| Age | Commit message (Collapse) | Author |
|---|
|
We can return a FileDescriptor object, which encapsulates logic needed
to write to a COW. This way, filesystem verfication action can use the
turend value directly for computing verity and hash the partition.
Test: th
Change-Id: Iafe9699ef0cc15961641fc94f8ad2820230a56e1
|
|
When DynamicPartitionControlAndroid is constructed, it initializes both
source and target slot to -1. These values get updated during
PreparePartitionsForUpdate call. And we only
PreparePartitionsForUpdate() when applying an OTA or applocating space
for an OTA(not when verifying OTA metadata). Which means if
VerifyPayloadApplicable() is called before any call two other APIs, we
could be using an "Uninitialiazed" dynamic partition control.
To mitigate this problem, we pass in source_slot at ctor of
DynamicPartitionControl, also make IsDynamicPartition() api take in a
slot number to avoid reading uninitialized member fields.
Bug: 181643302
Test: apply an OTA, abort, restart update_engine, verify a payload
Change-Id: I9a8a0fe8a9aca48e91241e15bdec33a1c1228553
|
|
If a slot doesn't support dynamic partitions, it's impossible to list
dynamic partitions on that slot. And we should just fall back to the
regular A/B in this case.
Bug: 180025432
Test: apply a retrofit package
Change-Id: I16c457b591e8c1d0cf1077a7be50dd9d8f61b8eb
|
|
This function returns whether Virtual A/B
compression is enabled for the ongoing update.
Test: TH
Bug: 178732971
Change-Id: I781802443afd1d8deac046a8da198658220f7ba0
|
|
Test: treehugger
Change-Id: I655e4d6114a25c58857679985cd5e08581c0faab
|
|
Test: treehugger
Change-Id: I393fc36372f9e979da1a6c7e49c7782a7372a5b9
|
|
During postinstall stage, update_engine needs all target
partitions(system_b, vendor_b) to be mounted. We add 2 APIs to dynamic
partition control, 1 for mounting all partitions, and UnmapAllPartitions for unmap all partitions.
Test: Treehugger
Bug: 168554689
Change-Id: I1047851f6cdae7f39f50bd1d50e6d3bfc0d9a7f5
|
|
Test: treehugger
Change-Id: I110c65e75a4594af51e68390e570ac31b50de8b5
|
|
Test: treehugger
Bug: 168554689
Change-Id: I732cd1ef55c5c4362ef88a640c2ca228f472af89
|
|
For partial updates, the metadata for untouched dynamic partitions
are just copied over to the target slot. So, verifying the extents
of these partitions in the target metadata should be sufficient for
correctness. This saves the work to read & hash the bytes on these
partitions for each resumed update.
Bug: 151088567
Test: unit tests pass, apply a partial update
Change-Id: I9d40ed2643e145a1546ea17b146fcdcfb91f213f
|
|
Implement the logic in PartitionUpdateGenerator. Here's the summary,
1. finds the a/b partitions (both static & dynamic) on the device
For partitions not included in the payload:
2. calculates the partition hash for filesystem verification
3. generates one SOURCE_COPY operation for each static partition
The order of the partitions are sorted. So the update will resume from
the correct operation in case it's interruptted.
Bug: 157778739
Test: run a partial OTA with boot & system_ext
Change-Id: I5683b85e3c6dab813a33d5144aceb996fd8163d4
|
|
If package schedules powerwash, snapshots on a Virtual A/B device
will be deleted. Tell SnapshotManager to handle this case.
Test: apply downgrade package
Test: apply package with wipe
Test: apply package then manually wipe
Bug: 152094219
Change-Id: I8601b661924abcc82956788e2f33cdb70d71f778
|
|
When ResetStatus() is called, delete snapshots to free up space on
VAB devices.
On regular A/B, there is no need to cancel the update completely so that
the update may be resumed next time. However, on VAB devices, a
cancelled update should not occupy previously allocated space.
Test: manually call update_engine_client --cancel when an update is in progress, then
call update_engine_client --reset_status, then
call snapshotctl dump to check that snapshots are deleted.
Then, re-apply the update.
Test: manually kill update_engine when an update is in progress, then
call update_engine_client --reset_status, then
call snapshotctl dump to check that snapshots are deleted.
Then, re-apply the update.
