| Age | Commit message (Collapse) | Author |
|---|
|
- also provide display_luts.h file for lut and lut entry
Bug: 377329333
Test: android.view.surfacecontrol.cts.ASurfaceControlTest#testSurfaceTransaction_setLuts
Flag: EXEMPT NDK
Change-Id: I4caa4572b5dffb3311c9743e0a2c25ca133d4834
|
|
|
|
This method should allow us to log more details of an event in case of
validation failures, without potential for infinite recursion in getter
methods.
Bug: 379368465
Test: add a test to MotionEventTest that calls
getHistoricalRawPointerCoords with an invalid pointer index, and
check the logs when run with `atest --host`
Flag: EXEMPT logs only
Change-Id: I9c7084cedbc7e6f6834cd1b401da04d07d22ce35
|
|
|
|
|
|
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3362726
Change-Id: I44268edb6478b5c293a311257dba905a2fdd4472
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3362726
Change-Id: Iac275cd94926d90ca0d5995284a56e95dfd6eeb4
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
This change makes the intial support check when
APerformanceHintManager starts pass the SupportInfo
object instead of just relying on the preferred rate
Bug: 367803904
Test: atest PerformanceHintNativeTestCases
Test: atest HintManagerServiceTest
Flag: EXEMPT refactor
Change-Id: Iba93310c58896a2b1d02c6156b4e67eaa766af5b
|
|
This patch implements a new way to create ADPF hint session,
with a new parcelable SessionCreationConfig. This object encapsulates
the required information for session creation, with a set of setters to
assign values to those fields in the session creation config.
This object can be reused on the client side in order to streamline
the session creation process.
Bug: 362801903
Bug: 367803904
Test: atest HintManagerServiceTest
atest PerformanceHintNativeTest
Flag: EXEMPT_NDK
Change-Id: Idd2ea87886c664b625f40066cc05c6df1bf0e6d4
|
|
Bug: 367803904
Change-Id: I1c3929b23f0543b18030f659fd681209fc8597c6
Test: n/a
Flag: EXEMPT owners
|
|
sessions" into main
|
|
8c297a8694
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3358418
Change-Id: Id7ee4e6f4acad5be797a92c7203430b0ba84fd16
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3358418
Change-Id: I0e365e14aa30ccab9549ede8dd9175a3e43f95be
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
Change-Id: Iddc5808654664888952b7c8848bd66e89e15a97a
Test: n/a
Bug: none
|
|
Add an API called "APerformanceHint_getSessionFromJava" for Java
clients to access the native APerformanceHintSession objects that
Java hint sessions wrap. This should will make the Java and Native APIs
substantially more interoperable.
This patch also makes the native code aware of when it is being used in
Java, to allow it to have slightly different behavior, especially when
it is being directly accessed through JNI. The main place where this
mostly matters is in preventing clients from closing Java-owned sessions
through the JNI, as that could cause weird behavior otherwise.
Bug: 367803904
Test: atest HintManagerServiceTest
Test: atest PerformanceHintNativeTestCases
Flag: EXEMPT trivial refactor
Change-Id: I9bae7f78b61e40037de63839103906c3e536a743
|
|
|
|
Bug: 367816690
Test: Alarm with and without USB headset
Flag: EXEMPT refactor
Change-Id: If707cd8abf5c248217d0855f42fb1a4900a1674d
|
|
Some of these tests create multiple input devices and mappers, which I
didn't have time to work out how to mimic using InputMapperUnitTest. The
rest have been migrated.
Notes:
* Split out Process_UnknownKey from Process_SimpleKeyPress, since it's a
separate behaviour to test
* As far as I can tell, Process_KeyRemapping was just testing the
remapping code in FakeEventHub, so I removed it
* Process_WhenNotOrientationAware_ShouldNotRotateDPad now checks for a
display ID of DISPLAY_ID, not INVALID. This is because the method
called by KeyboardInputMapper to check orientation
(getAssociatedViewport) will never return an orientation without also
returning a valid display ID; setting one without the other appears to
have been possible in the old testing infrastructure, but would only
have affected state somewhere outside of KeyboardInputMapper.
For similar reasons, removed the second half of
DisplayIdConfigurationChange_NotOrientationAware.
* MarkSupportedKeyCodes was removed, as the code to test it would have
been too bulky to test a trivial one-liner (and it would basically
have been a change detector).
