This replaces all occurrences of a hardcoded vboot path to the
VBOOT_SOURCE variable, that may be overridden from the command line,
witch fallback to the source from 3rdparty.
Change-Id: Ia57d498d38719cc71e17060b76b0162c4ab363ed
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Reviewed-on: https://review.coreboot.org/15825
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Use the ACPI generator for creating the Chrome OS gpio
package. Each mainboard has its own list of Chrome OS
gpios that are fed into a helper to generate the ACPI
external OIPG package. Additionally, the common
chromeos.asl is now conditionally included based on
CONFIG_CHROMEOS.
Change-Id: I1d3d951964374a9d43521879d4c265fa513920d2
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/15909
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
1. Currenty, boot reason is being added to elog only for some
ARM32/ARM64 platforms. Change this so that boot reason is logged by
default in elog for all devices which have CHROMEOS selected.
2. Add a new option to select ELOG_WATCHDOG_RESET for the devices that
want to add details about watchdog reset in elog. This requires a
special region WATCHDOG to be present in the memlayout.
3. Remove calls to elog add boot reason and watchdog reset from
mainboards.
BUG=chrome-os-partner:55639
Change-Id: I91ff5b158cfd2a0749e7fefc498d8659f7e6aa91
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15897
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
With VBOOT_VERIFY_FIRMWARE separated from CHROMEOS, move recovery and
developer mode check functions to vboot. Thus, get rid of the
BOOTMODE_STRAPS option which controlled these functions under src/lib.
BUG=chrome-os-partner:55639
Change-Id: Ia2571026ce8976856add01095cc6be415d2be22e
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15868
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
VBOOT_VERIFY_FIRMWARE should be independent of CHROMEOS. This allows use
of verified boot library without having to stick to CHROMEOS.
BUG=chrome-os-partner:55639
Change-Id: Ia2c328712caedd230ab295b8a613e3c1ed1532d9
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15867
Tested-by: build bot (Jenkins)
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
This renames the VB_SOURCE variable to VBOOT_SOURCE in the build system,
providing increased clarity about what it represents.
Since the submodule itself is called "vboot", it makes sense to use that
name in full instead of a very shortened (and confusing) version of it.
Change-Id: Ib343b6642363665ec1205134832498a59b7c4a26
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Reviewed-on: https://review.coreboot.org/15824
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
On some x86 platforms (skylake, apollolake), we observe reboots at
different steps during the FSP initialization. These additional reboots
result in loss of recovery request because vboot_reference library
clears recovery request on vbnv once verification is complete and it has
made a decision about which boot path to take(normal/dev, slot-a/slot-b,
recovery).
Provide a way to allow mainboards/chipsets to inform recovery module in
vboot2 to save recovery reason to survive unexpected reboots. The
recovery reason is set in vbnv after vboot_reference library completes
its verification and clears the reason in vbnv while jumping to
payload.
BUG=chrome-os-partner:55431
Change-Id: Ie96be9aeb42c8209d8215943409e6327d6a8bf98
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15802
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Add recovery module in vboot2 that checks if a recovery request is
pending and returns appropriate reason code:
1. Checks if recovery mode is initiated by EC.
2. Checks if recovery request is present in VBNV.
3. Checks if recovery request is present in handoff for post-cbmem
stages.
4. Checks if vboot verification is complete and looks up selected region
to identify if recovery is requested by vboot library.
BUG=chrome-os-partner:55431
Change-Id: I31e332a4d014a185df2434c3730954e08dc27281
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15800
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
1. Remove unused functions/structures.
2. Add checks for NULL return values.
3. Change prefixes to vb2 instead of vboot for functions used internally
within vboot2/
4. Get rid of vboot_handoff.h file and move the structure definition to
vboot_common.h
5. Rename all functions using handoff structure to have prefix
vboot_handoff_*. All the handoff functions can be run _only_ after cbmem
is online.
6. Organize vboot_common.h content according to different
functionalities.
