Commit c7b8809f155a ("soc/amd/common/block/gfx: Use TPM-stored hash
for vbios cache validation") replaced checking the vbios signature
(first two bytes) with checking against a TPM-stored hash, but there
exists an edge case where the empty cache can be hashed and therefore
never updated with the correct vbios data. To mitigate this, re-add
the signature check to ensure that an empty cache will never be hashed
to TPM.
BUG=b:255812886
BRANCH=skyrim
TEST=build/boot skyrim w/selective GOP enabled, flash full firmware
image, ensure GOP driver is run until cache updated with valid data
and hashed to TPM.
Change-Id: Id06a8cfaa44d346fb2eece53dcf74ee46f4a5352
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74525
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jason Glenesk <jason.glenesk@gmail.com>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
In the PPRs #57019 Rev 3.03 and #57396 Rev 3.04, SMITYPE_XHC3_PME,
SMITYPE_XHC4_PME and SMITYPE_CUR_TEMP_STATUS_5 are defined, so add those
defines. When doing the initial update for Phoenix, at least XHC3 and
XHC4 PME events were missing from the PPR. Those two are the PME events
of the two USB4 controllers. SMITYPE_XHC2_PME doesn't exist on this SoC.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Ic6fff9175b73cc9d0fd324d4a568a5761b92d078
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74655
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
Reviewed-by: Fred Reitberger <reitbergerfred@gmail.com>
When one of the General-Purpose PCIe bridges is not used, it doesn't
show up on the PCI bus at all, so coreboot notes it as an issue in the
devicetree. This happens even if the device is marked as off.
To solve this, we're marking the GPP bridge devices in devicetree as
hidden, so they'll only show up in devicetree if they're actually used
on a mainboard.
BUG=b:277997811
TEST=Build
Signed-off-by: Martin Roth <gaumless@gmail.com>
Change-Id: I7b7577baa2dbb0ea7ebbcdb1a8ae81770e61d76f
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74527
Reviewed-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
When one of the General-Purpose PCIe bridges is not used, it doesn't
show up on the PCI bus at all, so coreboot notes it as an issue in the
devicetree. This happens even if the device is marked as off.
To solve this, we're marking the GPP bridge devices in devicetree as
hidden, so they'll only show up in devicetree if they're actually used
on a mainboard.
BUG=None
TEST=Don't see the "PCI: Leftover static devices:" warning for these in
the boot console.
BRANCH=skyrim
Signed-off-by: Martin Roth <gaumless@gmail.com>
Change-Id: I517776e4dedc70e957a0c836ab3c2e5d49e156d2
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74526
Reviewed-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <eric_lai@quanta.corp-partner.google.com>
After the obsoletion of Processor() it is necessary to provide
_CST package to define P_LVLx IO addresses for C2/C3 transitions.
The latency values from _CST will always replace those in FADT.
Change-Id: I3230be719659fe9cdf9ed6ae73bc91b05093ab97
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74430
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Parts of this file were still a copy of the file from the Mendocino SoC,
so update the file to match the PPR #57019 Rev 3.03 and the chipset
devicetree of the Phoenix SoC. Phoenix has 4 GFX/GPP PCIe bridges/ports,
the numbering scheme of the GPP PCIe bridges/ports was changed so that
the numbers match the device and function numbers, and there are new
device functions for the IPU and the USB4 controller and router devices.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Ie9429c03839bb0199a04cd6cafe9a955ebdacc91
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74565
Reviewed-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-by: Fred Reitberger <reitbergerfred@gmail.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
CBFS library performs memory mapped access of the files during loading,
verification and de-compression. Even with MTRRs configured correctly,
first few file access through memory map are taking longer times to
load. Update the SPI DMA driver to load the files into CBFS cache, so
that they can be verified and de-compressed with less overhead. This
saves ~60 ms in boot time.
BUG=None
TEST=Build Skyrim BIOS image and boot to OS. Observe ~60 ms improvement
with the boot time. Performing additional test to confirm there are no
regressions.
