These result in weird command line invocations when used with RBE.
Test: Ran an RBE build locally and made sure it still worked as expected
after the change.
Bug: b/163618332
Change-Id: Ic8a8adb3fffb357754384ddbf13684d944a9767e
Allow hermetically running the test suites by packaging the JDK
into the output zip file.
Bug: 139202337
Test: m cts && zipinfo out/host/linux-x86/cts/android-cts.zip
Change-Id: I6a59bb97a33113c83df9c7162e9c65ca90a8c196
All uses of this property have been migrated to
LOCAL_SANITIZE_BLOCKLIST.
Update language to comply with Android’s inclusive language guidance
See https://source.android.com/setup/contribute/respectful-code for
reference
Bug: 161896447
Bug: 162245450
Test: Add the old property to an Android.mk and ensure error is thrown.
Change-Id: I797ff651f03b954b410f76f4d201a962ded717df
So we don't accidentally create phony rules like
`MODULES-IN-: MODULES-IN-` and `MODULES-IN-.: MODULES-IN-*`
Bug: 162918277
Test: m com.android.art
Test: ninja -t browse
Change-Id: Id0656d1b08bc933df57914df162853934352593b
This hasn't worked for a couple years, and continues to bitrot. Just
remove it.
Test: treehugger
Change-Id: Iea6caf3c08252a560155e095135c5ddaad712991
Merged-In: Iea6caf3c08252a560155e095135c5ddaad712991
Allow for relocated partition directories in out, mainly for
devices which do not possess e.g. /product but must
distribute it in /system/product.
Use argparse to cope with increased argument count.
Test: Device without physical /product partition
m out/target/product/<device>/shareduid_violation_modules.json
-> Observe addition of violating apks from product
partition in shareduid_violation_modules.json
Signed-off-by: Felix <google@ix5.org>
Change-Id: I2d7826b41fb5b924f15ae21a73178fa4eec6a879
It should happen after adjusting vendor and product variables.
Bug: 162476652
Test: see image variant with LOCAL_PROPRIETARY_MODULE := true
Change-Id: I071434916c9ad18572c559a26e3ef6abb8c47303
The directory structure of several ramdisks are different depending on
whether the target is A/B or non-A/B. When the target needs to be
compatible with the GKI however, which is always built as A/B, it is
necessary to build these ramdisks using the A/B file hierarchy. Provide
a flag to enable this override when necessary.
Bug: 161579556
Change-Id: I1cf9a0ef5a338b0b5152113533b62804f3088c83
Now that they're defined with prebuilt_build_tool, we don't need to set
them here.
In future changes we can replace more of these definitions with
prebuilt_build_tool, as it can centralize the selection of
build-from-source or prebuilt for Make, Soong, and user-defined
genrules.
Test: treehugger
Change-Id: I5821bbad1b655d561919245320d7c184a6eac737
This will let us quickly check the system image build type,
and modify *.rc behavior based on that.
Bug: 142430632
Test: adb shell getprop ro.sanitize.hwaddress in hwasan build
Change-Id: Id1738ebc94a7c29ea9902a063f5d8dd6deb48f1b
This improved NOTICE bundling now includes TARGET_OUT deps that Android
Test Suites packages need, especially those for .apk files.
Bug: b/141258651
Test: $ make cts && \
unzip -p out/host/linux-x86/cts/android-cts.zip \
android-cts/NOTICE.txt \
| less -N
Change-Id: I4c84d194ff5982fc57b6a8097602d4a9df68d61c
Merged-In: I4c84d194ff5982fc57b6a8097602d4a9df68d61c
(cherry picked from commit e73186d2ac)
- the NOTICE file is only taken from out/host.
- currently only enabled from cts by setting flag "test_suite_notice"
Test: $ make cts && unzip -l \
out/host/linux-x86/cts/android-cts.zip \
| grep NOTICE
$ make pts && \ unzip -l \
out/host/linux-x86/cts/android-cts.zip \
| grep NOTICE
Bug: b/141258651
Change-Id: I696b424e5f2b09d28d53ccbd9198e1f4f301ec01
Merged-In: I696b424e5f2b09d28d53ccbd9198e1f4f301ec01
(cherry picked from commit 93d2fbbcad)
In Make, clang-tidy and clang run in the same action, but only clang can
be remoted via RBE/GOMA. So to prevent running up to 500 clang-tidy
instances at the same time and overloading the machine, use the local
pool when using clang-tidy. This does limit the parallelism of the clang
actions attached to clang-tidy, but hopefully that's not that much of a
problem. Clang-tidy isn't enabled by default (opt-in per-build /
per-module), and Soong does not run into this problem.
Bug: 162615762
Test: m WITH_TIDY=1 nothing; build-aosp_flame.ninja is identical before/after
Test: use_rbe m WITH_TIDY=1; inspect build-aosp_flame.ninja, see local_pool
Change-Id: I7dd196fcf4183d175b9608d3d48cdcdf81b514ce
Merged-In: I7dd196fcf4183d175b9608d3d48cdcdf81b514ce
