Since Clang-4579689, crt libraries use i386 instead of i686. Update the
hardcoded paths to the new names.
Bug: 72942266
Test: N/A
Change-Id: I24f8f5742ac2fecd374fdaa3be8a11f410d31cbf
If PRODUCT_TREBLE_LINKER_NAMESPACES is not true when
BOARD_VNDK_VERSION is defined, report a build error.
To enable VNDK, linker namespace must be configured.
Bug: 72704254
Test: Try to build with one of them is missing
Change-Id: I804704e2f48fa44c8f998183508741d5cc40e4ce
BOARD_PROPERTY_OVERRIDES_SPLIT_ENABLED can be true only if early-mount of
partitions is supported. But the early-mount must be supported for full
treble products, and so BOARD_PROPERTY_OVERRIDES_SPLIT_ENABLED should be
set by default for full treble products.
Additionally this CL sets ro.vndk.version regardless of
BOARD_PROPERTY_OVERRIDES_SPLIT_ENABLED.
Bug: 73078796
Test: tested with walleye
Change-Id: Icfe1b6957b55b0f9261d34d7e92db3dc5073cb58
The LOCAL_COMPATIBILITY_SUITE tag is used to package tests and
auto-generate test configs for use in the test infrastructure
and locally by atest.
Rather than update hundreds of makefiles across the tree, this
change adds significant atest support to many unittests.
Bug: 73000943, 72969679, 71510529
Test: `atest NotificationStressTests` works as expected.
Change-Id: I1b0a779b8366aee4b4df98181267f72fcb68916a
The module name of system_$(VER) prebuilt file is system_sdk_v$(VER).
But this is inconsistent with sdk_v$(LOCAL_SDK_VERSION) in prebuilts/sdk.
So, system_sdk_v$(VER) is changed to sdk_vsystem_$(VER) to use the same
naming policy.
Bug: 72031391
Test: build
Change-Id: I379c0f9aa0de7407489eeec2b07ee7727e81d69d
Allows target modules to directly request host modules they
require to be built and installed as well.
Bug: 72761631
Test: Added LOCAL_HOST_REQUIRED_MODULES to a target apk test and
the host module was built.
Change-Id: I5aca11b3ba8b54676f068d337f1c9ba0e0b01997
Kati's find emulator doesn't understand "| sort", so use use the make
function instead so that we can emulate the find commands.
before: *kati*: shell time (regen): 1.316989 / 187
after: *kati*: shell time (regen): 1.249748 / 184
The time has a lot of variance, the important change is the reduction in
number of shell commands. Kati has other issues with some of these (it
doesn't support find's implicit -a), which is why it only goes down by
3. I'll be fixing Kati separately to handle those cases.
Test: diff out/build-aosp_taimen.ninja before&after
Change-Id: I5c8c2993f01a8a6af84963e903a8a0c186a61e82
This is a list of sepolicy versions that the framework supports.
Test: build framework compatibility matrix
Test: boots
Bug: 67920434
Change-Id: I82365742172b653eadb7b5315398e06cace4e000
This prevents them from mistakenly bleeding between test targets.
Bug: 72384841
Test: make Run_TeleService_robotests
Change-Id: Ie2c047be1450420f8e27b8b45e9416e96b477b45
core_current is a pseudo SDK version which is a core Java API subset of
the Android API. It is expected to be mainly used for external Java
projects which are agnostic to Android; such as junit, guava, etc.
A module built with this SDK version can only link to java modules of
the same kind. It can't depend on modules built with LOCAL_SDK_VERSION
:= current or without LOCAL_SDK_VERSION.
Bug: 72206056
Test: m -j
Change-Id: I34a9696393aa6704fd6684a40ea5b05d3fb46b23
CPU/ABIs related information is already in system/build.prop. It
should also be included in the vendor image so it can still be
examined outside of a device, build environment, or $OUT directory,
by a standalone utility to check its basic characteristics and if
it's compatible with some specific system image (usually GSI).
Bug: 72079894
Test: Built and verified aosp_x86(_64):
# On host
$ grep ro.vendor.product.cpu $OUT/vendor/build.prop
# On device
$ getprop | grep ro.vendor.product
$ su
# grep ro.vendor.product /vendor/build.prop
Change-Id: Ic9ac8065d5983710840e33a805f982de75d86ce3
This reverts commit 4a2926a44e.
