This CL has changes that allow us to enable a configurable
ramstage, and one change that allows us to minimize PCI
scanning. Minimal scanning is a frequently requested feature.
To enable it, we add two new variables to src/Kconfig
CONFIGURABLE_RAMSTAGE
is the overall variable controlling other options for minimizing the
ramstage.
MINIMAL_PCI_SCANNING is how we indicate we wish to enable minimal
PCI scanning.
Some devices must be scanned in all cases, such as 0:0.0.
To indicate which devices we must scan, we add a new mandatory
keyword to sconfig
It is used in place of on, off, or hidden, and indicates
a device is enabled and mandatory. Mandatory
devices are always scanned. When MINIMAL_PCI_SCANNING is enabled,
ONLY mandatory devices are scanned.
We further add support in src/device/pci_device.c to manage
both MINIMAL_PCI_SCANNING and mandatory devices.
Finally, to show how this works in practice, we add mandatory
keywords to 3 devices on the qemu-q35.
TEST=
1. This is tested and working on the qemu-q35 target.
2. On CML-Hatch
Before CL:
Total Boot time: ~685ms
After CL:
Total Boot time: ~615ms
Change-Id: I2073d9f8e9297c2b02530821ebb634ea2a5c758e
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/36221
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jeremy Soller <jeremy@system76.com>
This update is a combination of all four of the patches so that the
commit can be done without breaking parts of coreboot. This possible
breakage is because of the cross-dependencies between the original
separate patches would cause failure because of data structure changes.
security/intel/stm
This directory contains the functions that check and move the STM to the
MSEG, create its page tables, and create the BIOS resource list.
The STM page tables is a six page region located in the MSEG and are
pointed to by the CR3 Offset field in the MSEG header. The initial
page tables will identity map all memory between 0-4G. The STM starts
in IA32e mode, which requires page tables to exist at startup.
The BIOS resource list defines the resources that the SMI Handler is
allowed to access. This includes the SMM memory area where the SMI
handler resides and other resources such as I/O devices. The STM uses
the BIOS resource list to restrict the SMI handler's accesses.
The BIOS resource list is currently located in the same area as the
SMI handler. This location is shown in the comment section before
smm_load_module in smm_module_loader.c
Note: The files within security/intel/stm come directly from their
Tianocore counterparts. Unnecessary code has been removed and the
remaining code has been converted to meet coreboot coding requirements.
For more information see:
SMI Transfer Monitor (STM) User Guide, Intel Corp.,
August 2015, Rev 1.0, can be found at firmware.intel.com
include/cpu/x86:
Addtions to include/cpu/x86 for STM support.
cpu/x86:
STM Set up - The STM needs to be loaded into the MSEG during BIOS
initialization and the SMM Monitor Control MSR be set to indicate
that an STM is in the system.
cpu/x86/smm:
SMI module loader modifications needed to set up the
SMM descriptors used by the STM during its initialization
Change-Id: If4adcd92c341162630ce1ec357ffcf8a135785ec
Signed-off-by: Eugene D. Myers <edmyers@tycho.nsa.gov>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33234
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: ron minnich <rminnich@gmail.com>
This change adds support for allocating resources for PCI express hotplug
bridges when PCIEXP_HOTPLUG is selected. By default, this will add 32 PCI
subordinate numbers (buses), 256 MiB of prefetchable memory, 8 MiB of
non-prefetchable memory, and 8 KiB of I/O space to any device with the
PCI_EXP_SLTCAP_HPC bit set in the PCI_EXP_SLTCAP register, which
indicates hot-plugging capability. The resource allocation is configurable,
please see the PCIEXP_HOTPLUG_* variables in src/device/Kconfig.
In order to support the allocation of hotplugged PCI buses, a new field
is added to struct device called hotplug_buses. This is defaulted to
zero, but when set, it adds the hotplug_buses value to the subordinate
value of the PCI bridge. This allows devices to be plugged in and
unplugged after boot.
This code was tested on the System76 Darter Pro (darp6). Before this
change, there are not enough resources allocated to the Thunderbolt
PCI bridge to allow plugging in new devices after boot. This can be
worked around in the Linux kernel by passing a boot param such as:
pci=assign-busses,hpbussize=32,realloc
This change makes it possible to use Thunderbolt hotplugging without
kernel parameters, and attempts to match closely what our motherboard
manufacturer's firmware does by default.
Signed-off-by: Jeremy Soller <jeremy@system76.com>
Change-Id: I500191626584b83e6a8ae38417fd324b5e803afc
Reviewed-on: https://review.coreboot.org/c/coreboot/+/35946
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The names of each spi flash cause quite a bit of bloat in the text
size of each stage/program. Remove the name entirely from spi flash
in order to reduce overhead. In order to pack space as closely as
possible the previous 32-bit id and mask were split into 2 16-bit
ids and masks.
On Chrome OS build of Aleena there's a savings of >2.21KiB in each
of verstage, romstage, and ramstage.