BUG=chrome-os-partner:55431
Change-Id: I4c07d50327d88cddbdfbb0b6f82c264e2b8620eb
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15799
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
This fills up the RO_FRID, RW_FWID_A and RW_FWID_B FMAP sections with
the relevant version from KERNELVERSION, padded to the right size and
gap-filled with zeros.
Change-Id: I45c724555f8e41be02b92ef2990bf6710be805c2
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Reviewed-on: https://review.coreboot.org/15604
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
TPM1.2 is using the somewhat misnamed tlcl_set_global_lock() command
function to lock the hardware rollback counter. For TPM2 let's
implement and use the TPM2 command to lock an NV Ram location
(TPM2_NV_WriteLock).
BRANCH=none
BUG=chrome-os-partner:50645
TEST=verified that TPM2_NV_WriteLock command is invoked before RO
firmware starts RW, and succeeds.
Change-Id: I52aa8db95b908488ec4cf0843afeb6310dc7f38b
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 2f859335dfccfeea900f15bbb8c6cb3fd5ec8c77
Original-Change-Id: I62f22b9991522d4309cccc44180a5ebd4dca488d
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/358097
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Darren Krahn <dkrahn@chromium.org>
Reviewed-on: https://review.coreboot.org/15638
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
tlcl_force_clear() needs to be issued each time when the device mode
switches between normal/development/recovery.
This patch adds command implementation using TPM_Clear TPM2 command,
and also invokes it before factory initialization.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=verified that TPM_Clear command succeeds at factory startup and
the boot proceeds normally.
Change-Id: Ia431390870cbe448bc1b6f1755ed17953be9bdf1
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 347ff17b97da45fa4df547ff32f9dd2c8972cefd
Original-Change-Id: I2a0e62527ad46f9dd060afe5e75c7e4d56752849
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/358095
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Darren Krahn <dkrahn@chromium.org>
Reviewed-on: https://review.coreboot.org/15636
Tested-by: build bot (Jenkins)
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
The TPM2 specification allows defining NV ram spaces in a manner
that makes it impossible to remove the space until a certain PCR is in
a certain state.
This comes in handy when defining spaces for rollback counters: make
their removal depend on PCR0 being in the default state. Then extend
PCR0 to any value. This guarantees that the spaces can not be deleted.
Also, there is no need t create firmware and kernel rollback spaces
with different privileges: they both can be created with the same set of
properties, the firmware space could be locked by the RO firmware, and
the kernel space could be locked by the RW firmware thus providing
necessary privilege levels.
BRANCH=none
BUG=chrome-os-partner:50645, chrome-os-partner:55063
TEST=with the rest of the patches applied it is possible to boot into
Chrome OS maintaining two rollback counter spaces in the TPM NV
ram locked at different phases of the boot process.
Change-Id: I889b2c4c4831ae01c093f33c09b4d98a11d758da
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 36317f5e85107b1b2e732a5bb2a38295120560cd
Original-Change-Id: I69e5ada65a5f15a8c04be9def92a8e1f4b753d9a
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/358094
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/15635
Tested-by: build bot (Jenkins)
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
This patch adds a TPM2 specific path in the vboot2 initialization
sequence when the device is turned on in the factory for the first
time, namely two secure NVRAM spaces are created, with different
access privileges.
The higher privilege space can be modified only be the RO firmware,
and the lower privilege space can be modified by both RO and RW
firmware.
The API is being modified to hide the TPM implementation details from
the caller.
Some functions previously exported as global are in fact not used
anywhere else, they are being defined static.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=when this code is enabled the two secure spaces are successfully
created during factory initialization.
Original-Commit-Id: 5f082d6a9b095c3efc283b7a49eac9b4f2bcb6ec
Original-Change-Id: I917b2f74dfdbd214d7f651ce3d4b80f4a18def20
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/353916
Original-Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Original-Reviewed-by: Darren Krahn <dkrahn@chromium.org>
squashed:
mock tpm: drop unused functions
safe_write() and safe_define_space() functions are defined in
secdata_mock.c, but not used in mocked TPM mode.
