We have a general need for overriding more paths (e.g. "/tmp"), mostly
for testing purpose. Rename CacheLocation to Paths, and use that to
manage TEMPORARY_{INSTALL,LOG}_FILE.
Test: mmma -j bootable/recovery
Test: recovery_component_test
Change-Id: Ia8ce8e5695df37ca434f13ac4d3206de1e8e9396
Bug: 78469201
Test: Wipe from recovery menu, check that wipe is logged correctly
and boot works as expected.
Change-Id: I5bc8ef1b83d78de8b5edba6cc17882edcc744356
We used to skip the update installation if the battery doesn't reach the
threshold or the boot reason is in blacklist. Afterwards, we just cleared
the BCB and reboot/shutdown the device. However, this leaves the device
in a brick state if the update is a retry; because a subsequent reboot
won't resume the update after the BCB gets cleared.
This CL skips the battery and bootreason check on retries. An alternative
solution is to conduct the check and shutdown the device without clearing the
BCB on INSTALL_SKIPPED; but users who don't have context may reboot the
device manually anyway.
Bug: 65288661
Test: Trigger the INSTALL_SKIPPED with insufficient battery, and check
the behavior with retry_count.
Change-Id: I0bd8ae9cee0e35aeeec3df469a78cec9ee1b8db8
An interrupted update may stash extra blocks in /cache, leading to a
failure when checking the cache size. We can save the incremented
retry_count in the BCB before installing the update; and distinguish
a fresh update from an interrupted one this way.
Bug: 68679601
Test: An interrupted update reapplies successfully.
Change-Id: Ic1403e1fd25a937c91ef34c14b92a0f6c8f1c0f4
This reverts commit 26436d6d60 to re-land
"Move error_code.h into otautil.".
This way it stops requiring relative path ".." in LOCAL_C_INCLUDES
(uncrypt and edify). Soong doesn't accept non-local ".." in
"local_include_dirs".
This CL needs to land with device-specific module changes (e.g. adding
the dependency on libotautil).
Test: lunch aosp_{angler,bullhead,dragon,fugu,sailfish}-userdebug;
mmma bootable/recovery
Change-Id: If193241801af2dae73eccd31ce57cd2b81c9fd96
This way it stops requiring relative path ".." in LOCAL_C_INCLUDES
(uncrypt and edify). Soong doesn't accept non-local ".." in
"local_include_dirs".
Test: mmma bootable/recovery
Change-Id: Ia4649789cef2aaeb2785483660e9ea5a8b389c62
The earlier commit in 2dfc1a3898
unintentionally changed the behavior. It gives a different result when
looking up non-existent mount points (e.g. /cache on marlin).
The logic behind volume_for_path("/xyz") is unclear:
- It's fine to return non-null value if it's called by
ensure_path_mounted() before accessing that file "/xyz". (Just based
on the function name, we're not actually having this case.)
- It should return nullptr if the caller is interested in the existence
of that particular mount point "/xyz".
This CL renames the function to volume_for_mount_point(), which does an
exact match by querying the given mount point from libfs_mgr. The former
volume_for_path() has been moved down to function scope for serving
ensure_path_mounted() only.
Test: Build and boot into recovery on bullhead and marlin respectively.
'View recovery logs'.
Test: 'Mount /system'
Test: 'Apply update from ADB'
Change-Id: I1a16390f57540cae08a2b8f3d439d17886975217
Add a new option "Run locale test" to check the background text
images (i.e. texts for "erasing", "error", "no_command" and "installing"
with different locales.)
Use volume up/down button to cycle through all the locales embedded in
the png file, and power button to go back to recovery main menu.
Test: Run locale test with bullhead.
Change-Id: Ib16e119f372110cdb5e611ef497b0f9b9b418f51
This reverts commit 8be0f39fec to reland
the change that removes EXPAND/STRINGIFY macros.
It's error-prone by putting anything into a string (e.g.
EXPAND(RECOVERY_API_VERSION) would become "RECOVER_API_VERSION" if we
forgot to pass -DRECOVERY_API_VERSION=3).
