Standalone D8 invocations tend to be shorter-lived than R8 invocations
that involve deeper layers of bytecode optimization. Introduce a
separate set of JVM flags to pass to D8 invocations which limit the
runtime JIT tiering, reducing overall compile times and significantly
reducing total CPU consumption for D8 build workloads.
Representative incremental build times with this change:
* real 6m1.388s
* user 63m32.412s
and without:
* real 6m30.624s
* user 79m12.626s
Bug: 205303544
Test: time m
Change-Id: Ic568a4316325c6a4e0da41ffebbb07f0c5f7e8e3
Write `static_libs` and `libs` of Java library and Android app modules to module_bp_java_deps.json. This enables downstream tools to correctly set up the runtime environment. Note that while static libraries don't need to be on the Java classpath these modules could have non-static library dependencies that do need to be present.
Test: m out/soong/module_bp_java_deps.json
Bug: 227538646
Change-Id: I7c4aecb2fb03c890f0d2aaae80e619f6176809ef
Metalava increments the SDK level by one when it's not "REL", so we
temporarily force the build to be "REL" while we're still in the
process of finalizing it.
This CL must be reverted as part of actually declaring "REL".
Bug: 171506470
Test: Build
(cherry picked from commit 05597c0cc4)
Change-Id: I18405cdd27769598f5e7a6a968f0dd473f936f6c
Previously, the build applied the same filtering to remove
implementation details from the sdk snapshot's stub-flags.csv file as
it did for its all-flags.csv, i.e. removing the signatures that only
had a "blocked" flag. Unfortunately, that had no effect on the stub
flags as the implementation signatures had no flags, not a single
blocked flag. That meant that the sdk snapshot's
filtered-stub-flags.csv file contained a lot of implementation details.
This change removes signatures from stub-flags.csv that have no flags
which removes all implementation details from the sdk snapshot.
Bug: 194063708
Test: atest --host verify_overlaps_test
m out/soong/hiddenapi/hiddenapi-flags.csv
m art-module-sdk
# Check contents of its filtered-stub-flags.csv file
Change-Id: I30edc77348fad118ea732e787ae8e206c8841f84
Getting the first certificate will panic if there are no certificates,
which can happen when AllowMissingDependencies is set and the
certificate property is a module reference to a missing module.
Only get the first certificate if the list is not nil.
Bug: 228379411
Test: TestAppMissingCertificateAllowMissingDependencies
Change-Id: I046d75dbbd4f21f4a2b6851f558e430e9879fcff
The proguard test files are duplicated in apex and sysprop and will
be needed by app tests, move them to the java package.
Test: run all soong tests
Change-Id: Ie13817dcda8d98801d16a97ffceef1100c7d5380
Some java_* modules contain .logtag files in their srcs, but we should
separate these out into their own targets.
Test: enabled QuickSearchBox and build
Change-Id: I595722ca59686d450ff522987a15ce283d3607e0
Previously, verify_overlaps used positional arguments, the first was
the monolithi flags and the rest were the module flag pairs (filtered
flags file and signature patterns file). This change makes them use
named options to make the purpose of the arguments clearer on the
command line.
Bug: 194063708
Test: atest --host verify_overlaps_test
m out/soong/hiddenapi/hiddenapi-flags.csv
Change-Id: Ife0af0016eb0f91416e8330d5d98cb53c97d68a4
This reverts commit c6ef485370.
Reason for revert: Multiple systemui presubmits are broken due to this change
Bug: 228262695
Change-Id: I1efed519c49d8d0583815c7bf2b3801bf0ba12fa
Previously we were only generating used-by API-coverage for APEX modules.
This change adds support for APK-only modules such as NetworkStack and DocumentsUI.
Bug: b/216313756
Forrest Run: https://android-build.googleplex.com/builds/abtd/run/L10800000953846781
Test: TARGET_BUILD_VARIANT=userdebug PRODUCT=mainline_modules_x86 ./vendor/google/build/build_unbundled_coverage_mainline_module.sh -j16
Change-Id: Id17e4a55c2a52e9903632a654e778f8d54982dfc
Merged-In: Id17e4a55c2a52e9903632a654e778f8d54982dfc
(cherry picked from commit 56d75785bd)