Test: atest --host inputflinger_tests
Bug: 283812079
Flag: TEST_ONLY
Change-Id: Id6d2eb4ec2bae6bd0e9a263209169785e40a6024
|
|
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3349305
Change-Id: I87e0e1355890099b7b4ad1e13b3da03e80abae2d
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3349305
Change-Id: Ifa12011f30a58c8e0e2dd3e833af204efd25a0f1
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
Bug: 377329333
Change-Id: I5a0f6b406a0700ad0e9bbf8407316c98988ea2a8
Test: builds
|
|
into main
|
|
Test: None
Bug: None
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:960967767d222348825e91287255db7cff157325)
Merged-In: Id0dc26c5639417061f2676ab25e5c62b775e1a65
Change-Id: Id0dc26c5639417061f2676ab25e5c62b775e1a65
|
|
|
|
Bug: 370570306
Merged-In: I9be1254c3e2685b0aa950b314c581824f40ce26c
Change-Id: I35bc501a2b1d9eb100aaab25cd660cf2e0542f99
|
|
|
|
Revert submission 30360907-revert-29997970-load_hints-VOSJCKDBUL
Reason for revert: because the patch is fine, actually
Reverted changes: /q/submissionid:30360907-revert-29997970-load_hints-VOSJCKDBUL
Change-Id: I7395b8224c5d2cda2aa02b63e11c9575fbe0db23
|
|
into main
|
|
Revert submission 29997970-load_hints
Reason for revert: Droidmonitor created revert due to b/378700893. Will be verifying through ABTD before submission.
Reverted changes: /q/submissionid:29997970-load_hints
Change-Id: Ib42a3f0478b84bb36d0a49a311cc3844512f094a
|
|
batching" into main
|
|
Change-Id: Id0dc26c5639417061f2676ab25e5c62b775e1a65
Test: None
Bug: None
|
|
(cherry picked from https://partner-android-review.googlesource.com/q/commit:80f60ec95afc0c5995647aa19f87b5389f656c25)
Change-Id: I75f68a34b2557deeb9e5d760480e5ab4aa6af328
Bug: 373556678
Flag: NONE external OEM contribution only active in their devices
|
|
Test: None
Bug: 365920375
Flag: com.android.hardware.input.enable_new_25q2_keycodes
Change-Id: If8d07744844eff38226e022dfb04195321a0976b
|
|
|
|
Reimplemented Chromium's OneEuroFilter usage to InputConsumerNoResampling.
There are a few differences between Chromium's work and this CL. We
reimplemented One Euro filter an adaptive cutoff frequency low pass
made in this implementation as in the Chromium's implementation. The
class FilteredResampler filters the output of LegacyResampler using the
One Euro filter approach.
Here's the link to Chromium's to One Euro filter:
https://source.chromium.org/chromium/chromium/src/+/main:ui/base/prediction/one_euro_filter.h
Bug: 297226446
Flag: EXEMPT bugfix
Test: TEST=libinput_tests; m $TEST && $ANDROID_HOST_OUT/nativetest64/$TEST/$TEST
Change-Id: I0316cb1e81c73b1dc28dc809f55dee3a1cc0ebd2
|
|
|
|
Original change: https://android-review.googlesource.com/c/platform/frameworks/native/+/3325081
Change-Id: Ifc9da2b3df8eff909ea248cd18b5060d3e68aeba
Signed-off-by: Automerger Merge Worker <android-build-automerger-merge-worker@system.gserviceaccount.com>
|
|
This is implicit in Android's sys/cdefs.h and this header is also
built for the host, which doesn't have this header. The missing header
was previously hacked around by symlinking the header into a directory
that would be included by host builds, but the real header is
sensitive to the contents of sys/cdefs.h so that wasn't sound anyway
(and we in bionic have no interest in keeping it and its dependencies
compatible with non-Android).
It was included for __INTRODUCED_IN(), but that's already defined by a
command-line argument, so we can just remove the #include.
Bug: None
Test: treehugger
Change-Id: Ia9b057118ac70de1a8d2d6b8898b9fee00887a7e
|
|
- Adds ADPF public load hint NDK API
- Updates rate limiter to use a single budget for all load hints
- Adds support for load hint batching, to send multiple hints at a time
- Requires users of load hints to annotate their hints for better traces
Bug: 367803904
Test: atest HintManagerServiceTest
Test: atest PerformanceHintNativeTestCases
Flag: android.os.adpf_use_load_hints
Change-Id: I08502599aff0471bd73adc586245da2c924d24e0
|
|
In libandroid.
