| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch fixes a sanitizer issue in cgpt where a GPT entries array may
have been passed even though it was not loaded from disk (parsing an
uninitialized buffer). The GPT library seems to have been written with
the assumption that both headers and entries would always be loaded and
it could recover even if only the primary header and the secondary
entries were valid. In practice, this doesn't really work because the
caller doesn't know how to read entries for an invalid header.
Therefore, change the code so that entries are only assumed to be loaded
for valid headers.
Also fix some minor problems with loading GPTs by aligning sizes up (not
down) to the next sector boundary and making sure we always allocate the
maximum amount of space for entry arrays, even if the current header may
not need that much (in case a repair wants to overwrite it).
This practically reverts CL:276766 which becomes obsolete (and was
really just a dirty hack to hide an underlying problem).
BRANCH=none
BUG=chromium:1017797
TEST=make runtests
Change-Id: I86c601dc074261d53f013b98ae214efdc44f3563
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1885098
Reviewed-by: Mattias Nissler <mnissler@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also standardize on using hex for printing ASCII key values
across vboot_ui.c and vboot_ui_menu.c.
BUG=b:124141368
TEST=make clean && make runtests
BRANCH=none
Change-Id: Ib10288d95e29c248ebe807d99108aea75775b155
Signed-off-by: Joel Kitching <kitching@google.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1840191
Reviewed-by: Joel Kitching <kitching@chromium.org>
Tested-by: Joel Kitching <kitching@chromium.org>
Commit-Queue: Joel Kitching <kitching@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The TYPE_BASIC_DATA (called TYPE_LINUX_DATA before) is used by both
Windows and Linux systems, and has caused problems when dual-booting.
Modern Linux systems have been changed to TYPE_LINUX_FS.
In Chrome OS, we usually find the stateful partition by number (1)
instead of searching by type, so it should be fine simply replacing
default mapping in the cgpt tool.
BUG=chromium:944389
TEST=sudo emerge vboot_reference; make runtests
BRANCH=None
Change-Id: If18ff5180cbae5cdea8104f36203cffcf34db934
Signed-off-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1535456
Reviewed-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Matt Delco <delco@google.com>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Should have no extra line breaks in between local includes,
and should be sorted alphabetically.
BUG=b:124141368
TEST=make clean && make runtests
BRANCH=none
Change-Id: I83c25d30d7376712857314965a7d93f57190aa3f
Signed-off-by: Joel Kitching <kitching@google.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1776281
Tested-by: Joel Kitching <kitching@chromium.org>
Commit-Queue: Joel Kitching <kitching@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The drive details are the only place where block sizes are tracked,
so make sure we display them so the user knows the bytes-per-block.
Might as well dump all the other drive details while we're here.
BUG=None
TEST=CQ passes
BRANCH=None
Change-Id: I2f2546feec2d58a295ce69c2bfe89a5ecf392f46
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1773966
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Commit-Queue: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Standardize on the term "blocks" rather than "sectors" as this matches
our disk layout configuration files.
For cgpt show, split the specific field options out from the -i option.
The current text makes it seem like they're always required when using
the -i option, and it's not super clear that they only matter when the
-i flag is used.
BUG=None
TEST=CQ passes
BRANCH=None
Change-Id: Idfd939cfe9501da78e323bc1ce8c1fbf772fbaa9
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1773965
Reviewed-by: Julius Werner <jwerner@chromium.org>
Commit-Queue: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If people use -i0, the code runs as if the flag wasn't specified.
Since valid partition numbers are [1,128], and the lower layers
already reject values outside that range (except for 0), add an
explicit check to the CLI for -i0.
Trying to display specific fields w/out -i makes no sense, and
the lower layers just ignore it. Add an explicit check for it
so users don't try to do `cgpt show /dev/sda -s` and wonder why
the output is unchanged.
Passing more than one specific field selector like -s -b doesn't
work -- whatever flag was specified last wins. This isn't that
obvious to users, so throw an explicit error when it happens.
