For x86 systems, which resume through the boot reset vector, to
implement vboot verification of the memory init code one needs
check that the slot chosen on the resume path is the same as
the original boot path. That check is done by storing the
resulting hash of the slot. However, vb2api doesn't export
the resulting hash from vb2api_check_hash(). Thus, provide
a variant which saves the resulting digest in the supplied
buffer.
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.
Change-Id: Ic20be2024afedabc2d8bc767f1b794376348523c
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/323460
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
The VBOOT_OPROM_NEEDED flag is used for EC software sync when the
VBSD_EC_SLOW_UPDATE flag is set.
After a successful EC software sync vboot requests a reboot to disable
graphics but it is not clearing the VBNV flag first. With vboot1 this
was getting cleared as a side effect of calling VbInit in normal mode.
BUG=chrome-os-partner:49560
BRANCH=glados
TEST=Enable EC_SLOW_UPDATE on chell and test EC software sync in normal
mode and ensure that it reboots and does not do graphics init if the
update is successful.
Change-Id: I2aa0c4c3b1ad357a5b8ddc14539e264a1f5b76b2
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/322731
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Allow the AP to sync and verify the EC read only image after updating
the rewritable image.
BUG=chrome-os-partner:48703
BRANCH=none
TEST=manual
1. Update EC to a new version
2. rebuild EC code
3. Update AP firmware
4. Reboot and check that the RO image is updated after the RW image is
updated.
CQ-DEPEND=CL:319213
Change-Id: I774ef25320103f20d8c7d1c180a220dd0819c04d
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/320614
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Kernel TPM patches from 4.x moved /sys/class/misc/tpm0 to /sys/class/tpm/tpm0.
Support both paths in this utility.
BUG=chromium:573368
BRANCH=none
TEST=untested, not sure if this utility is still used.
Change-Id: Ib81476eee4c9de921502a3a47f6990b9e6b1968b
Reviewed-on: https://chromium-review.googlesource.com/320892
Commit-Ready: Bryan Freed <bfreed@chromium.org>
Tested-by: Bryan Freed <bfreed@chromium.org>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
This change will be used to support EC-RO software sync by allowing for
access to the readonly region of firmware. Currently only the writable
section is accessed by vboot using VB_SELECT_FIRMWARE_A and B.
BUG=chrome-os-partner:48703
BRANCH=none
TEST=built on jerry and check that the RO hash can be read and the image
can be updated.
CQ-DEPEND=CL:319185,CL:320425,CL:320598
Change-Id: Ic3942d86b65da3123798cfd11a78056f5dab6699
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/319213
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Musl complains about using sys/fcntl.h, and with -Werror that
gives a compilation failure. Every other file includes <fcntl.h>
so just use that.
Change-Id: Ibc743bbe335e9e9b91da7e4eacc01cc80cd8b35a
Reviewed-on: https://chromium-review.googlesource.com/313419
Commit-Ready: Doug Evans <dje@google.com>
Tested-by: Doug Evans <dje@google.com>
Reviewed-by: Nam Nguyen <namnguyen@chromium.org>
This flag will be used by the firmware updater to indicate that RO
software sync should be attempted.
BUG=chrome-os-partner:48703
BRANCH=None
TEST=make runtests
Change-Id: I42090ac47da45c724e66334648ab447ad3c21178
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/320621
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This is a standalone library for verifying the BDB structures in the
common boot flow document, and a bdb_create utility to create test BDB
structures. Eventually, creating these structures will be rolled into
futility.
BUG=chrome-os-partner:48448
BRANCH=none
TEST=cd bdb && make runtests
Change-Id: Ic57c26ca84137205da3b6c7d532f5324c93b4285
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/317275
Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
The standalone accessories are using a key name like this:
key_<product>.pem or key_<product>.vbprik2
when it doesn't exist, fallback using key.pem or key.vbprik2.
