* Switching Cronet tests to the orchestrator means
that we will be able to run tests exactly like how they are
run in Chromium, as seen in this CL, this has allowed running
previously disabled tests.
* This change comes with an introduced latency which is not hugely
impactful (Measured to be around 400ms per test case) which takes
our runtime from ~6 minutes to ~10 minutes on average. This is an
acceptable impact for better coverage. If it increases beyond that
then we can utilize sharding but that is not required at the moment
or in the foreseeable future unless we have passed 1000 test case.
Test: atest NetHttpCoverageTest
Bug: 329823447
Change-Id: I1fff900341d958b8f00fc777e909b3583a83cc6a
Revert submission 2980591-cronet_121_0_6167_71
Reason for revert: Seems to be the cause for b/328855816
Reverted changes: /q/submissionid:2980591-cronet_121_0_6167_71
Change-Id: Ia47515958eec58b851cfbb8e74e0e9ab58679323
Patrick Rohrauthored and Gerrit Code Reviewcommited 2024-03-11 19:08:45 +00:00
This CL is created as a best effort to migrate test targets
to the new android ownership model. If you find incorrect or unnecessary
attribution in this CL, please create a separate CL to fix that.
For more details please refer to the link below,
<add g3 doc link>
Bug: 304529413
Test: N/A
Change-Id: I243f17b3f0ad7af9ffa15ca242456e36688733f5
Merged-In: I243f17b3f0ad7af9ffa15ca242456e36688733f5
Locally created test http server was not terminated.
Use the test http server globally available instead.
Test: atest NetHttpCoverageTests --rerun-until-failure 3
Fix: 320964365
Change-Id: Ida01664df1640da7db22d03519d842f0c49309c7
* Push the test-data to the device in the expected destination, this is needed to get rid
of the divergance in TestFilesInstaller.
* Stop jarjaring tests Jni classes.
Test: TH
Bug: 304217500
Change-Id: I07fa7ed09f90bf286805479edc5238b6268d59f0
* After supporting netty, the certificates files are needed to start up the HTTP server. Those files must be moved to the emulator which happens through the changes in AndroidTest.xml, isolated-storage
has been disabled to allow the emulator to access the storage freely without restriction. Otherwise
I will have to change the destination of the data files which will result in divergence.
* Added the newly introduced test utility java classes including netty to the `jarjar_excludes` as they must not be jarjared because they are not part of the bootclasspath.
Test: atest NetHttpTests
Bug: 319629150
Bug: 304217500
Change-Id: I2331873a9e178783319f889a377ba7e390ecdd6f
This is a test-only utility class that must not be jarjared. If it is jarjared
then it becomes inaccessible to testing.
Test: atest NetHttpTests
Bug: 314275954
Change-Id: I6cb8eeca58b3dc506cd1e6f1fb6de00df9086509
some of the tests are completely new tests that have been added in the latest cronet import.
Fixing those tests on the fly while doing an import is going to be a lot of work, temporarily disable
those tests and created bugs to track them.
The rest of the tests were disabled through `@Ignore` annotation but that introduced divergence.
Bug: 314275954
Change-Id: I1ec2be189b5db4d08d410fa000a04c900f47113b
- Adds NetHttpCoverageTests to postsubmit. This would allow us
to run tests with @SkipPresubmit in post submit.
- Don't jarjar com.android.testutil so that the annotation can
still be found by the testrunner.
Bug: b/293843426
Test: th
Change-Id: I9e792d338515acba798092bbc4c615aef2cac47a
This fixes the flakiness. Because we call to a real server,
the server may not be reachable and our assumption would fail.
When this happens, the stream hangs and engine shutdown fails with
active request error. Hence cancel the stream before calling
shutdown.
Also @Ignored this test despite the fix as a safety measure to meet
the SLO. Will re-enable afterwards.
Bug: b/292298108
Test: atest NetHttpCoverageTests --test-filter
BidirectionalStreamTest#testBidirectionalStream_GetStream_CompletesSuccessfully
Change-Id: Iaae61fc62a148b58af86fb29c248399041785f9b
Change-Id: Iea9ef0b8468ec6fa05c306ea394485cec706b9ef
Calling callback.blockForDone() on a non-started stream's callback will
block until it timeouts (12s). Currently, only a single test within
BidirectionalStreamTest actually starts the stream. This means that
every other tests is calling callback.blockForDone() during teardown
for no reason at all and they will just get stuck for 12s.