Reason for revert: The original CL caused two issues:
1) Increased app startup because `profman` failed to verify boot
jars and created an empty profile for the boot image. All methods
in the framework therefore ran interpreted/JITed. New flag
"--skip-apk-verification" has been added to `profman` to fix
the issue in CL Iad614e47feada293218c11cff985a8012ca49a0f.
2) Caused a crash in Mockito CTS tests because jvmti did not
remove hiddenapi access flags from boot jars before using them.
This has been fixed with CL Ibc9a96a6b541c06844f276db009ac29514f7a3bb.
Bug: 64382372
Bug: 72550707
Test: make test-art-host
Change-Id: Ie8c58714455662f042f56b2977b01b75c5ec7d2f
Kati has a `--warn` option that adds some extra Kati-specific warnings
-- like using undefined make functions, undefined user functions, likely
bad variable lookups, etc. Some of these are bugs, others are fine doing
nothing. This fixes up all of them in the core build system.
It also complains about $(eval) usage in a recipe not being recommended.
Those aren't handled as part of this change.
Bug: 72661763
Test: build-aosp_arm.ninja is identical before and after
Change-Id: I8e00af142a7745236d3ad4efc9e91ec3ce71a511
Bug: http://b/72642679
Store missing profile files and the modules that refer to them, passed
via SOONG_MODULES_MISSING_PGO_PROFILE_FILE variable, to
$DIST_DIR/pgo_profile_file_missing.txt as part of the 'dist' target.
Test: 'm dist' and verify creation of pgo_profile_file_missing.txt.
Change-Id: I4ac0a1f413cbb99b33ad1903aeebd3ebcf3b9f59
Fixes lots of warnings of the form:
build/make/core/base_rules.mk:606: warning: overriding commands for target `out/target/product/generic_arm64/testcases/simpleperf_unit_test/testdata/perf.data'
build/make/core/base_rules.mk:606: warning: ignoring old commands for target `out/target/product/generic_arm64/testcases/simpleperf_unit_test/testdata/perf.data'
Test: lunch aosp_arm64-eng; m device-tests
Change-Id: I32575d25f3aef1d4000fcb63d6ffc025ad21f937
This reverts commit b7f00ed2af.
Bug: 72550707
Bug: 64382372
Test: atest CtsInlineMockingTestCases
Reason for revert: Causing invalid dex files to be passed to jvmti
agents. This causes CtsInlineMockingTestCases and
other tests to fail.
Change-Id: I50129e94791b9e2bef919fe398b4250ae461195b
Just like visual warnings about improper use of native libraries,
warnings about the usage of hidden APIs will be displayed in non-
final, non-user builds only.
Bug: 64382372
Test: make
Change-Id: Id0a4200f912ac3303026cb26b6d8974c47332828
ART buildbots are broken because they build against a pinned commit of
the framework and thus lack the build rules to generate hidden API lists
We aim to fix this by generating dummy API lists in the build folder.
This patch changes the location of the blacklist and dark greylist from
frameworks/base/config/ to ${TARGET_COMMON_OUT_INTERMEDIATES}/... . On
normal builds, the framework will copy its text files into the build
folder.
Test: make
Bug: 64382372
Change-Id: I9b55d2865599e367d9c4be4f834182f5c084c3bf
When building with OpenJDK 9's javac with -source 1.9 -target 1.9,
a runtime image (with system modules) is used via the --system
command line flag, instead of --bootclasspath.
The runtime image only contains the default libraries (libcore)
but is missing other libraries that might be on the bootclasspath
for a particular build target.
This CL fixes compilation for this case by adding the missing
jars onto the classpath instead.
This already used to work but was broken by CL [1]. I attempted
to conditionally revert some of that CL's changes to java_common.mk
for the case of language level 1.9, but couldn't get it to work.
Therefore this CL follows a different approach.
[1] http://r.android.com/519552
Bug: 38177569
Test: Treehugger
Test: Ran "EXPERIMENTAL_USE_OPENJDK9=true make checkbuild docs"
and spot-checked some javac invocations to confirm that the
system modules libs were not included in the bootclasspath,
but other libs were.