The actual functions have been redefined as static recently and their
declarations were removed from src/include/antirollback.h, which now
causes compilation problems when CONFIG_VBOOT2_MOCK_SECDATA is
defined.
Dropping the functions from secdata_mock.c solves the problem.
BRANCH=none
BUG=none
TEST=compilation in mock secdata mode does not fail any more.
Original-Commit-Id: c6d7824f52534ecd3b02172cb9078f03e318cb2b
Original-Change-Id: Ia781ce99630d759469d2bded40952ed21830e611
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/356291
Original-Tested-by: Caesar Wang <wxt@rock-chips.com>
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Randall Spangler <rspangler@chromium.org>
Change-Id: Icb686c5f9129067eb4bb3ea10bbb85a075b29955
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://review.coreboot.org/15571
Tested-by: build bot (Jenkins)
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
Reviewed-by: Furquan Shaikh <furquan@google.com>
If read_vbnv finds that the vbnv_copy is not valid, it initializes it
with the correct HEADER_SIGNATURE and other attributes. However, the
vbnv copy is checked for validity and initialized at the vboot layer as
well. Since, vboot is the owner of this data, it should be the one
initializing it. Thus, if read_vbnv sees that the data is not valid,
simply reset it to all 0s and let vboot layer take care of it. This also
removes the need for additional checks to ensure that the dirty vbnv
copy is properly updated on storage.
Change-Id: I6101ac41f31f720a6e357c9c56e571d62e0f2f47
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15498
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Currently, read_vbnv performs a reset of the vbnv cache if it is not
valid. However, this information is not passed up to the vboot layer,
thus resulting in missed write-back of vbnv cache to storage if vboot
does not update the cache itself.
Update read_vbnv to return a value depending upon whether it wants a
write-back to be performed when save is called.
Return value:
0 = No write-back required
1 = Write-back of VBNV cache is required.
Change-Id: I239939d5f9731d89a9d53fe662321b93fc1ab113
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/15457
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Until now it was assumed that all TPM devices were of the same type
(TCG 1.2 spec compliant) and x86 based boards had LPC connected TPMs
and all other boards had I2C connected TPMs.
With the advent of TPM2 specification there is a need to be able to
configure different combinations of TPM types (TPM or TPM2) and
interfaces (LPC, I2C and SPI).
This patch allows to do it. Picking Chrome OS still assumes that the
board has a TPM device, but adding MAINBOARD_HAS_TPM2 to the board's
Kconfig will trigger including of TPM2 instead.
MAINBOARD_HAS_LPC_TPM forces the interface to be set to LPC, adding
SPI_TPM to the board config switches interface choice to SPI, and if
neither of the two is defined, the interface is assumed to be I2C.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=verified that none of the generated board configurations change
as a result of this patch. With the rest of the stack in place it
is possible to configure different combinations of TPM types and
interfaces for ARM and x86 boards.
Change-Id: I24f2e3ee63636566bf2a867c51ed80a622672f07
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 5a25c1070560cd2734519f87dfbf401c135088d1
Original-Change-Id: I659e9301a4a4fe065ca6537ef1fa824a08d36321
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/349850
Original-Reviewed-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://review.coreboot.org/15294
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
The VBOOT_OPROM_MATTERS configuration option signals to vboot that the
board can skip display initialization in the normal boot path. It's name
is a left-over from a time when this could only happen by avoiding
loading the VGA option ROM on x86 devices. Now we have other
boards that can skip their native display initialization paths too, and
the effect to vboot is the same. (Really, we should rename oprom_matters
and oprom_loaded to display_skippable and display_initialized or
something, but I don't think that's worth the amount of repositories
this would need to touch.)
The only effect this still has in today's vboot is to reboot and
explicitly request display initialization for EC software sync on
VBOOT_EC_SLOW_UPDATE devices (which we haven't had yet on ARM). Still,
the vboot flag just declares the capability (for skipping display init),
and it should be set correctly regardless of whether that actually makes
a difference on a given platform (right now). This patch updates all
boards/SoCs that have a conditional path based on
display_init_required() accordingly.
