When disabling verity with make_dev_ssh.sh, a bug in
vbutil_kernel caused the re-signed kernel size to be the entire
kernel partition instead of just the necessary bits. Until we can
improve the test coverage, I'm rolling back the changes that
introduced this bug.
BUG=chromium:418647
BRANCH=ToT
TEST=manual
Created a new test image with these changes. You can install it
and disable dm-verity and it works (although there seems to be an
unrelated browser startup issue on ToT).
Change-Id: I48e8427b05e191c9894c42056429a79d57bfc78d
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/220935
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This adds a --strict mode to the show command, which requires
that all signatures be valid in order to exit cleanly. It also
creates a "verify" command, which is really just an alias for
"show --strict".
BUG=none
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: I1fed7db7fe7128191bcab0c615706ef4fe2709f5
Reviewed-on: https://chromium-review.googlesource.com/219732
Reviewed-by: Randall Spangler <rspangler@chromium.org>
BUG=none
BRANCH=ToT
TEST=make runtests
This also modifies the tests to compare the futility sign command
results against the vbutil_kernel results.
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: Ibc659f134cc83982e3f0c0bcc108cc0eddbe228e
Reviewed-on: https://chromium-review.googlesource.com/219730
Reviewed-by: Randall Spangler <rspangler@chromium.org>
We have been traversing things by passing a file descriptor. Now
the caller should mmap the file first. This will allow the caller
to determine the file type before traversing into it, so we can
check args.
BUG=none
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: If69799bde0133689dc9fb5111e6ecb5ac61639c7
Reviewed-on: https://chromium-review.googlesource.com/219649
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This adds new file types to prepare for signing kernel partitions
and raw firmware blobs (FW_MAIN_A/B).
BUG=none
BRANCH=ToT
TEST=make runtests
No new functionality yet.
Change-Id: Ic6b6b94bb99f00ab54609dfe1b753b53868abaca
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/219648
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The original vbutil_kernel command used file read and write to
make changes. Futility prefers to use memory-mapped files. This
rewrites cmd_vbutil_kernel.c to use that scheme.
BUG=none
BRANCH=ToT
TEST=make runtests
The original cmd_vbutil_kernel.c is renamed, and a test written
to ensure that the refactored version produces identical results.
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: Ic6c3e12429a5dcb271f8136a9edac70807d66120
Reviewed-on: https://chromium-review.googlesource.com/219647
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Previously, you could extract FMAP areas like so:
futility dump_fmap -x bios.bin FW_MAIN_A VBLOCK_A ...
This lets you decide what to name each area as it's extracted:
futility dump_fmap -x bios.bin FW_MAIN_A:/tmp/rw_a ../vblock
BUG=none
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: If02b57e03294b0b0b1dbc216ef57afdd3bdf2960
Reviewed-on: https://chromium-review.googlesource.com/219646
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Move the Debug() function into a common place instead of several
copies in different files, rename shared functions to start with
"futil_"
BUG=none
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: I6b844553dff95c24894dae611102716a8da5312d
Reviewed-on: https://chromium-review.googlesource.com/219645
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This provides help messages for the futility commands similar to
the way git does. These show the available commands:
futility
futility help
futility --help
While these show help for a specific command:
futility help COMMAND
futility --help COMMAND
futility COMMAND --help
BUG=none
BRANCH=ToT
TEST=manual
make runtests
And manually look at help messages for each command.
Change-Id: I1126471e242784c6ca7a2f11694fa7c505d833e8
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/219528
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Since all of the functionality of the resign_firmwarefd.sh script
is built in to futility, let's just make that script invoke
futility to do the work. We'll come back and remove the script
entirely, once all outside references to it have been changed to
do the right thing.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Also tested by editing tests/futility/test_resign_firmware.sh to
invoke the resign_firmwarefd.sh script instead of futility.
Everything passed.
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: Id068e551067a956cd7ddc3f9b9e23488261d8d94
Reviewed-on: https://chromium-review.googlesource.com/216716
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Just reporting that the parent process is "/bin/bash" doesn't
help much. Let's also report the cmdline args given to the parent
and the cwd. This will help us identify which shell script is
calling futility with the wrong args.
BUG=chromium:231547
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: I800995ff269ab8d8c56cad8827d8de48a53cd150
Reviewed-on: https://chromium-review.googlesource.com/216715
If we're re-signing a valid BIOS image, we want to be sure that
we preserve the original firmware preamble flags (RO_NORMAL and
so forth) if the --flags option does not specifically override
it.
