Change the cache-as-ram teardown to use invd instead of wbinvd.
Save the return and recover the call's return address in
chipset_teardown_car.
CAR teardown had been modified to use wbinvd to send CAR contents
to DRAM backing prior to teardown. This allowed CAR variables,
stack, and local variables to be preserved while running the
AMD_DISABLE_STACK macro.
Using the wbinvd instruction has the side effect of sending all
dirty cache contents to DRAM and not only our CAR data. This
would likely cause corruption, e.g. during S3 resume.
Stoney Ridge now uses a postcar stage and this is no longer a
requirement.
BUG=b:64768556
Change-Id: I8e6bcb3947f508b1db1a42fd0714bba70074837a
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/20967
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Insert a postcar stage for Stoney Ridge and move romstage's CAR
teardown there.
The AMD cache-as-ram teardown procedure currently uses a wbinvd
instruction to send CAR contents to DRAM backing. This allows
preserving stack contents and CAR globals after the teardown
happens, but likely results in memory corruption during S3 resume.
Due to the current base of the DCACHE region, reverting to an
invd instruction will break the detection mechanism for CAR
migrated variables. Using postcar avoids this problem.
The current behavior of AGESA is to set up all cores' MTRRs during
the AmdInitPost() entry point. This implementation takes control
back and causes postcar's _start to clear all settings and set
attributes only for the BIOS flash device, TSEG, and enough space
below cbmem_top to load and run ramstage.
BUG=b:64768556
Change-Id: I1045446655b81b806d75903d75288ab17b5e77d1
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/20966
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Relocate the call to AGESA in preparation for implementing postcar.
This change should have no net effect as long as the ordering is
maintained and AmdInitEnv stays later than CAR teardown.
BUG=b:66196801
Change-Id: I0e4a5fd979b06cf50907c62d51e55db63c5e00c5
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21613
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Now that soc/amd supports EARLY_CBMEM_INIT, put the HEAP into cbmem,
allowing better control of its cacheability in subsequent patches.
This relocates the heap initialization from the common directory into
a romstage cbmem hook. The conversion relies on cbmem_add() first
searching cbmem for the ID before adding a new entry.
Change-Id: I9ff35eefb2a68879ff44c6e29f58635831b19848
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21594
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Add an extra include file to northbridge.c for completeness. cpu/msr.h
is already included in cpu/amd/mtrr.h which allows the file to build
properly.
Remove include files that are no longer required for the file.
Change-Id: I3e5ab39fd0640d2983fc5b7b202fb65d42c5ce3d
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21615
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Add necessary features to allow mp_init_with_smm() to install and
relocate an SMM handler.
SMM region functions are added to easily identify the SMM attributes.
Adjust the neighboring cbmem_top() rounding downward to better reflect
the default TSEG size. Add relocation attributes to be set by each
core a relocation handler.
Modify the definition of smi_southbridge_handler() to match TSEG
prototype.
BUG=b:62103112
Change-Id: I4dc03ed27d0d109ab919a4f0861de9c7420d03ce
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21501
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Change the Stoney Ridge SOC to a more modern method for setting up
the multiple cores.
Add a new cpu.c file for most of the processor initiliazation. Build
mp_ops with the necessary callbacks. Note also that this patch removes
cpu_bus_scan. Rather than manually find CPUs and add them to the
devicetree, allow this to be done automatically in the generic
mp_init.c file.
SMM information is left blank in mp_ops to avoid having mp_init.c
install a handler at this time. A later patch will add TSEG SMM
capabilities for the APU.
This patch also contains a hack to mask the behavior of AGESA which
configures the MTRRs and Tom2ForceMemTypeWB coming out of AmdInitPost.
The hack immediately changes all WB variable MTRRs, on the BSP, to UC
so that all writes to memory space will make it to the DRAM.
BUG=b:66200075
Change-Id: Ie54295cb00c6835947456e8818a289b7eb260914
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21498
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
- Call into commmon SGX code to configure core PRMRR and follow
other SGX init seqeuence.
- Enable SOC_INTEL_COMMON_BLOCK_SGX for both GLK
- Enable SOC_INTEL_COMMON_BLOCK_CPU_MPINIT for GLK, as MP init needs
to be completed before calling into fsp-s for SGX.
Change-Id: I9331cf5b2cbc86431e2749b84a55f77f7f3c5960
Signed-off-by: Pratik Prajapati <pratikkumar.v.prajapati@intel.com>
Reviewed-on: https://review.coreboot.org/21276
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Subrata Banik <subrata.banik@intel.com>
Use a single define and set the CROS GPIO ASL device to match the
Stoney Ridge GPIO HID. Update the GPIO number to 142. Also, add a DDN
field in the GPIO ASL. This addresses the TEST indicated below.
BUG=b:65597554
BRANCH=none
TEST=grep ^ /sys/devices/platform/chromeos_acpi/GPIO.*/* reports AMD0030.
Change-Id: I1d6c42c6c9a0eef25e0e99aed6d838c767f5e01f
Signed-off-by: Marc Jones <marcj303@gmail.com>
Reviewed-on: https://review.coreboot.org/21614
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Refactor the GPIO functions to use GPIO numbers. This is more consistent
with other GPIO code in coreboot.
