To deprecate VBOOT_VBNV_CMOS [1], replace VBOOT_VBNV_CMOS with
VBOOT_VBNV_FLASH for boards using SOC_INTEL_BRASWELL.
Currently BOOT_DEVICE_SPI_FLASH_NO_EARLY_WRITES is selected for
CPU_INTEL_HASWELL, SOC_INTEL_BRASWELL and others (see [2]). However,
there seems to be no particular reason on those platforms. We've dropped
the config for haswell. Now do the same for SOC_INTEL_BRASWELL, so that
VBOOT_VBNV_FLASH can be enabled.
VBOOT_VBNV_FLASH is enabled for the following boards:
- facebook/fbg1701: A 0x2000 RW_NVRAM region is allocated, with the
FW_MAIN_A(CBFS) size reduced by 0x2000.
- google/cyan, intel/strago: Repurpose RW_UNUSED as RW_NVRAM.
[1] https://issuetracker.google.com/issues/235293589
[2] commit 6c2568f4f5
("drivers/spi: Add BOOT_DEVICE_SPI_FLASH_NO_EARLY_WRITES config")
BUG=b:235293589
TEST=./util/abuild/abuild -t FACEBOOK_FBG1701 -a (with VBOOT selected)
TEST=./util/abuild/abuild -x -t GOOGLE_CYAN -a
TEST=./util/abuild/abuild -x -t INTEL_STRAGO -a
Change-Id: I46542c2887b254f59245f20b8642b023a7871708
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71544
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Frans Hendriks <fhendriks@eltan.com>
Reviewed-by: Erik van den Bogaert <ebogaert@eltan.com>
When VBOOT is enabled, the COREBOOT region (of size 0x09B000) is not
large enough. Therefore, adjust vboot-rw.fmd (which is used only with
VBOOT) layout by moving 0x10000 space from FW_MAIN_A(CBFS) region to
COREBOOT(CBFS) region.
TEST=./util/abuild/abuild -t FACEBOOK_FBG1701 -a (with VBOOT selected)
Change-Id: I1bc0d6981b873ca631cc4cc0720ab212700a65aa
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71543
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Frans Hendriks <fhendriks@eltan.com>
Reviewed-by: Erik van den Bogaert <ebogaert@eltan.com>
There is an ongoing effort to deprecate VBOOT_VBNV_CMOS [1] and replace
with VBOOT_VBNV_FLASH. However, the rambi's CAR is too small for early
flash access in romstage:
/usr/local/google/home/yupingso/projects/coreboot/util/crossgcc/xgcc/bin/i386-elf-ld.bfd:
Cache as RAM area is too full
/usr/local/google/home/yupingso/projects/coreboot/util/crossgcc/xgcc/bin/i386-elf-ld.bfd:
section .car.mrc_var VMA [00000000fe008000,00000000fe00ffff] overlaps
section .car.data VMA [00000000fe000000,00000000fe008787]
make: *** [src/arch/x86/Makefile.inc:194:
coreboot-builds/GOOGLE_RAMBI/cbfs/fallback/romstage.debug] Error 1
More precisely, DCACHE_RAM_SIZE is 0x8000, and the current .car.data
size is 0x76c0. If BOOT_DEVICE_SPI_FLASH_NO_EARLY_WRITES is unselected,
then the _bss region will increase by 0x10c8 bytes (for global variables
such as `elog_mirror_buf` and `sfg`), so that .car.data will exceed
0x8000.
Since rambi has reached its AUE (2021-09-01), disable
MAINBOARD_HAS_CHROMEOS and VBOOT configs.
[1] https://issuetracker.google.com/issues/235293589
BUG=b:235293589
TEST=./util/abuild/abuild -t GOOGLE_RAMBI -a
TEST=./util/abuild/abuild -x -t GOOGLE_RAMBI -a
Change-Id: Id56795dd0653784b4d7141142ebef0b19a46ddc3
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71545
Reviewed-by: Yidi Lin <yidilin@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin L Roth <gaumless@gmail.com>
This patch adds a new section to the coding style which codifies
existing practices about how to handle errors and how to use the die()
and assert() macros. Also clean up some references to Linux-specific
facilities that do not exist in coreboot in the adjacent function return
type guidelines, and add a small blurb of documentation to the
definition of the assert() macro itself.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: Ice37ed9f995a56d69476e95a352209041b337284
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70775
Reviewed-by: Elyes Haouas <ehaouas@noos.fr>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin L Roth <gaumless@gmail.com>
Replace the intelblocks/gpio.h and soc/gpio.h includes with the
common gpio.h which will include soc/gpio.h which will include
intelblocks/gpio.h
BUG=b:261778357
TEST=Able to build and boot Google/brya.
Change-Id: Ia90a8ea7b4ee125657c7277e3e14018cfe5423a9
Signed-off-by: Dinesh Gehlot <digehlot@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71266
Reviewed-by: Subrata Banik <subratabanik@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Add devicetree and GPIO configuration for MTL-RVP
Changes include,
1. Add initial devicetree to support MTL-RVP board & variant
2. Add initial setup for ramstage gpio config
BRANCH=none
BUG=b:224325352
TEST=Able to build with the patch and boot the mtlrvp to chromeOS
using subsequent patches in the train.