Change-Id: Ie98f7e1c7d116c5d7b4bf78605f62fee89dee0a5
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38380
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
This patch creates a new commonlib/bsd subdirectory with a similar
purpose to the existing commonlib, with the difference that all files
under this subdirectory shall be licensed under the BSD-3-Clause license
(or compatible permissive license). The goal is to allow more code to be
shared with libpayload in the future.
Initially, I'm going to move a few files there that have already been
BSD-licensed in the existing commonlib. I am also exracting most
contents of the often-needed <commonlib/helpers.h> as long as they have
either been written by me (and are hereby relicensed) or have an
existing equivalent in BSD-licensed libpayload code. I am also
relicensing <commonlib/compression.h> (written by me) and
<commonlib/compiler.h> (same stuff exists in libpayload).
Finally, I am extracting the cb_err error code definitions from
<types.h> into a new BSD-licensed header so that future commonlib/bsd
code can build upon a common set of error values. I am making the
assumption here that the enum constants and the half-sentence fragments
of documentation next to them by themselves do not meet the threshold of
copyrightability.
Change-Id: I316cea70930f131e8e93d4218542ddb5ae4b63a2
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38420
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
When vboot was first integrated into CBFS it was still part of Google
vendorcode. So to not directly tie custom vendorcode into the core CBFS
library, the concept of cbfs_locator was introduced to decouple core
code from an arbitrary amount of platform-specific implementations that
want to decide where the CBFS can be found.
Nowadays vboot is a core coreboot feature itself, and the locator
concept isn't used by anything else anymore. This patch simplifies the
code by removing it and just calling vboot from the CBFS library
directly. That should make it easier to more closely integrate vboot
into CBFS in the future.
Change-Id: I7b9112adc7b53aa218c58b8cb5c85982dcc1dbc0
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38419
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
By grouping the spi flash parts by their {vendor, sector topology}
tuple one can use a common probe function for looking up the part
instead of having per-vendor probe functions. Additionally, by
grouping by the command set one can save more space as well. SST
is the exception that requires after_probe() function to unlock the
parts.
2KiB of savings in each of verstage, romstage, and ramstage
on Aleena Chrome OS Build.
Change-Id: I9cc20ca0f3d0a1b97154b000c95ff2e7e87f3375
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38379
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
To further drive to a common approach for describing the spi flash
parts in the drivers add spi_flash_part_id object. All the drivers
are updated to utilize the new object. Additionally, the driver_private
is also not needed in the spi_flash object.
A Chrome OS build of Aleena provides 960 byte saving of text. A subsequent
patch will save more memory.
Change-Id: I9c0cc75f188ac004ab647805b9551bf06a0c646b
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38378
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Put the write protection into its own object. This allows
for easier future reuse of objects in future consolidation
patches. It's also possible to eliminate the code implmementing
these in the future if the platform doesn't require it. For now
leave current behavior as-is.
The names of the callbacks were shortened as they are now in
the spi_flash_protection_ops object which is a new field in the
spi_flash object.
Change-Id: I2fec4e4430709fcf3e08a55dd36583211c035c08
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38376
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
In order to provide more consistent probing in future refactorings, pull out
the release from deep sleep path in STMicro's SPI flash probing function.
Call that function explicitly when RDID doesn't return anything at all.
The old STMicro parts, even if supporting RDID, won't decode that
instruction while in a deep power down state. Instead of re-issuing RDID after
the successful wake assume the id fixup is valid.
Change-Id: I46c47abcfb1376c1c3ce772f6f232857b8c54202
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38167
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Also change some of the types to match the register widths
of the controller. It is expected that these prototypes
will be used with SMBus host controllers inside AMD chipsets
as well, thus the change of location.
Change-Id: I88fe834f3eee7b7bfeff02f91a1c25bb5aee9b65
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38226
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Only smbuslib.c and spd_bin.c share the same prototypes for SMBUS
functions. Therefore, get_spd_smbus() currently only works with
soc/intel/.../smbuslib.c and can be implemented there locally.
This allows removal of <device/early_smbus.h>.
Change-Id: Ic2d9d83ede6388a01d40c6e4768f6bb6bf899c00
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/38121
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
The SPI flashes that support page programming mode had duplicated
the logic for writing in every driver. Add
spi_flash_cmd_write_page_program() and use the common implementation
to reduce code size that comes from duplication. The savings is
~2.5KiB per stage where the spi flash drivers are utilized.
Change-Id: Ie6db03fa8ad33789f1d07a718a769e4ca8bffe1d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/37963
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
There was one user of SPI_FLASH_SECTOR_ERASE_TIMEOUT_MS,
southbridge/intel/common/spi.c. Remove the define and encode
the 1 second timeout that it was wanting at the single use site.
Change-Id: If33a1a04bc4d3441e90bf0ca305ddf71c4f8bb88
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/37962
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Implement a free() that supports only the last malloc(). Rewind
the heap to the last allocation point if the ptr to be freed is
matching the end of heap before last malloc(). With current situation,
since free() is no-op, every call to malloc() is a memory leak.
BUG=b:140124451
TEST=Wrote a test function to do malloc and free operations.
Change-Id: I6d43cf54b79e6897cf6882335730b2310e4eae45
Signed-off-by: Bora Guvendik <bora.guvendik@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/37919
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