This change adds a test for that case, and makes it happen.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: I8cbde66abaf96ec82adf0205bedf57b1fd1b82a1
Reviewed-on: https://chromium-review.googlesource.com/216714
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The previous version of gbb_utility always zeros data before
writing new values and we should keep this behavior, to simplify
firmware hash calculation and potential security concern.
BRANCH=none
BUG=chromium:413066
TEST=emerge gbb_utility; factory/bin/gooftool get_firmware_hash bios.bin
Original-Change-Id: Ic97a118cefc9698d52d9370b627670ff103d5e23
Change-Id: If38e15f35ee491cc80f96b360c63ee25f71c1854
Reviewed-on: https://chromium-review.googlesource.com/217700
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Commit-Queue: Hung-Te Lin <hungte@chromium.org>
Tested-by: Hung-Te Lin <hungte@chromium.org>
(cherry picked from commit 3c8d29c204c6feb91ad951e2e1c5190d4ca98a13)
Reviewed-on: https://chromium-review.googlesource.com/217711
Commit-Queue: Bill Richardson <wfrichar@chromium.org>
Tested-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This allows the sign command to work on BIOS images with
invalid VBLOCK areas. When re-signing an existing image, the
length of the firmware body is part of the firmware preamble
in the VBLOCK areas. If those are invalid, the BIOS can
still be signed, but it will have to sign the entire FW_MAIN
area. That's a little slower to verify, so we'd prefer not
to do that, but it works.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: If58b5c86c5df12f004eabff72c22bfb1e84de7fd
Reviewed-on: https://chromium-review.googlesource.com/216229
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This adds a "load_fmap" command, which is pretty much the
opposite of the "dump_fmap -x" command. It allows you to
replace the content of any FMAP areas with new stuff, without
mucking around with dd. There's a test for it, too.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Change-Id: I5a9ab249c9e63a9bb1a9b26feeb3ed757cd294f1
Reviewed-on: https://chromium-review.googlesource.com/216228
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The "sign" command can perform the same operation as the old
resign_firmwarefd.sh script, only about 20 times faster. The
test for that will use the new command instead.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: Ie7f7a0ab6fc00d7e06cb263733bf6e7246fdb023
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/216227
Reviewed-by: Randall Spangler <rspangler@chromium.org>
It doesn't yet handle block devices, but it can display normal files
containing a entire BIOS image, a GBB, a VBLOCK, a .vbpubk, a .vblock,
and a firmware preamble (VbFirmwarePreambleHeader).
The command-line options are not well-documented.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: I181f6331ae23599302bbaee3f270e8af9586cf06
Reviewed-on: https://chromium-review.googlesource.com/216032
Commit-Queue: Bill Richardson <wfrichar@chromium.org>
Tested-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The futility.lds linker script was used to generate a table with
all the symbols in a specific section called .futil_cmds listed
in order under the symbol "futil_cmds". This allows the source files
to define a command and let the linker figure out the list of
compiled commands. Nevertheless, passing this linker script
makes the linker leave a gap of about 2MiB in the output ELF file.
Instead of mess up with linker scripts just to generate a table of
commands, this patch generates such table in the Makefile looking
at the included sources and compiling that table. The result is a
futility binary of about 88 KiB instead of the 2.1 MiB required
originally.
This patch also adds sys-boot/chromeos-u-boot to the list of ebuilds
tested by emerge_test.sh.
BUG=chromium:408926
BRANCH=None
TEST=BOARD=link ./emerge_test.sh
TEST=BOARD=daisy_spring ./emerge_test.sh
TEST=`readelf -S futility` shows no gap.
TEST=/usr/bin/futility shows no difference in the help output.
Change-Id: I9c0febc76140b404d48aa13e7f948e8ea77a41b5
Reviewed-on: https://chromium-review.googlesource.com/215496
Tested-by: Alex Deymo <deymo@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Commit-Queue: Alex Deymo <deymo@chromium.org>
This creates a disk image and verifies a kernel can be loaded from it.
It is roughly analogous to vb2_firmware_tests.sh, but at the kernel
step instead of the firmware step.
This will get more interesting in the near future, with the upcoming
addition of a streaming API to read the kernel.
BUG=chromium:408265
BRANCH=none
TEST=make runtests
Change-Id: Icc9e6d0e318c4bd38fc9ab1ad704da99232822e1
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/214508
Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
The functions that look for the FMAP and its entries should return more
useful values.