Bug: 362798998
Test: atest SetFrameRateTest
Flag: EXEMPT NDK
Change-Id: Id52288f77bf43f4c7573c6743e5307d849490852
|
|
LegacyResampler still resamples MotionEvents even when requestedFrameTime is
std::nullopt. The problem is that if the parameter requestedFrameTime
is std::nullopt, it is reassigned inside consumeBatchedInputEvents to
std::numeric_limits<nsecs_t>::max(). In spite of using max() as a
placeholder to indicate that everything should be consumed, it is still a valid time, and resampler uses it to resample MotionEvent. To fix this issue, having a valid requestedFrameTime should also be a condition to resample MotionEvents.
Bug: 374375203
Flag: EXEMPT refactor
Test: TEST=libinput_tests; m $TEST && $ANDROID_HOST_OUT/nativetest64/$TEST/$TEST
Change-Id: I6e53bdd0a35d359da8b50f10dd4aad9bddc8aa3f
|
|
Added map lookup operations to find pointers by ID when resampling them
inside LegacyResampler. The CL fixes the assumption that pointers should
appear in the same order between motion events. Moreover, tests from
LegacyResampler are updated to the new behavior.
Bug: 297226446
Flag: EXEMPT refactor
Test: TEST=libinput_tests; m $TEST && $ANDROID_HOST_OUT/nativetest64/$TEST/$TEST
Change-Id: I59c83d2e55bf4ab3cb7958cab333428499f6fee8
|
|
It turns out that 'finishInputEvent' was adding the fd back to the
looper. This caused an NPE because we were calling 'finishInputEvent' at
the end of the consumer destructor, thus leaving the callback in the
looper. The looper would hit an NPE whenever the fd signaled after that.
To fix this, we move the call to setFdEvents(0) to the very end of the
destructor. This way, we can be sure that the last thing that executes
is the removal of the fd from the Looper.
There is also a possibility that fd is signaled during the destructor
execution. Theoretically, this should be okay because the fd callbacks
are processed on the same thread as the one where destructor runs.
Therefore, the signals would only be processed after the destructor has
completed, which means the callback would be removed before it gets the
chance to execute.
Bug: 332613662
Test: TEST=libinput_tests; m $TEST && $ANDROID_HOST_OUT/nativetest64/$TEST/$TEST --gtest_filter="*InputConsumerTest*"
Test: TEST=libutils_test; m $TEST && $ANDROID_HOST_OUT/nativetest64/$TEST/$TEST
Flag: EXEMPT bugfix
Change-Id: If5ac7a8eaf96e842d5d8e44008b9c1bff74e674e
|
|
Currently, in InputDeviceInfo, we store the KeyCharacterMap object,
which is actually the original KeyCharacterMap from the EventHub. This
could potentially lead to issues where two threads operate on the same
KeyCharacterMap object simultaneously, resulting in thread safety
problems.
To avoid potential risks in the future, we make a copy of the original
KeyCharacterMap when generating InputDeviceInfo.
This change should not introduce any behavioral changes.
Bug: 373011069
Flag: EXEMPT refactor
Test: presubmit
Signed-off-by: Linnan Li <lilinnan@xiaomi.corp-partner.google.com>
(cherry picked from https://partner-android-review.googlesource.com/q/commit:1be9c1ce7400e38cf3f45921d7181e947929f91c)
Merged-In: I0d0155133f95b0f1dc925422eda0da04c6f196ea
Change-Id: I0d0155133f95b0f1dc925422eda0da04c6f196ea
|
|
|
|
|
|
This reverts commit f7f93f5fd5afa40858ac492402ec6cc85fe89275.
Reason for revert: fixed tests
Changes: the test for inconsistent sequence needed to be updated because
the injection will now always fail.
Original description:
The dispatcher currently crashes when certain inconsistent input events
are injected. This is affecting test stability negatively.
In this CL, we reject globally-injected inconsistent events. That may cause
some test flakiness, but should eliminate the crashes due to dispatcher
reaching bad state later.
Unfortunately, we can't currently reject all inconsistent injected
events. In the case of targeted injection, it is common for the caller
to leave pointers dangling. Since the injection happens into the
caller-owned windows only, at the end of those tests the windows get
cleaned up, so the dispatcher is still in a good state.
The eventual goal is to completely get rid of injection. Meanwhile,
however, this should help avoid at least some of the crashes.
Bug: 369935405
Flag: EXEMPT bugfix
Test: atest inputflinger_tests
Change-Id: Ic778677a63e544feadf01f3cc47b08ec3b207297
|
|
|
|
This reverts commit f6abbf4b919b121614384ab2286bcaf16bedd344.
Reason for revert: DroidMonitor: Potential culprit for http://b/372964183 - verifying through ABTD before revert submission. This is part of the standard investigation process, and does not mean your CL will be reverted
Change-Id: I9cdb4dfb4ca14a90d7527e5f546380ea6ec1d62f
|
|
|