Before:
=======
970:loading FSP-M
15:starting LZMA decompress (ignore for x86) 760,906 (60,035)
16:finished LZMA decompress (ignore for x86) 798,787 (37,881)
8:starting to load ramstage
17:starting LZ4 decompress (ignore for x86) 1,050,093 (13,790)
18:finished LZ4 decompress (ignore for x86) 1,054,086 (3,993)
971:loading FSP-S
17:starting LZ4 decompress (ignore for x86) 1,067,778 (3,313)
18:finished LZ4 decompress (ignore for x86) 1,068,022 (244)
90:starting to load payload
17:starting LZ4 decompress (ignore for x86) 1,302,155 (11,285)
18:finished LZ4 decompress (ignore for x86) 1,303,938 (1,783)
After:
======
970:loading FSP-M
15:starting LZMA decompress (ignore for x86) 709,542 (12,178)
16:finished LZMA decompress (ignore for x86) 739,379 (29,837)
8:starting to load ramstage
17:starting LZ4 decompress (ignore for x86) 1,001,316 (12,368)
18:finished LZ4 decompress (ignore for x86) 1,001,971 (655)
971:loading FSP-S
17:starting LZ4 decompress (ignore for x86) 1,016,514 (3,031)
18:finished LZ4 decompress (ignore for x86) 1,016,722 (207)
90:starting to load payload
17:starting LZ4 decompress (ignore for x86) 1,244,602 (10,313)
18:finished LZ4 decompress (ignore for x86) 1,244,831 (228)
Change-Id: Ie30b6324f9977261c60e55ed509e979ef290f1f1
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74334
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-by: Tim Van Patten <timvp@google.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Boards with SOC_INTEL_COMMON_BLOCK_ACPI_CPU_HYBRID have
special handling for the time being.
Change of aopen/dxplplusu is coupled with sb/intel/i82801dx.
Change of emulation/qemu-i440fx is coupled with intel/i82371eb.
For asus/p2b, this adds MADT LAPIC entries, even though platform
has ACPI_NO_MADT selected. Even previously ACPI_NO_MADT creates
the MADT, including an entry for LAPIC address.
Change-Id: I1f8d7ee9891553742d73a92b55a87c04fa95a132
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74316
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Printing the value of a variable is not informative for a normal user,
so decrease the value from BIOS_INFO to BIOS_DEBUG.
Fixes: b9caac74a3 ("soc/amd/mendocino: Reinterpret smu_power_and_thm_limit")
Change-Id: I22f6293fd47633dfdbdae37b7257f47a5a4bb29c
Signed-off-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74271
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Van Patten <timvp@google.com>
PCIe bridges need to provide the LTR (latency tolerance reporting)
maximum snoop/non-snoop values so that they are inherited by downstream
PCIe devices which support and enable LTR. Without this, downstream
devices cannot have LTR enabled, which is a requirement for supporting
PCIe L1 substates. Enabling L1ss without LTR has unpredictable behavior,
including some devices refusing to enter L1 low power modes at all.
Program the max snoop/non-snoop latency values for all PCIe bridges
using the same value used by AGESA/FSP, 1.049ms.
BUG=b:265890321
TEST=build/boot google/skyrim (multiple variants, NVMe drives), ensure
LTR is enabled, latency values are correctly set, and that device
power draw at idle is in the expected range (<25 mW).
Change-Id: Icf188e69cf5676be870873c56d175423d16704b4
Signed-off-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74288
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-by: Fred Reitberger <reitbergerfred@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Restrict DPTC to 15W boards, since we only have 15W values defined in
the devicetree. This will revert the 6W boards back to their default
values, rather than (incorrectly) configuring them with 15W values.
BUG=b:253301653
TEST=Verify DPTC values are set for 15W boards
TEST=Verify DPTC values are set not set for 6W boards
Change-Id: I94f3974fce6358e3cbb0c30c1af33eb7ecb29ad7
Signed-off-by: Tim Van Patten <timvp@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74127
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The FSP will return the TDP in the format 0xX0000, where 'X' is the
value we're interested in. For example: 0xF0000 (15W), 0x60000 (6W).
Re-interpret the value so the caller just sees the TDP directly, without
needing to re-interpret things themselves.
