When preparing transition of AGESA calls to romstage, I placed a comment
indicating the place to move a particular call. Now that the AGESA call
has been moved to romstage, the comment became obsolete.
BUG=b:116095766
TEST=none.
Change-Id: I2811657385ab088747e32d4c66b99fdd01e7315e
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/28687
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
FSP initializes the VT-d feature on Broadwell-DE and assigns an address
space to the MMIO range. coreboot's resource allocator needs to be aware
of this fixed resource as otherwise the address can be assigned to a
different PCI device. In this case addresses are overlapped and the VT-d
range is not accessible any more.
To deal with it the right way add a fixed MMIO resource to the resources
list if VT-d BAR is enabled.
TEST=Booted into Linux and checked coreboot log for resource assignment.
Change-Id: I626ac17420eadc0b49031e850f0f40b3b221a098
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/28672
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
The DMAR table generation depends on the VT-d feature which is
implemented in its own PCI device located in PCI:00:05.0 for
Broadwell-DE. Add a new PCI driver for this device and move
DMAR table generation to this device driver.
Change-Id: I103257c73f5e745e996a441a2535b885270bc204
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/28671
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
According to AMD, there exists an undocumented MSR which must be
written with the PSP's base address. Read the value from the PSP's
config space and sync each core's copy of the MSR to match.
BUG=b:76167350
TEST=boot Grunt and verify "rdrand: disabled" goes away from dmesg
Change-Id: I30027d3b0a6fbd540375e96001beb9c25bf3a678
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28608
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Default STAPM percentage causes a lot of thermal throttling on grunt.
AMD experimented with 80%, it works for grunt. This is initial code to
provide easy change path for other grunt based platforms.
BUG=b:111608748
TEST=build and boot grunt.
Change-Id: I22863f6ed76152bf872fce3e275f8a7fd8077504
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/28564
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Add definition for PCH_GPIO_PIRQ_INVERT, which is needed
for google/buddy, a to-be-merged variant of google/auron.
Taken from Chromium commit 70ee99b [buddy: change trigger type of gpio53]
Change-Id: I21448160cee791710df51d06efa32cdfecf38c0f
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/28611
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Mainly update headers to build.
Added option PMC_GLOBAL_RESET_ENABLE_LOCK to remove
function configuring the global reset through PMC base.
On denverton the global reset lock is not in PMC base
but in the PCI registers so this code cannot be shared.
Change-Id: I9ace70862cab63f8355252d034292596c7eab1fd
Signed-off-by: Julien Viard de Galbert <jviarddegalbert@online.net>
Reviewed-on: https://review.coreboot.org/25426
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Lijian Zhao <lijian.zhao@intel.com>
Reviewed-by: Evandro Luiz Hauenstein <kingsumos@gmail.com>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
After changing clock from 33.33Mhz to 1Ghz the UART divisor needs to be
recalculated. Return correct tlck frequency in uart_platform_refclk.
Change-Id: I2291e4198cf466a8334211c6c46bc3268fc979a9
Signed-off-by: Philipp Hug <philipp@hug.cx>
Reviewed-on: https://review.coreboot.org/28584
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
The RISC-V Privileged Architecture specification defines the Machine
Time Registers (mtime and mtimecmp) in section 3.1.15.
Makes it possible to use the generic udelay.
The timer is enabled using RISCV_USE_ARCH_TIMER for the lowrisc,
sifive and ucb soc.
Change-Id: I5139601226e6f89da69e302a10f2fb56b4b24f38
Signed-off-by: Philipp Hug <philipp@hug.cx>
Reviewed-on: https://review.coreboot.org/27434
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add a __always_inline macro that wraps __attribute__((always_inline))
and replace current users with the macro, excluding files under
src/vendorcode.
Change-Id: Ic57e474c1d2ca7cc0405ac677869f78a28d3e529
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/28587
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@google.com>
The documentation unfortunately doesn't match what SiFive uses in their FSBL.
Use the same values as in FSBL to make DDR RAM work.