BUG=None
TEST=CQ passes
BRANCH=None
Change-Id: I7c98822b79b389824b544b128ede93458b678342
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1773964
Reviewed-by: Julius Werner <jwerner@chromium.org>
Commit-Queue: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
BUG=b:124141368
TEST=make clean && make runtests
BRANCH=none
Change-Id: Id97f544da845f7070555e5e8cc6e782b2d45c300
Signed-off-by: Joel Kitching <kitching@google.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1758151
Tested-by: Joel Kitching <kitching@chromium.org>
Reviewed-by: Joel Kitching <kitching@chromium.org>
Commit-Queue: Joel Kitching <kitching@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch enables -ffunction-sections, -fdata-sections and
-Wl,--gc-sections for host builds. These flags already get passed by
firmware builds anyway, so having host builds match that behavior should
be a good idea in general. They may also occasionally help save a bit of
code size (though not much since vboot is a library, but I still get
about half a KB out of futility), and they will prevent clang from
omitting relocations for function calls inside the same file, which
means we don't have to splatter test_mockable all over our codebase
anymore. (We still need it for vb2_get_gbb() since that is so small that
both GCC and clang want to inline it, even if they are outputting a
discrete copy anyway.)
(Also add a comment about why GenerateGuid() has nothing do to with this
even though it is also a weak function, and why it is like that.)
BRANCH=None
BUG=chromium:991812
TEST=make runtests with both GCC and clang
Change-Id: Iede9d29e20b99b75a0c86bc7ecb907d2a0e5e3a1
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1754969
Legacy-Commit-Queue: Commit Bot <commit-bot@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These variables are all uint64_t, and so we need to use PRIu64 to
ensure they are printed with the correct format specifier.
BUG=none
TEST=make clean && make runtests
BRANCH=none
Change-Id: Idb8fee0ef525d224670a9d2b3a7915be1b19fd21
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 199873, 199878, 199889
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1737200
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use %zu, which is the format specifier for size_t.
BUG=none
TEST=make clean && make runtests
BRANCH=none
Change-Id: I3317c2f6a7b9d95c22a43ae3d786c7d7380342ad
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 199882
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/vboot_reference/+/1737706
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch adds a bunch of more warnings that are already enabled in
coreboot and thus already enabled for firmware builds anyway (because
coreboot just passes its CFLAGS through). Enabling it in the vboot
Makefile means they also apply to host utilities and tests, which sounds
desirable for consistency.
Fix enough of the cruft and bad coding practices that accumulated over
the years of not having warnings enabled to get it to build again (this
includes making functions static, removing dead code, cleaning up
prototypes, etc.).
Also remove -fno-strict-aliasing from the x86 firmware build options,
because it's not clear why it's there (coreboot isn't doing this, so
presumably it's not needed).
BRANCH=None
BUG=None
TEST=make runtests
Change-Id: Ie4a42083c4770a4eca133b22725be9ba85b24184
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1598721
Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Respect any existing IGNOREME signatures and do not attempt to make any
modifications until a user has explicitly removed this flag via cgpt
legacy or other method.
BRANCH=None
BUG=chromium:948742
TEST=chromeos-install on veyron_minnie with 4.19 kernel
Change-Id: I110a95ee0c136ebbe2274139deebcaacde712e80
Signed-off-by: Ryan Case <ryandcase@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1572440
Commit-Ready: Ryan Case <ryandcase@google.com>
Tested-by: Ryan Case <ryandcase@google.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
DriveOpen mallocs a couple of buffers, but DriveClose only freed them if a
particular flag was passed causing GptSave to be called. Move the free calls
out of GptSave so that the buffers are always freed, and add DriveClose calls
to a couple of cgpt functions that are missing them.
BUG=chromium:940543
TEST=precq passes, manual testing with cgpt/valgrind shows that memory leaks
for cgpt find, show are fixed.