BUG=b:253301653
TEST=Manually verify value is correct
Change-Id: I632e702d986a4ac85605040e09c1afab2bbdc59d
Signed-off-by: Tim Van Patten <timvp@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74126
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The i2c.c compilation unit is added to all stages in all cases, so use
the all target instead of adding it to all stages separately. Also order
the all targets alphabetically.
TEST=Timeless build on Mandolin results in identical image.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Ie90380075a3c87d226cdcb0f41f7e94275eaaa42
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74149
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Fred Reitberger <reitbergerfred@gmail.com>
tsc_freq.c gets built into all stages, but the tsc_freq_mhz function it
implements calls the get_pstate_0_reg function which was only built into
ramstage. Since tsc_freq_mhz was only called in ramstage, commit
2323acab6a ("soc/amd/stoneyridge: implement and use get_pstate_0_reg")
didn't cause the build to fail, but better factor out the P-state-
related utility functions into a separate compilation unit and include
it in all stages that also include tsc_freq.c.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Id3a3ee218f495be5e60a888944487704e7e8a1a1
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74145
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
monotonic_timer.c, tsc_freq.c and uart.c get added to all stage targets,
so just add those to the all stage targets. They still need to be added
to the smm stage target, since the all target doesn't add things to the
smm stage.
TEST=Timeless build results in identical image for Gardenia.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: I16c02bc0ff54553f212b94d110abef6a7bdedbb4
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74144
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
All AMD SoCs with Zen-based CPU cores are already using timestamps based
on the TSC counter, so use the existing common infrastructure instead of
reimplementing it in a similar way.
The behavior of the code changes slightly, but results in identical
timestamps. The timestamp_get implementation in soc/amd/common/block/cpu
divided the result of rdtscll() in timestamp_get by the result of
tsc_freq_mhz() and didn't override the weak timestamp_tick_freq_mhz
implementation that returns 1. The non AMD specific code returns the
result of rdtscll() in timestamp_get, but returns tsc_freq_mhz() instead
of 1 in timestamp_tick_freq_mhz, so we still get the correct timestamps.
TEST=The raw timestamps printed on the serial console are now multiplied
by the expected factor of the TSC frequency in MHz.
TEST=Normalized timestamps printed on the serial console by the x86 code
don't change significantly on Mandolin when comparing before and after
this patch. A slight variation in the timestamps is expected. An example
would be:
Before: CPU_CLUSTER: 0 init finished in 630 msecs
After: CPU_CLUSTER: 0 init finished in 629 msecs
TEST=The calculations of the time spent in verstage on PSP before
entering the bootblock on Guybrush result in similar times when
multiplying the value before the patch with the TSC frequency in the
case with the patch applied. The raw values printed on the serial
console by the verstage on PSP use the 1us time base, but the timestamp
logs that end up in CBMEM will be fixed up to use the same time base as
the x86 part of coreboot.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: I57b732e5c78222d278d3328b26bb8decb8f4783e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74016
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
In order for the code to find the correct VBIOS file in CBFS, remap the
revision ID in the RAVEN2_VBIOS_VID_DID case to the one that matches the
CBFS file name. This will make the code work as expected on devices with
the PCI ID RAVEN2_VBIOS_VID_DID and a revision != RAVEN2_VBIOS_REV.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: I94412dc2e778e7c4f74e475cd49114a00a81b2ce
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74045
Reviewed-by: Matt DeVillier <matt.devillier@amd.corp-partner.google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Instead of using the PSTATE SSDT generated by binaryPI, use the common
AMD code by selecting SOC_AMD_COMMON_BLOCK_ACPI_CPU_POWER_STATE. To
match the SSDT from binaryPI, set ACPI_SSDT_PSD_INDEPENDENT to n. There
are two differences to the binaryPI SSDT: Now coreboot includes the C1
state in the _CST package instead of just having the kernel add this due
to the ACPI_FADT_C1_SUPPORTED bit being set and the address of the
PS_STS_REG P state status MSR is written to the corresponding field of
the _PCT package instead of being 0.
TEST=On Careena the new P and C state ACPI packages are nearly identical
to the ones from the SSDT from binaryPI with the two functional
differences mentioned above.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Change-Id: Icdf6bc8f0e0363f185a294ab84edcb51322e7eb7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/74023
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martin.roth@amd.corp-partner.google.com>