Change-Id: I844cc41ed197333adeae495e71ea70b4a9603650
Signed-off-by: Philipp Hug <philipp@hug.cx>
Reviewed-on: https://review.coreboot.org/28582
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The SiFive UART on the HiFive Unleashed uses the tlclk as input clock
which runs at coreclk / 2.
The input frequency is configured in the board code depending on the
current stage. (bootblock + romstage run at 33.33Mhz, ramstage at 1Ghz)
Change-Id: Iaf66723dba3d308f809fde5b05dfc3e43f43bd42
Signed-off-by: Philipp Hug <philipp@hug.cx>
Reviewed-on: https://review.coreboot.org/27440
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
HAVE_INTEL_FIRMWARE is used to enable certain options that rely on a valid
Inter Flash Descriptor to exist. It does *not* identify platforms or boards
that are capable of running in descriptor mode if it's valid.
Refine the help text to make this clear.
Introduce a new option INTEL_DESCRIPTOR_MODE_CAPABLE that does simply
declare that IFD is supported by the platform. Select this value everywhere
instead of the HAVE_INTEL_FIRMWARE and default HAVE_INTEL_FIRMWARE to
y if INTEL_DESCRIPTOR_MODE_CAPABLE is selected.
Move the QEMU Q35 special case (deselection of HAVE_INTEL_FIRMWARE) to
the mainboard directory.
Change-Id: I4791fce03982bf0443bf0b8e26d9f4f06c6f2060
Signed-off-by: Stefan Tauner <stefan.tauner@gmx.at>
Reviewed-on: https://review.coreboot.org/28371
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Instead of gpio_num, gpio_address should be used as the address in
write32. This lets us also get rid of a few casts.
Commit c9ed3ee8d8 ("soc/amd/stoneyridge: Fix gpio_set function") fixed
one instance of this bug, but it was more widespread.
TEST=None
Change-Id: I0cf87aac2f1b87b6eac2b506515e48fe908c1f2b
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-on: https://review.coreboot.org/28570
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-by: Martin Roth <martinroth@google.com>
Add a duplicate ACPI_BERT symbol with a 'y' default setting and additional
help text.
BUG=b:65446699
TEST=inspect BERT region, and dmesg, on full patch stack. Use test
data plus a failing Grunt system.
Change-Id: I817111cbd3e81b93d8b02d0654ba68c8678b1bbe
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28479
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
Expand the BERT reserved region size setting to account for the
possibility of no TSEG configuration. This change is only for
completeness, as stoneyridge must always use TSEG.
Change-Id: I90753fa408cfac4de38aff08979c45349bb62a66
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28554
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
Fix the following compiler warning on the latest toolchain:
src/soc/sifive/fu540/otp.c:48:1: error: useless storage class specifier in empty declaration [-Werror]
} __packed;
^
Change-Id: Ice87c821de7650ac547394efa2a4bcc5ae1ea668
Signed-off-by: Philipp Hug <philipp@hug.cx>
Reviewed-on: https://review.coreboot.org/28553
Tested-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Fix compilation issue
clint.c/mtime.c is needed as well in ramstage due to CR 28372 and 28355
Change-Id: I7c7768744a165b97978bb8f7f95acf7b32ca4aa4
Signed-off-by: Philipp Hug <philipp@hug.cx>
Reviewed-on: https://review.coreboot.org/28551
Tested-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
This patch adds the support for CmdTriStateDis FSP upd in skylake
soc structure so that we can define it in devicetree.CmdTriStateDis
needed to be set for the skylake/kabylake based boards where LPDDR3
design is without RTT for CMD/CTRL.We need to set this bit for those
designs for the margin to be proper.
BUG=b:111812662
TEST=Run memtester app and also webgl fishtank on
the LPDDR3 kabylake boards and also check the
margin data is proper in FSP.