BRANCH=none
BUG=chrome-os-partner:47557
TEST=manual:
./scripts/image_signing/sign_official_build.sh accessory_usbpd ../ec/build/zinger/ec.bin tests/devkeys-acc /tmp/ec-zinger.TEST.SIGNED.bin
./scripts/image_signing/sign_official_build.sh accessory_rwsig ../ec/private/build/hadoken/keyboard_app.bin tests/devkeys-acc /tmp/ec-hadoken.TEST.SIGNED.bin
./scripts/image_signing/sign_official_build.sh accessory_rwsig ../ec/private/build/hadoken/keyboard_app.bin /tmp /tmp/ec-hadoken.TEST.SIGNED.bin
Change-Id: I68863664bdb9da1695e91b1986f3a0148af7da26
Reviewed-on: https://chromium-review.googlesource.com/312836
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
New devices have Depthcharge render vboot screens by calling
vboot_draw_screen. Thus, display initialization and backlight control should
not be duplicated. This patch prevents VbDisplayScreen from initializing
display and controlling backlight when vboot is rendering screens using GBB.
BUG=chrome-os-partner:43706,chromium:502066
BRANCH=tot
TEST=Tested on Glados
Change-Id: I50cd2decb7065af96779601b12f0fbf2554ff6ed
Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/312749
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
I don't even know what this is. It seems to have marked some kind of
debug buffer provided by H2C BIOS on pre-Daisy Chromebooks and has not
been touched since it was copied in here when crossystem was first
added. I can't find any references in our codebase so I doubt anybody
would miss it. Let's remove it so the '(error)' fields returned there on
any modern Chromebook stop confusing our vendors.
BRANCH=None
BUG=chromium:551715
TEST=Built for Falco and Jerry.
Change-Id: Ie2baec536b50bb192eb4cd3e48df212cce53561a
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/311346
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Bernie Thompson <bhthompson@chromium.org>
This field doesn't seem to be used for anyone and it keeps adding work
for people trying to bring up new platforms. If we ever needed something
like this again, we'd probably prefer to have it in mosys now anyway.
Let's get rid of it.
BRANCH=None
BUG=chromium:551715
TEST=Built for Falco and Jerry.
Change-Id: I6b96e255968fdd22a345d4a75bfdc1e79d3f5896
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/311345
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Bernie Thompson <bhthompson@chromium.org>
Move around the test keys used for standalone accessories signing
unittesting, so they are easier to use.
BRANCH=smaug, samus
BUG=chrome-os-partner:46635
TEST=run cros-signing unittests (./signing_unittests.py)
Change-Id: I02fcf444ec76bf6ebf6eb477e1236869aa877210
Reviewed-on: https://chromium-review.googlesource.com/309967
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
Add a new post-EC software sync API VbExEcVbootDone() to take actions
which normally need to happen after EC verification / sysjump.
BUG=chromium:537269
TEST=Manual on Glados. Set CHG_MW thresh to 20000, BAT_PCT to 50. Verify
that LIMIT_POWER host event is set until Zinger negotiates to 20V. Also
verify that we do not proceed with boot when Donette is plugged.
BRANCH=None
CQ-DEPEND=CL:307885,CL:309523
Change-Id: I77e6000aa8a44e3aca4fb5982e5b5f5191774989
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/307952
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
My previous patch using futility to re-sign standalone accessory
firmware images had a mistake in the key directory path : fix it.
Also add RSA-2048 'accessory' keys for signer unit testing.
BRANCH=smaug, samus
BUG=chrome-os-partner:46635
TEST=run cros-signing unittests (./signing_unittests.py)
Change-Id: Ia2f641c85337c67f81968be4730643a6ad5f22cf
Reviewed-on: https://chromium-review.googlesource.com/309530
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
FASTBOOT_FULL_CAP set
This change allows developers to boot dev-signed boot images in
unlocked mode if DEV_BOOT_FASTBOOT_FULL_CAP is set in VbNvStorage or
GBB_FLAG_FORCE_DEV_BOOT_FASTBOOT_FULL_CAP is set.