Change-Id: I48fd11aac9b310bfa58dee0f9cfb3ef33f10bca8
Some of the dependencies for target-files-package were
refactored but only partially. Since these are satisfied
by transitive dependencies, the problem wasn't noticed
earlier.
Bug: N/A
Test: target-files-package finds dependencies
Change-Id: Ibdd7587f87735148e55d3b9835478075b6547efc
Restrictions on usage of private APIs require encoding new information
into the dex files of the boot class path. ART now contains a new build
tool called `hiddenapi` which takes three lists of class member
signatures (blacklist, light and dark greylist), finds the class members
in their respective dex files and modifies their access flags in place.
This patch invokes the `hiddenapi` tool on all JARs in
PRODUCT_BOOT_JARS. For Java libraries built with Makefiles the tool is
invoked after the dexer directly on DEX files. For Soong-built
libraries, the build system has to unzip the JAR produced by Soong,
apply `hiddenapi` and rezip again. This is due to the fact that the
PRODUCT_BOOT_JARS variable is not available to Soong.
Bug: 64382372
Test: m
Change-Id: I6ce897d204459c8b6f46ed49e0909ff76c08a9ed
Add variables for the hidden API blacklist and dark and light greylist
files. These are text files with signatures of boot class path class
member with restricted access for non-platform code.
Bug: 64382372
Test: m
Change-Id: I5639c5269872ac626efc69bd3a374a7a7125d502
Earlier, we would use the prebuilt directly, which doesn't go through
build optimizations like uncompressing the dex files.
bug: 63920015
Test: m -j32 && verify priv-apps prebuilt's vdex don't contain
the dex code.
Change-Id: I1a69d3b6832e3a940616f71a95c13091a4220b4b
We need to prevent vendors from using old VNDK and SystemSDK versions
for newly launched device. Otherwise, deprecating old versions of the
APIs would be really difficult.
Specifically, it is enforced that
PRODUCT_SHIPPING_API_LEVEL <= BOARD_VNDK_VERSION
PRODUCT_SHIPPING_API_LEVEL <= min(BOARD_SYSTEMSDK_VERSIONS)
Bug: 72126206
Test: manually setting BOARD_VNDK_VERSION to 24 in
device/google/wahoo/BoardConfig.mk and choosecom to walleye
Test: m -j shows an error on BOARD_VNDK_VERSION
Test: BOARD_SYSTEMSDK_VERSION="25 26" m -j shows an error on
BOARD_SYSTEMSDK_VERSION
Change-Id: I17646487a9c77a6a5110749e22ba47f0f75920cf
This file is used to verify that all differences between /system
boot*.art and /data/dalvik-cache boot*.art only consist of relocation
changes. The goal is to mitigate /data/dalvik-cache boot*.art files
being a persistence vector.
Test: Build and ensure boot.art.rel files are on the image.
Bug: 66697305
Change-Id: I68c445da46b2d3099ee347a04af4fbb84da85b0d
This reverts commit d616e0e820.
Reason for revert: b/72207339 - user builds are failing with this change
Bug: 72207339
Bug: 66697305
Change-Id: Ia309d04d837851ae11518f05aca468ecca895e6f
Test: Before the revert, taimen-user does not build. After the revert taimen-user builds
PLATFORM_SYSTEMSDK_VERSIONS is the list of System SDK versions that the
platform is supporting. Contrary to the public SDK where platform
essentially supports all previous SDK versions, platform support only a
few recent System SDK versions, since some of old System APIs are
gradually deprecated, removed from the following SDKs and then finally
deleted from the platform. This will be part of the framework manifest.
The list can be specified by setting PLATFORM_SYSTEMSDK_MIN_VERSION. If
it is set to an old version number, then System SDKs from the version
to the current version (PLATFORM_SDK_VERSION) are considered to be
supported by the platform. If PLATFORM_SYSTEMSDK_MIN_VERSION is not set,
only the latest System SDK version is supported.
Next, BOARD_SYSTEMSDK_VERSIONS is the list of System SDK versions that
the device is using. This is put to the device compatibility matrix
device is using. The device and the platform is considered as compatible
only BOARD_SYSTEMSDK_VERSIONS in the device compatibility matrix are
in the PLATFORM_SYSTEMSDK_VERSIONS in the framework manifest.