BUG=none
BRANCH=ToT
TEST=make runtests
No functional changes.
Change-Id: I4b62ea0de972bceb3d58f4ee8eb82ad065ddcbae
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/214630
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Update the unit tests which use it to use futility. No functional
changes to it, just relocation.
Remove the futility test which checks the exact list of supported
commands. This doesn't have a good way of handling
conditionally-compiled commands, and will be even harder to maintain
as we add more commands in the future. Presence of sub-commands is
still ensured by the other tests which use them (such as
vb2_firmware_tests.sh)
BUG=chromium:231547
BRANCH=none
TEST=make runtests && VBOOT2=1 make runtests
Change-Id: Idddb639276e4c6449d023d40ac7977123113bd28
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/213191
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
This just reformats the futility sources to conform to the Linux kernel
coding style. No functional changes.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: I82df07dd3f8be2ad2f3df24cebe00a9a378b13f4
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/213915
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Add support for the 64bit arm architecture, which update_kernel.sh
passes in as "aarch64"
Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chrome-os-partner:31525
TEST=run vbutil_kernel --arch=aarch64
Check that it no longer complains about "Unknown architecture string: aarch64"
BRANCH=none
Change-Id: Iccd925e05baffb1953b229fc4150ca179d1d1e1c
Reviewed-on: https://chromium-review.googlesource.com/213706
Reviewed-by: David Riley <davidriley@chromium.org>
Tested-by: Benson Leung <bleung@chromium.org>
Commit-Queue: Benson Leung <bleung@chromium.org>
This change adds these formerly external utilities into the futility binary:
dev_sign_file
dump_kernel_config
gbb_utility
vbutil_firmware
vbutil_kernel
These target binaries will remain independent of futility, since they are
not directly related to verified boot:
cgpt
crossystem
tpm_init_temp_fix
tpmc
Also, dumpRSAPublicKey is removed from the target, since it is only used on
the build host to create new keypairs.
This change also add several additional tests.
BUG=chromium:224734
BRANCH=ToT
CQ-DEPEND=CL:210391,CL:210568,CL:210587
TEST=manual
make runtests
make clean
Also build and test:
- normal image
- test image
- recovery image
- firmware shellball
Note that this CL depends on simultaneous changes to the chromeos-initramfs
ebuild.
Change-Id: If791b5e9b5aac218ceafa9f45fc1785f16b91a64
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/210403
I don't think these utilities are needed any longer, so mark them as
deprecated. They will still be built and can be run via futility, but
invoking them directly will fail with a warning message.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: Ie704f2cecc3c37c91e4a0ffbcbcf94e2bf3ba05b
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/208775
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This cleans up the Makfile and test scripts a bit, and adds a new test for
the builtin commands.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: Ibf5aa867d4dcabc0e46daac6633036b035c99ac8
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/207718
The alignment rules were complicated and not always correct. This change
sorts pointers instead of structures, and aligns the pointer table better.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: I16c4e9b777fffe7537127aba15413b54e8e0f0a4
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/207717
Reviewed-by: Randall Spangler <rspangler@chromium.org>
There are a few utilities that print the full path of any file they open.
This isn't necessary, and it just has to be ignored when running regression
tests from different directories.
This just removes that extra noise.
BUG=chromium:224734
BRANCH=ToT
TEST=make runtests
Change-Id: I4291bca7952a0d7371f8682b7d57545361c6341c
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/207619
Reviewed-by: Randall Spangler <rspangler@chromium.org>
A few places in the code through up warnings when building with strict
compiler flags. Correct these.
BUG=chrome-os-partner:21115
BRANCH=pit
TEST=manual
Build with:
FEATURES=test emerge-peach_pit vboot_reference
and see that iot now succeeds. Warnings include:
host/arch/arm/lib/crossystem_arch.c: In function 'ReadFdtValue':
host/arch/arm/lib/crossystem_arch.c:93:8: error: ignoring return value of 'fread', declared with attribute warn_unused_result [-Werror=unused-result]
Change-Id: I765723636e5f8979b794925c7b610081b2849026
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/66174
It seems likely that this is causing the reported autoupdate failure,
although it will take a bit of time to understand the mechanism. For
now, a revert seems safest.
BUG=chromium:20939
BRANCH=none
TEST=none
This revert should be safe since it takes us back to the previous behavior.