BUG=chrome-os-partner:47002
BRANCH=None
TEST=Compiles successfully. make -j runtests
Change-Id: I56e3879594da1b57051dfe242ff347ac970c96bb
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://chromium-review.googlesource.com/309606
Commit-Ready: Furquan Shaikh <furquan@chromium.org>
Tested-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Use futility to re-sign standalone accessory firmware images either the
former "usbpd1" used by USB Power Delivery firmware generated from the
EC codebase or the new "rwsig" format.
BRANCH=smaug, samus
BUG=chrome-os-partner:46635
TEST=manual:
openssl genrsa -F4 -out key_zinger.pem 2048
openssl genrsa -F4 -out key_hadoken.pem 2048
futility create --desc="Hadoken fake MP key" key_hadoken.pem key_hadoken
./scripts/image_signing/sign_official_build.sh accessory_usbpd build/zinger/ec.bin . build/zinger/ec.SIGNED.bin
./scripts/image_signing/sign_official_build.sh accessory_rwsig build/hadoken/keyboard_app.bin . build/hadoken/keyboard_app.SIGNED.bin
and compare the re-signed files with the original files.
Change-Id: I586ba3e4349929782e734af1590f394824e7dd44
Reviewed-on: https://chromium-review.googlesource.com/306795
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
When a user hits esc+refresh+power to start recovery, the true recovery
reason will be lost after reboot. (It would always look like
VB2_RECOVERY_RO_MANUAL.) This patch makes VbBootRecovery save
the reason in the subcode area before entering the new 'broken' loop.
BUG=chromium:501060
BRANCH=tot
TEST=test_that -b veyron_jerry suite:faft_bios
Change-Id: Ib536daa0633721bfc975381782d348f122b3d337
Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/307586
Reviewed-by: Randall Spangler <rspangler@chromium.org>
VbExGetLocalizationCount is a callback function which is supposed to
return the number of screen locales supported by VbExDisplayScreen.
After this change, we still try to get the number of locales from GBB
first but when it fails, VbExGetLocalizationCount is called. The error
code from VbGbbReadBmpHeader will be masked, similarly to the error from
VbDislayScreenFromGBB.
BUG=chromium:502066
BRANCH=tot
TEST=Tested on Samus. make runtests
Change-Id: I04ef8bf1ea02b1aaa05e65673b57bcea1932d8b0
Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/304376
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Add support for PEM file containing a RSA Public key in futility "show"
and "create" commands.
When "futility create" is given a PEM file with only a RSA public key,
generate the proper .vbpubk2 rather than failing.
BRANCH=smaug
BUG=none
TEST=make runtests
and run manually
futility show tests/testkeys/key_rsa4096.pub.pem
futility show tests/testkeys/key_rsa4096.pem
Change-Id: I707ceca54c80ba21f53869ad86c86fa23b31e665
Reviewed-on: https://chromium-review.googlesource.com/306683
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Verify the size of the buffer read from the file before trying to use 1KB of
it for the new rwsig format detection.
Add a new test case with a short file containing only 4 bytes of unknown
data and run "futility show" on it.
BRANCH=smaug
BUG=none
TEST=futility show foobar.pub.pem
where foobar.pub.pem is a 451-byte file.
check that "make runtests" passes with the fix
and fails without it with the following message :
test_file_types.sh ... failed
FAIL: 13 / 14 passed
Change-Id: Ia9d68c6b528c2b3a595ea6791c907374616d051f
Reviewed-on: https://chromium-review.googlesource.com/306682
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
The "rwsig" type is used for independent device firmware (not
Chromebook BIOS) that need to verify themselves instead of using
software sync.
The expected use case is that a RO firmware contains a
vb2_public_key struct along with an FMAP or other pointers to a
slot for RW firmware. The RW firmware slot reserves room for a
vb2_signature struct.
This CL changes the args and behavior of the rwsig type, so that
the RW firmware can be [re]signed independently of the rest of
the image.