BRANCH=none
Change-Id: I58aeddfa6b8b4715ba4f8e064e95a660371a01c9
Reviewed-on: https://chromium-review.googlesource.com/1516413
Commit-Ready: Fletcher Woodruff <fletcherw@chromium.org>
Tested-by: Fletcher Woodruff <fletcherw@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
cgpt's C/C++ bindings use non-const char* parameters leading to
compiler errors if a user attempts to pass a const char* parameter
rather than creating an unnecessary mutable copy.
Since the code doesn't need to modify the parameters, change them to
const char* to make the library easier to use.
BUG=none
TEST=builds and test cgpt cli tool on-device.
CQ-DEPEND=CL:1460081
BRANCH=none
Change-Id: I6552db159e3dc4d9d07bb889a3f1e4e890b33cb0
Reviewed-on: https://chromium-review.googlesource.com/1459848
Commit-Ready: Fletcher Woodruff <fletcherw@chromium.org>
Tested-by: Fletcher Woodruff <fletcherw@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The lack of -D support in CgptEdit introduced a test failure. This
change adds support for -D.
BRANCH=none
BUG=chromium:605348
TEST=Verified that prior to this change the tests failed:
cros_workon --host start vboot_reference
sudo FEATURES=test emerge vboot_reference
The tests fail in a different area prior to CgptEdit, so I applied the
following temporary change to Makefile to see the relevant failure:
ifeq (${MINIMAL},)
# Bitmap utility isn't compiled for minimal variant
- test_targets:: runbmptests runfutiltests
+ test_targets:: runbmptests # runfutiltests
# Scripts don't work under qemu testing
With this change the tests pass.
Change-Id: Ia2127a3537c72e4ea6daf59c5c33b8701a89b0f6
Signed-off-by: Matt Delco <delco@google.com>
Reviewed-on: https://chromium-review.googlesource.com/1436496
Tested-by: Matt Delco <delco@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This change allows CgptEdit to be called via the API. Prior to this
change link fails in an app that uses CgptEdit due to undefined
reference. The underlying implementation wasn't checking set_unique so
I've fixed that as well.
BRANCH=none
BUG=None
TEST=Added CgptEdit(0) call to extern.c and verified that build failed.
Added cgpt_edit.c to Makefile and confirmed that build is now
successful. Successfully ran unit tests on both vboot_reference and the
app I'm working on that calls CgptEdit (which also has a unit test for
setting the drive ID).
Change-Id: Ie0a46ff96406eb83d0564d3f1eac978e0565ed76
Signed-off-by: Matt Delco <delco@google.com>
Reviewed-on: https://chromium-review.googlesource.com/1361948
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This change adds a command to cgpt to
change the GUID of the drive.
BRANCH=none
BUG=None
TEST=Compiled and ran utility to verify that GUID changes.
Also verified that the new and existing tests completed
successfully.
Change-Id: Ia8a815447509626312e2b06c6f293901290c73c3
Signed-off-by: Matt Delco <delco@google.com>
Reviewed-on: https://chromium-review.googlesource.com/1171834
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A later change I authored has the tests check the GUID of the drive.
When the primary table is ignored the GUID from the secondary wasn't
being displayed either. This change has the details of the secondary
table get displayed when the primary table is ignored.
BRANCH=none
BUG=None
TEST=Compiled. The change was runtime tested as part of a larger change though
I didn't unit test this particular change after it was split out into a
separate commit.
Change-Id: I300511cf65c67f4888e08ab49cd72c7acf234507
Signed-off-by: Matt Delco <delco@google.com>
Reviewed-on: https://chromium-review.googlesource.com/1173410
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Remove 512 sector block size restriction so that UFS, with sector block size 4096
or greater, can be used. The sector block size is queried from the kernel with
ioctl(BLKSSZGET) or queried from depthcharge with VbExDiskGetInfo().
BUG=b:77540192
BRANCH=none
TEST=manual
make runtests passed.
Tested firmware on Kevin and boot to kernel from disk.
Executed cgpt show /dev/mmcblk0 on eve device and verified output was correct.
Should be tested on device with sector block size greater than 512.
