This change allows the impl lib of java_sdk_library to call AddContext(...) instead of AddContextMap(...) and ensure that the behavior between the top level java_sdk_library and the impl lib is identical when creating the class loader context.
Test: CI && ABTD
Bug: 366572133
Change-Id: Ia0bc7558ca737fcd3ad4ed62b3190aa567aa103f
Add support to include font files into ravenwood_libgroup. Also make ravenwood data property to accept paths instead of module name.
Bug: 292141694
Flag: EXEMPT host side change only
Test: $ANDROID_BUILD_TOP/frameworks/base/ravenwood/scripts/run-ravenwood-tests.sh
Change-Id: Ida5ef0230d2cc59fa1209199168c6c35e4842fbd
So that we can decide the configuration after defaults are evaluated.
Instead, pass the enabled property on to submodules.
Bug: 361816274
Test: m nothing --no-skip-soong-tests
Change-Id: I853d557ed022f7434287cea0a2a19e684b59e145
There are some instances of .kt code in android platform that are not
fully compatible with kotlinc > 2. To account for this version skew when
generating xrefs, implicitly add `-language-version 1.9` to the args in
the .kzip file. The main kotlinc build rules remain the same.
This value has been picked manually by running
`external/kotlinc/bin/kotlinc -version` at ToT
Bug: 265428637
Test: verified that `-language-version 1.9` is added to the build
command of the .kzip files
Test: ran the indexer locally on the .kzip of `PermissionController-lib`
(this android_library contains .kt code that does not compile with
verison >2)
Change-Id: Icfb72800404d83f63204de23a1ec34b40ee05847
The property was introduced as an interim solution, and is currently
unused.
Test: m nothing --no-skip-soong-tests
Bug: 366071058
Change-Id: I57abdb64fabdb34fbbd1190851bc528dbb88c7f8
This CL creates the build rules to invoke the standalone kotlin
extractor and create .kzip files for each kotlin translation unit. These
will be indexed by the internal g3 kotlin indexers
Implementation details
- Create a `kotlinKytheExtract` static rule. Its inputs will be source
.kt/.srcjar files and other verbatim args passed to kotlinc. The
extrator will serialize this information into a .kzip file. The .kzip
file should contain the necessary information to "replay" the
compilation
- Create a xref_kotlin phony rule to build all .kzip files corresponding
to kotlin translation units.
This implementation has one limitation. Since the kotlin indexers
"replay" the compilation using its own version of kotlinc-jvm, there
might be indexing issues caused by kotlinc version skew between android
builds and indexing builds. `-kotlin-home` can likely solve that, but
this CL defers that for now.
Bug: 265428637
Test: Built the phony `xref_kotlin`, and ran the indexer locally using go/kythe-git-local-run#indexing-to-local-xrefs-server
Change-Id: Ifb370f2df8aa46f71df3fe1ae7d7aa5da773cc8b
I'm going to be adding some methods to this interface, give it
a name based on how it's going to be used, not based on what methods
it contains.
Bug: 323382414
Test: m nothing --no-skip-soong-tests
Change-Id: I9bba04ba756c4dbe00625e2d04af81e78a11cae9
Ravenizer and the other rules that modify the output jar also need to
replace the implementation jar depset with the modified output jar.
Bug: 308016794
Test: atest CtsOsTestCasesRavenwood
Flag: build.RELEASE_USE_TRANSITIVE_JARS_IN_CLASSPATH
Change-Id: I87ec143431c09ed878f8e1c3541b56d8077e23b5
There are two versions of aosp extservices - one for Tplus and one for
Sminus. When prebuilts are active, the `contents` validation of both
variants of extservices bootclasspath fragments should be disabled to
account for BCP jar skew.
This CL adds ApexAvailableName to ApexInfo to achive that. If an
api_domain matching the ApexAvailableName has been selected in
`all_apex_contributions`, the source apex variants will be disabled.
Bug: 361771739
Test: lunch cf_x86_64_phone-next-userdebug
Test: m nothing with ag/28851886
Change-Id: I505a7ad8295c258e2eeef7f183669f7a75a1c88e
... if prebuilts are selected.
This is a special case of b/361771739 where a apex starts contributing a
jar to the BCP. `disableSourceApexVariant` uses
`PrebuiltSelectionInfoProvider` and this gets populated via the
following dep chain
```
source --> prebuilt --> all_apex_contributions
```
This runs into issues if a prebuilt bootclasspath_fragment does not
exist.
This CL creates a dependency edge between the source module and
all_apex_contributions singleton module so that
`PrebuiltSelectionInfoProvider` is available for
`disableSourceApexVariant`
Bug: 361771739
Test: lunch cf_x86_64-next-userdebug
Test: m nothing with ag/28851886
Change-Id: I92d88199a27579d3c9879d40fd354653370efcce
I'm rewriting how globs work in soong to make them compatible with
hash-based ninja implementations. As part of this, I'm removing this
unnecessary usage of globs to make them simpler.