BUG=chrome-os-partner:46254
BRANCH=smaug,ToT
TEST=make runtests, manual
Create a keypair:
futility create --desc "Key One" tests/testkeys/key_rsa2048.pem foo
Sign a RW binary and build a complete image out of the parts:
futility sign --type rwsig --prikey foo.vbprik2 rw.bin sig.bin
dd if=/dev/zero bs=65536 count=1 of=image.bin
dd if=rw.bin of=image.bin conv=notrunc
dd if=sig.bin bs=$((65536 - 1024)) seek=1 of=image.bin conv=notrunc
Verify both the separate parts and the combined image:
futility show --type rwsig --pubkey foo.vbpubk2 rw.bin sig.bin
futility show --type rwsig --pubkey foo.vbpubk2 image.bin
Re-sign the combined image with a different keypair:
futility create --desc "Key Two" tests/testkeys/key_rsa1024.pem bar
futility sign --type rwsig --prikey bar.vbprik2 image.bin
Now the first key no longer verifies:
futility show --type rwsig --pubkey foo.vbpubk2 image.bin
But the second key does:
futility show --type rwsig --pubkey bar.vbpubk2 image.bin
Change-Id: Ifdddab08f218f506eb1dce28851b153d70140a7b
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/305980
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
This reverts commit 82dec09bd5.
This flag doesn't exist on Ubuntu Precise which is what the signers
are running atm. Until we get them upgraded to Trusty, back this
change out.
BUG=chromium:530730
TEST=`./signing_unittests.py` passes
BRANCH=None
Change-Id: I9ba508c1531dbb169fd020d06ab102f6576b7342
Reviewed-on: https://chromium-review.googlesource.com/306310
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: Amey Deshpande <ameyd@google.com>
This change makes VbDisplayScreen read the last saved locale from nvram
and pass it to VbExDisplayScreen so that it can draw locale dependent
screens.
BUG=chromium:502066
BRANCH=tot
TEST=Tested on Samus. make runtests.
CQ-DEPEND=CL:304382,CL:306100,CL:306110
Change-Id: I9782ec5a8a9f8393998aa8a0d64e88ad1809233b
Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/304375
This adds the "rwsig" type, with initial support for RO+RW
firmware images that need to verify themselves instead of using
software sync. This uses our vb2 structs instead of raw binary
blobs. That will help us locate, identify, and verify the keys
and signatures in the signed firmware images.
BUG=chrome-os-partner:46254
BRANCH=smaug,ToT
TEST=make runtests
I also hacked up a test board with the EC-side signature
verification routines from a preliminary CL and tested this
signing scheme with that. It works.
Additional work is needed to make this seamless, but you can try
it out like so:
futility create ./tests/testkeys/key_rsa2048.pem foo
futility sign --type rwsig --prikey foo.vbprik2 --pubkey foo.vbpubk2 ec.bin
Change-Id: I876ab312a2b0b36411c5f739fe3252529728d034
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/305394
Reviewed-by: Randall Spangler <rspangler@chromium.org>
In developer mode, this option will make the system try to boot into
a legacy OS first after the 30 second timeout. This removes the need to
press a key during boot to try legacy mode and the need to remove the
write protect screw to boot legacy as default.
BUG=chromium:310697
BRANCH=none
TEST=make runtests
Change-Id: I9a9f64c14ad015e21d08eec36e8fc187189cd2f2
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/304077
Reviewed-by: Randall Spangler <rspangler@chromium.org>
In the new recovery process, a user will see 'broken' screen
instead of 'remove' screen, where usb stick presence is no longer
detected. A user instead has to hit esc+refresh+power to proceed
to recovery mode.
BUG=chromium:501060
BRANCH=tot
TEST=make runtests
Change-Id: Icd511c1ca892628b96befbb0a34c2c84b881c857
Reviewed-on: https://chromium-review.googlesource.com/304404
Commit-Ready: Daisuke Nojiri <dnojiri@chromium.org>
Tested-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Specifically, this patch updates 'root_hexdigest' in legacy bootloader
templates in EFI system partition to match the signed rootfs.