BRANCH=None
BUG=chrome-os-partner:51145
TEST=Booted Oak, confirmed that there's no notable boot time impact.
Change-Id: Ic7c77dbd8356d67af7aee54e7869f9ac35241b99
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 9c242f7
Original-Change-Id: I75e5cdda2ba2d111ea50ed2c7cdf94322679f1cd
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/348786
Original-Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/15113
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Furquan Shaikh <furquan@google.com>
When the vboot cbfs selection runs in postcar stage it should be
utilizing cbmem to locate the vboot selected region.
Change-Id: I027ba19438468bd690d74ae55007393f051fde42
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14959
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Using a dedicated variable is slightly less readable and makes the code
less consistent, given that other test functions are called directly in
the if statements.
Change-Id: If52b2a4268acb1e2187574d15cc73a0c1d5fe9bb
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Reviewed-on: https://review.coreboot.org/14817
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
On modern x86 platforms like apollolake, pre-RAM stages verstage and
romstage run within the cache-as-ram region. Thus, we do not need to
pass in the --xip parameter to cbfstool while adding these
stages. Introduce a new Kconfig variable NO_XIP_EARLY_STAGES which is
default false for all x86 platforms. Apollolake selects this option
since it supports code execution with CAR.
Change-Id: I2848046472f40f09ce7fc230c258b0389851b2ea
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/14623
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
In the case where one of the FW_MAIN regions is empty, the last file
(empty) will also appear to be first and have a zero offset, making head
complain.
This is a very borderline use case, since the FW_MAIN_ regions should
have been filled previously, but an extra check doesn't hurt.
Change-Id: I15491c5b4a5e7d1f9fb369cc5fa4e3875e2dad3b
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Reviewed-on: https://review.coreboot.org/14472
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The codepath was untested and incomplete. It now determines the right
GBB region sizes and puts the data in.
BUG=chromium:595715
BRANCH=none
TEST=none
Change-Id: I2cc47ddd8aa7675375ca5ed5f75632c30c65dd1e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 36e026404ed049d61b677ef043a781c8c209dd93
Original-Change-Id: Ib872627740dbd8ac19fc3e2a01464457f38366ed
Original-Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/336358
Original-Commit-Ready: Patrick Georgi <pgeorgi@chromium.org>
Original-Tested-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14239
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The MT8173 hardware watchdog can assert an external signal which we use
to reset the TPM on Oak. Therefore we do not need to do the same
double-reset dance as on other Chromebooks to ensure that we reset in a
correct state.
Still, we have a situation where we need to reconfigure the watchdog
early in the bootblock in a way that will clear information about the
previous reboot from the status register, and we need that information
later in ramstage to log the right event. Let's reuse the same watchdog
tombstone mechanism from other boards, except that we don't perform a
second reset and the tombstone is simply used to communicate between
bootblock and ramstage within the same boot.
BRANCH=None
BUG=None
TEST=Run 'mem w 0x10007004 0x8' on Oak, observe how it reboots and how
'mosys eventlog list' shows a hardware watchdog reboot event afterwards.
Change-Id: I1ade018eba652af91814fdaec233b9920f2df01f
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 07af37e11499e86e730f7581862e8f0d67a04218
Original-Change-Id: I0b9c6b83b20d6e1362d650ac2ee49fff45b29767
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/334449
Original-Reviewed-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: https://review.coreboot.org/14234
Tested-by: build bot (Jenkins)
Reviewed-by: Julius Werner <jwerner@chromium.org>
When doing verification of memory init code in verstage vboot
should issue a TPM_Startup(ST_STATE) instead of TPM_Startup(ST_CLEAR)
in order to preserve the flags in TPM_STCLEAR_FLAGS which include
things like physical presence. In doing so we can also skip the rest
of the TPM init work in this function in the S3 resume path.
