The serw13 with the 4070 RTX sometimes fails to enumerate the dGPU.
Increasing the delays allows the dGPU to be enumerated consistently. The
values are arbitrary, but still less than the values from proprietary
firmware.
Change-Id: Ibbfda596a324df4b51d583af8d6a36b5cd53a561
Signed-off-by: Tim Crawford <tcrawford@system76.com>
Add a driver for laptops with NVIDIA Optimus (hybrid) graphics. The
driver provides ACPI support for dynamically powering on and off the
GPU, and a function for enabling the GPU power in romstage.
Tested on system76/gaze15.
References:
- DG-09845-001: NVIDIA GN20/QN20 Hardware Design Guide
- DG-09954-001: NVIDIA GN20/QN20 Software Design Guide
Change-Id: I2dec7aa2c8db7994f78a7cc1220502676e248465
Signed-off-by: Jeremy Soller <jeremy@system76.com>
Signed-off-by: Tim Crawford <tcrawford@system76.com>
Exposing the GPIOs via an ACPI PowerResource and the _CRS results in the
OS driver and ACPI thinking they own the GPIO. This can cause timing
problems because it's not clear which system should be controlling the
GPIO.
There's no reason to require explicit disablement however, so drop the
superfluous 'disable' flag, and change the _CRS generation to check if
the GPIOs will be exported via the 'has_power_resource' flag instead.
This mirrors the change made for drivers/i2c/generic.
TEST=untested, as no boards selected this option.
Change-Id: Icb60502a4a7c5e7a1fcf1ee60e23c77e00d6de7b
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71851
Reviewed-by: Martin L Roth <gaumless@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Exposing the GPIOs via an ACPI PowerResource and the _CRS results in the
OS driver and ACPI thinking they own the GPIO. This can cause timing
problems because it's not clear which system should be controlling the
GPIO.
There's no reason to require explicit disablement however, so drop the
superfluous 'disable' flag, and change the _CRS generation to check if
the GPIOs will be exported via the 'has_power_resource' flag instead.
This mirrors the change made for drivers/i2c/generic.
TEST=untested, as no boards selected this option.
Change-Id: I4f95d0e453d89b7e1978d3efac304518304495d1
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71850
Reviewed-by: Martin L Roth <gaumless@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Exposing the GPIOs via an ACPI PowerResource and the _CRS results in the
OS driver and ACPI thinking they own the GPIO. This can cause timing
problems because it's not clear which system should be controlling the
GPIO.
Previously, we flagged as an error any device which set the
'has_power_resource' flag but did not set 'disable_gpio_export_in_crs.'
There's no reason to require explicit disablement however, so drop the
superfluous 'disable' flag, and change the _CRS generation to check if
the GPIOs will be exported via the 'has_power_resource' flag instead.
BUG=b:265055477
TEST=build/boot skyrim, dump SSDT and verify touchscreen GPIOs only
listed under PRx, not under _CRS.
Change-Id: I837ae6c6fe4b8e1c4e10686406cba06bdb7759d2
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71849
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin L Roth <gaumless@gmail.com>
This patch introduces an early graphics driver which can be used in
romstage in cache-as-ram mode. The implementation relies on
`libgfxinit' and provide VGA text mode support.
SoCs wanting to take advantage of this driver must implement the
`early_graphics_soc_panel_init' function to set the panel power
sequence timing parameters.
BUG=b:252792591
BRANCH=firmware-brya-14505.B
TEST=Graphics bring up observed on skolas with extra patches
Change-Id: Ie4ad1215e5fadd0adc1271b6bd6ddb0ea258cb5b
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70299
Reviewed-by: Tarun Tuli <taruntuli@google.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Maulik Vaghela <maulikvaghela@google.com>
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
A mainboard port needs to:
- select `CONFIG_MAINBOARD_HAS_EARLY_LIBGFXINIT'
- implement the Ada package `GMA.Mainboard' with a single function
`ports' that returns a list of ports to be probed for displays.
- set the desired `GFX_GMA_DEFAULT_MMIO' IO memory address to use
in romstage (and ramstage) for the graphic device.
BUG=b:252792591
BRANCH=firmware-brya-14505.B
TEST=libgfxinit compiles in romstage.
libgfxinit successfully executes in romstage and ramstage using
the requested MMIO setting on skolas.
Change-Id: I3c2101de10dc5df54fe873e43bbe0f1c4dccff44
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70276
Reviewed-by: Tarun Tuli <taruntuli@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This is support for adding legacy VGA support into romstage.
Support for this is being provided by libgfxinit.
The current use case allows us to initialize the display
before memory init (prior to physical memory init) to inform
the user when lengthy memory training is needed.