BRANCH=None
BUG=chromium:512940
TEST=Ran sign_official_build.sh locally and booted the image on kvm
(using BIOS).
TEST=Ran signing_unittests.py by locally changing vboot_stable_hash to
include this patch.
$ ./sign_official_build.sh base chromiumos_base_image.bin \
../../tests/devkeys chromiumos_base_image_signed.bin
Change-Id: Ied021c4464b113a64508f5081605069bdcecbc1f
Reviewed-on: https://chromium-review.googlesource.com/301742
Commit-Ready: Amey Deshpande <ameyd@google.com>
Tested-by: Amey Deshpande <ameyd@google.com>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
The firmware for the USB Type-C power adapters uses raw binary
blobs for the public keys and signatures instead of
readily-identifiable structs. We've been able to sign these
firmware images for some time, but verifying the result generally
required testing them on hardware.
This CL adds some futilty support for recognizing and verifying
those images too. It just tries various sig and hash algorithms,
until it finds a combination for which the image is
self-consistent (where the pubkey blob verifies the signature
blob).
BUG=none
BRANCH=none
TEST=make runtests
This change also adds additional tests for usbpd1 images. We
ensure that we correctly recognize and verify an MP-signed
firmware, plus test signing and verifying usbpd1 images using
multiple signature and hash algorithms.
Change-Id: I4fbe8b37a694992f635d5469ae1c2449b1610dfd
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/302415
Reviewed-by: Randall Spangler <rspangler@chromium.org>
In recovery mode, the TPM may be bad / corrupt. This prevents access to
the soft developer switch stored in secdata. But it should not prevent
setting dev mode via GBB or context flags. Those flags may be set
during manufacturing or testing, and override the contents of secdata
anyway.
BUG=chrome-os-partner:45511
BRANCH=ryu
TEST=make runtests
Change-Id: I242714528203cc7cf78a714c660b7f8bbd0e04d0
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/300621
Commit-Ready: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@chromium.org>
BRANCH=None
BUG=chrome-os-partner:44227
TEST='sign_official_build.sh recovery_kernel boot.img keys
boot.img.recovery-signed' works fine and able to boot in locked recovery mode
using fastboot boot.
Change-Id: Iabde28bb2068b8294fc3d03f2f771c63368ecbb5
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://chromium-review.googlesource.com/300250
Commit-Ready: Furquan Shaikh <furquan@chromium.org>
Tested-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: David Riley <davidriley@chromium.org>
When a TPM goes from the disabled state to the enabled state, it must
reboot after being enabled, before it can be initialized. In vboot1,
TLCL was part of vboot and this was handled internally. In vboot2, the
caller must set a context flag, so that vboot can decide whether to
allow the reboot, or whether to go directly to recovery mode. This
check is necessary to handle the following cases:
1) The device is booting normally, but the TPM needs a reboot. This
should simply reboot, without going to recovery mode.
2) The device is booting in recovery mode, but the TPM needs a reboot.
If this is the first time it asked us, allow the reboot.
3) The TPM asked for a reboot last time, so we did. And it's still
asking. Don't reboot, because that runs the risk that whatever is wrong
won't be fixed next boot either, and we'll get stuck in a reboot loop
that will prevent recovery. Boot into recovery mode.
Add a new NvStorage bit to track whether the TPM requested a reboot on
the previous boot. That's better than what we did in vboot1, where we
used a special recovery request. Vboot1 couldn't track getting stuck in
a reboot loop in normal mode, only in recovery mode. The new code can
catch both.