BUG=chrome-os-partner:50633
BRANCH=glados
TEST=S3 resume on chell and ensure TPM is resumed instead of being
cleared and that 'tpmc getvf|getpf|getf' does not show any difference
in flags between boot and resume.
Change-Id: I7a48eaf7f57d2bc6ebc182178cbe60ceb2ad8863
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: f059f39a0f5c2f21e564b9554efacf26a41ad794
Original-Change-Id: I647869202d2f04328764155d3de4cad9edf10ae4
Original-Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Original-Previous-Reviewed-on: https://chromium-review.googlesource.com/332434
Original-(cherry picked from commit 5fc7792e4104523569140cd84ce313da721ec34b)
Original-Reviewed-on: https://chromium-review.googlesource.com/332542
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14107
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
If a platform does verification of the memory init step, and it must
resume with the same slot that it booted from then it needs to set
the vboot context flag when resuming instead of booting. This will
affect the slot that is selected to verify and resume from.
BUG=chromium:577269
BRANCH=glados
TEST=manually tested on chell:
1) ensure that booting from slot A resumes from slot A.
2) ensure that booting from slot B resumes from slot B.
3) do RW update while booted from slot A (so the flags are set to try
slot B) and ensure that suspend/resume still functions properly using
current slot A.
4) do RW update while booted from slot B (so the flags are set to try
slot A) and ensure that suspend/resume still functions properly using
current slot B.
Change-Id: I77e6320e36b4d2cbc308cfb39f0d4999e3497be3
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Original-Commit-Id: 4c84af7eae7b2a52a28cc3ef8a80649301215a68
Original-Change-Id: I395e5abaccd6f578111f242d1e85e28dced469ea
Original-Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/328775
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/13834
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This patch ports the LZ4 decompression code that debuted in libpayload
last year to coreboot for use in CBFS stages (upgrading the base
algorithm to LZ4's dev branch to access the new in-place decompression
checks). This is especially useful for pre-RAM stages in constrained
SRAM-based systems, which previously could not be compressed due to
the size requirements of the LZMA scratchpad and bounce buffer. The
LZ4 algorithm offers a very lean decompressor function and in-place
decompression support to achieve roughly the same boot speed gains
(trading compression ratio for decompression time) with nearly no
memory overhead.
For now we only activate it for the stages that had previously not been
compressed at all on non-XIP (read: non-x86) boards. In the future we
may also consider replacing LZMA completely for certain boards, since
which algorithm wins out on boot speed depends on board-specific
parameters (architecture, processor speed, SPI transfer rate, etc.).
BRANCH=None
BUG=None
TEST=Built and booted Oak, Jerry, Nyan and Falco. Measured boot time on
Oak to be about ~20ms faster (cutting load times for affected stages
almost in half).
Change-Id: Iec256c0e6d585d1b69985461939884a54e3ab900
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/13638
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
These files aren't updated (or updatable), and as such don't need to be
copied to the RW sections.
Change-Id: Ie78936792ad651fbf8500fc7e34f0899e33a904c
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/13633
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This adds a new kconfig option that will back up the VBNV data
from CMOS to flash, and restore it if the CMOS data is invalid
during boot.
This allows special flags to not get lost when power is lost,
RTC reset is triggered, or CMOS is corrupted.
BUG=chrome-os-partner:47915
BRANCH=glados
TEST=manually tested on chell:
1-boot and run "enable_dev_usb_boot"
2-reboot and check that it is enabled with crossystem
3-run "mosys nvram clear"
4-reboot and check that it is still enabled
Change-Id: I38103d100117da34471734a6dd31eb7058735c12
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 8a356e616c6885d5ae3b776691929675d48a28f9
Original-Change-Id: I06e7ddff7b272e579c704914a0cf8cc14d6994e8
Original-Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/324122
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/13600
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Certain platforms query the recovery mode switch more than just within
vboot during the boot flow. Therefore, it's important that the first call to
get_recovery_mode_switch() is consistent through memory training because
certain platforms use the recovery mode switch to take different action
for memory training. Therefore, defer the clearing of the rec mode
switch to a place when it's known that memory is up and online.