The initial attempt put RECOVERY_API_VERSION into common.h, which might
be included by device-specific codes but without defining that when
compiling the module. This CL avoids the issue by using a constant
in the header, with a static_assert in recovery.cpp that guards the
consistency.
Test: recovery_component_test
Test: Sideload OTAs on bullhead and sailfish respectively.
Change-Id: I12af3f73392a85554ba703f04970ec9d984ccbaa
This CL adds ScreenRecoveryUI::DrawWrappedTextLines() to better handle
long menu header texts. It does a word wrap at spaces, if available.
This avoids fixed-length menu headers being truncated on small screens.
Bug: 64293520
Test: On bullhead, boot into recovery with --prompt_and_wipe_data, and
check the prompt texts.
Change-Id: Ia22746583516dd230567a267584aca558429395e
- Changed to std::string based implementation (mostly moved from the
former make_parents() in updater/install.cpp);
- Removed the timestamp parameter, which is only neeed by file-based OTA;
- Changed the type of mode from int to mode_t;
- Renamed dirCreateHierarchy() to mkdir_recursively().
Test: recovery_unit_test passes.
Test: No external user of dirCreateHierarchy() in code search.
Change-Id: I71f8c4b29bab625513bbc3af6d0d1ecdc3a2719a
When there's no command specified when booting into debuggable builds
(such as using `adb reboot recovery`), we turn on the text mode (i.e.
recovery menu) directly. This CL fixes the issue to avoid showing the
background image in a flash while turning on the text mode.
Bug: 63985334
Test: `fastboot boot $OUT/recovery.img` and it shows the recovery menu
directly without the no command image in a flash.
Change-Id: Id86bbe346ab76c8defc95e2b423e695a86774b09
When /cache is unwritable, recovery hits a crash loop. Because it
passes nullptr to fileno(3) when writing back the locale file. This
prevents user from recovering a device - it cannot boot far enough to
recovery menu which allows wiping /cache.
Bug: 63927337
Test: Corrupt /cache and boot into recovery on bullhead:
1. m -j recoveryimage
2. fastboot erase cache
3. fastboot boot $OUT/recovery.img
4. recovery menu shows up.
Change-Id: I1407743f802049eb48add56a36298b665cb86139
Fixed by Loading the file_contexts specified in libselinux, whereas
previously recovery loaded /file_contexts which no longer exists.
Bug: 62587423
Test: build and flash recovery on Angler. Warning is gone.
Test: Wipe data and cache.
Test: sideload OTA
Change-Id: I11581c878b860ac5f412e6e8e7acde811f37870f
(cherry picked from commit 2330dd8733)
We have seen one case when bspatch failed likely due to patch
corruption. Since the package has passed verification before, we want
to reboot and retry the patch command again since there's no
alternative for users.
We won't delete the stash before reboot, and the src has passed SHA1
check. If there's an error on the patch, it will fail the package
verification during retry.
Bug: 37855643
Test: angler reboots and retries the update when bspatch fails.
Change-Id: I2ebac9621bd1f0649bb301b9a28a0dd079ed4e1d
We're not updating argc & argv during get_args(), so some boot
arguments missed when we set the boot message for retry.
Bug: 38383406
Test: boot command sets correctly during retry attempt.
Change-Id: Ie8583a22fad5e0084245e3431d4018518d508dfd
(cherry picked from commit 72449c9f99)
We're not updating argc & argv during get_args(), so some boot
arguments missed when we set the boot message for retry.
Bug: 38383406
Test: boot command sets correctly during retry attempt.
Change-Id: Ie8583a22fad5e0084245e3431d4018518d508dfd
This reverts commit ec9706738f.
Reason for revert: It's not a good idea to put RECOVERY_API_VERSION in
common.h, which might be included by device-specific codes (but with
RECOVERY_API_VERSION undefined).
Change-Id: I9feb9c64a5af3e9165164622a59b043aa28a8b8c
They are error-prone by putting anything into a string (e.g.
EXPAND(RECOVERY_API_VERSION) would become "RECOVER_API_VERSION" if we
forgot to pass -DRECOVERY_API_VERSION=3).
