The default partition for RRO is "product/" in Make, but it was
"system/" in Soong. This CL ports the logic from Make to Soong
To implement this, a new function PathForModuleInPartitionInstall is
created that enables callers to provide the relevant partition
Bug: 158407753
Test: from build/soong, ran go test ./java
Change-Id: I05b02eae7fe57189aaad5109c26cccc5823518ef
* changes:
Don't use unsafe_ignore_missing_latest_api in TestJavaSdkLibraryDist
Remove core_lib property from java_sdk_library
Ignore owner property when computing java_sdk_library dist subdirectory
Make the default java_sdk_library dist_group "unknown"
Use the --soong-only flag instead to create it the proper way with the
config make step. This fixes build problems arising from the incomplete
soong.variables settings.
This partially relands https://r.android.com/1535202 so that we'll use
TARGET_PRODUCT=mainline_sdk in module SDK builds everywhere.
build-aml-prebuilts.sh isn't removed completely because we still need
the special case to create build_number.txt.
Test: art/build/build-art-module.sh --skip-apex && \
packages/modules/ArtPrebuilt/update-art-module-prebuilts.py \
--local-dist out/dist
then check that the created prebuilt dex2oatd64 runs with the repro
in b/190194345
Test: build/soong/scripts/build-mainline-modules.sh
then
prebuilts/runtime/mainline/update.py --local-dist .../out/dist
in a master-art tree and check that art/build/build-art-module.sh
works there
Bug: 190194345
Bug: 174315599
Change-Id: I86327b082b61c292f18c0a6b315d9add08002832
I'm not sure why these targets are getting built as part of the sdk
build anyway, but we don't need them.
Test: treehugger
Bug: http://b/156513478
Bug: http://b/190394921
Change-Id: I1ea0e75f459ad7cad648f96b67a4f2424d36fda7
This is no longer required to find package.html files, so remove it.
Bug: 153703940
Test: m checkapi
Change-Id: I80ec219cb9ef8922094336a5d45652c8e59aa113
Metalava now expects html files to be specified explicitly. This
filtering was originally added because jdiff could not cope with html
inputs, but jdiff support has been removed in r.android.com/1455335.
Bug: 153703940
Test: m checkapi
Change-Id: I34164b633bab2892e3a1c0fd0257fcf98d537655
The previous --skip-kati flag could be interpreted as "do not run kati
to re-generate ninja file". Add a more specific flag for the "soong
only" build use-case, where we do not load the kati-generated ninja
files at all.
Bug: 189187214
Test: build/soong/soong_ui.bash \
--make-mode \
--soong-only
--skip-soong-tests \
TARGET_PRODUCT=mainline_sdk \
SOONG_ALLOW_MISSING_DEPENDENCIES=true \
SOONG_SDK_SNAPSHOT_VERSION=unversioned \
SOONG_SDK_SNAPSHOT_USE_SRCJAR=true \
out/soong/mainline-sdks/art-module-sdk.zip
Change-Id: I91abbd28af517d4b550ebc6d88fd64947caf9545
Merged-In: I91abbd28af517d4b550ebc6d88fd64947caf9545
(cherry picked from commit 546de4a1f3)
- Rename the "BuildX" variables to "RunX"
- Remove redundant comments
- Inline all the "what to do" based on config in build.go
- Inline some constants only used in one place
Bug: 189187214
Test: m nothing
Test: build/soong/build_test.bash
Change-Id: I111a69e642212d7938d4971283545e0d9acbb01a
Merged-In: I111a69e642212d7938d4971283545e0d9acbb01a
(cherry picked from commit d274ea9196)
Sourcepath is set to dir of the Android.bp, but that dir does not
necessarily have any inputs for the metalava invocation defined in that
Android.bp. This leads to problems when metalava tries to read a
non-existent dir, so create it first.
Bug: 153703940
Test: m sdk dist
Change-Id: I8c2ca45c8f066cabf3247b9652a7094cefc047a9
Merged-In: I8c2ca45c8f066cabf3247b9652a7094cefc047a9
(cherry picked from commit 7eff3a8fc3)
`reuseObjTag` is a dependency from the shared variant of a `cc_library`
to the static variant. When such dependency exist, the shared variant
would reuse the object files of the static variant.
Thus if the shared variant requires ABI dump, then the static variant
would need to generate ABI dump as well.
Bug: 190339664
Test: m findlsdump
Change-Id: I015328d74acbeb89ff7622397e59b68864af34a4
As long as we have boot jars defined in make, we should try to keep
global.UpdatableBootJars as a source of truth when it comes to classpath
configuration. However, for testing purposes there is a need to add
java libraries to bootclasspath that aren't present in
PRODUCT_UPDATABLE_BOOT_JARS variable.