BUG=chrome-os-partner:44827
BRANCH=glados
TEST=Three finger salute is honored on chell by retraining memory.
Change-Id: I26ea51de7ffa2fe75b9ef1401fe92f9aec2b4567
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 6b0de9369242e50c7ff3b164cf1ced0642c7b087
Original-Change-Id: Ia7709c7346d1222e314bf3ac7e4335a63e9a5144
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/325120
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/13604
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
This modifies the vbnv_flash driver to make it safe for use
in cache-as-ram by handling the global variables safely.
To make this cleaner all of the variables were moved into
one structure and referenced from there.
BUG=chrome-os-partner:47915
BRANCH=glados
TEST=build and boot on chell using following patches to
test backup and restore of vbnv_cmos into flash
Change-Id: I3a17fa51cfd754455502ac2e5f181dae35967f2a
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 48876561fa4fb61e1ec8f92596c5610d97135201
Original-Change-Id: Id9fda8467edcc55e5ed760ddab197ab97d1f3d25
Original-Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/324121
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/13599
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The VBNV region size is determined by vboot and is not really
configurable. Only the CMOS implementation defined this config
variable so switch it to use VBNV_BLOCK_SIZE defined by vboot
in vbnv_layout.h instead.
This requires updating the broadwell/skylake cmos reset functions
to use the right constant.
BUG=chrome-os-partner:47915
BRANCH=glados
TEST=manually tested on chell
Change-Id: I45e3efc2a22efcb1470bbbefbdae4eda33fc6c96
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: e2b803ff3ac30ab22d65d1e62aca623730999a1d
Original-Change-Id: I4896a1a5b7889d77ad00c4c8f285d184c4218e17
Original-Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/324520
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/13598
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Add a wrapper around the vbnv implementations and call into the different
backend functions from there. Also move some of the common functions to
the common code and simplify the backend drivers. This will allow some
of the code to be re-used so the CMOS backend can backup the data into
the flash backend.
One side effect of this is that the cache of VBNV was removed from CMOS
and EC backends and moved into the VBNV wrapper, but the flash backend
also still has a separate cache because it has more state and complexity
in the implementation. The wrapper cached data is not used for normal
vbnv_read/vbnv_write because some callers need the ability to force a
write if the backend storage is cleared (i.e. CMOS clear).
BUG=chrome-os-partner:47915
BRANCH=glados
TEST=build and boot on chell
Change-Id: I4d2e0e99af7e8a44aec77ad9991507401babcca6
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: c30f60434a64f6c0eb9ede45d48ddafff19dd24f
Original-Change-Id: Ia97f6607c5ad837b9aa10b45211137221ccb93a0
Original-Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/324120
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/13597
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Vboot keeps track of the size of the hashed region in each
RW slot. While that size was being used to calculate the hash
it wasn't being honored in restricting the access within the
FMAP region for that RW slot. To alleviate that create a sub
region that covers the hashed data for the region in which
we boot from while performing CBFS accesses.
BUG=chrome-os-partner:49764
BUG=chromium:445938
BRANCH=glados
TEST=Built and booted chell with cbfstool and dev-util patches.
Change-Id: I1a4f45573a6eb8d53a63bc4b2453592664c4f78b
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 4ac9e84af5b632e5735736d505bb2ca6dba4ce28
Original-Change-Id: Idca946926f5cfd2c87c4a740ad2108010b6b6973
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/324093
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/13586
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
When vboot_handoff_flag() is called in the bootblock or a separate
verstage there's no memory nor the possibility of dram coming online.
Therefore, don't bother to attempt call cbmem_find().
BUG=chrome-os-partner:44827
BRANCH=glados
TEST=Built chell with separate verstage which pulls in vboot_common.c
dependency. No more linking errors w/ cbmem_find() not being
around.