Change-Id: I8165c8ee4da68180eecc8d12b3fb501cc5c60a5d
Reviewed-on: https://chromium-review.googlesource.com/1007498
Commit-Ready: Sam Hurst <shurst@google.com>
Tested-by: Sam Hurst <shurst@google.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bit 0 in the GPT partition attributes is defined to indicate whether a
partition is required by the platform. This CL adds the support for
managing this bit to cgpt.
BUG=b:70807006
BRANCH=None
TEST=Run unit tests.
Change-Id: Iaf87c828438b3df6730de502ae420fcf4c61277b
Reviewed-on: https://chromium-review.googlesource.com/902196
Commit-Ready: Ben Chan <benchan@chromium.org>
Tested-by: Ben Chan <benchan@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The prioritize cmd expected primary gpt header and entries are valid.
Add a validity check to guarantee that.
BUG=b:35585016
TEST=Check on bob, the secondary gpt still valid after these commands:
dd if=/dev/zero of=/dev/mmcblk0 bs=512 count=64
cgpt prioritize /dev/mmcblk0
Change-Id: I853e25211a46fa3ce93cef1e5169b93b0a6430a9
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Reviewed-on: https://chromium-review.googlesource.com/604814
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Partition size on 4TB drive can overflow 32-bit integers. Running "cgpt
show /dev/sda" prints negative number for the state partition (sda1).
BRANCH=none
BUG=none
TEST=Run "cgpt show /dev/sda" with 4TB drive.
Change-Id: I56f3b43594028695745de8c5a1626d940a3b4c5b
Reviewed-on: https://chromium-review.googlesource.com/546879
Commit-Ready: Erdi Chen <erdi@google.com>
Tested-by: Erdi Chen <erdi@google.com>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A partition's name would always start with the disk name. And in
/proc/partitions, the partitions are always listed right after the
disk.
Let's filter out devices which are not followed by partitions when
go through the /proc/partitions.
BUG=chrome-os-partner:62955
TEST=run "cgpt find -t kernel" on kevin, no more this warning:
blk_update_request: I/O error, dev mmcblk0rpmb
Change-Id: If200a2476d26b1beaf644838d47ea2e60552855e
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Reviewed-on: https://chromium-review.googlesource.com/444492
Reviewed-by: Julius Werner <jwerner@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Originally, vboot1 code used VbExMalloc() and VbExFree() since it needed
to talk to EFI firmware that didn't have standard malloc() and free().
Now, coreboot and depthcharge implement them as wrappers around those
standard calls. vboot2 code already calls them directly, so let vboot1
code do that too.
BUG=chromium:611535
BRANCH=none
TEST=make runtests; emerge-kevin coreboot depthcharge
Change-Id: I49ad0e32e38d278dc3589bfaf494bcf0e4b0a4bd
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/400905
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
At this point, all that's left are a few constants in the cryptolib
header files, and they're only used by host-side code. So move them to
a host-side header file and get rid of cryptolib.
BUG=chromium:611535
BRANCH=none
TEST=make runtests; emerge-kevin coreboot depthcharge
Change-Id: I2235f0e84e13fef313afe54e749b73744b157884
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/400903
Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
find: A p is added betwen device name and partition whenever the last
character of a device is a number, as written in disk_name() in kernel
block/partition-generic.c file.
debug_vboot: Add regex for nvme device.
BUG=chromium:655192
BRANCH=none
TEST=Check that when a machine boots from NVMe, chromeos-setgoodkernel
set "successful" field properly.
Run " dev_debug_vboot --cleanup", check the NVMe device kernel
partitions are verified.
Change-Id: I6a9342c95500fa582f51f06e48c1ff90684c2a27
Signed-off-by: Gwendal Grignou <gwendal@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/398338
Reviewed-by: Mike Frysinger <vapier@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Assorted minor code issues, which we should fix so any new errors stand
out more.
BUG=chromium:643769
BRANCH=none
TEST=make runtests
Change-Id: I8fcf0c51e33d5dc49f650f4069f1579091cf188d
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/383713
Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bit 2 in the GPT partition attributes has been allocated as the legacy
bios boot (equivalent to the "active" or "boot" flag in MBR). If we
try to boot images on newer x86 systems, syslinux dies because it can't
find any GPT partition marked bootable.