When BOARD_SYSTEMSDK_VERSIONS is set, a Java app or library in vendor or
odm partitions which didn't specify LOCAL_SDK_VERSION is forced to use
System SDK. Also, the build system does the additional integrity check
to ensure that LOCAL_SDK_VERSION is within BOARD_SYSTEMSDK_VERSIONS or
PLATFORM_SYSTEMSDK_VERSIONS (if BOARD_SYSTEMSDK_VERSIONS isn't set).
Bug: 69088799
Test: m -j
Test: BOARD_SYSTEMSDK_VERSIONS=P m -j
Change-Id: If4d59f6030e4cc402e015701d0caf94aeec37263
- Move logic to uncompress dexs in an APK in dex_preopt_odex_install.mk
and definitions.mk.
- Explicitly mark nostripping cases where dexpreopt will not embed the dex
file in the APK.
bug: 63920015
Test: m
Test: verify priv-apps dexs are uncompressed and unstripped
Test: Verify a non priv-app APK with uncompressed dex doesn't get stripped.
Change-Id: I624a03e3d965cebc0cae43fd6f7a6260178e6b8a
PRODUCT_COMPATIBLE_PROPERTY will be set as true for products shipping
with Android P, and ro.actionable_compatible_property.enabled will be set
as a system default property accordingly.
But if PRODUCT_ACTIONABLE_COMPATIBLE_PROPERTY_DISABLE is set,
ro.actionable_compatible_property.enabled will be false.
Bug: 38146102
Test: tested on walleye with PRODUCT_COMPATIBLE_PROPERTY_OVERRIDE=true
Merged-In: Ifc1279a360b140c4d94edd32db7de3c6c7317297
Change-Id: Ifc1279a360b140c4d94edd32db7de3c6c7317297
(cherry picked from commit 2528cd26c0)
This file is used to verify that all differences between /system
boot*.art and /data/dalvik-cache boot*.art only consist of relocation
changes. The goal is to mitigate /data/dalvik-cache boot*.art files
being a persistence vector.
Test: Build and ensure boot.art.rel files are on the image.
Bug: 66697305
Change-Id: I41486dd541e429f2f2b4c4a521009ee5e81e50b9
This causes an unsatisfiable dependency on frameworks/base.
Bug: 67663308
Test: make
(cherry-picked from commit 83bae866e6)
Change-Id: Iba0c3893b49386591f84eb893abf0b577b9716a0
Merged-In: I06e45e8ca66ab6547bf1df25119dda80d44e50f6
Use profile based boot image instead of the compiled classes version.
Results using profile from ag/2815407
Storage savings:
Boot art+vdex+oat size: 78305556->61539326 (-17MB)
New system server profile has more methods (1.5MB -> 4MB oat) since it
has more coverage from use cases.
Ram savings (sample averages for CUJ test):
Average PSS of 3 runs (~100 samples total):
.Boot_vdex: 9359K -> 9902K
.oat_mmap: 21857K -> 13372K
.Heap: 29779K -> 29483K
.Boot_art: 15736K -> 18096K
.LinearAlloc: 9112K -> 8965K
Total PSS: 85846K -> 79819K (-6MB)
Average pages read from flash (512MB device) for CUJ tests (3 samples):
1508MB -> 1183MB
Bug: 37966211
Test: make and flash
(cherry-picked from commit bf9023ae7a)
Change-Id: Icf3039ef871b68d806ef7b7c8ad2220c3a37ce90
Merged-In: Id8f9b08490a50e2ccf39e8acc00b0c4e8ec782b0
Enables signed and unsigned integer overflow sanitization on-by-default
for modules in frameworks/ and system/ by using the integer_overflow
sanitization setting. This applies sanitization to dynamically linked
binaries and shared libraries, and comes with a default set of regex for
functions to exclude from sanitization.
(see build/soong/cc/config/integer_overflow_blacklist.txt)
Prepare to enable minimal runtime diagnostics for integer overflow
sanitization on userdebug and eng builds.
Adds an additional Make and product variable pair to apply integer
overflow sanitization by default to additional code paths.
Bug: 30969751
Bug: 63927620
Test: Included paths are being sanitized.
Test: CTS test suite run on Pixel, runtime errors resolved.