In this case, since we're already globbing the directory at analysis
time and causing soong to rerun, we can ensure the aapt2 action also
reruns by putting a hash of the glob results onto the command line.
Bug: 364749114
Test: m framework-res, touch frameworks/base/core/res/assets/test1.txt, m framework-res rebuilt framework-res.apk, rm frameworks/base/core/res/assets/test1.txt, m framework-res rebuilt framework-res.apk again
Change-Id: I4f666367a9a0fd0dfa42dc51ef3a788a02b41747
Skip combining jars into turbine-combined, combined, and withres jars
and instead collect transitive jars to use in the classpath and to
produce the final dexed jar.
This reduces the size of a `m checkbuild` in git_main by 11%, from
1300 KiB to 1154 KiB. It may also improve caching and reduce uplink
network bandwidth when building with RBE, as now the classpath inputs
to rules are themselves outputs of previous rules and so already in
the RBE CAS.
The downside is that the classpath inputs to each rule are now much
longer, increasing the Soong ninja file size 11%, from 4.6 GiB to
5.1 GiB. This could be mitigated in the future by supporting something
like depsets in the generated ninja file to reduce duplication.
Bug: 308016794
Test: TestSimple, TestKotlin, TestClasspath
Flag: build.RELEASE_USE_TRANSITIVE_JARS_IN_CLASSPATH
Change-Id: I2b7b4261375494370da70f98597c8719f1d561cf
The `m Run*RoboTests` test runner is no longer supported, remove the
code to generate the Make rules for it.
Also reduce the number of combining steps by passing the extra jars
that need to be combined into the compile step.
This relands Icf05079bf570bed7a10962cbf03459bd53c51f99 with a fix to
reorder the instrumented app classes after dependency classes.
Test: atest --host SystemUiRoboTests
Flag: EXEMPT refactor
Change-Id: I174ceb95af5557e1859eeb12a2f5fac2332975a8
This reverts commit ef5739ed64.
Reason for revert: Droidmonitor created revert due to b/364665936.
Change-Id: Ia18f642492eafdbe2e78d9f753d87f243d529037
The `m Run*RoboTests` test runner is no longer supported, remove the
code to generate the Make rules for it.
Also reduce the number of combining steps by passing the extra jars
that need to be combined into the compile step.
Test: atest --host SystemUiRoboTests
Flag: EXEMPT refactor
Change-Id: Icf05079bf570bed7a10962cbf03459bd53c51f99
collectTransitiveHeaderJarsForR8 visits direct dependencies and collects
all transitive libs and static_libs header jars. The semantics of the
collected jars are odd (it collects combined jars that contain the static
libs, but also the static libs, and it collects transitive libs dependencies
of static_libs), so these are only used to expand the --library arguments
to R8. Rename the method and all the fields it uses with a "ForR8"
suffix to avoid confusion when more transitive header jar collection is
added to support transitive classpaths.
Bug: 308016794
Test: all soong tests pass
Change-Id: I291a86fbbc2e2f088598fb309d1b992080d60941
Pass the name of Soong's checkbuild target for each module to
Make so that it can depend on it from the main checkbuild rule.
This will give better control over which files get built, allowing
checkbuild to skip the jar combining step when transitive classpath
jars are enabled. The per-module checkbuild targets are passed to
make instead of added directly as a dependency of checkbuild in order
to maintain the existing behavior of only building modules exposed
to make in checkbuild.
Also tweak the existing calls to CheckbuildFile and add
InstallFileWithoutCheckbuild to match the files that are
in the Make-based checkbuild.
Bug: 308016794
Test: m checkbuild
Change-Id: Ic5140819381d58f4d00f23a7a12447950c4cf268
Apimapper is a tool to instrument android tests to log potentail API calls at the run time.
The real tool is developed internally. We use a placeholder binary in
AOSP before the real tool is ready.
The tool will only take affect when the enviroment var EMMA_API_MAPPER
is set to true.
Test: m cts
Bug: 328699028
Change-Id: Iece53b8afdb9803334b7393527f4fa24e22f71a8
Some `java_sdk_library` and `droidstubs` modules already specify
`--error UnflaggedApi` but they are currently ignored because they are
added before the `--error-when-new UnflaggedApi` automatically added by
Soong and Metalava uses the last setting. That means adding
`--error-when-new UnflaggedApi` in change https://r.android.com/3248013
actually reduced the severity of the issue for those modules making it
more likely that they would be missed.
This change only adds the `--error-when-new UnflaggedApi` if it does
not already appear in the metalava command.
Bug: 361582214
Test: m checkapi
Change-Id: I7228bf9794b1c7e27d7acf7a461956b30f3335f5