Update the various parts of cgpt add & show to manage this bit. Now we
can run:
cgpt add -i 12 -B 1 chromiumos_image.bin
And the EFI partition will be marked bootable.
BUG=chromium:644845
TEST=vboot_reference unittests pass
TEST=booted an amd64-generic disk image via USB on a generic laptop
BRANCH=None
Change-Id: I78e17b8df5b0c61e9e2d8a3c703e6d5ad230fe92
Reviewed-on: https://chromium-review.googlesource.com/382411
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The documentation claims the -A option shows the raw 64-bit attributes
field when in reality it only shows the high reserved 16-bits. Change
the docs to match the code.
BUG=chromium:644845
TEST=vboot_reference unittests pass
BRANCH=None
Change-Id: If163896ddbca0dc27ac8205db313031e73a68fd7
Reviewed-on: https://chromium-review.googlesource.com/382431
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Most of the cmd funcs had the same logic copied & pasted multiple times
over. Unify them into a common header.
BUG=chromium:644845
TEST=precq passes
TEST=passing invalid args to some funcs is caught
BRANCH=None
Change-Id: Ib7212bcbb17da1135b2508a52910aac37ee8e6cd
Reviewed-on: https://chromium-review.googlesource.com/382691
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Assorted minor code issues, which we should fix so any new errors stand
out more.
BUG=chromium:643769
BRANCH=none
TEST=make runtests
Change-Id: I9e7ce2ba226993fc53d1745c98381cb7cfcb7712
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/380448
Reviewed-by: Stefan Reinauer <reinauer@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This change replaces all calls to the old vboot1 SHA library with their
vboot2 equivalents.
This is the first in a long series of changes to move the core vboot kernel
verification into vb2, and the control/display loop out to depthcharge.
BUG=chromium:611535
BRANCH=none
TEST=make runtests; build samus firmware and boot it
Change-Id: I31986eb766176c0e39a192c5ce15730471c3cf94
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/344342
Tested-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Now that we have support for the IGNOREME signature in cgpt, we need a
way to set it on an existing disk. The easiest option is to shoehorn
this into the cgpt legacy command, because that's already made to modify
GPT header signatures (really, it would be nice to rename it to cgpt
signature or something, but let's not break existing uses for now).
BRANCH=None
BUG=chrome-os-partner:52595
TEST=unit tests
Change-Id: If2835fec28a9c39373abd050e2e057f73e5ec700
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/340073
Reviewed-by: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch makes cgpt aware of a special "IGNOREME" GPT header signature
string that may appear in either the primary or the secondary GPT and
cause cgpt (and other cgptlib clients) to completely ignore that GPT. It
will continue to function correctly for all other purposes (using the
data from the non-ignored GPT), but never write any data back to the
ignored GPT.
BRANCH=None
BUG=chrome-os-partner:52595
TEST=unit tests
Change-Id: I7e53542385ae9d8d24dc25b75e91f4ff4917f66f
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/340072
Reviewed-by: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The point of having two GPTs is to always have a known good one if one
of them gets corrupted. One of the most obvious ways that could happen
is if the write stopped half-way through (e.g. due to a crash or random
power loss).
Unfortunately, the way we currently save modified GPTs can leave both
copies invalid if we stop writing at just the wrong time. Since a GPT
header contains a checksum over the GPT entries, we need to write both
the header and entries for one GPT (and make sure they're synced to
disk) before we start writing the other.
BRANCH=None
BUG=chrome-os-partner:52595
TEST=None
Change-Id: I2d4b56bcfba9a94395af5896f274ebade9e39081
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/340071
Reviewed-by: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The major() func is defined in the sys/sysmacros.h header, so include it
explicitly for the prototype. Upstream C libs are moving away from having
sys/types.h include it all the time implicitly.