Test: Performance impact in benchmarks acceptable.
Test: Boot-up successful on current Google devices.
Test: Teamfooded in diagnostics mode on Pixel for a month.
Test: Phone calls, camera photos + videos, bluetooth pairing.
Test: Wifi, work profiles, streaming videos, app installation.
Test: Split-screen, airplane mode, battery saver.
Test: Toggling accessibility settings.
Change-Id: Icc7a558c86f8655267afb4ca01b316773325c91a
Doclava now generates new API signature files. Add their filenames
to config.mk.
Bug: 64382372
Test: make
Change-Id: I6aed1b2e0575ddb8c89cea5fa7709023eb6f30a3
This build variable is unnecessary now that GCC is completely
unsupported for Android platform builds.
Bug: http://b/64032869
Test: Builds
Change-Id: I9d44ebd7129cb2bdfbb26d37922db19c6fb9efc9
libc++'s headers are annotated with availability macros that indicate
which version of Mac OS was the first to ship with a libc++ feature
available in its *system's* libc++.dylib. We do not use the system's
library, but rather ship our own. As such, these availability
attributes are meaningless for us but cause build breaks when we try
to use code that would not be available in the system's dylib.
Test: removed the darwin check and checked showcommands on linux
Bug: None
Change-Id: I5f69f337baa29fcdf84b5e2fe46bf4f85904d1cd
Specifying this option causes the make target's sources
and classpath dependencies to be patched into the specified
module via --patch-module (if the build toolchain implements
the module system, ie. when targeting 1.9). This allows tests
to compile that have classes in packages under test.
Currently, only java.base is on the module path (everything
else is on the classpath), and only when using -target 1.9
(ie. EXPERIMENTAL_USE_OPENJDK9=true).
Bug: 71699916
Test: Treehugger
Test: Checked that I was able to compile some ojluni & icu
test sources that had classes in (libcore) packages
under test.
Change-Id: I044fb9d904754ed281f574ab3dd52235cf0d7fec
This properly sets -format=gnu for all non-Darwin targets, because
llvm-ar is cross-platform (but defaults to the host environment).
Bug: http://b/71618641
Test: m checkbuild
Change-Id: Ib5d1c112a4ac792a11d1a0721ae6b0a5f9a42727
This architecture only existed for unbundled use, but even the NDK is
removing support in their r17 release, so just remove support for it.
Test: build/soong/build_test.bash -only-config
Test: check buildserver configs, I don't see anything still using armv5
Change-Id: Ic183b510c9ada94438bd4cc2b9362fa438a29ced
Currently we determine if a test artifact has multiple build
outputs by checking if it is NATIVE_TEST, NATIVE_BENCHMARK, or
LOCAL_MULTILIB and declare the module is_native.
This is technically not true therefore let's use a different
variable to determine if we need multiple architecture directories.
Bug: 71554249
Test: make sl4a -j ; Results in an autogenerated empty configuration file.
Change-Id: I17c8f3ef7e1cd502b57e25c71688f12a35a9787d
Static libraries using the system STL have type
native:ndk:system:none, not native:ndk:system:shared, since no
actually linking takes place. These are still allowed to be linked to
platform modules.
Test: make checkbuild
Bug: None
Change-Id: Ib999bb0f422ec3938daa740fca5681e153d6525a
llvm-ar is required for platform LTO build.
With GNU ar, we used crsPD flags, however 'P' flag is not supported by
llvm-ar. Since none of the platform archive files are built using third
party AR, the 'P' flag should not matter to us. However we've been
relying on one side-effect of 'P', where 'ar foo.a some/path/to/obj.o'
when foo.a already has a file named 'obj.o' will be an append operation,
regardless whether the existing one is the same (ar has no way of
telling the difference). We workaround this by always appending ('q'
flag) for llvm-ar. This may result in larger intermediate archive files
(more duplication) but will not affect the final build result.
Bug: 71618641
Test: m checkbuild
Change-Id: Id96a244cfe49cecfba08cc868dd18934ecb4ff23
* changes:
Clear resource_export_package outside conditionals
Fix adbs after move from build/tools to build/make/tools
Use proguard.jacoco.flags for EMMA_INSTRUMENT_STATIC
Yi Kong