Change-Id: I494c93adc1c00459fdfaa8ce535c6b4c884ed0fb
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 414ce6aeaff657dc90289b25e5c883562189b154
Original-Change-Id: I8a5f2d154026ce794a70e7ec38883fa3c28fb6e7
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/324070
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/13580
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Sometimes it's necessary for the platform to perform clean up
tasks prior to reboot when employing vboot. For example, x86 systems
that resume and do vboot verification may need to clear their
sleep control register prior to doing a cold reset so that the
next boot doesn't appear to be a resume. Allow that hook by
introducing vboot_platform_prepare_reboot().
BUG=chrome-os-partner:46049
BRANCH=glados
TEST=Ensure vboot_platform_prepare_reboot() called from vboot_reboot().
Change-Id: I622c9181d9fa3048204e3df3223d5dd4b458abca
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: f31ffc40bde002dec398fd4dd9d2ee9d65df0d7b
Original-Change-Id: I97318cec34494a7fc4b1ecf2cb22715d20e730ff
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/323501
Original-Reviewed-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/13575
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
For x86 systems which resume through the reset vector one needs to
ensure the the RW slot taken at resume time matches the one at
boot time. The reason is that any assets pulled out of the boot
media need to match how the platform previously booted. To do
that one needs obtain the hash digest of the chosen slot, and it
needs to be saved in a secure place on the normal boot path. On
resume one needs to retrieve the hash digest back to compare it
with the chosen slot. If they don't match resuming won't be
possible.
BUG=chrome-os-partner:46049
BRANCH=glados
TEST=Suspended and resumed on chell. Also, tested with an EC build
which returns a bad hash to ensure that is properly caught.
CQ-DEPEND=CL:323460
Change-Id: I90ce26813b67f46913aa4026b42d9490a564bb6c
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 01a42c0ecfc6d60d1d2e5e36a86781d91d5c47a9
Original-Change-Id: I6c6bdce7e06712bc06cc620a3d7a6a6250c59c95
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/323500
Original-Reviewed-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/13574
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This requires payload integration somewhere to be useful, because
without that, adding it will (hopefully) break the signature.
Change-Id: I67b8267e5040e26353df02d258e92a0610e19a52
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/13560
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
Use the flags to preset the GBB flags field. The Kconfig defaults are
chosen for a "developer" configuration.
Change-Id: Ifcc05aab10b92a2fc201b663df5ea47f92439a3f
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/13559
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The GBB contains hardware-specific data plus some configuration. The
latter isn't supported by this change yet and will come later.
The fields that are supported (hardware ID, bmpfv, vboot keys) are
configurable through Kconfig and point to Chrome OS-style default (eg.
developer keys).
While adding vboot keys, the two keys used to sign RW regions are also
added to Kconfig, even if not yet used.
Change-Id: Icfba6061ca83182df560cd36052fbb257826d4b0
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/13558
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Just rename the two scripts that are in the src/ tree to give them
a .sh extension. Since we generally expect files in the src directory
to be source files, this allows to identify these as scripts easily.
Change-Id: I0ab20a083880370164488d37a752ba2d5a192fdc
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/13432
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
This patch adds a check to the VPD parsing code to avoid reading the
whole thing if the first byte ('type' of the first VPD entry) is 0x00
or 0xff. These values match the TERMINATOR and IMPLICIT_TERMINATOR types
which should never occur as the first entry, so this usually means that
the VPD FMAP section has simply never been initialized correctly. This
early abort avoids wasting time to read the whole section from SPI flash
(which we'd otherwise have to since we're not going to find a Google VPD
2.0 header either).
BRANCH=None
BUG=None
TEST=Booted Oak, confirmed that VPD read times dropped from 100ms to
1.5ms.
Change-Id: I9fc473e06440aef4e1023238fb9e53d45097ee9d
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 20a726237e03941ad626a6146700170a45ee7720
Original-Change-Id: I09bfec3c24d24214fa4e9180878b58d00454f399
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/322897
Original-Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-on: https://review.coreboot.org/13467
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