BUG=None
TEST=precq passes
BRANCH=None
Change-Id: I56b84138f08ded2376193403f9c9db22c5f24f71
Reviewed-on: https://chromium-review.googlesource.com/339680
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When one of GPT headers is invalid the corresponding partition table
is not loaded and corresponding pointers in GptData are NULL.
GptRepair will try to memcpy one entries table to another which
results in SIGSEGV.
This change fixes it by freeing and then reallocating bad copy of
partition table. This potentially fixes problems which would occur
if two tables have different size.
Change that initially introduced this problem by not always allocating
secondary_entries:
https://chromium-review.googlesource.com/223800
TEST="cgpt repair" works where it previously didn't
TEST=make runtests
BUG=brillo:1203
BRANCH=none
Change-Id: Ibb2fcf33faa5ba157b0865d04c90ee3f26eee113
Reviewed-on: https://chromium-review.googlesource.com/276766
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Queue: Andrey Ulanov <andreyu@google.com>
Tested-by: Andrey Ulanov <andreyu@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This shows an error message when cgpt.bin failed to run, with its
reason. Without this patch, "cgpt" would just fail and return -1 in
that case making it difficult to know the reason of the failure.
BUG=chrome-os-partner:36061
TEST=replaced this binary in storm recovery initramfs 6699.0.0 and it
shows the error message if cgpt.bin is not installed.
BRANCH=None
Change-Id: I3ffaba5a63c491ac7d5b16086d5ae21005f40317
Reviewed-on: https://chromium-review.googlesource.com/251868
Reviewed-by: Alex Deymo <deymo@chromium.org>
Commit-Queue: Alex Deymo <deymo@chromium.org>
Trybot-Ready: Alex Deymo <deymo@chromium.org>
Tested-by: Alex Deymo <deymo@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The following works from a Mac with these changes:
make Q= ARCH=arm HAVE_MACOS=1 `pwd`/build/futility/futility
Only vbutil_keyblock and vbutil_kernel have been exercised.
BUG=none
TEST='make Q= ARCH=arm HAVE_MACOS=1 `pwd`/build/futility/futility'
BRANCH=none
Signed-off-by: David Riley <davidriley@chromium.org>
Change-Id: Ie69cfee0c650d4ff96be6322083a2fea1543ee39
Reviewed-on: https://chromium-review.googlesource.com/246773
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Tested-by: David Riley <davidriley@chromium.org>
Commit-Queue: David Riley <davidriley@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
I forgot to reset "ret" to zero in the case the two hashes come out the
same. This CL fixes that silly logic bug.
BUG=None
BRANCH=None
TEST=`cgpt show /dev/mtd0` should exit with status 0
Change-Id: I1c64ea215c861c8afb44e235d090c72a5d006d55
Reviewed-on: https://chromium-review.googlesource.com/243325
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@chromium.org>
Trybot-Ready: Nam Nguyen <namnguyen@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
flashrom spills out "Reading flash... SUCCESS" and so on to stdout. This
affects scripts calling to "cgpt". So this CL sets stdout to CLOEXEC
before exec'ing flashrom. It still leaves stderr and stdin unclosed.
BUG=None
BRANCH=None
TEST=cgpt show /dev/mtd0 2>/dev/null will not show any unnecessary text
Change-Id: Ide1414c56f63ffe8bc2385a797f166476dacd732
Reviewed-on: https://chromium-review.googlesource.com/242295
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Because we do not use "execvp", "cgpt.bin" is not resolved to the same
directory as "cgpt". So we need to resolve the original command to its
absolute path first, then append ".bin" to it.
BUG=None
BRANCH=None
TEST="cgpt" no longer fails.
Change-Id: Id22c2d97616867125e9744c00bbf527f8a176df4
Reviewed-on: https://chromium-review.googlesource.com/242294
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We always wrote back the rw_gpt file to NOR flash. This operation is too
slow. This CL compares if the original file has been changed by cgpt.bin
before writing the file back to NOR.
BUG=None
BRANCH=None
TEST=/usr/bin/cgpt show /dev/mtd0 now does not write back to NOR
Change-Id: I4c63f0d4da72f3674e06a896fa329f5fc964a885
Reviewed-on: https://chromium-review.googlesource.com/242293
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
flashrom has been fixed, we could enable this flag for better
performance.