I will work on repeating the problem and diagnosing it properly.
This reverts commit c1bbc75e3b.
Change-Id: I303fcbf45d835639b6d5a7f5f6423d0c18890ddf
Reviewed-on: https://gerrit.chromium.org/gerrit/62004
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Commit-Queue: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
New ARM boards use an FDTMAP, which is basically just an FDT. This means
that we don't have two potentially conflicting flash maps in the image.
Flashrom supports this without issue, but the firmware signer needs
dump_fmap to support it also.
It would be nice if we had libfmap implementing all this, but this is
apparently a long-running clean-up task. So the next best thing is to
deal with the flashmap in one program - i.e. flashrom.
So support FDTMAP in futility by redirecting 'dump_fmap -x' to flashrom.
BUG=chromium:256912
BRANCH=none
TEST=manual
With the flashrom -x option added, run this:
sudo futility dump_fmap -x /build/peach_pit/firmware/image-peach-pit.bin
gbb_utility --rootkey=rootkey.bin GBB
See that the various chunks of data from the image are output and that
gbb_utility completes without error.
Change-Id: Id8b2c774a16bdd07968765be5e6609b1b0661a0f
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/60862
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
This stops creating dump_fmap as a standalone utility and builds it into
futility. Since it was already invoked as a symlink, no user-visible changes
should be observed.
BUG=chromium:224734
BRANCH=none
TEST=manual, trybots
sudo FEATURES=test emerge vboot_reference
FEATURES=test emerge-$BOARD vboot_reference
Change-Id: I68d1bea0c1867043b2633e15509b95c2717009a7
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/47672
Reviewed-by: Randall Spangler <rspangler@chromium.org>
In almost every case we want the dynamically linked version of futility,
because it's smaller and the openssl functions require it (they use
dl_open() to invoke the correct RSA libraries).
However, the AU shellball requires three futility functions (crossystem,
gbb_utility, and dump_fmap). Those pretty much have to be built statically,
because they run from the new rootfs and packing all the dynamic libraries
into the shellball is way too large and complicated.
This change prepares to build both futility (full featured) and futility_s
(just those functions). The scripts that create the AU shellball will
already choose futility_s from /build/$BOARD/.
BUG=chromium:224734
BRANCH=none
TEST=none
CQ-DEPEND=CL:47589
Nothing to test just yet. The AU shellball is currently broken (it uses the
dynamic version of futility and doesn't copy the symlinks anyway), so this
should have no effect. We just need to ensure that the _s version doesn't go
into any of the other images.
Change-Id: I60b8dcd17e135f12a0d29ddacfb9fe8275567c70
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/47466
This tweaks the Makefile and adds a couple of placeholder tests to prepare
for testing the builtin futility operations. There aren't any useful builtin
functions yet, but this lets us start adding them along with the tests.
BUG=chromium:224734
BRANCH=none
TEST=none
This doesn't actually do anything yet.
Change-Id: Iff0ca514f7d26346f072bd80a3bcd04621284843
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/47432
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This moves all the old userspace utilities generated by vboot_reference into
a subdirectory not in $PATH, and replaces them with symlinks to a single
executable named 'futility'. At the moment that utility just execs the
original utilities (optionally logging that fact first).
Ultimately, the old utilities will be subsumed into a single binary instead
of multiple separate executables.
There is a matching CL needed to make the recovery image creation work.
BUG=chromium-os:37062
BRANCH=none
CQ-DEPEND=CL:44864
TEST=auto
To test, build everything, test everything. It should work as before in all
cases. I have built normal images, test images, factory installers, recovery
images; they all seem to work.
I've run trybots on daisy-paladin link-paladin lumpy-paladin and alex-paladin.
Change-Id: Ie93db676f2ed2a64e4b13b3b5dc6b65a77db0f8c
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/44871
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Cleaning up the Makefile to add correct dependencies, avoid using variables
before they're fully defined (which required .SECONDEXPANSION to work
around), generally improve readability.
There are so many changes that there's no point in trying to compare old
vs new - just look at the final result. And this still isn't perfect, but
it's a big step forward.
BUG=chromium-os:37062
BRANCH=none
TEST=manual
All these should continue to work:
make && make runtests
sudo emerge vboot_reference
emerge-$BOARD vboot_reference
Change-Id: Ic05a19231155d7e6427732e41d1682012d2dcdca
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/41768
Reviewed-by: Randall Spangler <rspangler@chromium.org>
