The vboot code tries reading rollback protection indices from the TPM,
and if the attempt to read returns TPM_E_BADINDEX, it decides that the
TPM has not yet been initialized for the Chromebook use, and needs to
be taken through the factory initialization sequence.
TPM_E_BADINDEX is an internal representation of the TPM error 0x28b,
generated on attempts to read a non existing NVMEM space.
If the space exists, but has never been written the TPM returns error
0x14a. This condition (the space exists but not written) could happen
if the previous factory initialization attempt was interrupted right
after the space was created.
Let's map this error to the same internal representation
(TPM_E_BADINDEX) so that the Chrome OS device could recover when this
condition occurs.
BRANCH=reef, gru
BUG=b:37443842
TEST=verified that the Pyro device stuck in TPM error state recovered
when this patch was applied.
Change-Id: I6ff976c839efcd23ae26cef3ee428e7ae02e68f8
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://review.coreboot.org/20299
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The upcoming Cr50 firmware changes will require the AP to enable the
previously downloaded Cr50 firmware update(s).
A new vendor command (TPM2_CR50_SUB_CMD_TURN_UPDATE_ON) is used for
that. The command accepts one parameter - a timeout value in range of
0 to 1000 ms.
When processing the command the Cr50 checks if the alternative RO or
RW image(s) need to be enabled, and if so - enables them and returns
to the host the number of enabled headers.
If the vendor command requested a non-zero timeout, the Cr50 starts
a timer to trigger system reboot after the requested timeout expires.
The host acts on the number of enabled headers - if the number is
nonzero, the host prepares the device to be reset and waits for the
Cr50 to reboot the device after timeout expires.
This patch also adds more formal vendor command
marshaling/unmarshaling to make future additions easier.
BRANCH=gru,reef
BUG=b:35580805
TEST=with the actual user of this code in the next patch verified that
the cr50 update is enabled as expected.
Change-Id: Ic76d384d637c0eeaad206e0a8242cbb8e2b19b37
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://review.coreboot.org/18945
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
cr50 by default delays nvmem commits internally from the point of
reset to accumulate change state. However, the factory process can
put a board into dev mode through the recovery screen. This state
is stored in the TPM's nvmem space. When the factory process is
complete a disable_dev_request and battery_cutoff_request is performed.
This leads to disabling the dev mode in TPM, but the battery is
subsequently cut off so the nvmem contents never stick. Therefore,
whenever antirollback_write_space_firmware() is called we know there
was a change in secdata so request cr50 to immediately enable nvmem
commits going forward. This allows state changes to happen immediately.
The fallout from this is that when secdata is changed that current
boot will take longer because every transaction that writes to TPM
nvmem space will perform a write synchronously. All subsequent boots
do not have that effect.
It should also be noted that this approach to the implementation is
a pretty severe layering violation. However, the current TPM APIs
don't lend themselves well to extending commands or re-using code
outside of the current routines which inherently assume all knowledge
of every command (in conflict with vendor commands since those are
vendor-specific by definition).
BUG=b:35775104
BRANCH=reef
TEST=Confirmed disablement of dev mode sticks in the presence of:
crossystem disable_dev_request=1; crossystem
battery_cutoff_request=1; reboot;
Change-Id: I3395db9cbdfea45da1f5cb994c6570978593b944
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/18681
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Tested-by: build bot (Jenkins)
On Chrome OS devices that use TPM2 parts the platform hierarchy
is disabled by the boot loader, depthcharge. Since the bootloader
isn't involved in resuming a suspended machine there's no equivalent
action in coreboot to disable the platform hierarchy. Therefore, to
ensure consistent state in resume the platform hierarchy in the TPM2
needs to be disabled as well. For systems that resume using the
firmware the platform hierarchy is disabled when utilizing
TPM2 devices.
BUG=chrome-os-partner:61097
BRANCH=reef
TEST=Suspend and resume. Confirmed 'stop trunksd; tpmc getvf; start
trunksd' shows that phEnable is 0.