Signed-off-by: Harsha B R <harsha.b.r@intel.com>
Change-Id: I3173c3f32b36d24467431df3652badd70efeab93
Reviewed-on: https://review.coreboot.org/c/coreboot/+/69396
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Usha P <usha.p@intel.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
This patch adds initial bootblock code. This also configures required
GPIOs for early board initialization.
1. Add bootblock file for MTL-RVP
2. Add early gpio config for MTL-P variant in gpio.c
BRANCH=none
BUG=b:224325352
TEST=Able to build with the patch and boot the mtlrvp platform to
ChromeOS with the subsequent patches in the train
Signed-off-by: Harsha B R <harsha.b.r@intel.com>
Change-Id: I9c0893e52036147c5f6bbfafc6d818e9d3460bed
Signed-off-by: Jamie Ryu <jamie.m.ryu@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/69886
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Usha P <usha.p@intel.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
On Marasov, the ISH is running closed source firmware, so the ChromeOS
security requirements specify it must be behind an IOMMU. Add
DmaProperty to the ISH _DSD on Marasov.
TEST=Kernel marks ISH (PCI device 12.0) as untrusted, and changes the
IOMMU group type to "DMA". Also, device still goes to S0i3.
Before:
$ cat /sys/devices/pci0000\:00/0000\:00\:12.0/untrusted
0
After:
$ cat /sys/devices/pci0000\:00/0000\:00\:12.0/untrusted
1
Change-Id: I4b65b8909c41b06852fe7771375029bd2e76e111
Signed-off-by: Subrata Banik <subratabanik@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71263
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
1. Replace all pcidev_path_on_root() and is_dev_enabled()
functions combination with is_devfn_enabled().
2. Remove unused local variable of device structure type
(struct device *).
3. Replace pcidev_path_on_root() and dev->enabled check with
is_devfn_enabled() call.
TEST=Able to build and boot without any regression seen on MTL.
Port of 'commit 50134eccbd ("soc/intel/alderlake: Make use
of is_devfn_enabled() function")'
Signed-off-by: Dinesh Gehlot <digehlot@google.com>
Change-Id: I54bbd2bdba69a19e0559738035916fa7ac60faaa
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71161
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
This patch moves API "smbios_cpu_get_max_speed_mhz()"
to common code from board specific. This API was made
generic in 'commit d34364bdea ("soc/intel/alderlake:
Utilize `CPU_BCLK_MHZ` over dedicated macro")'
BUG=NONE
TEST=Boot and verified that SMBIOS max speed value is
correct on brya and rex.
(brya) dmidecode -t : "Max Speed: 4400 MHz"
(rex) dmidecode -t : "Max Speed: 3400 MHz"
Signed-off-by: Dinesh Gehlot <digehlot@google.com>
Change-Id: I87040ab23319097287e191d7fc9579f16d716e62
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70879
Reviewed-by: Johnny Lin <Johnny_Lin@wiwynn.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
python3-setuptools installed via apt-get is not currently working to
build dtc from git.kernel.org. Falling back to setuptools version
58.2.0 allows it to build again.
The failure message was:
/usr/lib/python3/dist-packages/setuptools/installer.py:27:
SetuptoolsDeprecationWarning: setuptools.installer is deprecated.
Requirements should be satisfied by a PEP 517 installer.
Signed-off-by: Martin Roth <gaumless@gmail.com>
Change-Id: I04cb6d776c3748f9a4b0cfc4ffd4f46458560d3d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71500
Reviewed-by: Felix Singer <felixsinger@posteo.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
With enabling FSP Notify Phase APIs, it has chance to issue a global
reset in FSP after CSE EOP (with selecting SOC_INTEL_CSE_SEND_EOP_EARLY
), which CSE already in idle mode and cause failure. For this reason we
should drop SOC_INTEL_CSE_SEND_EOP_EARLY in all ADL sku and select
SOC_INTEL_CSE_SEND_EOP_LATE instead.
BUG=b:261544011
BRANCH=firmware-brya-14505.B
TEST=tested and verified on Marasov, make sure this kind of global
reset can be executed successfully.
Signed-off-by: Tracy Wu <tracy.wu@intel.corp-partner.google.com>
Change-Id: I29736ca8efee64dd03feb48404241ee6295b7c72
Reviewed-on: https://review.coreboot.org/c/coreboot/+/71224
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Contents of unsigned AMDFW in RW sections are verified twice in PSP
verstage - first time by vboot verifying the firmware body, second time
by CBFS verification while the file is loaded to update PSP about the
boot region. This redundant verification adds to boot time. Minimize the
redundancy by splitting the EFS header from the AMDFW body and keep them
as 2 separate CBFS files. This helps to improve the boot time by another
25 ms.
BUG=None
TEST=Build Skyrim BIOS image and boot to OS. Observe boot time
improvement of ~25ms.
Before:
6:end of verified boot 363,676 (16)
11:start of bootblock 641,392 (277,716)
After:
6:end of verified boot 361,655 (16)
11:start of bootblock 616,967 (255,312)
Change-Id: Ib18a4f5c6781e5a7868e9395c0f1212da0823100
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70839
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add support to write EFS and AMDFW body to separate files. This is done
through passing an optional --body-location parameter to the amdfwtool.
If that option is not passed, then EFS will be written in the same file
as the AMDFW body. This will help to keep the minimum data to be
loaded/mapped from CBFS in PSP verstage.
BUG=None
TEST=Build and boot to OS in Skyrim.
Change-Id: I79325c81394cf8a0c663752d094adf6660896127
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/70778
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