RECOVERY_API_VERSION is the only user (in bootable/recovery) that gets
stringified. Assign it to a typed var and sanity check the value.
Don't see other reference to the macros from device-specific recovery
directories (they can still define that locally if really needed).
Test: recovery_component_test
Test: Sideload an OTA on angler and marlin respectively.
Change-Id: I358bbdf8f0a99db5ce4c7bc2fdcafe8013501b64
It's already a global declared in common.h which is included by
adb_install.cpp.
Remove '#include "minadbd/fuse_adb_provider.h"' that's not needed by
adb_install.cpp (minadbd takes care of that).
Test: mmma bootable/recovery
Change-Id: I6d08b7abc706b4b05de2ef46a57ced2204ad297e
Bug: 37401320
Test: build and push OTA and hit adb reboot recovery,quiescent. The screen should remain off throughout the upgrade process.
(cherry picked from commit 8706a98aa6)
Change-Id: I79789a151f6faafda8ecc6198c2182cc2a91da70
recovery.cpp doesn't have a direct dependency on adb.h (only minadbd
does). 'adb.h' was once needed for DEFAULT_ADB_PORT, until being killed
by commit acb2a2fa4c.
Test: mmma bootable/recovery
Change-Id: I84f87988d01c3f0c3d1383cc3cffde2e7d5e7aea
Users can't do much after the install fails with the device showing
"error" under recovery. So our best choice is to reboot the device
since sometimes the system image is still bootable (i.e. on package
verification failure). At worst the device would stuck in a boot loop
where the users need the same professional knowledge to recover as
before.
Behaviors after installation failure (including data wipe):
If recovery text is visible:
No change.
If recovery text is not visible:
Old behavior: Wait under "error" screen. Reboot after UI timeout (120s)
if not connected to usb charger.
New behavior: Wait for 5s (shortens from the 120s timeout) under "error"
screen and reboot (w or w/o charger).
sideload-auto-reboot (only available for userdebug):
Old behavior: Reboot immediately after installation failure.
New behavior: Wait for 5s under "error" screen and reboot.
Bug: 35386985
Test: On angler user, device auto reboots 5s after a failing OTA.
Change-Id: I3ff0ead86e2ccec9445d6a2865bc2c463855f23c
Bug: 37401320
Test: build and push OTA and hit adb reboot recovery,quiescent. The screen should remain off throughout the upgrade process.
Change-Id: Ibed3795c09e26c4fa73684d40b94e40c78394d3f
Switch the locale header in the png files from Locale.toString() to
Locale.toLanguageTag(). For example, en_US --> en-us and sr__#Latn
--> sr-Latn. Also clean up recovery a bit to expect the new locale
format.
Bug: 35215015
Test: sr-Latn shows correctly under graphic tests && recovery tests pass
Change-Id: Ic62bab7756cdc6e5f98f26076f7c2dd046f811db
The call to read_metadata_from_package() is broken due to being passed
an invalid pointer (ZipArchiveHandle vs ZipArchiveHandle*). It's
introduced when switching from minzip to libziparchive. Compiler didn't
complain, because ZipArchiveHandle is typedef'd as void*, which legitly
accepts a void**.
Also clean up secure_wipe_partition() logs a bit by using android-base
logging.
Bug: 36427762
Test: Send a wipe package.
Change-Id: I791a0f09a066f1c257dae890e7ae13d02a02e78b
It used to return a REBOOT action on timeout, until the CL in commit
daefc1d442 that redefined the return value
of get_menu_selection() (changing from action to a menu index).
Prior to this CL, it was returning 0 (i.e. the value of Device::REBOOT)
to trigger the reboot. This CL specifies a return value of -1 to
indicate the timeout.
Test: Boot into a user build recovery; it reboots automatically on
timeout (120 sec).
Change-Id: I4aedb7a4628bf258017078fe73eb8b48a21d0ea8
Most of the changes are trivial.
Also update a dead reference to device_handle_key() in device.h comment,
and add some comments to get_menu_selection() function.
Test: `mmma bootable/recovery`
Change-Id: I59ef9213ec88ab35c0e7b8a7813ccf9c56dbd5c5