Ideally, we would unconditionally add all contents to the config for
test apexes, however, I don't fully understand apex merging to support
it.
Bug: 180105615
Test: atest sdkextensions_classpaths_e2e_tests
Change-Id: I552f93289c27b3713e75dcda1faed909b0bfcef2
This allows it to be used for other modules types and other properties
(e.g. static_libs & exclude_static_libs).
Test: go test soong tests
Bug: 188497994
Change-Id: I40ab16e3b540ece0a6684558b32f7e8e25df6f24
This is to simplify the process of resolving label + exclude labels
across the various configuration axes we have and across the various
properties/modules that use this behavior.
Test: ci/bp2build.sh && ci/mixed_droid.sh
Change-Id: I8efae3e75ddb365384f5caaf5bb504a5206618d3
The local diffing behavior is currently flagged off so we can land
this in stages.
Test: pytest cc
Test: treehugger
Test: development/tools/update_ndk_abi.sh
Test: m ndk
Bug: http://b/156513478
Change-Id: Iccb314411bc74ea3ddfea8b85b0539709295f65a
Adds support for snapshotting Rust rlibs. This allows us
vendor-specific code that uses rlib-only linkage until dylib
snapshot support is added.
Bug: 184042776
Test: m nothing # new Soong tests pass
Test: Example test Rust vendor module builds
Test: m dist vendor-snapshot # includes rlibs
Change-Id: I4976d3e1efec0ee778cc97730d45be471dffb678
Export additional fuctions, structs, and interfaces from cc for use in
the rust package to allow for rlib snapshotting.
Bug: 184042776
Test: m nothing
Change-Id: I4c53b9378d5d5b5973dbd23ab692cdfb2ede60b9
Adds support for platform vendor_available Rust FFI libraries and
binaries to be included in the vendor snapshot.
Because rlib and dylibs are not yet in snapshots, libstd cannot be
included in a vendor snapshot. As a result, vendor-specific Rust code
can't be guaranteed to work with the platform-provided vendor_available
modules built with a newer toolchain. For now, a check is added
indicating vendor-specific Rust code is unsupported.
This changes the linkage for vendor variants of these modules to default
to rlib linkage since dylibs cannot be included in the snapshot yet.
Bug: 184042776
Test: m nothing # new Soong tests pass
Change-Id: I502eaa4bb962eb87ff868fcf49b435f0d2f982e6
This CL exports and refactors some cc vendor-snapshot related functions
so they can be reused by rust modules to support vendor snapshotting.
Bug: 184042776
Test: m nothing
Change-Id: I12706e62ce0ac3b2b4298085fafc1d77b8e0a0c4
This CL adds support for cc_library to correctly split c, as and cpp
srcs in shared/static nested props, as well as splitting
the *filegroup* deps in those props, where each filegroup is expanded
into its own c, cpp and as srcs filegroups. This ensures that the
correct sources go into cc_library_static's underlying cc_libraries for
c, cpp and as sources respectively.
See the bp2build conversion test for a better visualization.
Bug: 183064430
Test: TH
Change-Id: I29add5140672d042adff65527d8b65f4a5f0a05b
Using broader premissions for archived files allows them to be more
easily used when extracted. For example, defaulting regular files to
0644 will allow other tooling to use a file without the need to change
permissions manually.
Bug: 189919409
Test: m dist and inspected file permissions of archived files
Change-Id: I4a0f8075206391254c639ecf865639bb9e8df0bf
Use FixtureWithLastReleaseApis to create the necessary files instead
of disabling the check with unsafe_ignore_missing_latest_api.
Bug: 186723288
Test: TestJavaSdkLibraryDist
Change-Id: I162b0f68eec596274f8d98dca9d3f0500ab13f5d
The owner property is no longer used, the dist subdirectory is determined
by the dist_group property.
Bug: 186723288
Test: TestJavaSdkLibraryDist
Change-Id: Id6d997eef05f6511070677974219674f248cb754
Change the default dist_group from "android" to "unknown" to prevent
accidentally including java_sdk_library stubs that do not set
dist_group or owner in the public SDK.
Bug: 186723288
Test: TestJavaSdkLibraryDist
Change-Id: I9aae2a16254ac1a8d444acfa63bc571d1ef4b045
This script expects to build uncompressed APEX and use them as
prebuilts.
Bug: 189926000
Test: build_mainline_modules.sh
Change-Id: If0f62c2baa3db3c079b2913fcdc3dafd66394011
* changes:
Support dist_group property instead of owner for setting sdk dist subdirectory
Add test for java_sdk_library dist properties
Move java_sdk_library tests to sdk_library_test.go