BUG=None
BRANCH=None
TEST=cgpt show /dev/mtd0 works much faster now.
Change-Id: I7a6f5b8649b4293d9b4b4878ae8e599ea1c35289
Reviewed-on: https://chromium-review.googlesource.com/242292
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When working on NAND, we do not actually work with one device name. We
work on a temporary file instead. Moreover, depending on the type of the
partition, we need to show different devices.
BUG=None
BRANCH=None
TEST=All commands must be run on storm_nand
TEST=/usr/bin/cgpt.bin find -t kernel should print out /dev/mtd2
TEST=/usr/bin/cgpt.bin find -t rootfs should print out /dev/ubiblock5_0
TEST=/usr/bin/cgpt.bin find -t data should print out /dev/ubi1_0
Change-Id: Ia36777ffa6a9cfc7c8ec4b128e49ece140428238
Reviewed-on: https://chromium-review.googlesource.com/242291
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This CL adds a "cgpt_wrapper" binary that will forward to "cgpt" binary
as needed to transparently support NAND.
The idea is that if we are working with an MTD device, this binary will
exec out to "flashrom" to read in the GPT section, then exec out to the
actual "cgpt" binary to work on that file, and finally write it back
with "flashrom".
This CL introduces a make target "install_mtd" to install this wrapper.
The corresponding ebuild should use this make target when "mtd" USE flag
is on.
BUG=chromium:442518
BRANCH=none
CQ-DEPEND=CL:239573
TEST=unittest and some trial runs with mtdram
TEST=test on a real device with NOR flash and NAND
Change-Id: I54534afa9a970ec858f313f698c0eb210c827477
Reviewed-on: https://chromium-review.googlesource.com/239580
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Reviewed-by: Daniel Ehrenberg <dehrenberg@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch makes some small modifications to cgpt and vboot to
root out the last vestigates of a fixed 128-entry GPT:
- Get rid of the TOTAL_ENTRIES_SIZE constant and all users.
- Reduce MAX_NUMBER_OF_ENTRIES to 128 (which is what the GPT
spec specifies) so that this can be used for things like memory
allocations without additional overhead.
- Base the amount of GPT read/written on the number of entries
specified in the GPT header on disk/flash.
BUG=chromium:433433
TEST=make runalltests
TEST=Modified fmap to make an 8k RW_GPT, wrote a GPT with cgpt, then
rebooted and found that the GPT was correctly read after restarting
and the appropriate mtd partitions were present.
BRANCH=none
Change-Id: I45317377da20259caf04a7a4fa077a892b03c45f
Signed-off-by: Dan Ehrenberg <dehrenberg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/238245
Reviewed-by: Randall Spangler <rspangler@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As pointed out by Daniel in CL:234996, the half_size_sectors should have
been checked in both places.
This CL reworks that part of the code to read easier.
BUG=none
BRANCH=none
TEST=unittest
Change-Id: I8faea3b094c375e4fd1a604a8fe759af88943fdf
Reviewed-on: https://chromium-review.googlesource.com/235792
Reviewed-by: Daniel Ehrenberg <dehrenberg@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The standard says that entries table must be at least 16384 bytes. On
some of our devices, the NOR section is only 8 KiB and used to store
both primary and secondary tables. On this device, we can only store 24
entries.
Therefore, this CL adds support for non-standard entry table. It adjusts
the MIN_NUMBER_OF_ENTRIES to 16, and replaces GPT_ENTRIES_SECTORS with
CalculateEntriesSectors.
BUG=chromium:441812
BRANCH=none
TEST=unittest
Change-Id: I6b85b35ce5612c7abb22142f8252bd0d45b676c5
Reviewed-on: https://chromium-review.googlesource.com/234996
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Commit-Queue: Nam Nguyen <namnguyen@chromium.org>
Tested-by: Nam Nguyen <namnguyen@chromium.org>
|