BUG=chrome-os-partner:45462
BRANCH=ryu
TEST=make runtests
Change-Id: I2ee54af107275ccf64a6cb41132b7a0fc02bb983
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/300572
Tested-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
(resubmit)
Previously crossystem assumed that mosys was located
in /usr/sbin. In Android mosys is currently located
in /system/bin. Using fixed paths as opposed to
'which' to prevent attacks where attacker could insert
mosys in PATH.
difference from previous commit:
Removed the allocation of duplicate arrays. Kept
with simplicity of original version, just returning
correct constant depending on detected platform.
BUG=chromium:527484
BRANCH=none
TEST=ran crossystem, crossystem fw_try_count/
fw_try_next, crossystem fw_try_count/fw_try_next=x
on smaug and daisy.
Change-Id: I923206db1411a9a35c9c8e3f9ede5016f49b5f26
Signed-off-by: Shelley Chen <shchen@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/299801
Reviewed-by: danny chan <dchan@chromium.org>
Often the partitions we extract have extra space in them, but the dd
utility will still write out the excess zeros. That can mean we write
out hundreds of megs of data which could otherwise be skipped. We thus
waste a good amount of I/O and storage.
For now, only use this flag when extracting a partition to a new file
as this should be safe (there's no pre-existing data to clobber/merge).
BUG=chromium:530730
TEST=`./signing_unittests.py` passes
BRANCH=None
Change-Id: Ic32665cf7c38fc0a5efc3f8b227fa8ff408ca9e3
Reviewed-on: https://chromium-review.googlesource.com/299450
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: David Riley <davidriley@chromium.org>
This should speed up the copies significantly by using less disk
storage & I/O when the unpacked file is not sparse already. This
option has been in cp for a long time, and works in Ubuntu Precise
(coreutils-8.13) & Trusty (coreutils-8.21).
BUG=chromium:530730
TEST=`./signing_unittests.py` passes
BRANCH=None
Change-Id: I82192455a623eabf96abf4f25296f3dc0c129ca2
Reviewed-on: https://chromium-review.googlesource.com/299440
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: David Riley <davidriley@chromium.org>
Reviewed-by: Amey Deshpande <ameyd@google.com>
Rather than use the existence of the output as a marker for running the
gbb step, key off the loem index. We want to run it the first time and
not bother after that.
BUG=chrome-os-partner:44227
BRANCH=None
TEST=signer can still sign loem keysets
Change-Id: I26e9ccaf1333f769d6993a8e0d84c63644bb2597
Reviewed-on: https://chromium-review.googlesource.com/298980
Reviewed-by: David Riley <davidriley@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Signing test of firmware for non-t210 targets fails when PKC
keys are present so separate devkeys into two keysets where only
one has the PKC key.
BUG=chrome-os-partner:44227
TEST=sign_official_build.sh nv_lp0_firmware tegra_lp0_resume.fw tests/devkeys-pkc tegra_lp0_resume.fw.signed versions.default
BRANCH=signer
Change-Id: Ie0670638d30d17692f1113b5a9f0309ea610d72a
Signed-off-by: David Riley <davidriley@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/298091
Reviewed-by: Mike Frysinger <vapier@chromium.org>
The "function" keyword is not portable -- use the normal function style.
The awk command uses a non-portable regex (the word anchor \>). Rework
it to avoid regexes entirely.
BUG=chromium:475101
TEST=keyset_version_check.sh works on a POSIX system
BRANCH=None
Change-Id: I5446f63aa9181d06da1898aafb8fab17f5042989
Reviewed-on: https://chromium-review.googlesource.com/296562
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: David Riley <davidriley@chromium.org>
Signer was calling sign_official_build.sh in a manner that wasn't
being accepted correctly. Also add test keys from firmware branch.
BUG=chrome-os-partner:44227
TEST=sign_official_build.sh nv_lp0_firmware tegra_lp0_resume.fw tests/devkeys tegra_lp0_resume.fw.signed versions.default
BRANCH=signer
Change-Id: Icd298ac75e3da746220826dc2fb9cc2466e41f1d
Signed-off-by: David Riley <davidriley@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/297802
Reviewed-by: Furquan Shaikh <furquan@chromium.org>