BUG=b:66462235
BRANCH=none
TEST=Build and boot Kahlee
Change-Id: I6d6af7f6a0ed9ba1230342e1ca024535c4f34d47
Signed-off-by: Marc Jones <marcj303@gmail.com>
Reviewed-on: https://review.coreboot.org/21684
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
I triggered a bug, when I try to debug riscv code by spike.
This bug is caused by an instruction exception[csrwi 0x320,7].
This is operate for mcounteren. This address is error. 0x306
is right. scounteren is not need to be set, because S-mode
code controls it.
Change-Id: If0bea4bf52d8ad2fb2598724d6feb59dc1b3084a
Signed-off-by: wxjstz<wxjstz@126.com>
Reviewed-on: https://review.coreboot.org/20043
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
The EXT_CONF_SUPPORT symbol doesn't exist for the Stoney Ridge SoC.
Clean up northbridge.c by removing the check for the config value set.
Remove the CPU initialization code that clears the EnableCf8ExtCfg
bit. The location where it was set was removed in
c1d72942 Disable PCI_CFG_EXT_IO
BUG=b:66202622
Change-Id: Ic58c47fc5f568d17f5027c96d4152b0e5b3e1d14
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21497
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Delete the obselete Kconfig symbols regarding the memory hole. Integrate
the hole check into domain_set_resources(). The hardware configuration
is done by AGESA, so only discover the setting and adjust the mmio_basek
accordingly.
BUG=b:66202887
TEST=Check settings with HDT
Change-Id: Id15a88897e29bff28ab7c498dc4d3818834f08b2
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/21496
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
A regression is hidden in commit 7eb0157fca
(device/dram/ddr2.c: Decoding byte[12] bit7 as self refresh flag ), which
breaks the build procedure when CONFIG_DEBUG_RAM_SETUP is set.
Maybe it had better implement "printram" and the like as inline
functions instead of macros.
Change-Id: If956435bd0c39b1f3e722c2cfc48d2d0f35abc9b
Signed-off-by: Bill XIE <persmule@gmail.com>
Reviewed-on: https://review.coreboot.org/21673
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
New AGESA support files will be used for binaryPI
platforms as well. Furthermore, some of those should
move from split nb/ sb/ directories to soc/, so move
support files for the API under drivers/.
Change-Id: I549788091de91f61de8b9adc223d52ffb5732235
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/21455
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Except for family15, all AGESA boards have moved
away from AGESA_LEGACY_WRAPPER, thus they all
have POSTCAR_STAGE now.
AGESA family15 boards remain at AGESA_LEGACY=y, but
those boards have per-board romstage.c files and
are not touched here.
Change-Id: If750766cc7a9ecca4641a8f14e1ab15e9abb7ff5
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/18632
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Move all boards that have moved away from AGESA_LEGACY_WRAPPER
or BINARYPI_LEGACY_WRAPPER to use POSTCAR_STAGE.
We use POSTCAR_STAGE as a conditional in CAR teardown to tell
our MTRR setup is prepared such that invalidation without
writeback is a valid operation.
Change-Id: I3f4e2170054bdb84c72d2f7c956f8d51a6d7f0ca
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/21384
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
This file mostly mimics Porting.h and should be removed.
For now, move it and use it consistently with incorrect form
as #include "cbtypes.h".
Change-Id: Ifaee2694f9f33a4da6e780b03d41bdfab9e2813e
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/21663
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Fix regression of IDS debugging after commit
1210026 AGESA buildsystem: Reduce include path exposure
Mainboard directory was removed from libagesa includes
path here, and this resulted with fam15tn and fam16kb using
a template OptionsIds.h file under vendorcode/ instead.
Add mainboard directory back to include path of libagesa
and remove those (empty) template files.
Change-Id: Iee4341a527b4c152269565cac85e52db44503ea6
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/21661
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
It's already implemented like this with binaryPI API header.
That implementation is essentially the same with 'const' qualifier
just being ignored in the build process for PI blob.
For open-source AGESA build, work around -Werror=discarded-qualifier
using a simple but ugly cast.
Change-Id: Ib84eb9aa40f1f4442f7aeaa8c15f6f1cbc6ca295
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/21630
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Since snappy PCB may have over 9 SKU and current GPIO board ID GP16/GP17
is insufficient to use.
Using VPD to control could prevent H/W change.
BUG=b:65339688
BRANCH=reef
TEST=emerge-snappy coreboot
Change-Id: I55ab741354797e022dd945da9c8499ee5e041316
Signed-off-by: Kevin Chiu <Kevin.Chiu@quantatw.com>
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/21533
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
variant_board_sku() callback exists to allow some of the variants to
report the sku id differently based on board implementation. However,
there are cases where there are multiple ways to encode the sku id, but
the original way should be used as a fallback. As such expose a helper
function, sku_strapping_value(), such that there isn't code duplication
for the common fallback case.
BUG=b:65339688
Change-Id: I1e917733eb89aebc41a483e2001a02acfda31bf4
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/21645
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
