes for all affected boards. All USB ports still work
on asus/p8x7x-series/v/p8z77-m.
Change-Id: I38378c7ee0701abc434b030dd97873f2af63e6b0
Signed-off-by: Keith Hui <buurin@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81881
Reviewed-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This board has used the USB current map from asus/p8z77-m_pro since it
first landed in coreboot, which actually doesn't match vendor firmware.
Apply values obtained from hardware while running vendor firmware
to both native and MRC config.
Change-Id: I7ce13493c3ecac8154460c1fedf05e2d70a8e394
Signed-off-by: Keith Hui <buurin@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82756
Reviewed-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
For each sandybridge boards with option to use MRC or native platform
init code, add a copy of the board's USB port config, consolidated between
both code paths, into the southbridge devicetree, using special values
allocated for this consolidation.
These get hooked up in a separate patch.
Change-Id: I53efca3d29b3c5d4d5b7e3d6dc3e6ce6c34201e6
Signed-off-by: Keith Hui <buurin@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81880
Reviewed-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