Change-Id: I060252f338c8fd68389273224ee58caa99881de8
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/18096
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
TPM PCRs are used in Chrome OS for two purposes: to communicate
crucial information from RO firmware and to protect FW and kernel
rollback counters from being deleted.
As implemented in a TPM1 compatible way, the PCR extension command
requires a prebuilt digest to calculate a new PCR value.
TPM2 specification introduces a PCR_Event command, where the TPM
itself calculates the digest of an arbitrary length string, and then
uses the calculated digest for PCR extension. PCR_Event could be a
better option for Chrome OS, this needs to be investigated separately.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=verified that the two PCRs are successfully extended before the
RW firmware is called.
Change-Id: I38fc88172de8ec8bef56fec026f83058480c8010
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 73388139db3ffaf61a3d9027522c5ebecb3ad051
Original-Change-Id: I1a9bab7396fdb652e2e3bc8529b828ea3423d851
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/358098
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Darren Krahn <dkrahn@chromium.org>
Reviewed-on: https://review.coreboot.org/15639
Tested-by: build bot (Jenkins)
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
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>
The command is sent in session mode, but has no parameters associated
with it.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=with the following patches verified that TPM_Clear command is
handled successfully by the TPM.
Change-Id: I3c9151e336084160acd3bb1f36f45b4d5efd4a33
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 503ad5e72fd5bd902325d74fd680c17c7c590e36
Original-Change-Id: Ida19e75166e1282732810cf45be21e59515d88e2
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/357973
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/15634
Tested-by: build bot (Jenkins)
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
Some structures were included in tpm2_tlcl_structures.h that are not
needed for tpm2 commands used by coreboot. Drop them from the include
file.
BRANCH=none
BUG=none
TEST=coreboot image for gru/kevin still builds fine.
Change-Id: Id3a01f7afbddc98b4d14125452ae6a571f1b19cb
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 9375eef5a3f5ed2ba216b1cc8a4ce5c78ebe53d8
Original-Change-Id: I89b46900e5356989f2683d671552ecca5103ef90
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/358093
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/15608
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
TPM2 structure definitions use pointers instead of buffers where
possible. One structure was left behind. Replace that buffer definition
with a pointer to be consistent.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=compilation succeeds, the code using the changed structure in the
upcoming patches allows to successfully boot chrome OS on Kevin
Change-Id: Iea59943aa0ad6e42fcd479765a9ded0d7a1680d7
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: 02b2909b1875ba65cd56cf8e3697a2b67ddaea07
Original-Change-Id: I9856ac516be13f5892ba8af0526708409a297033
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/358771
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://review.coreboot.org/15609
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
This is the first approximation of implementing TPM2 support in
coreboot. It is very clearly incomplete, some of the larger missing
pieces being:
- PCR(s) modification
- protection NVRAM spaces from unauthorized deletion/modification.
- resume handling
- cr50 specific factory initialization
The existing TPM1.2 firmware API is being implemented for TPM2. Some
functions are not required at all, some do not map fully, but the API
is not yet being changed, many functions are just stubs.
An addition to the API is the new tlcl_define_space() function. It
abstracts TMP internals allowing the caller to specify the privilege
level of the space to be defined. Two privilege levels are defined,
higher for the RO firmware and lower for RW firmware, they determine
who can write into the spaces.
BRANCH=none
BUG=chrome-os-partner:50645
TEST=with the rest of the patches applied Kevin/Gru devices can
initialize and use firmware and kernel spaces
Change-Id: Ife3301cf161ce38d61f11e4b60f1b43cab9a4eba
Signed-off-by: Martin Roth <martinroth@chromium.org>
Original-Commit-Id: bcc8e62604c705798ca106e7995a0960b92b3f35
Original-Change-Id: Ib340fa8e7db51c10e5080973c16a19b0ebbb61e6
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/353914
Original-Commit-Ready: Martin Roth <martinroth@chromium.org>
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://review.coreboot.org/15569
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@googlemail.com>