Test: manually apply an update until UPDATED_NEED_REBOOT, then
call update_engine_client --reset_status, then
call snapshotctl dump to check that snapshots are deleted.
Then, re-apply the update.
Test: manually apply an update until UPDATED_NEED_REBOOT, then
kill update_engine, then
call update_engine_client --reset_status, then
call snapshotctl dump to check that snapshots are deleted.
Then, re-apply the update.
Bug: 147696014
Change-Id: Ic7049772091a34e0e666fd7ae361ef474d5a28aa
Merged-In: Ic7049772091a34e0e666fd7ae361ef474d5a28aa
|
|
For SOURCE_COPY operations like
563412 -> 123456
OptimizeOperation optimizes it to
5612 -> 1256
and skip writing blocks that does not need to be written for snapshot
partitions.
Bug: 148623880
Test: update_engine_unittests
Test: apply incremental OTA
Change-Id: Ifd2c3851f703f272a74c8f0e9a1c9a82dbcce3e3
|
|
It is replaced with GetCleanupPreviousUpdateAction now.
Test: builds
Change-Id: I8cb281bf811052d1dfea6c14c8685a90915cafa8
|
|
On update_engine starts, schedule CleanupPreviousUpdateAction that calls
CleanupSuccessfulUpdate to do necessary cleanup as soon as possible.
In the good case, update_engine initiates merge when
sys.boot_completed, and clean up snapshots.
If the update is
rolled back or partitions are flashed, the following happens (on
a Virtual A/B device):
- UpdateAttempterAndroid::CleanupSuccessfulUpdate is called
- DynamicPartitionControlAndroid::CleanupSuccessfulUpdate is called
- SnapshotManager::InitiateMergeAndWait is called
- SnapshotManager::RemoveAllUpdateState(before_cancel) is called
- before_cancel is called,
DeltaPerformer::ResetUpdateProgress is called
- All update states in update_engine is reset.
- SnapshotManager proceeds to delete snapshots
- All update states in SnapshotManager is reset.
Hence, on an VAB device, when an update is rolled back or partitions
are flashed, the whole update needs to be re-applied
(while in A/B, it skips writing and directly start verifying hashes of
the target partitions because the update markers are still there).
Bug: 147696014
Test: apply OTA then reboot, inspect logs and do `snapshotctl dump`
Change-Id: I0fc5e7768dfb53e4fd474f2d8d85d2a1b615a88b
|
|
This is a wrapper over SnapshotManager::WaitForMerge. It waits until the
previous update is merged, then return.
Bug: 138808328
Test: manual with update_engine_client
Change-Id: If44854810f37dd959ffdf3f62f26528867a71fc8
|
|
Add out parameter required_size to PreparePartitionsForUpdate
to indicate the total size required on /userdata in order
to apply the update.
Bug: 138808058
Test: update_engine_unittests
Change-Id: I2768d13671e212fd24a1a22811b50c9738834459
|
|
On Virtual A/B devices, don't skip SOURCE_COPY on
static partitions.
Test: update_engine_unittest
Test: incremental update to self
Change-Id: I5c93b501e09f50f559151eb77d83052373c90d0d
|
|
For some devices, some InstallOperations can be skipped during the
update process. An example is the SOURCE_COPY operation with same
source and destination locations for Virtual A/B devices.
This patch extents the DynamicPartitionControl interface with the
ShouldSkipOperation() method to query if given operation should be
skipped or not.
Bug: 141207436
Test: build
Change-Id: I94dbf1d33d531944e04cb2ffcba274106e866d5d
Signed-off-by: Alessio Balsini <balsini@google.com>
|
|
Add BootControlInterface::GetDynamicPartitionControl, which
exposes the internal DynamicPartitionControlInterface object.
BootControlStub / FakeBootControl / BootControlChromeOS uses
DynamicPartitionControlStub (all functions succeeds).
BootControlAndroid uses DynamicPartitionControlAndroid.
GetPartitionDevice is exposed so that BootControlAndroid can use it.
Follow-up CLs delete duplicated PreparePartitionsForUpdate
and Cleanup from BootControlInterface so that BootControlAndroid remains
a thin wrapper of the HAL (+GetPartitionDevice, which exists before
dynamic partitions.)
Test: update_engine_unittests
Change-Id: Ifc2aa2ee8a63ef581c8ebc562ec158794ac51dfd
|