BUG=b:252792591
BRANCH=firmware-brya-14505.B
TEST=VGA code compiles for romstage
Change-Id: I81309871e8db71657b2a9816708141f121d767d3
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70278
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Tarun Tuli <taruntuli@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch provides an API to allow users to output multi-line
messages using VGA framebuffer.
The current limitation with multiline message is that,
vga_line_write() function is unable to understand newline character
hence, eventually output multiple lines separated with a newline
character with a single line statement.
This patch ensures to parse the entire string and split it into
multiple lines based on the newline character and print each line
separately to the VFG framebuffer.
User can choose to align the output video message as per given choice
between left/center/right of the screen
(i.e. enum VGA_TEXT_ALIGNMENT ).
Additionally, added macros to define the horizontal screen alignment
as well. Ideally if user would like to print the video message at the
middle of the screen then the vertical alignment would be
`VGA_TEXT_CENTER` and horizontal alignment would be
`VGA_TEXT_HORIZONTAL_MIDDLE`.
TEST=Able to build and boot Google/Taeko.
While output a video message such as :
"Your device is finishing an update. This may take 1-2 minutes.\nPlease
do not turn off your device."
Without this patch:
Your device is finishing an update. This may take 1-2 minutes. nPlease
do not turn off your device.
With this patch:
(in Left Alignment):
Your device is finishing an update. This may take 1-2 minutes.
Please do not turn off your device.
(in Right Alignment):
Your device is finishing an update. This may take 1-2 minutes.
Please do not turn off your device.
(in Center Alignment):
Your device is finishing an update. This may take 1-2 minutes.
Please do not turn off your device.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: Ib837e4deeba9b84038a91c93a68f03cee3474f9b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71265
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tarun Tuli <taruntuli@google.com>
When PTN3460_EARLY_INIT is selected, the PTN3460 (DP-2-LVDS-bridge) will
be initialized before all devices are initialized. This is necessary to
get a valid EDID data set into the PTN3460 before the graphic controller
is initialized in order to be able to show a splash screen.
For ptn3460_init() to work properly the I2C bus this bridge is connected
to needs to be initialized. As this I2C bus initialization would be done
too late in the normal flow, it needs to be called here explicitly
before ptn3460_init() to initialize the I2C bus with the needed
conditions. Otherwise the default I2C settings of the controller will be
used which results in a clock rate too high for this mainboard.
Test=Measure I2C bus signals and make sure that the clock is <= 400 kHz.
Change-Id: I1775fb7c2d29f765224d0e7c7ff9fcd4dbf847c5
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71226
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch fills in a generic platform info structure based on the
inputs from the SoC dptf header file (soc/dptf.h).
It will help to make things common and drop unnecessary back and
forth call between common code and SoC code.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I3521882495485cef686655abd65337515bae5faa
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71111
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Reviewed-by: Sumeet R Pawnikar <sumeet.r.pawnikar@intel.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch makes the SoC specific callback code more readable by adding
`soc_` prefix into the `get_dptf_platform_info()`.
In nutshell this patch renames `get_dptf_platform_info()` to
`soc_get_dptf_platform_info()`.
TEST=Able to build Google/Rex without any compilation issue.
Signed-off-by: Subrata Banik <subratabanik@google.com>
Change-Id: I27d6a146d5928e1742f82f85f51ad42656f46344
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71096
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sumeet R Pawnikar <sumeet.r.pawnikar@intel.com>
This patch implements mps2_noop_get_number_of_processors() API with
minimal information required for Intel MTL FSP to utilise the
`MP_SERVICES_PPI_V2_NOOP` config.
The major difference between Intel ADL and MTL FSP in terms of doing
CPU feature programming aka utilizing MP PPI wrapper code is that,
starting with MTL, FSP has dropped the `SkipMpInit` UPD.
It means now, coreboot doesn't have any way to skip FSP doing MP Init
operation. But during ADL, coreboot had introduced the
MP_SERVICES_PPI_V2_NOOP config that is used to skip FSP about
actually running any CPU feature programming on APs.
The idea is to use the same config even in MTL to provide only the
must have information (to bypass any assert in FSP during debug image)
to FSP.
Passing `FSP_UNSUPPORTED` from mps2_noop_get_number_of_processors()
results in `assert` while compiling FSP in debug mode hence,
implementing the function to pass only the information about BSP being
the active processor along with passing `FSP_SUCCESS` (eventually it
makes FSP happy and doesn't run into any issue in debug and/or release
mode).
TEST=Able to build and boot Google/Rex and Google/Kano while
coreboot skip calling into FSP for doing MP init.
Change-Id: I75d7e151699782210e86be564b0055d572cacc3f
Signed-off-by: Subrata Banik <subratabanik@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70555
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kapil Porwal <kapilporwal@google.com>