Change-Id: Ida69e443aa6ea4b524bd3ea2dcf26f4e63010291
Signed-off-by: Shaunak Saha <shaunak.saha@intel.com>
Reviewed-on: https://review.coreboot.org/28424
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Nick Vaccaro <nvaccaro@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add a Boot Error Record Table to the ACPI information. Avoid a driver
error message by skipping the table altogether when no errors are found,
or support isn't built in.
BUG=b:65446699
TEST=inspect BERT region, and dmesg, on full patch stack. Use test
data plus a failing Grunt system.
Change-Id: I6fe38eefacaad0bc73d0cb4ae44a339a45857128
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28478
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add functions to build a Boot Error Record Table region based on
settings found in the MCA registers.
Two entries are reported for each error due to the nature of the ACPI
driver. The first is a Generic Processor Error, which the OS recognizes
and parses. Generic errors cannot convey much error description or
processor context. Therefore an IA32/X64 Processor Error is also added,
which allows reporting the values found in the MCA MSR registers.
Follow-on work could decode the MC errors more precisely, and better
completing the Generic Error and the Check structure. The current
level of support is sufficient to identify a (i.e., human readable)
problem in dmesg, and provides adequate context information for
analysis.
BUG=b:65446699
TEST=inspect BERT region, and dmesg, on full patch stack. Use test
data plus a failing Grunt system.
Change-Id: I4d4ce29ddefa22aa29e6d3184f1adeaea1d5f837
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28477
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Convert the Machine Check reporting to use a newly defined structure.
This will facilitate later patches that will pass pointers to the MSR
values.
BUG=b:65446699
TEST=inspect BERT region, and dmesg, on full patch stack. Use test
data plus a failing Grunt system.
Change-Id: I0a98aecc83a0fa1c5ca7926849a89145a595d9ff
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28476
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Move the process of interrogating the Machine Check registers into
its own file. This rearranges source code in preparation of supporting
a Boot Error Record Table, which stoneyridge will use to report latent
MC errors to the OS.
BUG=b:65446699
TEST=inspect BERT region, and dmesg, on full patch stack. Use test
data plus a failing Grunt system.
Change-Id: Ia3275e9135dc96ba4a717c9371f38843fa1e3e64
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28475
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Carve out memory to be reported to the OS as reserved. This makes
room for a region usable for Boot Error Record Table information.
The BERT region reserved size is larger than likely requried, however
the SMM region's base must be on a boundary matching the granularity
of its size.
BUG=b:65446699
TEST=inspect BERT region, and dmesg, on full patch stack. Use test
data plus a failing Grunt system.
Change-Id: I0958f6b6bab3fe9dae36c83e1fd9ae6ed0290a18
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/28474
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
this enables spi console for wedge100s with broadwell_de. the console
size is 64kb. enabling spi console in `board.fmd` enables code which
calls into `timer_monotonic_get` (from `spi_flash_cmd_poll_bit`) and
`udelay` (from `ich_status_poll`). this patch selects `TSC_CONSTANT_RATE`
in fsp_broadwell_de's Kconfig to satisfy that.
Change-Id: Ib925c5aee88b65c46a81534405c364dd5649f8e8
Signed-off-by: Okash Khawaja <okash.khawaja@gmail.com>
Reviewed-on: https://review.coreboot.org/28528
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
Add a function to set the Bios Interface Lock Down bit (bit 31)
in RTC Configuration register (0x3400). This bit when set prevents
the top swap enable bit (bit 0) in the RTC BUC register (0x3414)
from being changed.
Change-Id: Iacaeeb0d6cabcf0c2c46a58948457ab832351476
Signed-off-by: Rizwan Qureshi <rizwan.qureshi@intel.com>
Reviewed-on: https://review.coreboot.org/28057
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
BOOTBLOCK/ROMSTAGE run in CAR/scratchpad. When RAMSTAGE begins
execution will enable cache, then CAR will disappear. So the
Stack will be separated.
Change-Id: I37a0c1928052cabf61ba5c25b440363b75726782
Signed-off-by: Xiang Wang <wxjstz@126.com>
Reviewed-on: https://review.coreboot.org/28383
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
