ActivityMetricsLoggers build the proto with the compilation reason first
and then the compilation filter. Instead of changing the logging code,
we prefer to change the proto field order: this will facilitate the
speed up the resolution without the need to purge data or wait for
further release builds.
It is ok, because both fields are in fact int32 proto values so the ABI
does not change.
Test: m
Bug: 178108751
(cherry picked from commit cf676e8028c2611461dc3e009dd9fe771f44bdb2)
Merged-In: I36ec354d273dceace5245d7634c68ac63dc5a8cd
Change-Id: I03c701efe2181ae7796547267902f884007593eb
Historically, we havent held a lock on ForceClearCache and
ClearCacheIfNecessary. This is not thread safe because kAllPullAtomInfo
must be accessed in a lock, especially now that pullers can be
registered/unregistered.
Test: atest statsd_test, wrote a repro cl.
Bug: 168156854
Bug: 173552790
Change-Id: I47d53a6d9d274bca4c78dbfd87e0097091b7b8cb
Merged-In: I47d53a6d9d274bca4c78dbfd87e0097091b7b8cb
Merged-In: I8bee7a0a6acecc1274d5acc0adb44c5dde8862e4
(cherry picked from commit f9a4bb18ff)
mUidFieldIndex, mExclusiveStateFieldIndex,
mAttributionChain{Start|End}Index are stored using int8_t.
These values can overflow when mValues.size() exceeds INT8_MAX.
Add proper bounds checking to ensure these values don't exceed
INT8_MAX.
Bug: 174488848
Bug: 174485572
Test: m
Test: statsd_test
Test: TreeHugger
Change-Id: I416e2de06ca2935017e15d4e91000652831e6b6c
Bug: 172181610
Test: (1) run statsd_testdrive 277, and check the
metric output correctness after content switching.
(2) run statsd_testdrive 90, and check the metric
output correctness after launching the output dialog
via Sound Setting / Volume Rocker / Media Object Chip.
(3) atest MediaOutputControllerTest
(4) atest MediaOutputDialogTest
(5) atest MediaOutputBaseDialogTest
Change-Id: I4094a510b5483a456522da9d0d8a6eee1526f687
Merged-In: I0fea0c59afcb44fc1d22920036107524f221a7be
Add free memory / swap as well as the kill reason to better understand
and diagnose lmkd behaviour in the field
Test: build with corresponding lmkd change
Bug: 168117803
Merged-In: I190f70758daa442b8605e17960c74574676fa3ae
Change-Id: I190f70758daa442b8605e17960c74574676fa3ae
Before this change, it was possible for the code to suffer an out of
bound error.
Bug: 150706572
Test: make
Change-Id: I3e8d37f2ee3c942bc9b176edee043557b005c757
(cherry picked from commit 8ff5315e98)
android_uiautomator is distributed as part of the SDK and should use
-target 8 for compatibility with old JDKs.
Bug: 150153790
Bug: 162522383
Test: javap -v -cp out/target/common/obj/JAVA_LIBRARIES/android_uiautomator_intermediates/classes.jar com.android.uiautomator.core.Configurator | grep "major version"
Change-Id: I6871ab5436232ad3bffff6bd26adf2de6b3c06ff
Add a Telecom command to stop the number blocking suppression that
happens after an emergency call.
Bug: 158798240
Fixes: 160667708
Test: atest CtsTelecomTestCases
Change-Id: I122de1962fb6aa48a3092c2aa323dcbdcbb8f42e
Define a new pulled atom which will be logged in MediaProvider.
The atom collects general access stats for external storage:
* Total number of accesses
* Total number of file path accesses
* Total number of secondary storage accesses
* List of accesses mime types
Test: build
Test: make statsd_testdrive ; ./out/host/linux-x86/bin/statsd_testdrive \
-p com.google.android.providers.media.module 10085
Bug: 149669087
Change-Id: Ic0ebb040a05ec06603f1a617f290af6a2f21916d
Right now, if an alarm fires exactly on time (within 1 second), or
exactly on a multiple of its period, statsd will reset the alarm for the
same time, which can result a loop of statsd setting an alarm, the alarm
firing, statsd setting an alarm at the same time, etc.
The fix is that if an alarm fires on time, we should move to the next
period.
Test: atest statsd_test (added to a test, which would fail without the
code change)
Bug: 159098080
Change-Id: I150538cb1d3d5a2b943df2a32d92f8f84bb96626
Introduce strategy which uniformly samples OPs, without biasing in
favour of specific ops.
Use this strategy half of the time.
Test: atest RuntimeMessageCollectionTest && statsd_testdrive 10069
Bug: 143519689
Change-Id: I481d3a9ef9c66f04577b57432cecdf2c02055e58
Without this, RANDOM_ONE_SAMPLE would result in statsd dropping all bar the
first shard. Adding the shard id as a dimension field preserves all the
data.
Bug: 158294266
Test: manual build & test
Change-Id: I94ffb610ef3651f460600e3df6b32c87fc430594
This is for migrating keystore logging to use statsd.
Test: Adding tests for logging is yet to be decided.
Change-Id: I66fbc2b8ed98a34f243387f8550b6523b13cb9a2
Bug: 157664923
Previously, if the condition was unknown for the entire bucket, the
bucket would be annotated with a NO_DATA rather than being dropped
because CONDITION_UNKNOWN. This updates all the buckets to be dropped
due to CONDITION_UNKNOWN. Note that this will make the condition unknown
problem appear worse, when in reality it is making our bucket drop
reasons more correct and should not have any impact on the true
CONDITION_UNKNOWN.
Also fix an issue where a test was failing due to a previous test
leaving data persisted to disk. This fix is a hack for now, and should
be properly fixed later.
Bug: 158897179
Test: atest statsd_test
Change-Id: Ic703600d37aa1873ec67cb1a279dcf349d48fc9f