Change-Id: Ia40cbc371b4e2450dcd92a3ad429ff4351931a21
MTS tests shouldn't load Cronet from the platform (through the
bootclasspath). They should instead bundle it directly.
Things still need to be jarjar'ed to avoid clashing with classnames on
the bootclasspath (which would otherwise take precedence). Though, this
will no longer be needed once the android.net.http -> org.chromium.net
wrapper lands (MTS will then target org.chromium.net directly, which
isn't exposed through the bootclasspath).
Bug: 288886953
Test: m & th
Change-Id: I1533b357889e13e5e1c739c8a8419d21b10face3
CtsNetHttpTestCases were missing a DevSdkIgnoreRule after updating the
min_sdk version to R+ (which made it installable on R where the cronet
classes are not in the bootclasspath.)
Test: atest CtsNetHttpTestCases
Bug: 259632936
Change-Id: I706ff92376227c1759fd6e542066fcff3a7cf052
Cronet tests do not use a jarjar prefix, so do not jarjar their J/N
implementation.
Test: atest NetHttpTests
Bug: 278639767
Change-Id: Ie4b8bfa13ecbbca014e22d653fa0d39a5554c10e
Since cronet_java_tests (previously cronet_aml_javatests_sources) is now
an android_library directly in external/cronet (to support "mm" in that
folder), the dependencies can be removed.
A form of android_library is still required to wrap cronet_java_tests,
so the AndroidManifest (in this folder) is forwarded properly to the
common coverage test target.
Test: TH
Bug: 278639767
Change-Id: I25c6371c64e6cfcbb48e9e3f2976d47279791f21
With the release cut of the July train and the recent automerger
cutover, tm-mainline-prod is now officially an abandoned branch.
This change deletes (most) infrastructure that was put in place to
disable cronet on tm-mainline-prod.
Test: builds
Change-Id: I078f2114b736a634f08d8f704c19beb2224ef645
These CTS tests make use of a non-hermetic server. This means that:
* Test failures are not reproducible
* Test run are affected by the enviroment they run in (e.g., network
configuration of the device running the tests)
Ideally we would have a local QUIC test server, but setting that up is
non-trivial. So, for the time being (while I come up with a QUIC test
server setup), simply enable QUIC and test that things keep working.
Bug: 274369142
Change-Id: Icc8b04c1671189b77f7d8995b721fb52afea371d
aosp/2384137 added cronet to framework-tethering.
But framework-connectivity is a better place to put cronet since
cronet does not work on R devices (b/270049141) and
framework-tethering is R+, framework-connectivity is S+.
Followup CLs will move some modules (e.g. CronetJavaPrejarjarDefaults)
that use the branch dependent soong variables to framework/Android.bp
Test: TH
Bug: 278070640
Change-Id: I6bc10116759fb9e083c02147908e53022dab740a
This test needs Truth library which is part of the CronetTestJavaDefaults target
to work.
Test: atest NetHttpCoverageTests
Bug: 267353182
Change-Id: Ic3f084651af91be71832936c4fc0fdd241159198
This reverts commit 5f44ed3e7e.
Reason for revert: The symbol not found error was a transient build failure.
Change-Id: Iada1a0bc9c2282286c1e3b972de42e6387c901d6
This is needed for android.net.TrafficStats dependency
Test: atest NetHttpCoverageTests
Bug: 267353182
Change-Id: Ic426a2c7844e432c6444eee56515de58b8eccb92
Chidera Olibie
authored and
Gerrit Code Review
commited 
George Verhaegen
authored and
Lisa (LeeWei) Liu
commited 
Mohannad Farrag
commited 
Patrick Rohr
authored and
Aditya Choudhary
commited 
Junyu Lai
authored and
Treehugger Robot
authored and
Tongbo Liu
commited 
Sam Saccone
commited 
Stefano Duo
authored and
Motomu Utsumi
commited 
Dan Stahr
authored and